Strategies for efficient resolution analysis in full-waveform inversion

NASA Astrophysics Data System (ADS)

Fichtner, A.; van Leeuwen, T.; Trampert, J.

2016-12-01

Full-waveform inversion is developing into a standard method in the seismological toolbox. It combines numerical wave propagation for heterogeneous media with adjoint techniques in order to improve tomographic resolution. However, resolution becomes increasingly difficult to quantify because of the enormous computational requirements. Here we present two families of methods that can be used for efficient resolution analysis in full-waveform inversion. They are based on the targeted extraction of resolution proxies from the Hessian matrix, which is too large to store and to compute explicitly. Fourier methods rest on the application of the Hessian to Earth models with harmonic oscillations. This yields the Fourier spectrum of the Hessian for few selected wave numbers, from which we can extract properties of the tomographic point-spread function for any point in space. Random probing methods use uncorrelated, random test models instead of harmonic oscillations. Auto-correlating the Hessian-model applications for sufficiently many test models also characterises the point-spread function. Both Fourier and random probing methods provide a rich collection of resolution proxies. These include position- and direction-dependent resolution lengths, and the volume of point-spread functions as indicator of amplitude recovery and inter-parameter trade-offs. The computational requirements of these methods are equivalent to approximately 7 conjugate-gradient iterations in full-waveform inversion. This is significantly less than the optimisation itself, which may require tens to hundreds of iterations to reach convergence. In addition to the theoretical foundations of the Fourier and random probing methods, we show various illustrative examples from real-data full-waveform inversion for crustal and mantle structure.

Frequency Domain Full-Waveform Inversion in Imaging Thrust Related Features

NASA Astrophysics Data System (ADS)

Jaiswal, P.; Zelt, C. A.

2010-12-01

Seismic acquisition in rough terrain such as mountain belts suffers from problems related to near-surface conditions such as statics, inconsistent energy penetration, rapid decay of signal, and imperfect receiver coupling. Moreover in the presence of weakly compacted soil, strong ground roll may obscure the reflection arrivals at near offsets further diminishing the scope of estimating a reliable near surface image though conventional processing. Traveltime and waveform inversion not only overcome the simplistic assumptions inherent in conventional processing such as hyperbolic moveout and convolution model, but also use parts of the seismic coda, such as the direct arrival and refractions, that are discarded in the latter. Traveltime and waveform inversion are model-based methods that honour the physics of wave propagation. Given the right set of preconditioned data and starting model, waveform inversion in particular has been realized as a powerful tool for velocity model building. This paper examines two case studies on waveform inversion using real data from the Naga Thrust Belt in the Northeast India. Waveform inversion in this paper is performed in the frequency domain and is multiscale in nature i.e., the inversion progressively ascends from the lower to the higher end of the frequency spectra increasing the wavenumber content of the recovered model. Since the real data are band limited, the success of waveform inversion depends on how well the starting model can account for the missing low wavenumbers. In this paper it is observed that the required starting model can be prepared using the regularized inversion of direct and reflected arrival times.

Seismic waveform inversion best practices: regional, global and exploration test cases

NASA Astrophysics Data System (ADS)

Modrak, Ryan; Tromp, Jeroen

2016-09-01

Reaching the global minimum of a waveform misfit function requires careful choices about the nonlinear optimization, preconditioning and regularization methods underlying an inversion. Because waveform inversion problems are susceptible to erratic convergence associated with strong nonlinearity, one or two test cases are not enough to reliably inform such decisions. We identify best practices, instead, using four seismic near-surface problems, one regional problem and two global problems. To make meaningful quantitative comparisons between methods, we carry out hundreds of inversions, varying one aspect of the implementation at a time. Comparing nonlinear optimization algorithms, we find that limited-memory BFGS provides computational savings over nonlinear conjugate gradient methods in a wide range of test cases. Comparing preconditioners, we show that a new diagonal scaling derived from the adjoint of the forward operator provides better performance than two conventional preconditioning schemes. Comparing regularization strategies, we find that projection, convolution, Tikhonov regularization and total variation regularization are effective in different contexts. Besides questions of one strategy or another, reliability and efficiency in waveform inversion depend on close numerical attention and care. Implementation details involving the line search and restart conditions have a strong effect on computational cost, regardless of the chosen nonlinear optimization algorithm.

Inversion of ocean-bottom seismometer (OBS) waveforms for oceanic crust structure: a synthetic study

NASA Astrophysics Data System (ADS)

Li, Xueyan; Wang, Yanbin; Chen, Yongshun John

2016-08-01

The waveform inversion method is applied—using synthetic ocean-bottom seismometer (OBS) data—to study oceanic crust structure. A niching genetic algorithm (NGA) is used to implement the inversion for the thickness and P-wave velocity of each layer, and to update the model by minimizing the objective function, which consists of the misfit and cross-correlation of observed and synthetic waveforms. The influence of specific NGA method parameters is discussed, and suitable values are presented. The NGA method works well for various observation systems, such as those with irregular and sparse distribution of receivers as well as single receiver systems. A strategy is proposed to accelerate the convergence rate by a factor of five with no increase in computational complexity; this is achieved using a first inversion with several generations to impose a restriction on the preset range of each parameter and then conducting a second inversion with the new range. Despite the successes of this method, its usage is limited. A shallow water layer is not favored because the direct wave in water will suppress the useful reflection signals from the crust. A more precise calculation of the air-gun source signal should be considered in order to better simulate waveforms generated in realistic situations; further studies are required to investigate this issue.

Full waveform inversion in the frequency domain using classified time-domain residual wavefields

NASA Astrophysics Data System (ADS)

Son, Woohyun; Koo, Nam-Hyung; Kim, Byoung-Yeop; Lee, Ho-Young; Joo, Yonghwan

2017-04-01

We perform the acoustic full waveform inversion in the frequency domain using residual wavefields that have been separated in the time domain. We sort the residual wavefields in the time domain according to the order of absolute amplitudes. Then, the residual wavefields are separated into several groups in the time domain. To analyze the characteristics of the residual wavefields, we compare the residual wavefields of conventional method with those of our residual separation method. From the residual analysis, the amplitude spectrum obtained from the trace before separation appears to have little energy at the lower frequency bands. However, the amplitude spectrum obtained from our strategy is regularized by the separation process, which means that the low-frequency components are emphasized. Therefore, our method helps to emphasize low-frequency components of residual wavefields. Then, we generate the frequency-domain residual wavefields by taking the Fourier transform of the separated time-domain residual wavefields. With these wavefields, we perform the gradient-based full waveform inversion in the frequency domain using back-propagation technique. Through a comparison of gradient directions, we confirm that our separation method can better describe the sub-salt image than the conventional approach. The proposed method is tested on the SEG/EAGE salt-dome model. The inversion results show that our algorithm is better than the conventional gradient based waveform inversion in the frequency domain, especially for deeper parts of the velocity model.

High resolution tsunami inversion for 2010 Chile earthquake

NASA Astrophysics Data System (ADS)

Wu, T.-R.; Ho, T.-C.

2011-12-01

We investigate the feasibility of inverting high-resolution vertical seafloor displacement from tsunami waveforms. An inversion method named "SUTIM" (small unit tsunami inversion method) is developed to meet this goal. In addition to utilizing the conventional least-square inversion, this paper also enhances the inversion resolution by Grid-Shifting method. A smooth constraint is adopted to gain stability. After a series of validation and performance tests, SUTIM is used to study the 2010 Chile earthquake. Based upon data quality and azimuthal distribution, we select tsunami waveforms from 6 GLOSS stations and 1 DART buoy record. In total, 157 sub-faults are utilized for the high-resolution inversion. The resolution reaches 10 sub-faults per wavelength. The result is compared with the distribution of the aftershocks and waveforms at each gauge location with very good agreement. The inversion result shows that the source profile features a non-uniform distribution of the seafloor displacement. The highly elevated vertical seafloor is mainly concentrated in two areas: one is located in the northern part of the epicentre, between 34° S and 36° S; the other is in the southern part, between 37° S and 38° S.

Frequency domain, waveform inversion of laboratory crosswell radar data

USGS Publications Warehouse

Ellefsen, Karl J.; Mazzella, Aldo T.; Horton, Robert J.; McKenna, Jason R.

2010-01-01

A new waveform inversion for crosswell radar is formulated in the frequency-domain for a 2.5D model. The inversion simulates radar waves using the vector Helmholtz equation for electromagnetic waves. The objective function is minimized using a backpropagation method suitable for a 2.5D model. The inversion is tested by processing crosswell radar data collected in a laboratory tank. The estimated model is consistent with the known electromagnetic properties of the tank. The formulation for the 2.5D model can be extended to inversions of acoustic and elastic data.

Virtual Seismic Observation (VSO) with Sparsity-Promotion Inversion

NASA Astrophysics Data System (ADS)

Tiezhao, B.; Ning, J.; Jianwei, M.

2017-12-01

Large station interval leads to low resolution images, sometimes prevents people from obtaining images in concerned regions. Sparsity-promotion inversion, a useful method to recover missing data in industrial field acquisition, can be lent to interpolate seismic data on none-sampled sites, forming Virtual Seismic Observation (VSO). Traditional sparsity-promotion inversion suffers when coming up with large time difference in adjacent sites, which we concern most and use shift method to improve it. The procedure of the interpolation is that we first employ low-pass filter to get long wavelength waveform data and shift the waveforms of the same wave in different seismograms to nearly same arrival time. Then we use wavelet-transform-based sparsity-promotion inversion to interpolate waveform data on none-sampled sites and filling a phase in each missing trace. Finally, we shift back the waveforms to their original arrival times. We call our method FSIS (Filtering, Shift, Interpolation, Shift) interpolation. By this way, we can insert different virtually observed seismic phases into none-sampled sites and get dense seismic observation data. For testing our method, we randomly hide the real data in a site and use the rest to interpolate the observation on that site, using direct interpolation or FSIS method. Compared with directly interpolated data, interpolated data with FSIS can keep amplitude better. Results also show that the arrival times and waveforms of those VSOs well express the real data, which convince us that our method to form VSOs are applicable. In this way, we can provide needed data for some advanced seismic technique like RTM to illuminate shallow structures.

Reflection full-waveform inversion using a modified phase misfit function

NASA Astrophysics Data System (ADS)

Cui, Chao; Huang, Jian-Ping; Li, Zhen-Chun; Liao, Wen-Yuan; Guan, Zhe

2017-09-01

Reflection full-waveform inversion (RFWI) updates the low- and highwavenumber components, and yields more accurate initial models compared with conventional full-waveform inversion (FWI). However, there is strong nonlinearity in conventional RFWI because of the lack of low-frequency data and the complexity of the amplitude. The separation of phase and amplitude information makes RFWI more linear. Traditional phase-calculation methods face severe phase wrapping. To solve this problem, we propose a modified phase-calculation method that uses the phase-envelope data to obtain the pseudo phase information. Then, we establish a pseudophase-information-based objective function for RFWI, with the corresponding source and gradient terms. Numerical tests verify that the proposed calculation method using the phase-envelope data guarantees the stability and accuracy of the phase information and the convergence of the objective function. The application on a portion of the Sigsbee2A model and comparison with inversion results of the improved RFWI and conventional FWI methods verify that the pseudophase-based RFWI produces a highly accurate and efficient velocity model. Moreover, the proposed method is robust to noise and high frequency.

Centroid-moment tensor inversions using high-rate GPS waveforms

NASA Astrophysics Data System (ADS)

O'Toole, Thomas B.; Valentine, Andrew P.; Woodhouse, John H.

2012-10-01

Displacement time-series recorded by Global Positioning System (GPS) receivers are a new type of near-field waveform observation of the seismic source. We have developed an inversion method which enables the recovery of an earthquake's mechanism and centroid coordinates from such data. Our approach is identical to that of the 'classical' Centroid-Moment Tensor (CMT) algorithm, except that we forward model the seismic wavefield using a method that is amenable to the efficient computation of synthetic GPS seismograms and their partial derivatives. We demonstrate the validity of our approach by calculating CMT solutions using 1 Hz GPS data for two recent earthquakes in Japan. These results are in good agreement with independently determined source models of these events. With wider availability of data, we envisage the CMT algorithm providing a tool for the systematic inversion of GPS waveforms, as is already the case for teleseismic data. Furthermore, this general inversion method could equally be applied to other near-field earthquake observations such as those made using accelerometers.

Extracting Low-Frequency Information from Time Attenuation in Elastic Waveform Inversion

NASA Astrophysics Data System (ADS)

Guo, Xuebao; Liu, Hong; Shi, Ying; Wang, Weihong

2017-03-01

Low-frequency information is crucial for recovering background velocity, but the lack of low-frequency information in field data makes inversion impractical without accurate initial models. Laplace-Fourier domain waveform inversion can recover a smooth model from real data without low-frequency information, which can be used for subsequent inversion as an ideal starting model. In general, it also starts with low frequencies and includes higher frequencies at later inversion stages, while the difference is that its ultralow frequency information comes from the Laplace-Fourier domain. Meanwhile, a direct implementation of the Laplace-transformed wavefield using frequency domain inversion is also very convenient. However, because broad frequency bands are often used in the pure time domain waveform inversion, it is difficult to extract the wavefields dominated by low frequencies in this case. In this paper, low-frequency components are constructed by introducing time attenuation into the recorded residuals, and the rest of the method is identical to the traditional time domain inversion. Time windowing and frequency filtering are also applied to mitigate the ambiguity of the inverse problem. Therefore, we can start at low frequencies and to move to higher frequencies. The experiment shows that the proposed method can achieve a good inversion result in the presence of a linear initial model and records without low-frequency information.

Infrasound Waveform Inversion and Mass Flux Validation from Sakurajima Volcano, Japan

NASA Astrophysics Data System (ADS)

Fee, D.; Kim, K.; Yokoo, A.; Izbekov, P. E.; Lopez, T. M.; Prata, F.; Ahonen, P.; Kazahaya, R.; Nakamichi, H.; Iguchi, M.

2015-12-01

Recent advances in numerical wave propagation modeling and station coverage have permitted robust inversion of infrasound data from volcanic explosions. Complex topography and crater morphology have been shown to substantially affect the infrasound waveform, suggesting that homogeneous acoustic propagation assumptions are invalid. Infrasound waveform inversion provides an exciting tool to accurately characterize emission volume and mass flux from both volcanic and non-volcanic explosions. Mass flux, arguably the most sought-after parameter from a volcanic eruption, can be determined from the volume flux using infrasound waveform inversion if the volcanic flow is well-characterized. Thus far, infrasound-based volume and mass flux estimates have yet to be validated. In February 2015 we deployed six infrasound stations around the explosive Sakurajima Volcano, Japan for 8 days. Here we present our full waveform inversion method and volume and mass flux estimates of numerous high amplitude explosions using a high resolution DEM and 3-D Finite Difference Time Domain modeling. Application of this technique to volcanic eruptions may produce realistic estimates of mass flux and plume height necessary for volcanic hazard mitigation. Several ground-based instruments and methods are used to independently determine the volume, composition, and mass flux of individual volcanic explosions. Specifically, we use ground-based ash sampling, multispectral infrared imagery, UV spectrometry, and multigas data to estimate the plume composition and flux. Unique tiltmeter data from underground tunnels at Sakurajima also provides a way to estimate the volume and mass of each explosion. In this presentation we compare the volume and mass flux estimates derived from the different methods and discuss sources of error and future improvements.

A Gauss-Newton full-waveform inversion in PML-truncated domains using scalar probing waves

NASA Astrophysics Data System (ADS)

Pakravan, Alireza; Kang, Jun Won; Newtson, Craig M.

2017-12-01

This study considers the characterization of subsurface shear wave velocity profiles in semi-infinite media using scalar waves. Using surficial responses caused by probing waves, a reconstruction of the material profile is sought using a Gauss-Newton full-waveform inversion method in a two-dimensional domain truncated by perfectly matched layer (PML) wave-absorbing boundaries. The PML is introduced to limit the semi-infinite extent of the half-space and to prevent reflections from the truncated boundaries. A hybrid unsplit-field PML is formulated in the inversion framework to enable more efficient wave simulations than with a fully mixed PML. The full-waveform inversion method is based on a constrained optimization framework that is implemented using Karush-Kuhn-Tucker (KKT) optimality conditions to minimize the objective functional augmented by PML-endowed wave equations via Lagrange multipliers. The KKT conditions consist of state, adjoint, and control problems, and are solved iteratively to update the shear wave velocity profile of the PML-truncated domain. Numerical examples show that the developed Gauss-Newton inversion method is accurate enough and more efficient than another inversion method. The algorithm's performance is demonstrated by the numerical examples including the case of noisy measurement responses and the case of reduced number of sources and receivers.

FWT2D: A massively parallel program for frequency-domain full-waveform tomography of wide-aperture seismic data—Part 1: Algorithm

NASA Astrophysics Data System (ADS)

Sourbier, Florent; Operto, Stéphane; Virieux, Jean; Amestoy, Patrick; L'Excellent, Jean-Yves

2009-03-01

This is the first paper in a two-part series that describes a massively parallel code that performs 2D frequency-domain full-waveform inversion of wide-aperture seismic data for imaging complex structures. Full-waveform inversion methods, namely quantitative seismic imaging methods based on the resolution of the full wave equation, are computationally expensive. Therefore, designing efficient algorithms which take advantage of parallel computing facilities is critical for the appraisal of these approaches when applied to representative case studies and for further improvements. Full-waveform modelling requires the resolution of a large sparse system of linear equations which is performed with the massively parallel direct solver MUMPS for efficient multiple-shot simulations. Efficiency of the multiple-shot solution phase (forward/backward substitutions) is improved by using the BLAS3 library. The inverse problem relies on a classic local optimization approach implemented with a gradient method. The direct solver returns the multiple-shot wavefield solutions distributed over the processors according to a domain decomposition driven by the distribution of the LU factors. The domain decomposition of the wavefield solutions is used to compute in parallel the gradient of the objective function and the diagonal Hessian, this latter providing a suitable scaling of the gradient. The algorithm allows one to test different strategies for multiscale frequency inversion ranging from successive mono-frequency inversion to simultaneous multifrequency inversion. These different inversion strategies will be illustrated in the following companion paper. The parallel efficiency and the scalability of the code will also be quantified.

Imaging Crustal Structure with Waveform and HV Ratio of Body-wave Receiver Function

NASA Astrophysics Data System (ADS)

Chong, J.; Chu, R.; Ni, S.; Meng, Q.; Guo, A.

2017-12-01

It is known that receiver function has less constraint on the absolute velocity, and joint inversion of receiver function and surface wave dispersion has been widely applied to reduce the non-uniqueness of velocity and interface depth. However, some studies indicate that the receiver function itself is capable for determining the absolute shear wave velocity. In this study, we propose to measure the receiver function HV ratio which takes advantage of the amplitude information of the radial and vertical receiver functions to constrain the shear-wave velocity. Numerical analysis indicates that the receiver function HV ratio is sensitive to the average shear wave velocity in the depth range it samples, and can help to reduce the non-uniqueness of receiver function waveform inversion. A joint inversion scheme has been developed, and both synthetic tests and real data application proved the feasibility of the joint inversion. The method has been applied to the dense seismic array of ChinArray program in SE Tibet during the time period from August 2011 to August 2012 in SE Tibet (ChinArray-Himalaya, 2011). The measurements of receiver function HV ratio reveals the lateral variation of the tectonics in of the study region. And main features of the velocity structure imagined by the new joint inversion method are consistent with previous studies. KEYWORDS: receiver function HV ratio, receiver function waveform inversion, crustal structure ReferenceChinArray-Himalaya. 2011. China Seismic Array waveform data of Himalaya Project. Institute of Geophysics, China Earthquake Administration. doi:10.12001/ChinArray.Data. Himalaya. Jiajun Chong, Risheng Chu*, Sidao Ni, Qingjun Meng, Aizhi Guo, 2017. Receiver Function HV Ratio, a New Measurement for Reducing Non-uniqueness of Receiver Function Waveform Inversion. (under revision)

Crustal velocity structure of central Gansu Province from regional seismic waveform inversion using firework algorithm

NASA Astrophysics Data System (ADS)

Chen, Yanyang; Wang, Yanbin; Zhang, Yuansheng

2017-04-01

The firework algorithm (FWA) is a novel swarm intelligence-based method recently proposed for the optimization of multi-parameter, nonlinear functions. Numerical waveform inversion experiments using a synthetic model show that the FWA performs well in both solution quality and efficiency. We apply the FWA in this study to crustal velocity structure inversion using regional seismic waveform data of central Gansu on the northeastern margin of the Qinghai-Tibet plateau. Seismograms recorded from the moment magnitude ( M W) 5.4 Minxian earthquake enable obtaining an average crustal velocity model for this region. We initially carried out a series of FWA robustness tests in regional waveform inversion at the same earthquake and station positions across the study region, inverting two velocity structure models, with and without a low-velocity crustal layer; the accuracy of our average inversion results and their standard deviations reveal the advantages of the FWA for the inversion of regional seismic waveforms. We applied the FWA across our study area using three component waveform data recorded by nine broadband permanent seismic stations with epicentral distances ranging between 146 and 437 km. These inversion results show that the average thickness of the crust in this region is 46.75 km, while thicknesses of the sedimentary layer, and the upper, middle, and lower crust are 3.15, 15.69, 13.08, and 14.83 km, respectively. Results also show that the P-wave velocities of these layers and the upper mantle are 4.47, 6.07, 6.12, 6.87, and 8.18 km/s, respectively.

Full-waveform inversion of GPR data for civil engineering applications

NASA Astrophysics Data System (ADS)

van der Kruk, Jan; Kalogeropoulos, Alexis; Hugenschmidt, Johannes; Klotzsche, Anja; Busch, Sebastian; Vereecken, Harry

2014-05-01

Conventional GPR ray-based techniques are often limited in their capability to image complex structures due to the pertaining approximations. Due to the increased computational power, it is becoming more easy to use modeling and inversion tools that explicitly take into account the detailed electromagnetic wave propagation characteristics. In this way, new civil engineering application avenues are opening up that enable an improved high resolution imaging of quantitative medium properties. In this contribution, we show recent developments that enable the full-waveform inversion of off-ground, on-ground and crosshole GPR data. For a successful inversion, a proper start model must be used that generates synthetic data that overlaps the measured data with at least half a wavelength. In addition, the GPR system must be calibrated such that an effective wavelet is obtained that encompasses the complexity of the GPR source and receiver antennas. Simple geometries such as horizontal layers can be described with a limited number of model parameters, which enable the use of a combined global and local search using the Simplex search algorithm. This approach has been implemented for the full-waveform inversion of off-ground and on-ground GPR data measured over horizontally layered media. In this way, an accurate 3D frequency domain forward model of Maxwell's equation can be used where the integral representation of the electric field is numerically evaluated. The full-waveform inversion (FWI) for a large number of unknowns uses gradient-based optimization methods where a 3D to 2D conversion is used to apply this method to experimental data. Off-ground GPR data, measured over homogeneous concrete specimens, were inverted using the full-waveform inversion. In contrast to traditional ray-based techniques we were able to obtain quantitative values for the permittivity and conductivity and in this way distinguish between moisture and chloride effects. For increasing chloride content increasing frequency-dependent conductivity values were obtained. The off-ground full-waveform inversion was extended to invert for positive and negative gradients in conductivity and the conductivity gradient direction could be correctly identified. Experimental specimen containing gradients were generated by exposing a concrete slab to controlled wetting-drying cycles using a saline solution. Full-waveform inversion of the measured data correctly identified the conductivity gradient direction which was confirmed by destructive analysis. On-ground CMP GPR data measured over a concrete layer overlying a metal plate show interfering multiple reflections, which indicates that the structure acts as a waveguide. Calculation of the phase-velocity spectrum shows the presence of several higher order modes. Whereas the dispersion inversion returns the thickness and layer height, the full-waveform inversion was also able to estimate quantitative conductivity values. This abstract is a contribution to COST Action TU1208

Efficient calculation of full waveform time domain inversion for electromagnetic problem using fictitious wave domain method and cascade decimation decomposition

NASA Astrophysics Data System (ADS)

Imamura, N.; Schultz, A.

2016-12-01

Recently, a full waveform time domain inverse solution has been developed for the magnetotelluric (MT) and controlled-source electromagnetic (CSEM) methods. The ultimate goal of this approach is to obtain a computationally tractable direct waveform joint inversion to solve simultaneously for source fields and earth conductivity structure in three and four dimensions. This is desirable on several grounds, including the improved spatial resolving power expected from use of a multitude of source illuminations, the ability to operate in areas of high levels of source signal spatial complexity, and non-stationarity. This goal would not be obtainable if one were to adopt the pure time domain solution for the inverse problem. This is particularly true for the case of MT surveys, since an enormous number of degrees of freedom are required to represent the observed MT waveforms across a large frequency bandwidth. This means that for the forward simulation, the smallest time steps should be finer than that required to represent the highest frequency, while the number of time steps should also cover the lowest frequency. This leads to a sensitivity matrix that is computationally burdensome to solve a model update. We have implemented a code that addresses this situation through the use of cascade decimation decomposition to reduce the size of the sensitivity matrix substantially, through quasi-equivalent time domain decomposition. We also use a fictitious wave domain method to speed up computation time of the forward simulation in the time domain. By combining these refinements, we have developed a full waveform joint source field/earth conductivity inverse modeling method. We found that cascade decimation speeds computations of the sensitivity matrices dramatically, keeping the solution close to that of the undecimated case. For example, for a model discretized into 2.6x105 cells, we obtain model updates in less than 1 hour on a 4U rack-mounted workgroup Linux server, which is a practical computational time for the inverse problem.

Time-lapse seismic waveform inversion for monitoring near-surface microbubble injection

NASA Astrophysics Data System (ADS)

Kamei, R.; Jang, U.; Lumley, D. E.; Mouri, T.; Nakatsukasa, M.; Takanashi, M.

2016-12-01

Seismic monitoring of the Earth provides valuable information regarding the time-varying changes in subsurface physical properties that are caused by natural or man-made processes. However, the resulting changes in subsurface properties are often small both in terms of magnitude and spatial extent, leading to seismic data differences that are difficult to detect at typical non-repeatable noise levels. In order to better extract information from the time-lapse data, exploiting the full seismic waveform information can be critical, since detected amplitude or traveltime changes may be minimal. We explore methods of waveform inversion that estimate an optimal model of time-varying elastic parameters at the wavelength scale to fit the observed time-lapse seismic data with modelled waveforms based on numerical solutions of the wave equation. We apply acoustic waveform inversion to time-lapse cross-well monitoring surveys of 64-m well intervals, and estimate the velocity changes that occur during the injection of microbubble water into shallow unconsolidated Quaternary sediments in the Kanto basin of Japan at a depth of 25 m below the surface. Microbubble water is comprised of water infused with air bubbles of a diameter less than 0.1mm, and may be useful to improve resistance to ground liquefaction during major earthquakes. Monitoring the space-time distribution and physical properties of microbubble injection is therefore important to understanding the full potential of the technique. Repeated monitoring surveys (>10) reveal transient behaviours in waveforms during microbubble injection. Time-lapse waveform inversion detects changes in P-wave velocity of less than 1 percent, initially as velocity increases and subsequently as velocity decreases. The velocity changes are mainly imaged within a thin (1 m) layer between the injection and the receiver well, inferring the fluid-flow influence of the fluvial sediment depositional environment. The resulting velocity models fit the observed waveforms very well, supporting the validity of the estimated velocity changes. In order to further improve the estimation of velocity changes, we investigate the limitations of acoustic waveform inversion, and apply elastic waveform inversion to the time-lapse data set.

Full Waveform Inversion with Multisource Frequency Selection of Marine Streamer Data

DOE Office of Scientific and Technical Information (OSTI.GOV)

Huang, Yunsong; Schuster, Gerard T.

The theory and practice of multisource full waveform inversion of marine supergathers are described with a frequency-selection strategy. The key enabling property of frequency selection is that it eliminates the crosstalk among sources, thus overcoming the aperture mismatch of marine multisource inversion. Tests on multisource full waveform inversion of synthetic marine data and Gulf of Mexico data show speedups of 4× and 8×, respectively, compared to conventional full waveform inversion.

Full Waveform Inversion with Multisource Frequency Selection of Marine Streamer Data

DOE PAGES

Huang, Yunsong; Schuster, Gerard T.

2017-10-26

The theory and practice of multisource full waveform inversion of marine supergathers are described with a frequency-selection strategy. The key enabling property of frequency selection is that it eliminates the crosstalk among sources, thus overcoming the aperture mismatch of marine multisource inversion. Tests on multisource full waveform inversion of synthetic marine data and Gulf of Mexico data show speedups of 4× and 8×, respectively, compared to conventional full waveform inversion.

Localized time-lapse elastic waveform inversion using wavefield injection and extrapolation: 2-D parametric studies

NASA Astrophysics Data System (ADS)

Yuan, Shihao; Fuji, Nobuaki; Singh, Satish; Borisov, Dmitry

2017-06-01

We present a methodology to invert seismic data for a localized area by combining source-side wavefield injection and receiver-side extrapolation method. Despite the high resolving power of seismic full waveform inversion, the computational cost for practical scale elastic or viscoelastic waveform inversion remains a heavy burden. This can be much more severe for time-lapse surveys, which require real-time seismic imaging on a daily or weekly basis. Besides, changes of the structure during time-lapse surveys are likely to occur in a small area rather than the whole region of seismic experiments, such as oil and gas reservoir or CO2 injection wells. We thus propose an approach that allows to image effectively and quantitatively the localized structure changes far deep from both source and receiver arrays. In our method, we perform both forward and back propagation only inside the target region. First, we look for the equivalent source expression enclosing the region of interest by using the wavefield injection method. Second, we extrapolate wavefield from physical receivers located near the Earth's surface or on the ocean bottom to an array of virtual receivers in the subsurface by using correlation-type representation theorem. In this study, we present various 2-D elastic numerical examples of the proposed method and quantitatively evaluate errors in obtained models, in comparison to those of conventional full-model inversions. The results show that the proposed localized waveform inversion is not only efficient and robust but also accurate even under the existence of errors in both initial models and observed data.

Full waveform inversion using a decomposed single frequency component from a spectrogram

NASA Astrophysics Data System (ADS)

Ha, Jiho; Kim, Seongpil; Koo, Namhyung; Kim, Young-Ju; Woo, Nam-Sub; Han, Sang-Mok; Chung, Wookeen; Shin, Sungryul; Shin, Changsoo; Lee, Jaejoon

2018-06-01

Although many full waveform inversion methods have been developed to construct velocity models of subsurface, various approaches have been presented to obtain an inversion result with long-wavelength features even though seismic data lacking low-frequency components were used. In this study, a new full waveform inversion algorithm was proposed to recover a long-wavelength velocity model that reflects the inherent characteristics of each frequency component of seismic data using a single-frequency component decomposed from the spectrogram. We utilized the wavelet transform method to obtain the spectrogram, and the decomposed signal from the spectrogram was used as transformed data. The Gauss-Newton method with the diagonal elements of an approximate Hessian matrix was used to update the model parameters at each iteration. Based on the results of time-frequency analysis in the spectrogram, numerical tests with some decomposed frequency components were performed using a modified SEG/EAGE salt dome (A-A‧) line to demonstrate the feasibility of the proposed inversion algorithm. This demonstrated that a reasonable inverted velocity model with long-wavelength structures can be obtained using a single frequency component. It was also confirmed that when strong noise occurs in part of the frequency band, it is feasible to obtain a long-wavelength velocity model from the noise data with a frequency component that is less affected by the noise. Finally, it was confirmed that the results obtained from the spectrogram inversion can be used as an initial velocity model in conventional inversion methods.

The Modularized Software Package ASKI - Full Waveform Inversion Based on Waveform Sensitivity Kernels Utilizing External Seismic Wave Propagation Codes

NASA Astrophysics Data System (ADS)

Schumacher, F.; Friederich, W.

2015-12-01

We present the modularized software package ASKI which is a flexible and extendable toolbox for seismic full waveform inversion (FWI) as well as sensitivity or resolution analysis operating on the sensitivity matrix. It utilizes established wave propagation codes for solving the forward problem and offers an alternative to the monolithic, unflexible and hard-to-modify codes that have typically been written for solving inverse problems. It is available under the GPL at www.rub.de/aski. The Gauss-Newton FWI method for 3D-heterogeneous elastic earth models is based on waveform sensitivity kernels and can be applied to inverse problems at various spatial scales in both Cartesian and spherical geometries. The kernels are derived in the frequency domain from Born scattering theory as the Fréchet derivatives of linearized full waveform data functionals, quantifying the influence of elastic earth model parameters on the particular waveform data values. As an important innovation, we keep two independent spatial descriptions of the earth model - one for solving the forward problem and one representing the inverted model updates. Thereby we account for the independent needs of spatial model resolution of forward and inverse problem, respectively. Due to pre-integration of the kernels over the (in general much coarser) inversion grid, storage requirements for the sensitivity kernels are dramatically reduced.ASKI can be flexibly extended to other forward codes by providing it with specific interface routines that contain knowledge about forward code-specific file formats and auxiliary information provided by the new forward code. In order to sustain flexibility, the ASKI tools must communicate via file output/input, thus large storage capacities need to be accessible in a convenient way. Storing the complete sensitivity matrix to file, however, permits the scientist full manual control over each step in a customized procedure of sensitivity/resolution analysis and full waveform inversion.

Joint Inversion of Source Location and Source Mechanism of Induced Microseismics

NASA Astrophysics Data System (ADS)

Liang, C.

2014-12-01

Seismic source mechanism is a useful property to indicate the source physics and stress and strain distribution in regional, local and micro scales. In this study we jointly invert source mechanisms and locations for microseismics induced in fluid fracturing treatment in the oil and gas industry. For the events that are big enough to see waveforms, there are quite a few techniques can be applied to invert the source mechanism including waveform inversion, first polarity inversion and many other methods and variants based on these methods. However, for events that are too small to identify in seismic traces such as the microseismics induced by the fluid fracturing in the Oil and Gas industry, a source scanning algorithms (SSA for short) with waveform stacking are usually applied. At the same time, a joint inversion of location and source mechanism are possible but at a cost of high computation budget. The algorithm is thereby called Source Location and Mechanism Scanning Algorithm, SLMSA for short. In this case, for given velocity structure, all possible combinations of source locations (X,Y and Z) and source mechanism (Strike, Dip and Rake) are used to compute travel-times and polarities of waveforms. Correcting Normal moveout times and polarities, and stacking all waveforms, the (X, Y, Z , strike, dip, rake) combination that gives the strongest stacking waveform is identified as the solution. To solve the problem of high computation problem, CPU-GPU programing is applied. Numerical datasets are used to test the algorithm. The SLMSA has also been applied to a fluid fracturing datasets and reveal several advantages against the location only method: (1) for shear sources, the source only program can hardly locate them because of the canceling out of positive and negative polarized traces, but the SLMSA method can successfully pick up those events; (2) microseismic locations alone may not be enough to indicate the directionality of micro-fractures. The statistics of source mechanisms can certainly provide more knowledges on the orientation of fractures; (3) in our practice, the joint inversion method almost always yield more events than the source only method and for those events that are also picked by the SSA method, the stacking power of SLMSA are always higher than the ones obtained in SSA.

Waveform inversion of acoustic waves for explosion yield estimation

DOE PAGES

Kim, K.; Rodgers, A. J.

2016-07-08

We present a new waveform inversion technique to estimate the energy of near-surface explosions using atmospheric acoustic waves. Conventional methods often employ air blast models based on a homogeneous atmosphere, where the acoustic wave propagation effects (e.g., refraction and diffraction) are not taken into account, and therefore, their accuracy decreases with increasing source-receiver distance. In this study, three-dimensional acoustic simulations are performed with a finite difference method in realistic atmospheres and topography, and the modeled acoustic Green's functions are incorporated into the waveform inversion for the acoustic source time functions. The strength of the acoustic source is related to explosionmore » yield based on a standard air blast model. The technique was applied to local explosions (<10 km) and provided reasonable yield estimates (<~30% error) in the presence of realistic topography and atmospheric structure. In conclusion, the presented method can be extended to explosions recorded at far distance provided proper meteorological specifications.« less

Waveform inversion of acoustic waves for explosion yield estimation

DOE Office of Scientific and Technical Information (OSTI.GOV)

Kim, K.; Rodgers, A. J.

We present a new waveform inversion technique to estimate the energy of near-surface explosions using atmospheric acoustic waves. Conventional methods often employ air blast models based on a homogeneous atmosphere, where the acoustic wave propagation effects (e.g., refraction and diffraction) are not taken into account, and therefore, their accuracy decreases with increasing source-receiver distance. In this study, three-dimensional acoustic simulations are performed with a finite difference method in realistic atmospheres and topography, and the modeled acoustic Green's functions are incorporated into the waveform inversion for the acoustic source time functions. The strength of the acoustic source is related to explosionmore » yield based on a standard air blast model. The technique was applied to local explosions (<10 km) and provided reasonable yield estimates (<~30% error) in the presence of realistic topography and atmospheric structure. In conclusion, the presented method can be extended to explosions recorded at far distance provided proper meteorological specifications.« less

The New Method of Tsunami Source Reconstruction With r-Solution Inversion Method

NASA Astrophysics Data System (ADS)

Voronina, T. A.; Romanenko, A. A.

2016-12-01

Application of the r-solution method to reconstructing the initial tsunami waveform is discussed. This methodology is based on the inversion of remote measurements of water-level data. The wave propagation is considered within the scope of a linear shallow-water theory. The ill-posed inverse problem in question is regularized by means of a least square inversion using the truncated Singular Value Decomposition method. As a result of the numerical process, an r-solution is obtained. The method proposed allows one to control the instability of a numerical solution and to obtain an acceptable result in spite of ill posedness of the problem. Implementation of this methodology to reconstructing of the initial waveform to 2013 Solomon Islands tsunami validates the theoretical conclusion for synthetic data and a model tsunami source: the inversion result strongly depends on data noisiness, the azimuthal and temporal coverage of recording stations with respect to the source area. Furthermore, it is possible to make a preliminary selection of the most informative set of the available recording stations used in the inversion process.

Workflows for Full Waveform Inversions

NASA Astrophysics Data System (ADS)

Boehm, Christian; Krischer, Lion; Afanasiev, Michael; van Driel, Martin; May, Dave A.; Rietmann, Max; Fichtner, Andreas

2017-04-01

Despite many theoretical advances and the increasing availability of high-performance computing clusters, full seismic waveform inversions still face considerable challenges regarding data and workflow management. While the community has access to solvers which can harness modern heterogeneous computing architectures, the computational bottleneck has fallen to these often manpower-bounded issues that need to be overcome to facilitate further progress. Modern inversions involve huge amounts of data and require a tight integration between numerical PDE solvers, data acquisition and processing systems, nonlinear optimization libraries, and job orchestration frameworks. To this end we created a set of libraries and applications revolving around Salvus (http://salvus.io), a novel software package designed to solve large-scale full waveform inverse problems. This presentation focuses on solving passive source seismic full waveform inversions from local to global scales with Salvus. We discuss (i) design choices for the aforementioned components required for full waveform modeling and inversion, (ii) their implementation in the Salvus framework, and (iii) how it is all tied together by a usable workflow system. We combine state-of-the-art algorithms ranging from high-order finite-element solutions of the wave equation to quasi-Newton optimization algorithms using trust-region methods that can handle inexact derivatives. All is steered by an automated interactive graph-based workflow framework capable of orchestrating all necessary pieces. This naturally facilitates the creation of new Earth models and hopefully sparks new scientific insights. Additionally, and even more importantly, it enhances reproducibility and reliability of the final results.

Modularized seismic full waveform inversion based on waveform sensitivity kernels - The software package ASKI

NASA Astrophysics Data System (ADS)

Schumacher, Florian; Friederich, Wolfgang; Lamara, Samir; Gutt, Phillip; Paffrath, Marcel

2015-04-01

We present a seismic full waveform inversion concept for applications ranging from seismological to enineering contexts, based on sensitivity kernels for full waveforms. The kernels are derived from Born scattering theory as the Fréchet derivatives of linearized frequency-domain full waveform data functionals, quantifying the influence of elastic earth model parameters and density on the data values. For a specific source-receiver combination, the kernel is computed from the displacement and strain field spectrum originating from the source evaluated throughout the inversion domain, as well as the Green function spectrum and its strains originating from the receiver. By storing the wavefield spectra of specific sources/receivers, they can be re-used for kernel computation for different specific source-receiver combinations, optimizing the total number of required forward simulations. In the iterative inversion procedure, the solution of the forward problem, the computation of sensitivity kernels and the derivation of a model update is held completely separate. In particular, the model description for the forward problem and the description of the inverted model update are kept independent. Hence, the resolution of the inverted model as well as the complexity of solving the forward problem can be iteratively increased (with increasing frequency content of the inverted data subset). This may regularize the overall inverse problem and optimizes the computational effort of both, solving the forward problem and computing the model update. The required interconnection of arbitrary unstructured volume and point grids is realized by generalized high-order integration rules and 3D-unstructured interpolation methods. The model update is inferred solving a minimization problem in a least-squares sense, resulting in Gauss-Newton convergence of the overall inversion process. The inversion method was implemented in the modularized software package ASKI (Analysis of Sensitivity and Kernel Inversion), which provides a generalized interface to arbitrary external forward modelling codes. So far, the 3D spectral-element code SPECFEM3D (Tromp, Komatitsch and Liu, 2008) and the 1D semi-analytical code GEMINI (Friederich and Dalkolmo, 1995) in both, Cartesian and spherical framework are supported. The creation of interfaces to further forward codes is planned in the near future. ASKI is freely available under the terms of the GPL at www.rub.de/aski . Since the independent modules of ASKI must communicate via file output/input, large storage capacities need to be accessible conveniently. Storing the complete sensitivity matrix to file, however, permits the scientist full manual control over each step in a customized procedure of sensitivity/resolution analysis and full waveform inversion. In the presentation, we will show some aspects of the theory behind the full waveform inversion method and its practical realization by the software package ASKI, as well as synthetic and real-data applications from different scales and geometries.

Eruption mass estimation using infrasound waveform inversion and ash and gas measurements: Evaluation at Sakurajima Volcano, Japan

NASA Astrophysics Data System (ADS)

Fee, David; Izbekov, Pavel; Kim, Keehoon; Yokoo, Akihiko; Lopez, Taryn; Prata, Fred; Kazahaya, Ryunosuke; Nakamichi, Haruhisa; Iguchi, Masato

2017-12-01

Eruption mass and mass flow rate are critical parameters for determining the aerial extent and hazard of volcanic emissions. Infrasound waveform inversion is a promising technique to quantify volcanic emissions. Although topography may substantially alter the infrasound waveform as it propagates, advances in wave propagation modeling and station coverage permit robust inversion of infrasound data from volcanic explosions. The inversion can estimate eruption mass flow rate and total eruption mass if the flow density is known. However, infrasound-based eruption flow rates and mass estimates have yet to be validated against independent measurements, and numerical modeling has only recently been applied to the inversion technique. Here we present a robust full-waveform acoustic inversion method, and use it to calculate eruption flow rates and masses from 49 explosions from Sakurajima Volcano, Japan. Six infrasound stations deployed from 12-20 February 2015 recorded the explosions. We compute numerical Green's functions using 3-D Finite Difference Time Domain modeling and a high-resolution digital elevation model. The inversion, assuming a simple acoustic monopole source, provides realistic eruption masses and excellent fit to the data for the majority of the explosions. The inversion results are compared to independent eruption masses derived from ground-based ash collection and volcanic gas measurements. Assuming realistic flow densities, our infrasound-derived eruption masses for ash-rich eruptions compare favorably to the ground-based estimates, with agreement ranging from within a factor of two to one order of magnitude. Uncertainties in the time-dependent flow density and acoustic propagation likely contribute to the mismatch between the methods. Our results suggest that realistic and accurate infrasound-based eruption mass and mass flow rate estimates can be computed using the method employed here. If accurate volcanic flow parameters are known, application of this technique could be broadly applied to enable near real-time calculation of eruption mass flow rates and total masses. These critical input parameters for volcanic eruption modeling and monitoring are not currently available.

3D Acoustic Full Waveform Inversion for Engineering Purpose

NASA Astrophysics Data System (ADS)

Lim, Y.; Shin, S.; Kim, D.; Kim, S.; Chung, W.

2017-12-01

Seismic waveform inversion is the most researched data processing technique. In recent years, with an increase in marine development projects, seismic surveys are commonly conducted for engineering purposes; however, researches for application of waveform inversion are insufficient. The waveform inversion updates the subsurface physical property by minimizing the difference between modeled and observed data. Furthermore, it can be used to generate an accurate subsurface image; however, this technique consumes substantial computational resources. Its most compute-intensive step is the calculation of the gradient and hessian values. This aspect gains higher significance in 3D as compared to 2D. This paper introduces a new method for calculating gradient and hessian values, in an effort to reduce computational overburden. In the conventional waveform inversion, the calculation area covers all sources and receivers. In seismic surveys for engineering purposes, the number of receivers is limited. Therefore, it is inefficient to construct the hessian and gradient for the entire region (Figure 1). In order to tackle this problem, we calculate the gradient and the hessian for a single shot within the range of the relevant source and receiver. This is followed by summing up of these positions for the entire shot (Figure 2). In this paper, we demonstrate that reducing the area of calculation of the hessian and gradient for one shot reduces the overall amount of computation and therefore, the computation time. Furthermore, it is proved that the waveform inversion can be suitably applied for engineering purposes. In future research, we propose to ascertain an effective calculation range. This research was supported by the Basic Research Project(17-3314) of the Korea Institute of Geoscience and Mineral Resources(KIGAM) funded by the Ministry of Science, ICT and Future Planning of Korea.

Resolution analysis of marine seismic full waveform data by Bayesian inversion

NASA Astrophysics Data System (ADS)

Ray, A.; Sekar, A.; Hoversten, G. M.; Albertin, U.

2015-12-01

The Bayesian posterior density function (PDF) of earth models that fit full waveform seismic data convey information on the uncertainty with which the elastic model parameters are resolved. In this work, we apply the trans-dimensional reversible jump Markov Chain Monte Carlo method (RJ-MCMC) for the 1D inversion of noisy synthetic full-waveform seismic data in the frequency-wavenumber domain. While seismic full waveform inversion (FWI) is a powerful method for characterizing subsurface elastic parameters, the uncertainty in the inverted models has remained poorly known, if at all and is highly initial model dependent. The Bayesian method we use is trans-dimensional in that the number of model layers is not fixed, and flexible such that the layer boundaries are free to move around. The resulting parameterization does not require regularization to stabilize the inversion. Depth resolution is traded off with the number of layers, providing an estimate of uncertainty in elastic parameters (compressional and shear velocities Vp and Vs as well as density) with depth. We find that in the absence of additional constraints, Bayesian inversion can result in a wide range of posterior PDFs on Vp, Vs and density. These PDFs range from being clustered around the true model, to those that contain little resolution of any particular features other than those in the near surface, depending on the particular data and target geometry. We present results for a suite of different frequencies and offset ranges, examining the differences in the posterior model densities thus derived. Though these results are for a 1D earth, they are applicable to areas with simple, layered geology and provide valuable insight into the resolving capabilities of FWI, as well as highlight the challenges in solving a highly non-linear problem. The RJ-MCMC method also presents a tantalizing possibility for extension to 2D and 3D Bayesian inversion of full waveform seismic data in the future, as it objectively tackles the problem of model selection (i.e., the number of layers or cells for parameterization), which could ease the computational burden of evaluating forward models with many parameters.

Waveform inversion for 3-D earth structure using the Direct Solution Method implemented on vector-parallel supercomputer

NASA Astrophysics Data System (ADS)

Hara, Tatsuhiko

2004-08-01

We implement the Direct Solution Method (DSM) on a vector-parallel supercomputer and show that it is possible to significantly improve its computational efficiency through parallel computing. We apply the parallel DSM calculation to waveform inversion of long period (250-500 s) surface wave data for three-dimensional (3-D) S-wave velocity structure in the upper and uppermost lower mantle. We use a spherical harmonic expansion to represent lateral variation with the maximum angular degree 16. We find significant low velocities under south Pacific hot spots in the transition zone. This is consistent with other seismological studies conducted in the Superplume project, which suggests deep roots of these hot spots. We also perform simultaneous waveform inversion for 3-D S-wave velocity and Q structure. Since resolution for Q is not good, we develop a new technique in which power spectra are used as data for inversion. We find good correlation between long wavelength patterns of Vs and Q in the transition zone such as high Vs and high Q under the western Pacific.

2D Seismic Imaging of Elastic Parameters by Frequency Domain Full Waveform Inversion

NASA Astrophysics Data System (ADS)

Brossier, R.; Virieux, J.; Operto, S.

2008-12-01

Thanks to recent advances in parallel computing, full waveform inversion is today a tractable seismic imaging method to reconstruct physical parameters of the earth interior at different scales ranging from the near- surface to the deep crust. We present a massively parallel 2D frequency-domain full-waveform algorithm for imaging visco-elastic media from multi-component seismic data. The forward problem (i.e. the resolution of the frequency-domain 2D PSV elastodynamics equations) is based on low-order Discontinuous Galerkin (DG) method (P0 and/or P1 interpolations). Thanks to triangular unstructured meshes, the DG method allows accurate modeling of both body waves and surface waves in case of complex topography for a discretization of 10 to 15 cells per shear wavelength. The frequency-domain DG system is solved efficiently for multiple sources with the parallel direct solver MUMPS. The local inversion procedure (i.e. minimization of residuals between observed and computed data) is based on the adjoint-state method which allows to efficiently compute the gradient of the objective function. Applying the inversion hierarchically from the low frequencies to the higher ones defines a multiresolution imaging strategy which helps convergence towards the global minimum. In place of expensive Newton algorithm, the combined use of the diagonal terms of the approximate Hessian matrix and optimization algorithms based on quasi-Newton methods (Conjugate Gradient, LBFGS, ...) allows to improve the convergence of the iterative inversion. The distribution of forward problem solutions over processors driven by a mesh partitioning performed by METIS allows to apply most of the inversion in parallel. We shall present the main features of the parallel modeling/inversion algorithm, assess its scalability and illustrate its performances with realistic synthetic case studies.

Source mechanism of the 2006 M5.1 Wen'an Earthquake determined from a joint inversion of local and teleseismic broadband waveform data

NASA Astrophysics Data System (ADS)

Huang, J.; Ni, S.; Niu, F.; Fu, R.

2007-12-01

On July 4th, 2006, a magnitude 5.1 earthquake occurred at Wen'an, {~}100 km south of Beijing, which was felt at Beijing metropolitan area. To better understand the regional tectonics, we have inverted local and teleseismic broadband waveform data to determine the focal mechanism of this earthquake. We selected waveform data of 9 stations from the recently installed Beijing metropolitan digital Seismic Network (BSN). These stations are located within 600 km and cover a good azimuthal range to the earthquake. To better fit the lower amplitude P waveform, we employed two different weights for the P wave and surface wave arrivals, respectively. A grid search method was employed to find the strike, dip and slip of the earthquake that best fits the P and surface waveforms recorded at all the three components (the tangential component of the P-wave arrivals was not used). Synthetic waveforms were computed with an F-K method. Two crustal velocity models were used in the synthetic calculation to reflect a rapid east-west transition in crustal structure observed by seismic and geological studies in the study area. The 3D grid search results in reasonable constraints on the fault geometry and the slip vector with a less well determined focal depth. As such we combined teleseismic waveform data from 8 stations of the Global Seismic Network in a joint inversion. Clearly identifiable depth phases (pP, sP) recorded in the teleseismic stations obviously provided a better constraint on the resulting source depth. Results from the joint inversion indicate that the Wen'an earthquake is mainly a right-lateral strike slip event (-150°) which occurred at a near vertical (dip, 80° ) NNE trend (210°º) fault. The estimated focal depth is {~}14- 15km, and the moment magnitude is 5.1. The estimated fault geometry here agrees well with aftershock distribution and is consistent with the major fault systems in the area which were developed under a NNE-SSW oriented compressional stress field. Key word: waveform modeling method, source mechanism, grid search method, cut and paste method, aftershocks distribution

Micro-seismic waveform matching inversion based on gravitational search algorithm and parallel computation

NASA Astrophysics Data System (ADS)

Jiang, Y.; Xing, H. L.

2016-12-01

Micro-seismic events induced by water injection, mining activity or oil/gas extraction are quite informative, the interpretation of which can be applied for the reconstruction of underground stress and monitoring of hydraulic fracturing progress in oil/gas reservoirs. The source characterises and locations are crucial parameters that required for these purposes, which can be obtained through the waveform matching inversion (WMI) method. Therefore it is imperative to develop a WMI algorithm with high accuracy and convergence speed. Heuristic algorithm, as a category of nonlinear method, possesses a very high convergence speed and good capacity to overcome local minimal values, and has been well applied for many areas (e.g. image processing, artificial intelligence). However, its effectiveness for micro-seismic WMI is still poorly investigated; very few literatures exits that addressing this subject. In this research an advanced heuristic algorithm, gravitational search algorithm (GSA) , is proposed to estimate the focal mechanism (angle of strike, dip and rake) and source locations in three dimension. Unlike traditional inversion methods, the heuristic algorithm inversion does not require the approximation of green function. The method directly interacts with a CPU parallelized finite difference forward modelling engine, and updating the model parameters under GSA criterions. The effectiveness of this method is tested with synthetic data form a multi-layered elastic model; the results indicate GSA can be well applied on WMI and has its unique advantages. Keywords: Micro-seismicity, Waveform matching inversion, gravitational search algorithm, parallel computation

Sensitivity analyses of acoustic impedance inversion with full-waveform inversion

NASA Astrophysics Data System (ADS)

Yao, Gang; da Silva, Nuno V.; Wu, Di

2018-04-01

Acoustic impedance estimation has a significant importance to seismic exploration. In this paper, we use full-waveform inversion to recover the impedance from seismic data, and analyze the sensitivity of the acoustic impedance with respect to the source-receiver offset of seismic data and to the initial velocity model. We parameterize the acoustic wave equation with velocity and impedance, and demonstrate three key aspects of acoustic impedance inversion. First, short-offset data are most suitable for acoustic impedance inversion. Second, acoustic impedance inversion is more compatible with the data generated by density contrasts than velocity contrasts. Finally, acoustic impedance inversion requires the starting velocity model to be very accurate for achieving a high-quality inversion. Based upon these observations, we propose a workflow for acoustic impedance inversion as: (1) building a background velocity model with travel-time tomography or reflection waveform inversion; (2) recovering the intermediate wavelength components of the velocity model with full-waveform inversion constrained by Gardner’s relation; (3) inverting the high-resolution acoustic impedance model with short-offset data through full-waveform inversion. We verify this workflow by the synthetic tests based on the Marmousi model.

ℓ1-Regularized full-waveform inversion with prior model information based on orthant-wise limited memory quasi-Newton method

NASA Astrophysics Data System (ADS)

Dai, Meng-Xue; Chen, Jing-Bo; Cao, Jian

2017-07-01

Full-waveform inversion (FWI) is an ill-posed optimization problem which is sensitive to noise and initial model. To alleviate the ill-posedness of the problem, regularization techniques are usually adopted. The ℓ1-norm penalty is a robust regularization method that preserves contrasts and edges. The Orthant-Wise Limited-Memory Quasi-Newton (OWL-QN) method extends the widely-used limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) method to the ℓ1-regularized optimization problems and inherits the efficiency of L-BFGS. To take advantage of the ℓ1-regularized method and the prior model information obtained from sonic logs and geological information, we implement OWL-QN algorithm in ℓ1-regularized FWI with prior model information in this paper. Numerical experiments show that this method not only improve the inversion results but also has a strong anti-noise ability.

Acoustic and elastic waveform inversion best practices

NASA Astrophysics Data System (ADS)

Modrak, Ryan T.

Reaching the global minimum of a waveform misfit function requires careful choices about the nonlinear optimization, preconditioning and regularization methods underlying an inversion. Because waveform inversion problems are susceptible to erratic convergence, one or two test cases are not enough to reliably inform such decisions. We identify best practices instead using two global, one regional and four near-surface acoustic test problems. To obtain meaningful quantitative comparisons, we carry out hundreds acoustic inversions, varying one aspect of the implementation at a time. Comparing nonlinear optimization algorithms, we find that L-BFGS provides computational savings over nonlinear conjugate gradient methods in a wide variety of test cases. Comparing preconditioners, we show that a new diagonal scaling derived from the adjoint of the forward operator provides better performance than two conventional preconditioning schemes. Comparing regularization strategies, we find that projection, convolution, Tikhonov regularization, and total variation regularization are effective in different contexts. Besides these issues, reliability and efficiency in waveform inversion depend on close numerical attention and care. Implementation details have a strong effect on computational cost, regardless of the chosen material parameterization or nonlinear optimization algorithm. Building on the acoustic inversion results, we carry out elastic experiments with four test problems, three objective functions, and four material parameterizations. The choice of parameterization for isotropic elastic media is found to be more complicated than previous studies suggests, with "wavespeed-like'' parameters performing well with phase-based objective functions and Lame parameters performing well with amplitude-based objective functions. Reliability and efficiency can be even harder to achieve in transversely isotropic elastic inversions because rotation angle parameters describing fast-axis direction are difficult to recover. Using Voigt or Chen-Tromp parameters avoids the need to include rotation angles explicitly and provides an effective strategy for anisotropic inversion. The need for flexible and portable workflow management tools for seismic inversion also poses a major challenge. In a final chapter, the software used to the carry out the above experiments is described and instructions for reproducing experimental results are given.

Full-waveform Inversion of Crosshole GPR Data Collected in Strongly Heterogeneous Chalk: Challenges and Pitfalls

NASA Astrophysics Data System (ADS)

Keskinen, Johanna; Looms, Majken C.; Nielsen, Lars; Klotzsche, Anja; van der Kruk, Jan; Moreau, Julien; Stemmerik, Lars; Holliger, Klaus

2015-04-01

Chalk is an important reservoir rock for hydrocarbons and for groundwater resources for many major cities. Therefore, this rock type has been extensively investigated using both geological and geophysical methods. Many applications of crosshole GPR tomography rely on the ray approximation and corresponding inversions of first break traveltimes and/or maximum first-cycle amplitudes. Due to the inherent limitations associated with such approaches, the resulting models tend to be overly smooth and cannot adequately capture the small-scale heterogeneities. In contrast, the full-waveform inversion uses all the information contained in the data and is able to provide significantly improved images. Here, we apply full-waveform inversion to crosshole GPR data to image strong heterogeneity of the chalk related to changes in lithology and porosity. We have collected a crosshole tomography dataset in an old chalk quarry in Eastern Denmark. Based on core data (including plug samples and televiewer logging data) collected in our four ~15-m-deep boreholes and results from previous related studies, it is apparent that the studied chalk is strongly heterogeneous. The upper ~7 m consist of variable coarse-grained chalk layers with numerous flint nodules. The lower half of the studied section appears to be finer-grained and contains less flint. However, still significant porosity variations are also detected in the lower half. In general, the water-saturated (watertable depth ~2 m) chalk is characterized by high porosities, and thus low velocities and high attenuation, while the flint is essentially non-porous and has correspondingly high velocities and low attenuation. Together these characteristics form a strongly heterogeneous medium, which is challenging for the full-waveform inversion to recover. Here, we address the importance of (i) adequate starting models, both in terms of the dielectric permittivity and the electrical conductivity, (ii) the estimation of the source wavelet, (iii) and the effects of data sampling density when imaging this rock type. Moreover, we discuss the resolution of the bedding recovered by the full-waveform approach. Our results show that with proper estimates of the above-mentioned prior parameters, crosshole GPR full-waveform tomography provides high-resolution images capturing a high degree of variability that standard methods cannot resolve in chalk. This in turn makes crosshole full-waveform inversion a promising tool to support time-lapse flow modelling.

3-D characterization of high-permeability zones in a gravel aquifer using 2-D crosshole GPR full-waveform inversion and waveguide detection

NASA Astrophysics Data System (ADS)

Klotzsche, Anja; van der Kruk, Jan; Linde, Niklas; Doetsch, Joseph; Vereecken, Harry

2013-11-01

Reliable high-resolution 3-D characterization of aquifers helps to improve our understanding of flow and transport processes when small-scale structures have a strong influence. Crosshole ground penetrating radar (GPR) is a powerful tool for characterizing aquifers due to the method's high-resolution and sensitivity to porosity and soil water content. Recently, a novel GPR full-waveform inversion algorithm was introduced, which is here applied and used for 3-D characterization by inverting six crosshole GPR cross-sections collected between four wells arranged in a square configuration close to the Thur River in Switzerland. The inversion results in the saturated part of this gravel aquifer reveals a significant improvement in resolution for the dielectric permittivity and electrical conductivity images compared to ray-based methods. Consistent structures where acquisition planes intersect indicate the robustness of the inversion process. A decimetre-scale layer with high dielectric permittivity was revealed at a depth of 5-6 m in all six cross-sections analysed here, and a less prominent zone with high dielectric permittivity was found at a depth of 7.5-9 m. These high-permittivity layers act as low-velocity waveguides and they are interpreted as high-porosity layers and possible zones of preferential flow. Porosity estimates from the permittivity models agree well with estimates from Neutron-Neutron logging data at the intersecting diagonal planes. Moreover, estimates of hydraulic permeability based on flowmeter logs confirm the presence of zones of preferential flow in these depth intervals. A detailed analysis of the measured data for transmitters located within the waveguides, revealed increased trace energy due to late-arrival elongated wave trains, which were observed for receiver positions straddling this zone. For the same receiver positions within the waveguide, a distinct minimum in the trace energy was visible when the transmitter was located outside the waveguide. A novel amplitude analysis was proposed to explore these maxima and minima of the trace energy. Laterally continuous low-velocity waveguides and their boundaries were identified in the measured data alone. In contrast to the full-waveform inversion, this method follows a simple workflow and needs no detailed and time consuming processing or inversion of the data. Comparison with the full-waveform inversion results confirmed the presence of the waveguides illustrating that full-waveform inversion return reliable results at the highest resolution currently possible at these scales. We envision that full-waveform inversion of GPR data will play an important role in a wide range of geological, hydrological, glacial and periglacial studies in the critical zone.

Full Waveform Inversion of Diving & Reflected Waves based on Scale Separation for Velocity and Impedance Imaging

NASA Astrophysics Data System (ADS)

Brossier, Romain; Zhou, Wei; Operto, Stéphane; Virieux, Jean

2015-04-01

Full Waveform Inversion (FWI) is an appealing method for quantitative high-resolution subsurface imaging (Virieux et al., 2009). For crustal-scales exploration from surface seismic, FWI generally succeeds in recovering a broadband of wavenumbers in the shallow part of the targeted medium taking advantage of the broad scattering-angle provided by both reflected and diving waves. In contrast, deeper targets are often only illuminated by short-spread reflections, which favor the reconstruction of the short wavelengths at the expense of the longer ones, leading to a possible notch in the intermediate part of the wavenumber spectrum. To update the velocity macromodel from reflection data, image-domain strategies (e.g., Symes & Carazzone, 1991) aim to maximize a semblance criterion in the migrated domain. Alternatively, recent data-domain strategies (e.g., Xu et al., 2012, Ma & Hale, 2013, Brossier et al., 2014), called Reflection FWI (RFWI), inspired by Chavent et al. (1994), rely on a scale separation between the velocity macromodel and prior knowledge of the reflectivity to emphasize the transmission regime in the sensitivity kernel of the inversion. However, all these strategies focus on reflected waves only, discarding the low-wavenumber information carried out by diving waves. With the current development of very long-offset and wide-azimuth acquisitions, a significant part of the recorded energy is provided by diving waves and subcritical reflections, and high-resolution tomographic methods should take advantage of all types of waves. In this presentation, we will first review the issues of classical FWI when applied to reflected waves and how RFWI is able to retrieve the long wavelength of the model. We then propose a unified formulation of FWI (Zhou et al., 2014) to update the low wavenumbers of the velocity model by the joint inversion of diving and reflected arrivals, while the impedance model is updated thanks to reflected wave only. An alternate inversion of high wavenumber impedance model and low wavenumber velocity model is performed to iteratively improve subsurface models. References : Brossier, R., Operto, S. & Virieux, J., 2014. Velocity model building from seismic reflection data by full waveform inversion, Geophysical Prospecting, doi:10.1111/1365-2478.12190 Chavent, G., Clément, F. & Gomez, S., 1994.Automatic determination of velocities via migration-based traveltime waveform inversion: A synthetic data example, SEG Technical Program Expanded Abstracts 1994, pp. 1179--1182. Ma, Y. & Hale, D., 2013. Wave-equation reflection traveltime inversion with dynamic warping and full waveform inversion, Geophysics, 78(6), R223--R233. Symes, W.W. & Carazzone, J.J., 1991. Velocity inversion by differential semblance optimization, Geophysics, 56, 654--663. Virieux, J. & Operto, S., 2009. An overview of full waveform inversion in exploration geophysics, Geophysics, 74(6), WCC1--WCC26. Xu, S., Wang, D., Chen, F., Lambaré, G. & Zhang, Y., 2012. Inversion on reflected seismic wave, SEG Technical Program Expanded Abstracts 2012, pp. 1--7. Zhou, W., Brossier, R., Operto, S., & Virieux, J., 2014. Acoustic multiparameter full-waveform inversion through a hierachical scheme, in SEG Technical Program Expanded Abstracts 2014, pp. 1249--1253

Wavelet-based multiscale adjoint waveform-difference tomography using body and surface waves

NASA Astrophysics Data System (ADS)

Yuan, Y. O.; Simons, F. J.; Bozdag, E.

2014-12-01

We present a multi-scale scheme for full elastic waveform-difference inversion. Using a wavelet transform proves to be a key factor to mitigate cycle-skipping effects. We start with coarse representations of the seismogram to correct a large-scale background model, and subsequently explain the residuals in the fine scales of the seismogram to map the heterogeneities with great complexity. We have previously applied the multi-scale approach successfully to body waves generated in a standard model from the exploration industry: a modified two-dimensional elastic Marmousi model. With this model we explored the optimal choice of wavelet family, number of vanishing moments and decomposition depth. For this presentation we explore the sensitivity of surface waves in waveform-difference tomography. The incorporation of surface waves is rife with cycle-skipping problems compared to the inversions considering body waves only. We implemented an envelope-based objective function probed via a multi-scale wavelet analysis to measure the distance between predicted and target surface-wave waveforms in a synthetic model of heterogeneous near-surface structure. Our proposed method successfully purges the local minima present in the waveform-difference misfit surface. An elastic shallow model with 100~m in depth is used to test the surface-wave inversion scheme. We also analyzed the sensitivities of surface waves and body waves in full waveform inversions, as well as the effects of incorrect density information on elastic parameter inversions. Based on those numerical experiments, we ultimately formalized a flexible scheme to consider both body and surface waves in adjoint tomography. While our early examples are constructed from exploration-style settings, our procedure will be very valuable for the study of global network data.

Full waveform inversion for ultrasonic flaw identification

NASA Astrophysics Data System (ADS)

Seidl, Robert; Rank, Ernst

2017-02-01

Ultrasonic Nondestructive Testing is concerned with detecting flaws inside components without causing physical damage. It is possible to detect flaws using ultrasound measurements but usually no additional details about the flaw like position, dimension or orientation are available. The information about these details is hidden in the recorded experimental signals. The idea of full waveform inversion is to adapt the parameters of an initial simulation model of the undamaged specimen by minimizing the discrepancy between these simulated signals and experimentally measured signals of the flawed specimen. Flaws in the structure are characterized by a change or deterioration in the material properties. Commonly, full waveform inversion is mostly applied in seismology on a larger scale to infer mechanical properties of the earth. We propose to use acoustic full waveform inversion for structural parameters to visualize the interior of the component. The method is adapted to US NDT by combining multiple similar experiments on the test component as the typical small amount of sensors is not sufficient for a successful imaging. It is shown that the combination of simulations and multiple experiments can be used to detect flaws and their position, dimension and orientation in emulated simulation cases.

Source encoding in multi-parameter full waveform inversion

NASA Astrophysics Data System (ADS)

Matharu, Gian; Sacchi, Mauricio D.

2018-04-01

Source encoding techniques alleviate the computational burden of sequential-source full waveform inversion (FWI) by considering multiple sources simultaneously rather than independently. The reduced data volume requires fewer forward/adjoint simulations per non-linear iteration. Applications of source-encoded full waveform inversion (SEFWI) have thus far focused on monoparameter acoustic inversion. We extend SEFWI to the multi-parameter case with applications presented for elastic isotropic inversion. Estimating multiple parameters can be challenging as perturbations in different parameters can prompt similar responses in the data. We investigate the relationship between source encoding and parameter trade-off by examining the multi-parameter source-encoded Hessian. Probing of the Hessian demonstrates the convergence of the expected source-encoded Hessian, to that of conventional FWI. The convergence implies that the parameter trade-off in SEFWI is comparable to that observed in FWI. A series of synthetic inversions are conducted to establish the feasibility of source-encoded multi-parameter FWI. We demonstrate that SEFWI requires fewer overall simulations than FWI to achieve a target model error for a range of first-order optimization methods. An inversion for spatially inconsistent P - (α) and S-wave (β) velocity models, corroborates the expectation of comparable parameter trade-off in SEFWI and FWI. The final example demonstrates a shortcoming of SEFWI when confronted with time-windowing in data-driven inversion schemes. The limitation is a consequence of the implicit fixed-spread acquisition assumption in SEFWI. Alternative objective functions, namely the normalized cross-correlation and L1 waveform misfit, do not enable SEFWI to overcome this limitation.

Waveform inversion in the frequency domain for the simultaneous determination of earthquake source mechanism and moment function

NASA Astrophysics Data System (ADS)

Nakano, M.; Kumagai, H.; Inoue, H.

2008-06-01

We propose a method of waveform inversion to rapidly and routinely estimate both the moment function and the centroid moment tensor (CMT) of an earthquake. In this method, waveform inversion is carried out in the frequency domain to obtain the moment function more rapidly than when solved in the time domain. We assume a pure double-couple source mechanism in order to stabilize the solution when using data from a small number of seismic stations. The fault and slip orientations are estimated by a grid search with respect to the strike, dip and rake angles. The moment function in the time domain is obtained from the inverse Fourier transform of the frequency components determined by the inversion. Since observed waveforms used for the inversion are limited in a particular frequency band, the estimated moment function is a bandpassed form. We develop a practical approach to estimate the deconvolved form of the moment function, from which we can reconstruct detailed rupture history and the seismic moment. The source location is determined by a spatial grid search using adaptive grid spacings, which are gradually decreased in each step of the search. We apply this method to two events that occurred in Indonesia by using data from a broad-band seismic network in Indonesia (JISNET): one northeast of Sulawesi (Mw = 7.5) on 2007 January 21, and the other south of Java (Mw = 7.5) on 2006 July 17. The source centroid locations and mechanisms we estimated for both events are consistent with those determined by the Global CMT Project and the National Earthquake Information Center of the U.S. Geological Survey. The estimated rupture duration of the Sulawesi event is 16 s, which is comparable to a typical duration for earthquakes of this magnitude, while that of the Java event is anomalously long (176 s), suggesting that this event was a tsunami earthquake. Our application demonstrates that this inversion method has great potential for rapid and routine estimations of both the CMT and the moment function, and may be useful for identification of tsunami earthquakes.

Potency backprojection

NASA Astrophysics Data System (ADS)

Okuwaki, R.; Kasahara, A.; Yagi, Y.

2017-12-01

The backprojection (BP) method has been one of the powerful tools of tracking seismic-wave sources of the large/mega earthquakes. The BP method projects waveforms onto a possible source point by stacking them with the theoretical-travel-time shifts between the source point and the stations. Following the BP method, the hybrid backprojection (HBP) method was developed to enhance depth-resolution of projected images and mitigate the dummy imaging of the depth phases, which are shortcomings of the BP method, by stacking cross-correlation functions of the observed waveforms and theoretically calculated Green's functions (GFs). The signal-intensity of the BP/HBP image at a source point is related to how much of observed waveforms was radiated from that point. Since the amplitude of the GF associated with the slip-rate increases with depth as the rigidity increases with depth, the intensity of the BP/HBP image inherently has depth dependence. To make a direct comparison of the BP/HBP image with the corresponding slip distribution inferred from a waveform inversion, and discuss the rupture properties along the fault drawn from the waveforms in high- and low-frequencies with the BP/HBP methods and the waveform inversion, respectively, it is desirable to have the variants of BP/HBP methods that directly image the potency-rate-density distribution. Here we propose new formulations of the BP/HBP methods, which image the distribution of the potency-rate density by introducing alternative normalizing factors in the conventional formulations. For the BP method, the observed waveform is normalized with the maximum amplitude of P-phase of the corresponding GF. For the HBP method, we normalize the cross-correlation function with the squared-sum of the GF. The normalized waveforms or the cross-correlation functions are then stacked for all the stations to enhance the signal to noise ratio. We will present performance-tests of the new formulations by using synthetic waveforms and the real data of the Mw 8.3 2015 Illapel Chile earthquake, and further discuss the limitations of the new BP/HBP methods proposed in this study when they are used for exploring the rupture properties of the earthquakes.

Micro-seismic imaging using a source function independent full waveform inversion method

NASA Astrophysics Data System (ADS)

Wang, Hanchen; Alkhalifah, Tariq

2018-03-01

At the heart of micro-seismic event measurements is the task to estimate the location of the source micro-seismic events, as well as their ignition times. The accuracy of locating the sources is highly dependent on the velocity model. On the other hand, the conventional micro-seismic source locating methods require, in many cases manual picking of traveltime arrivals, which do not only lead to manual effort and human interaction, but also prone to errors. Using full waveform inversion (FWI) to locate and image micro-seismic events allows for an automatic process (free of picking) that utilizes the full wavefield. However, full waveform inversion of micro-seismic events faces incredible nonlinearity due to the unknown source locations (space) and functions (time). We developed a source function independent full waveform inversion of micro-seismic events to invert for the source image, source function and the velocity model. It is based on convolving reference traces with these observed and modeled to mitigate the effect of an unknown source ignition time. The adjoint-state method is used to derive the gradient for the source image, source function and velocity updates. The extended image for the source wavelet in Z axis is extracted to check the accuracy of the inverted source image and velocity model. Also, angle gathers is calculated to assess the quality of the long wavelength component of the velocity model. By inverting for the source image, source wavelet and the velocity model simultaneously, the proposed method produces good estimates of the source location, ignition time and the background velocity for synthetic examples used here, like those corresponding to the Marmousi model and the SEG/EAGE overthrust model.

Permittivity and conductivity parameter estimations using full waveform inversion

NASA Astrophysics Data System (ADS)

Serrano, Jheyston O.; Ramirez, Ana B.; Abreo, Sergio A.; Sadler, Brian M.

2018-04-01

Full waveform inversion of Ground Penetrating Radar (GPR) data is a promising strategy to estimate quantitative characteristics of the subsurface such as permittivity and conductivity. In this paper, we propose a methodology that uses Full Waveform Inversion (FWI) in time domain of 2D GPR data to obtain highly resolved images of the permittivity and conductivity parameters of the subsurface. FWI is an iterative method that requires a cost function to measure the misfit between observed and modeled data, a wave propagator to compute the modeled data and an initial velocity model that is updated at each iteration until an acceptable decrease of the cost function is reached. The use of FWI with GPR are expensive computationally because it is based on the computation of the electromagnetic full wave propagation. Also, the commercially available acquisition systems use only one transmitter and one receiver antenna at zero offset, requiring a large number of shots to scan a single line.

Angular velocity of gravitational radiation from precessing binaries and the corotating frame

NASA Astrophysics Data System (ADS)

Boyle, Michael

2013-05-01

This paper defines an angular velocity for time-dependent functions on the sphere and applies it to gravitational waveforms from compact binaries. Because it is geometrically meaningful and has a clear physical motivation, the angular velocity is uniquely useful in helping to solve an important—and largely ignored—problem in models of compact binaries: the inverse problem of deducing the physical parameters of a system from the gravitational waves alone. It is also used to define the corotating frame of the waveform. When decomposed in this frame, the waveform has no rotational dynamics and is therefore as slowly evolving as possible. The resulting simplifications lead to straightforward methods for accurately comparing waveforms and constructing hybrids. As formulated in this paper, the methods can be applied robustly to both precessing and nonprecessing waveforms, providing a clear, comprehensive, and consistent framework for waveform analysis. Explicit implementations of all these methods are provided in accompanying computer code.

Visco-elastic controlled-source full waveform inversion without surface waves

NASA Astrophysics Data System (ADS)

Paschke, Marco; Krause, Martin; Bleibinhaus, Florian

2016-04-01

We developed a frequency-domain visco-elastic full waveform inversion for onshore seismic experiments with topography. The forward modeling is based on a finite-difference time-domain algorithm by Robertsson that uses the image-method to ensure a stress-free condition at the surface. The time-domain data is Fourier-transformed at every point in the model space during the forward modeling for a given set of frequencies. The motivation for this approach is the reduced amount of memory when computing kernels, and the straightforward implementation of the multiscale approach. For the inversion, we calculate the Frechet derivative matrix explicitly, and we implement a Levenberg-Marquardt scheme that allows for computing the resolution matrix. To reduce the size of the Frechet derivative matrix, and to stabilize the inversion, an adapted inverse mesh is used. The node spacing is controlled by the velocity distribution and the chosen frequencies. To focus the inversion on body waves (P, P-coda, and S) we mute the surface waves from the data. Consistent spatiotemporal weighting factors are applied to the wavefields during the Fourier transform to obtain the corresponding kernels. We test our code with a synthetic study using the Marmousi model with arbitrary topography. This study also demonstrates the importance of topography and muting surface waves in controlled-source full waveform inversion.

Computing the Sensitivity Kernels for 2.5-D Seismic Waveform Inversion in Heterogeneous, Anisotropic Media

NASA Astrophysics Data System (ADS)

Zhou, Bing; Greenhalgh, S. A.

2011-10-01

2.5-D modeling and inversion techniques are much closer to reality than the simple and traditional 2-D seismic wave modeling and inversion. The sensitivity kernels required in full waveform seismic tomographic inversion are the Fréchet derivatives of the displacement vector with respect to the independent anisotropic model parameters of the subsurface. They give the sensitivity of the seismograms to changes in the model parameters. This paper applies two methods, called `the perturbation method' and `the matrix method', to derive the sensitivity kernels for 2.5-D seismic waveform inversion. We show that the two methods yield the same explicit expressions for the Fréchet derivatives using a constant-block model parameterization, and are available for both the line-source (2-D) and the point-source (2.5-D) cases. The method involves two Green's function vectors and their gradients, as well as the derivatives of the elastic modulus tensor with respect to the independent model parameters. The two Green's function vectors are the responses of the displacement vector to the two directed unit vectors located at the source and geophone positions, respectively; they can be generally obtained by numerical methods. The gradients of the Green's function vectors may be approximated in the same manner as the differential computations in the forward modeling. The derivatives of the elastic modulus tensor with respect to the independent model parameters can be obtained analytically, dependent on the class of medium anisotropy. Explicit expressions are given for two special cases—isotropic and tilted transversely isotropic (TTI) media. Numerical examples are given for the latter case, which involves five independent elastic moduli (or Thomsen parameters) plus one angle defining the symmetry axis.

Geophysical characterization of peatlands using crosshole GPR full-waveform inversion: Case study from a bog in northwestern Germany

NASA Astrophysics Data System (ADS)

Schmäck, J.; Klotzsche, A.; Van Der Kruk, J.; Vereecken, H.; Bechtold, M.

2017-12-01

The characterization of peatlands is of particular interest, since areas with peat soils represent global hotspots for the exchange of greenhouse gases. Their effect on global warming depends on several parameters, like mean annual water level and land use. Models of greenhouse gas emissions and carbon accumulation in peatlands can be improved by including small-scale soil properties that e.g. act as gas traps and periodically release gases to the atmosphere during ebullition events. Ground penetrating radar (GPR) is well suited to non- or minimal invasively characterize and improve our understanding of dynamic processes that take place in the critical zone. It uses high frequency electromagnetic waves to image and characterize the dielectric permittivity and electrical conductivity of the critical zone, which can be related to hydrogeological properties like porosity, soil water content, salinity and clay content. In the last decade, the full-waveform inversion of crosshole GPR data has proved to be a powerful tool to improve the image resolution compared to standard ray-based methods. This approach was successfully applied to several different aquifers and was able to provide decimeter-scale resolution images including small-scale high contrast layers that can be related to zones of high porosity, zones of preferential flow or clay lenses. The comparison to independently measured e.g. logging data proved the reliability of the method. Here, for the first time crosshole GPR full-waveform inversion is used to image three peatland plots with different land use that are part of the "Ahlen-Falkenberger Moor peat bog complex" in northwestern Germany. The full-waveform inversion of the acquired data returned higher resolution images than standard ray-based GPR methods, and, is able to improve our understanding of subsurface structures. The comparison of the different plots is expected to provide new insights into gas content and gas trapping structures across different land uses. Additionally, season-related changes of peatland soil properties are investigated. The crosshole GPR full-waveform inversion was successfully applied to several datasets and the results show the utility and credibility of GPR FWI to analyze peatland properties.

Magnetic Resonance Elastography: Measurement of Hepatic Stiffness Using Different Direct Inverse Problem Reconstruction Methods in Healthy Volunteers and Patients with Liver Disease.

PubMed

Saito, Shigeyoshi; Tanaka, Keiko; Hashido, Takashi

2016-02-01

The purpose of this study was to compare the mean hepatic stiffness values obtained by the application of two different direct inverse problem reconstruction methods to magnetic resonance elastography (MRE). Thirteen healthy men (23.2±2.1 years) and 16 patients with liver diseases (78.9±4.3 years; 12 men and 4 women) were examined for this study using a 3.0 T-MRI. The healthy volunteers underwent three consecutive scans, two 70-Hz waveform and a 50-Hz waveform scans. On the other hand, the patients with liver disease underwent scanning using the 70-Hz waveform only. The MRE data for each subject was processed twice for calculation of the mean hepatic stiffness (Pa), once using the multiscale direct inversion (MSDI) and once using the multimodel direct inversion (MMDI). There were no significant differences in the mean stiffness values among the scans obtained with two 70-Hz and different waveforms. However, the mean stiffness values obtained with the MSDI technique (with mask: 2895.3±255.8 Pa, without mask: 2940.6±265.4 Pa) were larger than those obtained with the MMDI technique (with mask: 2614.0±242.1 Pa, without mask: 2699.2±273.5 Pa). The reproducibility of measurements obtained using the two techniques was high for both the healthy volunteers [intraclass correlation coefficients (ICCs): 0.840-0.953] and the patients (ICC: 0.830-0.995). These results suggest that knowledge of the characteristics of different direct inversion algorithms is important for longitudinal liver stiffness assessments such as the comparison of different scanners and evaluation of the response to fibrosis therapy.

The Collaborative Seismic Earth Model: Generation 1

NASA Astrophysics Data System (ADS)

Fichtner, Andreas; van Herwaarden, Dirk-Philip; Afanasiev, Michael; SimutÄ--, SaulÄ--; Krischer, Lion; ćubuk-Sabuncu, Yeşim; Taymaz, Tuncay; Colli, Lorenzo; Saygin, Erdinc; Villaseñor, Antonio; Trampert, Jeannot; Cupillard, Paul; Bunge, Hans-Peter; Igel, Heiner

2018-05-01

We present a general concept for evolutionary, collaborative, multiscale inversion of geophysical data, specifically applied to the construction of a first-generation Collaborative Seismic Earth Model. This is intended to address the limited resources of individual researchers and the often limited use of previously accumulated knowledge. Model evolution rests on a Bayesian updating scheme, simplified into a deterministic method that honors today's computational restrictions. The scheme is able to harness distributed human and computing power. It furthermore handles conflicting updates, as well as variable parameterizations of different model refinements or different inversion techniques. The first-generation Collaborative Seismic Earth Model comprises 12 refinements from full seismic waveform inversion, ranging from regional crustal- to continental-scale models. A global full-waveform inversion ensures that regional refinements translate into whole-Earth structure.

Measuring the misfit between seismograms using an optimal transport distance: application to full waveform inversion

NASA Astrophysics Data System (ADS)

Métivier, L.; Brossier, R.; Mérigot, Q.; Oudet, E.; Virieux, J.

2016-04-01

Full waveform inversion using the conventional L2 distance to measure the misfit between seismograms is known to suffer from cycle skipping. An alternative strategy is proposed in this study, based on a measure of the misfit computed with an optimal transport distance. This measure allows to account for the lateral coherency of events within the seismograms, instead of considering each seismic trace independently, as is done generally in full waveform inversion. The computation of this optimal transport distance relies on a particular mathematical formulation allowing for the non-conservation of the total energy between seismograms. The numerical solution of the optimal transport problem is performed using proximal splitting techniques. Three synthetic case studies are investigated using this strategy: the Marmousi 2 model, the BP 2004 salt model, and the Chevron 2014 benchmark data. The results emphasize interesting properties of the optimal transport distance. The associated misfit function is less prone to cycle skipping. A workflow is designed to reconstruct accurately the salt structures in the BP 2004 model, starting from an initial model containing no information about these structures. A high-resolution P-wave velocity estimation is built from the Chevron 2014 benchmark data, following a frequency continuation strategy. This estimation explains accurately the data. Using the same workflow, full waveform inversion based on the L2 distance converges towards a local minimum. These results yield encouraging perspectives regarding the use of the optimal transport distance for full waveform inversion: the sensitivity to the accuracy of the initial model is reduced, the reconstruction of complex salt structure is made possible, the method is robust to noise, and the interpretation of seismic data dominated by reflections is enhanced.

Lithospheric layering in the North American craton revealed by including Short Period Constraints in Full Waveform Tomography

NASA Astrophysics Data System (ADS)

Roy, C.; Calo, M.; Bodin, T.; Romanowicz, B. A.

2017-12-01

Recent receiver function studies of the North American craton suggest the presence of significant layering within the cratonic lithosphere, with significant lateral variations in the depth of the velocity discontinuities. These structural boundaries have been confirmed recently using a transdimensional Markov Chain Monte Carlo approach (TMCMC), inverting surface wave dispersion data and converted phases simultaneously (Calò et al., 2016; Roy and Romanowicz 2017). The lateral resolution of upper mantle structure can be improved with a high density of broadband seismic stations, or with a sparse network using full waveform inversion based on numerical wavefield computation methods such as the Spectral Element Method (SEM). However, inverting for discontinuities with strong topography such as MLDS's or LAB, presents challenges in an inversion framework, both computationally, due to the short periods required, and from the point of view of stability of the inversion. To overcome these limitations, and to improve resolution of layering in the upper mantle, we are developing a methodology that combines full waveform inversion tomography and information provided by short period seismic observables. We have extended the 30 1D radially anisotropic shear velocity profiles of Calò et al. 2016 to several other stations, for which we used a recent shear velocity model (Clouzet et al., 2017) as constraint in the modeling. These 1D profiles, including both isotropic and anisotropic discontinuities in the upper mantle (above 300 km depth) are then used to build a 3D starting model for the full waveform tomographic inversion. This model is built after 1) homogenization of the layered 1D models and 2) interpolation between the 1D smooth profiles and the model of Clouzet et al. 2017, resulting in a smooth 3D starting model. Waveforms used in the inversion are filtered at periods longer than 30s. We use the SEM code "RegSEM" for forward computations and a quasi-Newton inversion approach in which kernels are computed using normal mode perturbation theory. The resulting volumetric velocity perturbations around the homogenized starting model are then added to the discontinuous 3D starting model by dehomogenizing the model. We present here the first results of such an approach for refining structure in the North American continent.

Flexible kinematic earthquake rupture inversion of tele-seismic waveforms: Application to the 2013 Balochistan, Pakistan earthquake

NASA Astrophysics Data System (ADS)

Shimizu, K.; Yagi, Y.; Okuwaki, R.; Kasahara, A.

2017-12-01

The kinematic earthquake rupture models are useful to derive statistics and scaling properties of the large and great earthquakes. However, the kinematic rupture models for the same earthquake are often different from one another. Such sensitivity of the modeling prevents us to understand the statistics and scaling properties of the earthquakes. Yagi and Fukahata (2011) introduces the uncertainty of Green's function into the tele-seismic waveform inversion, and shows that the stable spatiotemporal distribution of slip-rate can be obtained by using an empirical Bayesian scheme. One of the unsolved problems in the inversion rises from the modeling error originated from an uncertainty of a fault-model setting. Green's function near the nodal plane of focal mechanism is known to be sensitive to the slight change of the assumed fault geometry, and thus the spatiotemporal distribution of slip-rate should be distorted by the modeling error originated from the uncertainty of the fault model. We propose a new method accounting for the complexity in the fault geometry by additionally solving the focal mechanism on each space knot. Since a solution of finite source inversion gets unstable with an increasing of flexibility of the model, we try to estimate a stable spatiotemporal distribution of focal mechanism in the framework of Yagi and Fukahata (2011). We applied the proposed method to the 52 tele-seismic P-waveforms of the 2013 Balochistan, Pakistan earthquake. The inverted-potency distribution shows unilateral rupture propagation toward southwest of the epicenter, and the spatial variation of the focal mechanisms shares the same pattern as the fault-curvature along the tectonic fabric. On the other hand, the broad pattern of rupture process, including the direction of rupture propagation, cannot be reproduced by an inversion analysis under the assumption that the faulting occurred on a single flat plane. These results show that the modeling error caused by simplifying the fault model is non-negligible in the tele-seismic waveform inversion of the 2013 Balochistan, Pakistan earthquake.

Identification of complex stiffness tensor from waveform reconstruction

NASA Astrophysics Data System (ADS)

Leymarie, N.; Aristégui, C.; Audoin, B.; Baste, S.

2002-03-01

An inverse method is proposed in order to determine the viscoelastic properties of composite-material plates from the plane-wave transmitted acoustic field. Analytical formulations of both the plate transmission coefficient and its first and second derivatives are established, and included in a two-step inversion scheme. Two objective functions to be minimized are then designed by considering the well-known maximum-likelihood principle and by using an analytic signal formulation. Through these innovative objective functions, the robustness of the inversion process against high level of noise in waveforms is improved and the method can be applied to a very thin specimen. The suitability of the inversion process for viscoelastic property identification is demonstrated using simulated data for composite materials with different anisotropy and damping degrees. A study of the effect of the rheologic model choice on the elastic property identification emphasizes the relevance of using a phenomenological description considering viscosity. Experimental characterizations show then the good reliability of the proposed approach. Difficulties arise experimentally for particular anisotropic media.

Mini-batch optimized full waveform inversion with geological constrained gradient filtering

NASA Astrophysics Data System (ADS)

Yang, Hui; Jia, Junxiong; Wu, Bangyu; Gao, Jinghuai

2018-05-01

High computation cost and generating solutions without geological sense have hindered the wide application of Full Waveform Inversion (FWI). Source encoding technique is a way to dramatically reduce the cost of FWI but subject to fix-spread acquisition setup requirement and slow convergence for the suppression of cross-talk. Traditionally, gradient regularization or preconditioning is applied to mitigate the ill-posedness. An isotropic smoothing filter applied on gradients generally gives non-geological inversion results, and could also introduce artifacts. In this work, we propose to address both the efficiency and ill-posedness of FWI by a geological constrained mini-batch gradient optimization method. The mini-batch gradient descent optimization is adopted to reduce the computation time by choosing a subset of entire shots for each iteration. By jointly applying the structure-oriented smoothing to the mini-batch gradient, the inversion converges faster and gives results with more geological meaning. Stylized Marmousi model is used to show the performance of the proposed method on realistic synthetic model.

A frozen Gaussian approximation-based multi-level particle swarm optimization for seismic inversion

DOE Office of Scientific and Technical Information (OSTI.GOV)

Li, Jinglai, E-mail: jinglaili@sjtu.edu.cn; Lin, Guang, E-mail: lin491@purdue.edu; Computational Sciences and Mathematics Division, Pacific Northwest National Laboratory, Richland, WA 99352

2015-09-01

In this paper, we propose a frozen Gaussian approximation (FGA)-based multi-level particle swarm optimization (MLPSO) method for seismic inversion of high-frequency wave data. The method addresses two challenges in it: First, the optimization problem is highly non-convex, which makes hard for gradient-based methods to reach global minima. This is tackled by MLPSO which can escape from undesired local minima. Second, the character of high-frequency of seismic waves requires a large number of grid points in direct computational methods, and thus renders an extremely high computational demand on the simulation of each sample in MLPSO. We overcome this difficulty by threemore » steps: First, we use FGA to compute high-frequency wave propagation based on asymptotic analysis on phase plane; Then we design a constrained full waveform inversion problem to prevent the optimization search getting into regions of velocity where FGA is not accurate; Last, we solve the constrained optimization problem by MLPSO that employs FGA solvers with different fidelity. The performance of the proposed method is demonstrated by a two-dimensional full-waveform inversion example of the smoothed Marmousi model.« less

Inversion for slip distribution using teleseismic P waveforms: North Palm Springs, Borah Peak, and Michoacan earthquakes

USGS Publications Warehouse

Mendoza, C.; Hartzell, S.H.

1988-01-01

We have inverted the teleseismic P waveforms recorded by stations of the Global Digital Seismograph Network for the 8 July 1986 North Palm Springs, California, the 28 October 1983 Borah Peak, Idaho, and the 19 September 1985 Michoacan, Mexico, earthquakes to recover the distribution of slip on each of the faults using a point-by-point inversion method with smoothing and positivity constraints. Results of the inversion indicate that the Global digital Seismograph Network data are useful for deriving fault dislocation models for moderate to large events. However, a wide range of frequencies is necessary to infer the distribution of slip on the earthquake fault. Although the long-period waveforms define the size (dimensions and seismic moment) of the earthquake, data at shorter period provide additional constraints on the variation of slip on the fault. Dislocation models obtained for all three earthquakes are consistent with a heterogeneous rupture process where failure is controlled largely by the size and location of high-strength asperity regions. -from Authors

2-D traveltime and waveform inversion for improved seismic imaging: Naga Thrust and Fold Belt, India

NASA Astrophysics Data System (ADS)

Jaiswal, Priyank; Zelt, Colin A.; Bally, Albert W.; Dasgupta, Rahul

2008-05-01

Exploration along the Naga Thrust and Fold Belt in the Assam province of Northeast India encounters geological as well as logistic challenges. Drilling for hydrocarbons, traditionally guided by surface manifestations of the Naga thrust fault, faces additional challenges in the northeast where the thrust fault gradually deepens leaving subtle surface expressions. In such an area, multichannel 2-D seismic data were collected along a line perpendicular to the trend of the thrust belt. The data have a moderate signal-to-noise ratio and suffer from ground roll and other acquisition-related noise. In addition to data quality, the complex geology of the thrust belt limits the ability of conventional seismic processing to yield a reliable velocity model which in turn leads to poor subsurface image. In this paper, we demonstrate the application of traveltime and waveform inversion as supplements to conventional seismic imaging and interpretation processes. Both traveltime and waveform inversion utilize the first arrivals that are typically discarded during conventional seismic processing. As a first step, a smooth velocity model with long wavelength characteristics of the subsurface is estimated through inversion of the first-arrival traveltimes. This velocity model is then used to obtain a Kirchhoff pre-stack depth-migrated image which in turn is used for the interpretation of the fault. Waveform inversion is applied to the central part of the seismic line to a depth of ~1 km where the quality of the migrated image is poor. Waveform inversion is performed in the frequency domain over a series of iterations, proceeding from low to high frequency (11-19 Hz) using the velocity model from traveltime inversion as the starting model. In the end, the pre-stack depth-migrated image and the waveform inversion model are jointly interpreted. This study demonstrates that a combination of traveltime and waveform inversion with Kirchhoff pre-stack depth migration is a promising approach for the interpretation of geological structures in a thrust belt.

Viscoacoustic anisotropic full waveform inversion

NASA Astrophysics Data System (ADS)

Qu, Yingming; Li, Zhenchun; Huang, Jianping; Li, Jinli

2017-01-01

A viscoacoustic vertical transverse isotropic (VTI) quasi-differential wave equation, which takes account for both the viscosity and anisotropy of media, is proposed for wavefield simulation in this study. The finite difference method is used to solve the equations, for which the attenuation terms are solved in the wavenumber domain, and all remaining terms in the time-space domain. To stabilize the adjoint wavefield, robust regularization operators are applied to the wave equation to eliminate the high-frequency component of the numerical noise produced during the backward propagation of the viscoacoustic wavefield. Based on these strategies, we derive the corresponding gradient formula and implement a viscoacoustic VTI full waveform inversion (FWI). Numerical tests verify that our proposed viscoacoustic VTI FWI can produce accurate and stable inversion results for viscoacoustic VTI data sets. In addition, we test our method's sensitivity to velocity, Q, and anisotropic parameters. Our results show that the sensitivity to velocity is much higher than that to Q and anisotropic parameters. As such, our proposed method can produce acceptable inversion results as long as the Q and anisotropic parameters are within predefined thresholds.

A Synthetic Study on the Resolution of 2D Elastic Full Waveform Inversion

NASA Astrophysics Data System (ADS)

Cui, C.; Wang, Y.

2017-12-01

Gradient based full waveform inversion is an effective method in seismic study, it makes full use of the information given by seismic records and is capable of providing a more accurate model of the interior of the earth at a relatively low computational cost. However, the strong non-linearity of the problem brings about many difficulties in the assessment of its resolution. Synthetic inversions are therefore helpful before an inversion based on real data is made. Checker-board test is a commonly used method, but it is not always reliable due to the significant difference between a checker-board and the true model. Our study aims to provide a basic understanding of the resolution of 2D elastic inversion by examining three main factors that affect the inversion result respectively: 1. The structural characteristic of the model; 2. The level of similarity between the initial model and the true model; 3. The spacial distribution of sources and receivers. We performed about 150 synthetic inversions to demonstrate how each factor contributes to quality of the result, and compared the inversion results with those achieved by checker-board tests. The study can be a useful reference to assess the resolution of an inversion in addition to regular checker-board tests, or to determine whether the seismic data of a specific region is sufficient for a successful inversion.

Application of Adjoint Method and Spectral-Element Method to Tomographic Inversion of Regional Seismological Structure Beneath Japanese Islands

NASA Astrophysics Data System (ADS)

Tsuboi, S.; Miyoshi, T.; Obayashi, M.; Tono, Y.; Ando, K.

2014-12-01

Recent progress in large scale computing by using waveform modeling technique and high performance computing facility has demonstrated possibilities to perform full-waveform inversion of three dimensional (3D) seismological structure inside the Earth. We apply the adjoint method (Liu and Tromp, 2006) to obtain 3D structure beneath Japanese Islands. First we implemented Spectral-Element Method to K-computer in Kobe, Japan. We have optimized SPECFEM3D_GLOBE (Komatitsch and Tromp, 2002) by using OpenMP so that the code fits hybrid architecture of K-computer. Now we could use 82,134 nodes of K-computer (657,072 cores) to compute synthetic waveform with about 1 sec accuracy for realistic 3D Earth model and its performance was 1.2 PFLOPS. We use this optimized SPECFEM3D_GLOBE code and take one chunk around Japanese Islands from global mesh and compute synthetic seismograms with accuracy of about 10 second. We use GAP-P2 mantle tomography model (Obayashi et al., 2009) as an initial 3D model and use as many broadband seismic stations available in this region as possible to perform inversion. We then use the time windows for body waves and surface waves to compute adjoint sources and calculate adjoint kernels for seismic structure. We have performed several iteration and obtained improved 3D structure beneath Japanese Islands. The result demonstrates that waveform misfits between observed and theoretical seismograms improves as the iteration proceeds. We now prepare to use much shorter period in our synthetic waveform computation and try to obtain seismic structure for basin scale model, such as Kanto basin, where there are dense seismic network and high seismic activity. Acknowledgements: This research was partly supported by MEXT Strategic Program for Innovative Research. We used F-net seismograms of the National Research Institute for Earth Science and Disaster Prevention.

Probabilistic source mechanism estimation based on body-wave waveforms through shift and stack algorithm

NASA Astrophysics Data System (ADS)

Massin, F.; Malcolm, A. E.

2017-12-01

Knowing earthquake source mechanisms gives valuable information for earthquake response planning and hazard mitigation. Earthquake source mechanisms can be analyzed using long period waveform inversion (for moderate size sources with sufficient signal to noise ratio) and body-wave first motion polarity or amplitude ratio inversion (for micro-earthquakes with sufficient data coverage). A robust approach that gives both source mechanisms and their associated probabilities across all source scales would greatly simplify the determination of source mechanisms and allow for more consistent interpretations of the results. Following previous work on shift and stack approaches, we develop such a probabilistic source mechanism analysis, using waveforms, which does not require polarity picking. For a given source mechanism, the first period of the observed body-waves is selected for all stations, multiplied by their corresponding theoretical polarity and stacked together. (The first period is found from a manually picked travel time by measuring the central period where the signal power is concentrated, using the second moment of the power spectral density function.) As in other shift and stack approaches, our method is not based on the optimization of an objective function through an inversion. Instead, the power of the polarity-corrected stack is a proxy for the likelihood of the trial source mechanism, with the most powerful stack corresponding to the most likely source mechanism. Using synthetic data, we test our method for robustness to the data coverage, coverage gap, signal to noise ratio, travel-time picking errors and non-double couple component. We then present results for field data in a volcano-tectonic context. Our results are reliable when constrained by 15 body-wavelets, with gap below 150 degrees, signal to noise ratio over 1 and arrival time error below a fifth of the period (0.2T) of the body-wave. We demonstrate that the source scanning approach for source mechanism analysis has similar advantages to waveform inversion (full waveform data, no manual intervention, probabilistic approach) and similar applicability to polarity inversion (any source size, any instrument type).

Estimating uncertainty of Full Waveform Inversion with Ensemble-based methods

NASA Astrophysics Data System (ADS)

Thurin, J.; Brossier, R.; Métivier, L.

2017-12-01

Uncertainty estimation is one key feature of tomographic applications for robust interpretation. However, this information is often missing in the frame of large scale linearized inversions, and only the results at convergence are shown, despite the ill-posed nature of the problem. This issue is common in the Full Waveform Inversion community.While few methodologies have already been proposed in the literature, standard FWI workflows do not include any systematic uncertainty quantifications methods yet, but often try to assess the result's quality through cross-comparison with other results from seismic or comparison with other geophysical data. With the development of large seismic networks/surveys, the increase in computational power and the more and more systematic application of FWI, it is crucial to tackle this problem and to propose robust and affordable workflows, in order to address the uncertainty quantification problem faced for near surface targets, crustal exploration, as well as regional and global scales.In this work (Thurin et al., 2017a,b), we propose an approach which takes advantage of the Ensemble Transform Kalman Filter (ETKF) proposed by Bishop et al., (2001), in order to estimate a low-rank approximation of the posterior covariance matrix of the FWI problem, allowing us to evaluate some uncertainty information of the solution. Instead of solving the FWI problem through a Bayesian inversion with the ETKF, we chose to combine a conventional FWI, based on local optimization, and the ETKF strategies. This scheme allows combining the efficiency of local optimization for solving large scale inverse problems and make the sampling of the local solution space possible thanks to its embarrassingly parallel property. References:Bishop, C. H., Etherton, B. J. and Majumdar, S. J., 2001. Adaptive sampling with the ensemble transform Kalman filter. Part I: Theoretical aspects. Monthly weather review, 129(3), 420-436.Thurin, J., Brossier, R. and Métivier, L. 2017,a.: Ensemble-Based Uncertainty Estimation in Full Waveform Inversion. 79th EAGE Conference and Exhibition 2017, (12 - 15 June, 2017)Thurin, J., Brossier, R. and Métivier, L. 2017,b.: An Ensemble-Transform Kalman Filter - Full Waveform Inversion scheme for Uncertainty estimation; SEG Technical Program Expanded Abstracts 2012

Towards a Full Waveform Ambient Noise Inversion

NASA Astrophysics Data System (ADS)

Sager, K.; Ermert, L. A.; Boehm, C.; Fichtner, A.

2015-12-01

Noise tomography usually works under the assumption that the inter-station ambient noise correlation is equal to a scaled version of the Green's function between the two receivers. This assumption, however, is only met under specific conditions, for instance, wavefield diffusivity and equipartitioning, zero attenuation, etc., that are typically not satisfied in the Earth. This inconsistency inhibits the exploitation of the full waveform information contained in noise correlations regarding Earth structure and noise generation. To overcome this limitation we attempt to develop a method that consistently accounts for noise distribution, 3D heterogeneous Earth structure and the full seismic wave propagation physics in order to improve the current resolution of tomographic images of the Earth. As an initial step towards a full waveform ambient noise inversion we develop a preliminary inversion scheme based on a 2D finite-difference code simulating correlation functions and on adjoint techniques. With respect to our final goal, a simultaneous inversion for noise distribution and Earth structure, we address the following two aspects: (1) the capabilities of different misfit functionals to image wave speed anomalies and source distribution and (2) possible source-structure trade-offs, especially to what extent unresolvable structure could be mapped into the inverted noise source distribution and vice versa.

Seismic waveform tomography with shot-encoding using a restarted L-BFGS algorithm.

PubMed

Rao, Ying; Wang, Yanghua

2017-08-17

In seismic waveform tomography, or full-waveform inversion (FWI), one effective strategy used to reduce the computational cost is shot-encoding, which encodes all shots randomly and sums them into one super shot to significantly reduce the number of wavefield simulations in the inversion. However, this process will induce instability in the iterative inversion regardless of whether it uses a robust limited-memory BFGS (L-BFGS) algorithm. The restarted L-BFGS algorithm proposed here is both stable and efficient. This breakthrough ensures, for the first time, the applicability of advanced FWI methods to three-dimensional seismic field data. In a standard L-BFGS algorithm, if the shot-encoding remains unchanged, it will generate a crosstalk effect between different shots. This crosstalk effect can only be suppressed by employing sufficient randomness in the shot-encoding. Therefore, the implementation of the L-BFGS algorithm is restarted at every segment. Each segment consists of a number of iterations; the first few iterations use an invariant encoding, while the remainder use random re-coding. This restarted L-BFGS algorithm balances the computational efficiency of shot-encoding, the convergence stability of the L-BFGS algorithm, and the inversion quality characteristic of random encoding in FWI.

Source mechanism of long-period events at Kusatsu-Shirane Volcano, Japan, inferred from waveform inversion of the effective excitation functions

USGS Publications Warehouse

Nakano, M.; Kumagai, H.; Chouet, B.A.

2003-01-01

We investigate the source mechanism of long-period (LP) events observed at Kusatsu-Shirane Volcano, Japan, based on waveform inversions of their effective excitation functions. The effective excitation function, which represents the apparent excitation observed at individual receivers, is estimated by applying an autoregressive filter to the LP waveform. Assuming a point source, we apply this method to seven LP events the waveforms of which are characterized by simple decaying and nearly monochromatic oscillations with frequency in the range 1-3 Hz. The results of the waveform inversions show dominant volumetric change components accompanied by single force components, common to all the events analyzed, and suggesting a repeated activation of a sub-horizontal crack located 300 m beneath the summit crater lakes. Based on these results, we propose a model of the source process of LP seismicity, in which a gradual buildup of steam pressure in a hydrothermal crack in response to magmatic heat causes repeated discharges of steam from the crack. The rapid discharge of fluid causes the collapse of the fluid-filled crack and excites acoustic oscillations of the crack, which produce the characteristic waveforms observed in the LP events. The presence of a single force synchronous with the collapse of the crack is interpreted as the release of gravitational energy that occurs as the slug of steam ejected from the crack ascends toward the surface and is replaced by cooler water flowing downward in a fluid-filled conduit linking the crack and the base of the crater lake. ?? 2003 Elsevier Science B.V. All rights reserved.

Total variation regularization for seismic waveform inversion using an adaptive primal dual hybrid gradient method

NASA Astrophysics Data System (ADS)

Yong, Peng; Liao, Wenyuan; Huang, Jianping; Li, Zhenchuan

2018-04-01

Full waveform inversion is an effective tool for recovering the properties of the Earth from seismograms. However, it suffers from local minima caused mainly by the limited accuracy of the starting model and the lack of a low-frequency component in the seismic data. Because of the high velocity contrast between salt and sediment, the relation between the waveform and velocity perturbation is strongly nonlinear. Therefore, salt inversion can easily get trapped in the local minima. Since the velocity of salt is nearly constant, we can make the most of this characteristic with total variation regularization to mitigate the local minima. In this paper, we develop an adaptive primal dual hybrid gradient method to implement total variation regularization by projecting the solution onto a total variation norm constrained convex set, through which the total variation norm constraint is satisfied at every model iteration. The smooth background velocities are first inverted and the perturbations are gradually obtained by successively relaxing the total variation norm constraints. Numerical experiment of the projection of the BP model onto the intersection of the total variation norm and box constraints has demonstrated the accuracy and efficiency of our adaptive primal dual hybrid gradient method. A workflow is designed to recover complex salt structures in the BP 2004 model and the 2D SEG/EAGE salt model, starting from a linear gradient model without using low-frequency data below 3 Hz. The salt inversion processes demonstrate that wavefield reconstruction inversion with a total variation norm and box constraints is able to overcome local minima and inverts the complex salt velocity layer by layer.

Waveform inversion of volcano-seismic signals for an extended source

USGS Publications Warehouse

Nakano, M.; Kumagai, H.; Chouet, B.; Dawson, P.

2007-01-01

We propose a method to investigate the dimensions and oscillation characteristics of the source of volcano-seismic signals based on waveform inversion for an extended source. An extended source is realized by a set of point sources distributed on a grid surrounding the centroid of the source in accordance with the source geometry and orientation. The source-time functions for all point sources are estimated simultaneously by waveform inversion carried out in the frequency domain. We apply a smoothing constraint to suppress short-scale noisy fluctuations of source-time functions between adjacent sources. The strength of the smoothing constraint we select is that which minimizes the Akaike Bayesian Information Criterion (ABIC). We perform a series of numerical tests to investigate the capability of our method to recover the dimensions of the source and reconstruct its oscillation characteristics. First, we use synthesized waveforms radiated by a kinematic source model that mimics the radiation from an oscillating crack. Our results demonstrate almost complete recovery of the input source dimensions and source-time function of each point source, but also point to a weaker resolution of the higher modes of crack oscillation. Second, we use synthetic waveforms generated by the acoustic resonance of a fluid-filled crack, and consider two sets of waveforms dominated by the modes with wavelengths 2L/3 and 2W/3, or L and 2L/5, where W and L are the crack width and length, respectively. Results from these tests indicate that the oscillating signature of the 2L/3 and 2W/3 modes are successfully reconstructed. The oscillating signature of the L mode is also well recovered, in contrast to results obtained for a point source for which the moment tensor description is inadequate. However, the oscillating signature of the 2L/5 mode is poorly recovered owing to weaker resolution of short-scale crack wall motions. The triggering excitations of the oscillating cracks are successfully reconstructed. Copyright 2007 by the American Geophysical Union.

Real-time digital signal recovery for a multi-pole low-pass transfer function system.

PubMed

Lee, Jhinhwan

2017-08-01

In order to solve the problems of waveform distortion and signal delay by many physical and electrical systems with multi-pole linear low-pass transfer characteristics, a simple digital-signal-processing (DSP)-based method of real-time recovery of the original source waveform from the distorted output waveform is proposed. A mathematical analysis on the convolution kernel representation of the single-pole low-pass transfer function shows that the original source waveform can be accurately recovered in real time using a particular moving average algorithm applied on the input stream of the distorted waveform, which can also significantly reduce the overall delay time constant. This method is generalized for multi-pole low-pass systems and has noise characteristics of the inverse of the low-pass filter characteristics. This method can be applied to most sensors and amplifiers operating close to their frequency response limits to improve the overall performance of data acquisition systems and digital feedback control systems.

Lane marking detection based on waveform analysis and CNN

NASA Astrophysics Data System (ADS)

Ye, Yang Yang; Chen, Hou Jin; Hao, Xiao Li

2017-06-01

Lane markings detection is a very important part of the ADAS to avoid traffic accidents. In order to obtain accurate lane markings, in this work, a novel and efficient algorithm is proposed, which analyses the waveform generated from the road image after inverse perspective mapping (IPM). The algorithm includes two main stages: the first stage uses an image preprocessing including a CNN to reduce the background and enhance the lane markings. The second stage obtains the waveform of the road image and analyzes the waveform to get lanes. The contribution of this work is that we introduce local and global features of the waveform to detect the lane markings. The results indicate the proposed method is robust in detecting and fitting the lane markings.

High resolution aquifer characterization using crosshole GPR full-waveform tomography

NASA Astrophysics Data System (ADS)

Gueting, N.; Vienken, T.; Klotzsche, A.; Van Der Kruk, J.; Vanderborght, J.; Caers, J.; Vereecken, H.; Englert, A.

2016-12-01

Limited knowledge about the spatial distribution of aquifer properties typically constrains our ability to predict subsurface flow and transport. Here, we investigate the value of using high resolution full-waveform inversion of cross-borehole ground penetrating radar (GPR) data for aquifer characterization. By stitching together GPR tomograms from multiple adjacent crosshole planes, we are able to image, with a decimeter scale resolution, the dielectric permittivity and electrical conductivity of an alluvial aquifer along cross-sections of 50 m length and 10 m depth. A logistic regression model is employed to predict the spatial distribution of lithological facies on the basis of the GPR results. Vertical profiles of porosity and hydraulic conductivity from direct-push, flowmeter and grain size data suggest that the GPR predicted facies classification is meaningful with regard to porosity and hydraulic conductivity, even though the distributions of individual facies show some overlap and the absolute hydraulic conductivities from the different methods (direct-push, flowmeter, grain size) differ up to approximately one order of magnitude. Comparison of the GPR predicted facies architecture with tracer test data suggests that the plume splitting observed in a tracer experiment was caused by a hydraulically low-conductive sand layer with a thickness of only a few decimeters. Because this sand layer is identified by GPR full-waveform inversion but not by conventional GPR ray-based inversion we conclude that the improvement in spatial resolution due to full-waveform inversion is crucial to detect small-scale aquifer structures that are highly relevant for solute transport.

Moment tensor inversions using strong motion waveforms of Taiwan TSMIP data, 1993–2009

USGS Publications Warehouse

Chang, Kaiwen; Chi, Wu-Cheng; Gung, Yuancheng; Dreger, Douglas; Lee, William H K.; Chiu, Hung-Chie

2011-01-01

Earthquake source parameters are important for earthquake studies and seismic hazard assessment. Moment tensors are among the most important earthquake source parameters, and are now routinely derived using modern broadband seismic networks around the world. Similar waveform inversion techniques can also apply to other available data, including strong-motion seismograms. Strong-motion waveforms are also broadband, and recorded in many regions since the 1980s. Thus, strong-motion data can be used to augment moment tensor catalogs with a much larger dataset than that available from the high-gain, broadband seismic networks. However, a systematic comparison between the moment tensors derived from strong motion waveforms and high-gain broadband waveforms has not been available. In this study, we inverted the source mechanisms of Taiwan earthquakes between 1993 and 2009 by using the regional moment tensor inversion method using digital data from several hundred stations in the Taiwan Strong Motion Instrumentation Program (TSMIP). By testing different velocity models and filter passbands, we were able to successfully derive moment tensor solutions for 107 earthquakes of Mw >= 4.8. The solutions for large events agree well with other available moment tensor catalogs derived from local and global broadband networks. However, for Mw = 5.0 or smaller events, we consistently over estimated the moment magnitudes by 0.5 to 1.0. We have tested accelerograms, and velocity waveforms integrated from accelerograms for the inversions, and found the results are similar. In addition, we used part of the catalogs to study important seismogenic structures in the area near Meishan Taiwan which was the site of a very damaging earthquake a century ago, and found that the structures were dominated by events with complex right-lateral strike-slip faulting during the recent decade. The procedures developed from this study may be applied to other strong-motion datasets to compliment or fill gaps in catalogs from regional broadband networks and teleseismic networks.

Nonlinear 1D and 2D waveform inversions of SS precursors and their applications in mantle seismic imaging

NASA Astrophysics Data System (ADS)

Dokht, R.; Gu, Y. J.; Sacchi, M. D.

2016-12-01

Seismic velocities and the topography of mantle discontinuities are crucial for the understanding of mantle structure, dynamics and mineralogy. While these two observables are closely linked, the vast majority of high-resolution seismic images are retrieved under the assumption of horizontally stratified mantle interfaces. This conventional correction-based process could lead to considerable errors due to the inherent trade-off between velocity and discontinuity depth. In this study, we introduce a nonlinear joint waveform inversion method that simultaneously recovers discontinuity depths and seismic velocities using the waveforms of SS precursors. Our target region is the upper mantle and transition zone beneath Northeast Asia. In this region, the inversion outcomes clearly delineate a westward dipping high-velocity structure in association with the subducting Pacific plate. Above the flat part of the slab west of the Japan sea, our results show a shear wave velocity reduction of 1.5% in the upper mantle and 10-15 km depression of the 410 km discontinuity beneath the Changbaishan volcanic field. We also identify the maximum correlation between shear velocity and transition zone thickness at an approximate slab dip of 30 degrees, which is consistent with previously reported values in this region.To validate the results of the 1D waveform inversion of SS precursors, we discretize the mantle beneath the study region and conduct a 2D waveform tomographic survey using the same nonlinear approach. The problem is simplified by adopting the discontinuity depths from the 1D inversion and solving only for perturbations in shear velocities. The resulting models obtained from the 1D and 2D approaches are self-consistent. Low-velocities beneath the Changbai intraplate volcano likely persist to a depth of 500 km. Collectively, our seismic observations suggest that the active volcanoes in eastern China may be fueled by a hot thermal anomaly originating from the mantle transition zone.

Source mechanism analysis of central Aceh earthquake July 2, 2013 Mw 6.2 using moment tensor inversion with BMKG waveform data

DOE Office of Scientific and Technical Information (OSTI.GOV)

Prasetyo, Retno Agung, E-mail: prasetyo.agung@bmkg.go.id; Heryandoko, Nova; Afnimar

The source mechanism of earthquake on July 2, 2013 was investigated by using moment tensor inversion. The result also compared by the field observation. Five waveform data of BMKG’s seismic network used to estimate the mechanism of earthquake, namely ; KCSI, MLSI, LASI, TPTI and SNSI. Main shock data taken during 200 seconds and filtered by using Butterworth band pass method from 0.03 to 0.05 Hz of frequency. Moment tensor inversion method is applied based on the point source assumption. Furthermore, the Green function calculated using the extended reflectivity method which modified by Kohketsu. The inversion result showed a strike-slipmore » faulting, where the nodal plane strike/dip/rake (124/80.6/152.8) and minimum variance value 0.3285 at a depth of 6 km (centroid). It categorized as a shallow earthquake. Field observation indicated that the building orientated to the east. It can be related to the southwest of dip direction which has 152 degrees of slip. As conclusion, the Pressure (P) and Tension (T) axis described dominant compression is happen from the south which is caused by pressure of the Indo-Australian plate.« less

Investigation of the reconstruction accuracy of guided wave tomography using full waveform inversion

NASA Astrophysics Data System (ADS)

Rao, Jing; Ratassepp, Madis; Fan, Zheng

2017-07-01

Guided wave tomography is a promising tool to accurately determine the remaining wall thicknesses of corrosion damages, which are among the major concerns for many industries. Full Waveform Inversion (FWI) algorithm is an attractive guided wave tomography method, which uses a numerical forward model to predict the waveform of guided waves when propagating through corrosion defects, and an inverse model to reconstruct the thickness map from the ultrasonic signals captured by transducers around the defect. This paper discusses the reconstruction accuracy of the FWI algorithm on plate-like structures by using simulations as well as experiments. It was shown that this algorithm can obtain a resolution of around 0.7 wavelengths for defects with smooth depth variations from the acoustic modeling data, and about 1.5-2 wavelengths from the elastic modeling data. Further analysis showed that the reconstruction accuracy is also dependent on the shape of the defect. It was demonstrated that the algorithm maintains the accuracy in the case of multiple defects compared to conventional algorithms based on Born approximation.

Including Short Period Constraints In the Construction of Full Waveform Tomographic Models

NASA Astrophysics Data System (ADS)

Roy, C.; Calo, M.; Bodin, T.; Romanowicz, B. A.

2015-12-01

Thanks to the introduction of the Spectral Element Method (SEM) in seismology, which allows accurate computation of the seismic wavefield in complex media, the resolution of regional and global tomographic models has improved in recent years. However, due to computational costs, only long period waveforms are considered, and only long wavelength structure can be constrained. Thus, the resulting 3D models are smooth, and only represent a small volumetric perturbation around a smooth reference model that does not include upper-mantle discontinuities (e.g. MLD, LAB). Extending the computations to shorter periods, necessary for the resolution of smaller scale features, is computationally challenging. In order to overcome these limitations and to account for layered structure in the upper mantle in our full waveform tomography, we include information provided by short period seismic observables (receiver functions and surface wave dispersion), sensitive to sharp boundaries and anisotropic structure respectively. In a first step, receiver functions and dispersion curves are used to generate a number of 1D radially anisotropic shear velocity profiles using a trans-dimensional Markov-chain Monte Carlo (MCMC) algorithm. These 1D profiles include both isotropic and anisotropic discontinuities in the upper mantle (above 300 km depth) beneath selected stationsand are then used to build a 3D starting model for the full waveform tomographic inversion. This model is built after 1) interpolation between the available 1D profiles, and 2) homogeneization of the layered 1D models to obtain an equivalent smooth 3D starting model in the period range of interest for waveform inversion. The waveforms used in the inversion are collected for paths contained in the region of study and filtered at periods longer than 40s. We use the spectral element code "RegSEM" (Cupillard et al., 2012) for forward computations and a quasi-Newton inversion approach in which kernels are computed using normal mode perturbation theory. We present here the first reults of such an approach after successive iterations of a full waveform tomography of the North American continent.

Comparison of the Cut-and-Paste and Full Moment Tensor Methods for Estimating Earthquake Source Parameters

NASA Astrophysics Data System (ADS)

Templeton, D.; Rodgers, A.; Helmberger, D.; Dreger, D.

2008-12-01

Earthquake source parameters (seismic moment, focal mechanism and depth) are now routinely reported by various institutions and network operators. These parameters are important for seismotectonic and earthquake ground motion studies as well as calibration of moment magnitude scales and model-based earthquake-explosion discrimination. Source parameters are often estimated from long-period three- component waveforms at regional distances using waveform modeling techniques with Green's functions computed for an average plane-layered models. One widely used method is waveform inversion for the full moment tensor (Dreger and Helmberger, 1993). This method (TDMT) solves for the moment tensor elements by performing a linearized inversion in the time-domain that minimizes the difference between the observed and synthetic waveforms. Errors in the seismic velocity structure inevitably arise due to either differences in the true average plane-layered structure or laterally varying structure. The TDMT method can account for errors in the velocity model by applying a single time shift at each station to the observed waveforms to best match the synthetics. Another method for estimating source parameters is the Cut-and-Paste (CAP) method. This method breaks the three-component regional waveforms into five windows: vertical and radial component Pnl; vertical and radial component Rayleigh wave; and transverse component Love waves. The CAP method performs a grid search over double-couple mechanisms and allows the synthetic waveforms for each phase (Pnl, Rayleigh and Love) to shift in time to account for errors in the Green's functions. Different filtering and weighting of the Pnl segment relative to surface wave segments enhances sensitivity to source parameters, however, some bias may be introduced. This study will compare the TDMT and CAP methods in two different regions in order to better understand the advantages and limitations of each method. Firstly, we will consider the northeastern China/Korean Peninsula region where average plane-layered structure is well known and relatively laterally homogenous. Secondly, we will consider the Middle East where crustal and upper mantle structure is laterally heterogeneous due to recent and ongoing tectonism. If time allows we will investigate the efficacy of each method for retrieving source parameters from synthetic data generated using a three-dimensional model of seismic structure of the Middle East, where phase delays are known to arise from path-dependent structure.

Towards seismic waveform inversion of long-offset Ocean-Bottom Seismic data for deep crustal imaging offshore Western Australia

NASA Astrophysics Data System (ADS)

Monnier, S.; Lumley, D. E.; Kamei, R.; Goncharov, A.; Shragge, J. C.

2016-12-01

Ocean Bottom Seismic datasets have become increasingly used in recent years to develop high-resolution, wavelength-scale P-wave velocity models of the lithosphere from waveform inversion, due to their recording of long-offset transmitted phases. New OBS surveys evolve towards novel acquisition geometries involving longer offsets (several hundreds of km), broader frequency content (1-100 Hz), while receiver sampling often remains sparse (several km). Therefore, it is critical to assess the effects of such geometries on the eventual success and resolution of waveform inversion velocity models. In this study, we investigate the feasibility of waveform inversion on the Bart 2D OBS profile acquired offshore Western Australia, to investigate regional crustal and Moho structures. The dataset features 14 broadband seismometers (0.01-100 Hz) from AuScope's national OBS fleet, offsets in excess of 280 km, and a sparse receiver sampling (18 km). We perform our analysis in four stages: (1) field data analysis, (2) 2D P-wave velocity model building, synthetic data (3) modelling, and (4) waveform inversion. Data exploration shows high-quality active-source signal down to 2Hz, and usable first arrivals to offsets greater than 100 km. The background velocity model is constructed by combining crustal and Moho information in continental reference models (e.g., AuSREM, AusMoho). These low-resolution studies suggest a crustal thickness of 20-25 km along our seismic line and constitute a starting point for synthetic modelling and inversion. We perform synthetic 2D time-domain modelling to: (1) evaluate the misfit between synthetic and field data within the usable frequency band (2-10 Hz); (2) validate our velocity model; and (3) observe the effects of sparse OBS interval on data quality. Finally, we apply 2D acoustic frequency-domain waveform inversion to the synthetic data to generate velocity model updates. The inverted model is compared to the reference model to investigate the improved crustal resolution and Moho boundary delineation that could be realized using waveform inversion, and to evaluate the effects of the acquisition parameters. The inversion strategies developed through the synthetic tests will help the subsequent inversion of sparse, long-offset OBS field data.

A multi-frequency receiver function inversion approach for crustal velocity structure

NASA Astrophysics Data System (ADS)

Li, Xuelei; Li, Zhiwei; Hao, Tianyao; Wang, Sheng; Xing, Jian

2017-05-01

In order to constrain the crustal velocity structures better, we developed a new nonlinear inversion approach based on multi-frequency receiver function waveforms. With the global optimizing algorithm of Differential Evolution (DE), low-frequency receiver function waveforms can primarily constrain large-scale velocity structures, while high-frequency receiver function waveforms show the advantages in recovering small-scale velocity structures. Based on the synthetic tests with multi-frequency receiver function waveforms, the proposed approach can constrain both long- and short-wavelength characteristics of the crustal velocity structures simultaneously. Inversions with real data are also conducted for the seismic stations of KMNB in southeast China and HYB in Indian continent, where crustal structures have been well studied by former researchers. Comparisons of inverted velocity models from previous and our studies suggest good consistency, but better waveform fitness with fewer model parameters are achieved by our proposed approach. Comprehensive tests with synthetic and real data suggest that the proposed inversion approach with multi-frequency receiver function is effective and robust in inverting the crustal velocity structures.

Laplace-domain waveform modeling and inversion for the 3D acoustic-elastic coupled media

NASA Astrophysics Data System (ADS)

Shin, Jungkyun; Shin, Changsoo; Calandra, Henri

2016-06-01

Laplace-domain waveform inversion reconstructs long-wavelength subsurface models by using the zero-frequency component of damped seismic signals. Despite the computational advantages of Laplace-domain waveform inversion over conventional frequency-domain waveform inversion, an acoustic assumption and an iterative matrix solver have been used to invert 3D marine datasets to mitigate the intensive computing cost. In this study, we develop a Laplace-domain waveform modeling and inversion algorithm for 3D acoustic-elastic coupled media by using a parallel sparse direct solver library (MUltifrontal Massively Parallel Solver, MUMPS). We precisely simulate a real marine environment by coupling the 3D acoustic and elastic wave equations with the proper boundary condition at the fluid-solid interface. In addition, we can extract the elastic properties of the Earth below the sea bottom from the recorded acoustic pressure datasets. As a matrix solver, the parallel sparse direct solver is used to factorize the non-symmetric impedance matrix in a distributed memory architecture and rapidly solve the wave field for a number of shots by using the lower and upper matrix factors. Using both synthetic datasets and real datasets obtained by a 3D wide azimuth survey, the long-wavelength component of the P-wave and S-wave velocity models is reconstructed and the proposed modeling and inversion algorithm are verified. A cluster of 80 CPU cores is used for this study.

Teleseismic tomography for imaging Earth's upper mantle

NASA Astrophysics Data System (ADS)

Aktas, Kadircan

Teleseismic tomography is an important imaging tool in earthquake seismology, used to characterize lithospheric structure beneath a region of interest. In this study I investigate three different tomographic techniques applied to real and synthetic teleseismic data, with the aim of imaging the velocity structure of the upper mantle. First, by applying well established traveltime tomographic techniques to teleseismic data from southern Ontario, I obtained high-resolution images of the upper mantle beneath the lower Great Lakes. Two salient features of the 3D models are: (1) a patchy, NNW-trending low-velocity region, and (2) a linear, NE-striking high-velocity anomaly. I interpret the high-velocity anomaly as a possible relict slab associated with ca. 1.25 Ga subduction, whereas the low-velocity anomaly is interpreted as a zone of alteration and metasomatism associated with the ascent of magmas that produced the Late Cretaceous Monteregian plutons. The next part of the thesis is concerned with adaptation of existing full-waveform tomographic techniques for application to teleseismic body-wave observations. The method used here is intended to be complementary to traveltime tomography, and to take advantage of efficient frequency-domain methodologies that have been developed for inverting large controlled-source datasets. Existing full-waveform acoustic modelling and inversion codes have been modified to handle plane waves impinging from the base of the lithospheric model at a known incidence angle. A processing protocol has been developed to prepare teleseismic observations for the inversion algorithm. To assess the validity of the acoustic approximation, the processing procedure and modelling-inversion algorithm were tested using synthetic seismograms computed using an elastic Kirchhoff integral method. These tests were performed to evaluate the ability of the frequency-domain full-waveform inversion algorithm to recover topographic variations of the Moho under a variety of realistic scenarios. Results show that frequency-domain full-waveform tomography is generally successful in recovering both sharp and discontinuous features. Thirdly, I developed a new method for creating an initial background velocity model for the inversion algorithm, which is sufficiently close to the true model so that convergence is likely to be achieved. I adapted a method named Deformable Layer Tomography (DLT), which adjusts interfaces between layers rather than velocities within cells. I applied this method to a simple model comprising a single uniform crustal layer and a constant-velocity mantle, separated by an irregular Moho interface. A series of tests was performed to evaluate the sensitivity of the DLT algorithm; the results show that my algorithm produces useful results within a realistic range of incident-wave obliquity, incidence angle and signal-to-noise level. Keywords. Teleseismic tomography, full waveform tomography, deformable layer tomography, lower Great Lakes, crust and upper mantle.

Joint Inversion of Earthquake Source Parameters with local and teleseismic body waves

NASA Astrophysics Data System (ADS)

Chen, W.; Ni, S.; Wang, Z.

2011-12-01

In the classical source parameter inversion algorithm of CAP (Cut and Paste method, by Zhao and Helmberger), waveform data at near distances (typically less than 500km) are partitioned into Pnl and surface waves to account for uncertainties in the crustal models and different amplitude weight of body and surface waves. The classical CAP algorithms have proven effective for resolving source parameters (focal mechanisms, depth and moment) for earthquakes well recorded on relatively dense seismic network. However for regions covered with sparse stations, it is challenging to achieve precise source parameters . In this case, a moderate earthquake of ~M6 is usually recorded on only one or two local stations with epicentral distances less than 500 km. Fortunately, an earthquake of ~M6 can be well recorded on global seismic networks. Since the ray paths for teleseismic and local body waves sample different portions of the focal sphere, combination of teleseismic and local body wave data helps constrain source parameters better. Here we present a new CAP mothod (CAPjoint), which emploits both teleseismic body waveforms (P and SH waves) and local waveforms (Pnl, Rayleigh and Love waves) to determine source parameters. For an earthquake in Nevada that is well recorded with dense local network (USArray stations), we compare the results from CAPjoint with those from the traditional CAP method involving only of local waveforms , and explore the efficiency with bootstraping statistics to prove the results derived by CAPjoint are stable and reliable. Even with one local station included in joint inversion, accuracy of source parameters such as moment and strike can be much better improved.

Full seismic waveform tomography for upper-mantle structure in the Australasian region using adjoint methods

NASA Astrophysics Data System (ADS)

Fichtner, Andreas; Kennett, Brian L. N.; Igel, Heiner; Bunge, Hans-Peter

2009-12-01

We present a full seismic waveform tomography for upper-mantle structure in the Australasian region. Our method is based on spectral-element simulations of seismic wave propagation in 3-D heterogeneous earth models. The accurate solution of the forward problem ensures that waveform misfits are solely due to as yet undiscovered Earth structure and imprecise source descriptions, thus leading to more realistic tomographic images and source parameter estimates. To reduce the computational costs, we implement a long-wavelength equivalent crustal model. We quantify differences between the observed and the synthetic waveforms using time-frequency (TF) misfits. Their principal advantages are the separation of phase and amplitude misfits, the exploitation of complete waveform information and a quasi-linear relation to 3-D Earth structure. Fréchet kernels for the TF misfits are computed via the adjoint method. We propose a simple data compression scheme and an accuracy-adaptive time integration of the wavefields that allows us to reduce the storage requirements of the adjoint method by almost two orders of magnitude. To minimize the waveform phase misfit, we implement a pre-conditioned conjugate gradient algorithm. Amplitude information is incorporated indirectly by a restricted line search. This ensures that the cumulative envelope misfit does not increase during the inversion. An efficient pre-conditioner is found empirically through numerical experiments. It prevents the concentration of structural heterogeneity near the sources and receivers. We apply our waveform tomographic method to ~1000 high-quality vertical-component seismograms, recorded in the Australasian region between 1993 and 2008. The waveforms comprise fundamental- and higher-mode surface and long-period S body waves in the period range from 50 to 200 s. To improve the convergence of the algorithm, we implement a 3-D initial model that contains the long-wavelength features of the Australasian region. Resolution tests indicate that our algorithm converges after around 10 iterations and that both long- and short-wavelength features in the uppermost mantle are well resolved. There is evidence for effects related to the non-linearity in the inversion procedure. After 11 iterations we fit the data waveforms acceptably well; with no significant further improvements to be expected. During the inversion the total fitted seismogram length increases by 46 per cent, providing a clear indication of the efficiency and consistency of the iterative optimization algorithm. The resulting SV-wave velocity model reveals structural features of the Australasian upper mantle with great detail. We confirm the existence of a pronounced low-velocity band along the eastern margin of the continent that can be clearly distinguished against Precambrian Australia and the microcontinental Lord Howe Rise. The transition from Precambrian to Phanerozoic Australia (the Tasman Line) appears to be sharp down to at least 200 km depth. It mostly occurs further east of where it is inferred from gravity and magnetic anomalies. Also clearly visible are the Archean and Proterozoic cratons, the northward continuation of the continent and anomalously low S-wave velocities in the upper mantle in central Australia. This is, to the best of our knowledge, the first application of non-linear full seismic waveform tomography to a continental-scale problem.

Elastic full waveform inversion based on the homogenization method: theoretical framework and 2-D numerical illustrations

NASA Astrophysics Data System (ADS)

Capdeville, Yann; Métivier, Ludovic

2018-05-01

Seismic imaging is an efficient tool to investigate the Earth interior. Many of the different imaging techniques currently used, including the so-called full waveform inversion (FWI), are based on limited frequency band data. Such data are not sensitive to the true earth model, but to a smooth version of it. This smooth version can be related to the true model by the homogenization technique. Homogenization for wave propagation in deterministic media with no scale separation, such as geological media, has been recently developed. With such an asymptotic theory, it is possible to compute an effective medium valid for a given frequency band such that effective waveforms and true waveforms are the same up to a controlled error. In this work we make the link between limited frequency band inversion, mainly FWI, and homogenization. We establish the relation between a true model and an FWI result model. This relation is important for a proper interpretation of FWI images. We numerically illustrate, in the 2-D case, that an FWI result is at best the homogenized version of the true model. Moreover, it appears that the homogenized FWI model is quite independent of the FWI parametrization, as long as it has enough degrees of freedom. In particular, inverting for the full elastic tensor is, in each of our tests, always a good choice. We show how the homogenization can help to understand FWI behaviour and help to improve its robustness and convergence by efficiently constraining the solution space of the inverse problem.

High resolution aquifer characterization using crosshole GPR full-waveform tomography: Comparison with direct-push and tracer test data

NASA Astrophysics Data System (ADS)

Gueting, Nils; Vienken, Thomas; Klotzsche, Anja; van der Kruk, Jan; Vanderborght, Jan; Caers, Jef; Vereecken, Harry; Englert, Andreas

2017-01-01

Limited knowledge about the spatial distribution of aquifer properties typically constrains our ability to predict subsurface flow and transport. Here we investigate the value of using high resolution full-waveform inversion of cross-borehole ground penetrating radar (GPR) data for aquifer characterization. By stitching together GPR tomograms from multiple adjacent crosshole planes, we are able to image, with a decimeter scale resolution, the dielectric permittivity and electrical conductivity of an alluvial aquifer along cross sections of 50 m length and 10 m depth. A logistic regression model is employed to predict the spatial distribution of lithological facies on the basis of the GPR results. Vertical profiles of porosity and hydraulic conductivity from direct-push, flowmeter and grain size data suggest that the GPR predicted facies classification is meaningful with regard to porosity and hydraulic conductivity, even though the distributions of individual facies show some overlap and the absolute hydraulic conductivities from the different methods (direct-push, flowmeter, grain size) differ up to approximately one order of magnitude. Comparison of the GPR predicted facies architecture with tracer test data suggests that the plume splitting observed in a tracer experiment was caused by a hydraulically low-conductive sand layer with a thickness of only a few decimeters. Because this sand layer is identified by GPR full-waveform inversion but not by conventional GPR ray-based inversion we conclude that the improvement in spatial resolution due to full-waveform inversion is crucial to detect small-scale aquifer structures that are highly relevant for solute transport.

Towards Full-Waveform Ambient Noise Inversion

NASA Astrophysics Data System (ADS)

Sager, Korbinian; Ermert, Laura; Afanasiev, Michael; Boehm, Christian; Fichtner, Andreas

2017-04-01

Noise tomography usually works under the assumption that the inter-station ambient noise correlation is equal to a scaled version of the Green function between the two receivers. This assumption, however, is only met under specific conditions, e.g. wavefield diffusivity and equipartitioning, or the isotropic distribution of both mono- and dipolar uncorrelated noise sources. These assumptions are typically not satisfied in the Earth. This inconsistency inhibits the exploitation of the full waveform information contained in noise correlations in order to constrain Earth structure and noise generation. To overcome this limitation, we attempt to develop a method that consistently accounts for the distribution of noise sources, 3D heterogeneous Earth structure and the full seismic wave propagation physics. This is intended to improve the resolution of tomographic images, to refine noise source distribution, and thereby to contribute to a better understanding of both Earth structure and noise generation. First, we develop an inversion strategy based on a 2D finite-difference code using adjoint techniques. To enable a joint inversion for noise sources and Earth structure, we investigate the following aspects: i) the capability of different misfit functionals to image wave speed anomalies and source distribution and ii) possible source-structure trade-offs, especially to what extent unresolvable structure can be mapped into the inverted noise source distribution and vice versa. In anticipation of real-data applications, we present an extension of the open-source waveform modelling and inversion package Salvus (http://salvus.io). It allows us to compute correlation functions in 3D media with heterogeneous noise sources at the surface and the corresponding sensitivity kernels for the distribution of noise sources and Earth structure. By studying the effect of noise sources on correlation functions in 3D, we validate the aforementioned inversion strategy and prepare the workflow necessary for the first application of full waveform ambient noise inversion to a global dataset, for which a model for the distribution of noise sources is already available.

SeisFlows-Flexible waveform inversion software

NASA Astrophysics Data System (ADS)

Modrak, Ryan T.; Borisov, Dmitry; Lefebvre, Matthieu; Tromp, Jeroen

2018-06-01

SeisFlows is an open source Python package that provides a customizable waveform inversion workflow and framework for research in oil and gas exploration, earthquake tomography, medical imaging, and other areas. New methods can be rapidly prototyped in SeisFlows by inheriting from default inversion or migration classes, and code can be tested on 2D examples before application to more expensive 3D problems. Wave simulations must be performed using an external software package such as SPECFEM3D. The ability to interface with external solvers lends flexibility, and the choice of SPECFEM3D as a default option provides optional GPU acceleration and other useful capabilities. Through support for massively parallel solvers and interfaces for high-performance computing (HPC) systems, inversions with thousands of seismic traces and billions of model parameters can be performed. So far, SeisFlows has run on clusters managed by the Department of Defense, Chevron Corp., Total S.A., Princeton University, and the University of Alaska, Fairbanks.

Low frequency full waveform seismic inversion within a tree based Bayesian framework

NASA Astrophysics Data System (ADS)

Ray, Anandaroop; Kaplan, Sam; Washbourne, John; Albertin, Uwe

2018-01-01

Limited illumination, insufficient offset, noisy data and poor starting models can pose challenges for seismic full waveform inversion. We present an application of a tree based Bayesian inversion scheme which attempts to mitigate these problems by accounting for data uncertainty while using a mildly informative prior about subsurface structure. We sample the resulting posterior model distribution of compressional velocity using a trans-dimensional (trans-D) or Reversible Jump Markov chain Monte Carlo method in the wavelet transform domain of velocity. This allows us to attain rapid convergence to a stationary distribution of posterior models while requiring a limited number of wavelet coefficients to define a sampled model. Two synthetic, low frequency, noisy data examples are provided. The first example is a simple reflection + transmission inverse problem, and the second uses a scaled version of the Marmousi velocity model, dominated by reflections. Both examples are initially started from a semi-infinite half-space with incorrect background velocity. We find that the trans-D tree based approach together with parallel tempering for navigating rugged likelihood (i.e. misfit) topography provides a promising, easily generalized method for solving large-scale geophysical inverse problems which are difficult to optimize, but where the true model contains a hierarchy of features at multiple scales.

Strategies to Enhance the Model Update in Regions of Weak Sensitivities for Use in Full Waveform Inversion

NASA Astrophysics Data System (ADS)

Nuber, André; Manukyan, Edgar; Maurer, Hansruedi

2014-05-01

Conventional methods of interpreting seismic data rely on filtering and processing limited portions of the recorded wavefield. Typically, either reflections, refractions or surface waves are considered in isolation. Particularly in near-surface engineering and environmental investigations (depths less than, say 100 m), these wave types often overlap in time and are difficult to separate. Full waveform inversion is a technique that seeks to exploit and interpret the full information content of the seismic records without the need for separating events first; it yields models of the subsurface at sub-wavelength resolution. We use a finite element modelling code to solve the 2D elastic isotropic wave equation in the frequency domain. This code is part of a Gauss-Newton inversion scheme which we employ to invert for the P- and S-wave velocities as well as for density in the subsurface. For shallow surface data the use of an elastic forward solver is essential because surface waves often dominate the seismograms. This leads to high sensitivities (partial derivatives contained in the Jacobian matrix of the Gauss-Newton inversion scheme) and thus large model updates close to the surface. Reflections from deeper structures may also include useful information, but the large sensitivities of the surface waves often preclude this information from being fully exploited. We have developed two methods that balance the sensitivity distributions and thus may help resolve the deeper structures. The first method includes equilibrating the columns of the Jacobian matrix prior to every inversion step by multiplying them with individual scaling factors. This is expected to also balance the model updates throughout the entire subsurface model. It can be shown that this procedure is mathematically equivalent to balancing the regularization weights of the individual model parameters. A proper choice of the scaling factors required to balance the Jacobian matrix is critical. We decided to normalise the columns of the Jacobian based on their absolute column sum, but defining an upper threshold for the scaling factors. This avoids particularly small and therefore insignificant sensitivities being over-boosted, which would produce unstable results. The second method proposed includes adjusting the inversion cell size with depth. Multiple cells of the forward modelling grid are merged to form larger inversion cells (typical ratios between forward and inversion cells are in the order of 1:100). The irregular inversion grid is adapted to the expected resolution power of full waveform inversion. Besides stabilizing the inversion, this approach also reduces the number of model parameters to be recovered. Consequently, the computational costs and the memory consumption are reduced significantly. This is particularly critical when Gauss-Newton type inversion schemes are employed. Extensive tests with synthetic data demonstrated that both methods stabilise the inversion and improve the inversion results. The two methods have some redundancy, which can be seen when both are applied simultaneously, that is, when scaling of the Jacobian matrix is applied to an irregular inversion grid. The calculated scaling factors are quite balanced and span a much smaller range than in the case of a regular inversion grid.

Full Waveform Inversion for Seismic Velocity And Anelastic Losses in Heterogeneous Structures

DOE Office of Scientific and Technical Information (OSTI.GOV)

Askan, A.; /Carnegie Mellon U.; Akcelik, V.

2009-04-30

We present a least-squares optimization method for solving the nonlinear full waveform inverse problem of determining the crustal velocity and intrinsic attenuation properties of sedimentary valleys in earthquake-prone regions. Given a known earthquake source and a set of seismograms generated by the source, the inverse problem is to reconstruct the anelastic properties of a heterogeneous medium with possibly discontinuous wave velocities. The inverse problem is formulated as a constrained optimization problem, where the constraints are the partial and ordinary differential equations governing the anelastic wave propagation from the source to the receivers in the time domain. This leads to amore » variational formulation in terms of the material model plus the state variables and their adjoints. We employ a wave propagation model in which the intrinsic energy-dissipating nature of the soil medium is modeled by a set of standard linear solids. The least-squares optimization approach to inverse wave propagation presents the well-known difficulties of ill posedness and multiple minima. To overcome ill posedness, we include a total variation regularization functional in the objective function, which annihilates highly oscillatory material property components while preserving discontinuities in the medium. To treat multiple minima, we use a multilevel algorithm that solves a sequence of subproblems on increasingly finer grids with increasingly higher frequency source components to remain within the basin of attraction of the global minimum. We illustrate the methodology with high-resolution inversions for two-dimensional sedimentary models of the San Fernando Valley, under SH-wave excitation. We perform inversions for both the seismic velocity and the intrinsic attenuation using synthetic waveforms at the observer locations as pseudoobserved data.« less

Magnetotelluric inversion via reverse time migration algorithm of seismic data

DOE Office of Scientific and Technical Information (OSTI.GOV)

Ha, Taeyoung; Shin, Changsoo

2007-07-01

We propose a new algorithm for two-dimensional magnetotelluric (MT) inversion. Our algorithm is an MT inversion based on the steepest descent method, borrowed from the backpropagation technique of seismic inversion or reverse time migration, introduced in the middle 1980s by Lailly and Tarantola. The steepest descent direction can be calculated efficiently by using the symmetry of numerical Green's function derived from a mixed finite element method proposed by Nedelec for Maxwell's equation, without calculating the Jacobian matrix explicitly. We construct three different objective functions by taking the logarithm of the complex apparent resistivity as introduced in the recent waveform inversionmore » algorithm by Shin and Min. These objective functions can be naturally separated into amplitude inversion, phase inversion and simultaneous inversion. We demonstrate our algorithm by showing three inversion results for synthetic data.« less

Fully probabilistic seismic source inversion - Part 2: Modelling errors and station covariances

NASA Astrophysics Data System (ADS)

Stähler, Simon C.; Sigloch, Karin

2016-11-01

Seismic source inversion, a central task in seismology, is concerned with the estimation of earthquake source parameters and their uncertainties. Estimating uncertainties is particularly challenging because source inversion is a non-linear problem. In a companion paper, Stähler and Sigloch (2014) developed a method of fully Bayesian inference for source parameters, based on measurements of waveform cross-correlation between broadband, teleseismic body-wave observations and their modelled counterparts. This approach yields not only depth and moment tensor estimates but also source time functions. A prerequisite for Bayesian inference is the proper characterisation of the noise afflicting the measurements, a problem we address here. We show that, for realistic broadband body-wave seismograms, the systematic error due to an incomplete physical model affects waveform misfits more strongly than random, ambient background noise. In this situation, the waveform cross-correlation coefficient CC, or rather its decorrelation D = 1 - CC, performs more robustly as a misfit criterion than ℓp norms, more commonly used as sample-by-sample measures of misfit based on distances between individual time samples. From a set of over 900 user-supervised, deterministic earthquake source solutions treated as a quality-controlled reference, we derive the noise distribution on signal decorrelation D = 1 - CC of the broadband seismogram fits between observed and modelled waveforms. The noise on D is found to approximately follow a log-normal distribution, a fortunate fact that readily accommodates the formulation of an empirical likelihood function for D for our multivariate problem. The first and second moments of this multivariate distribution are shown to depend mostly on the signal-to-noise ratio (SNR) of the CC measurements and on the back-azimuthal distances of seismic stations. By identifying and quantifying this likelihood function, we make D and thus waveform cross-correlation measurements usable for fully probabilistic sampling strategies, in source inversion and related applications such as seismic tomography.

A Joint Method of Envelope Inversion Combined with Hybrid-domain Full Waveform Inversion

NASA Astrophysics Data System (ADS)

CUI, C.; Hou, W.

2017-12-01

Full waveform inversion (FWI) aims to construct high-precision subsurface models by fully using the information in seismic records, including amplitude, travel time, phase and so on. However, high non-linearity and the absence of low frequency information in seismic data lead to the well-known cycle skipping problem and make inversion easily fall into local minima. In addition, those 3D inversion methods that are based on acoustic approximation ignore the elastic effects in real seismic field, and make inversion harder. As a result, the accuracy of final inversion results highly relies on the quality of initial model. In order to improve stability and quality of inversion results, multi-scale inversion that reconstructs subsurface model from low to high frequency are applied. But, the absence of very low frequencies (< 3Hz) in field data is still bottleneck in the FWI. By extracting ultra low-frequency data from field data, envelope inversion is able to recover low wavenumber model with a demodulation operator (envelope operator), though the low frequency data does not really exist in field data. To improve the efficiency and viability of the inversion, in this study, we proposed a joint method of envelope inversion combined with hybrid-domain FWI. First, we developed 3D elastic envelope inversion, and the misfit function and the corresponding gradient operator were derived. Then we performed hybrid-domain FWI with envelope inversion result as initial model which provides low wavenumber component of model. Here, forward modeling is implemented in the time domain and inversion in the frequency domain. To accelerate the inversion, we adopt CPU/GPU heterogeneous computing techniques. There were two levels of parallelism. In the first level, the inversion tasks are decomposed and assigned to each computation node by shot number. In the second level, GPU multithreaded programming is used for the computation tasks in each node, including forward modeling, envelope extraction, DFT (discrete Fourier transform) calculation and gradients calculation. Numerical tests demonstrated that the combined envelope inversion + hybrid-domain FWI could obtain much faithful and accurate result than conventional hybrid-domain FWI. The CPU/GPU heterogeneous parallel computation could improve the performance speed.

Investigating source processes of isotropic events

NASA Astrophysics Data System (ADS)

Chiang, Andrea

This dissertation demonstrates the utility of the complete waveform regional moment tensor inversion for nuclear event discrimination. I explore the source processes and associated uncertainties for explosions and earthquakes under the effects of limited station coverage, compound seismic sources, assumptions in velocity models and the corresponding Green's functions, and the effects of shallow source depth and free-surface conditions. The motivation to develop better techniques to obtain reliable source mechanism and assess uncertainties is not limited to nuclear monitoring, but they also provide quantitative information about the characteristics of seismic hazards, local and regional tectonics and in-situ stress fields of the region . This dissertation begins with the analysis of three sparsely recorded events: the 14 September 1988 US-Soviet Joint Verification Experiment (JVE) nuclear test at the Semipalatinsk test site in Eastern Kazakhstan, and two nuclear explosions at the Chinese Lop Nor test site. We utilize a regional distance seismic waveform method fitting long-period, complete, three-component waveforms jointly with first-motion observations from regional stations and teleseismic arrays. The combination of long period waveforms and first motion observations provides unique discrimination of these sparsely recorded events in the context of the Hudson et al. (1989) source-type diagram. We examine the effects of the free surface on the moment tensor via synthetic testing, and apply the moment tensor based discrimination method to well-recorded chemical explosions. These shallow chemical explosions represent rather severe source-station geometry in terms of the vanishing traction issues. We show that the combined waveform and first motion method enables the unique discrimination of these events, even though the data include unmodeled single force components resulting from the collapse and blowout of the quarry face immediately following the initial explosion. In contrast, recovering the announced explosive yield using seismic moment estimates from moment tensor inversion remains challenging but we can begin to put error bounds on our moment estimates using the NSS technique. The estimation of seismic source parameters is dependent upon having a well-calibrated velocity model to compute the Green's functions for the inverse problem. Ideally, seismic velocity models are calibrated through broadband waveform modeling, however in regions of low seismicity velocity models derived from body or surface wave tomography may be employed. Whether a velocity model is 1D or 3D, or based on broadband seismic waveform modeling or the various tomographic techniques, the uncertainty in the velocity model can be the greatest source of error in moment tensor inversion. These errors have not been fully investigated for the nuclear discrimination problem. To study the effects of unmodeled structures on the moment tensor inversion, we set up a synthetic experiment where we produce synthetic seismograms for a 3D model (Moschetti et al., 2010) and invert these data using Green's functions computed with a 1D velocity mode (Song et al., 1996) to evaluate the recoverability of input solutions, paying particular attention to biases in the isotropic component. The synthetic experiment results indicate that the 1D model assumption is valid for moment tensor inversions at periods as short as 10 seconds for the 1D western U.S. model (Song et al., 1996). The correct earthquake mechanisms and source depth are recovered with statistically insignificant isotropic components as determined by the F-test. Shallow explosions are biased by the theoretical ISO-CLVD tradeoff but the tectonic release component remains low, and the tradeoff can be eliminated with constraints from P wave first motion. Path-calibration to the 1D model can reduce non-double-couple components in earthquakes, non-isotropic components in explosions and composite sources and improve the fit to the data. When we apply the 3D model to real data, at long periods (20-50 seconds), we see good agreement in the solutions between the 1D and 3D models and slight improvement in waveform fits when using the 3D velocity model Green's functions. (Abstract shortened by ProQuest.).

Mach-Zehnder interferometry method for acoustic shock wave measurements in air and broadband calibration of microphones.

PubMed

Yuldashev, Petr; Karzova, Maria; Khokhlova, Vera; Ollivier, Sébastien; Blanc-Benon, Philippe

2015-06-01

A Mach-Zehnder interferometer is used to measure spherically diverging N-waves in homogeneous air. An electrical spark source is used to generate high-amplitude (1800 Pa at 15 cm from the source) and short duration (50 μs) N-waves. Pressure waveforms are reconstructed from optical phase signals using an Abel-type inversion. It is shown that the interferometric method allows one to reach 0.4 μs of time resolution, which is 6 times better than the time resolution of a 1/8-in. condenser microphone (2.5 μs). Numerical modeling is used to validate the waveform reconstruction method. The waveform reconstruction method provides an error of less than 2% with respect to amplitude in the given experimental conditions. Optical measurement is used as a reference to calibrate a 1/8-in. condenser microphone. The frequency response function of the microphone is obtained by comparing the spectra of the waveforms resulting from optical and acoustical measurements. The optically measured pressure waveforms filtered with the microphone frequency response are in good agreement with the microphone output voltage. Therefore, an optical measurement method based on the Mach-Zehnder interferometer is a reliable tool to accurately characterize evolution of weak shock waves in air and to calibrate broadband acoustical microphones.

Optimization of contrast resolution by genetic algorithm in ultrasound tissue harmonic imaging.

PubMed

Ménigot, Sébastien; Girault, Jean-Marc

2016-09-01

The development of ultrasound imaging techniques such as pulse inversion has improved tissue harmonic imaging. Nevertheless, no recommendation has been made to date for the design of the waveform transmitted through the medium being explored. Our aim was therefore to find automatically the optimal "imaging" wave which maximized the contrast resolution without a priori information. To overcome assumption regarding the waveform, a genetic algorithm investigated the medium thanks to the transmission of stochastic "explorer" waves. Moreover, these stochastic signals could be constrained by the type of generator available (bipolar or arbitrary). To implement it, we changed the current pulse inversion imaging system by including feedback. Thus the method optimized the contrast resolution by adaptively selecting the samples of the excitation. In simulation, we benchmarked the contrast effectiveness of the best found transmitted stochastic commands and the usual fixed-frequency command. The optimization method converged quickly after around 300 iterations in the same optimal area. These results were confirmed experimentally. In the experimental case, the contrast resolution measured on a radiofrequency line could be improved by 6% with a bipolar generator and it could still increase by 15% with an arbitrary waveform generator. Copyright © 2016 Elsevier B.V. All rights reserved.

Seismic moment tensor inversion using 3D velocity model and its application to the 2013 Lushan earthquake sequence

NASA Astrophysics Data System (ADS)

Zhu, Lupei; Zhou, Xiaofeng

2016-10-01

Source inversion of small-magnitude events such as aftershocks or mine collapses requires use of relatively high frequency seismic waveforms which are strongly affected by small-scale heterogeneities in the crust. In this study, we developed a new inversion method called gCAP3D for determining general moment tensor of a seismic source using Green's functions of 3D models. It inherits the advantageous features of the ;Cut-and-Paste; (CAP) method to break a full seismogram into the Pnl and surface-wave segments and to allow time shift between observed and predicted waveforms. It uses grid search for 5 source parameters (relative strengths of the isotropic and compensated-linear-vector-dipole components and the strike, dip, and rake of the double-couple component) that minimize the waveform misfit. The scalar moment is estimated using the ratio of L2 norms of the data and synthetics. Focal depth can also be determined by repeating the inversion at different depths. We applied gCAP3D to the 2013 Ms 7.0 Lushan earthquake and its aftershocks using a 3D crustal-upper mantle velocity model derived from ambient noise tomography in the region. We first relocated the events using the double-difference method. We then used the finite-differences method and reciprocity principle to calculate Green's functions of the 3D model for 20 permanent broadband seismic stations within 200 km from the source region. We obtained moment tensors of the mainshock and 74 aftershocks ranging from Mw 5.2 to 3.4. The results show that the Lushan earthquake is a reverse faulting at a depth of 13-15 km on a plane dipping 40-47° to N46° W. Most of the aftershocks occurred off the main rupture plane and have similar focal mechanisms to the mainshock's, except in the proximity of the mainshock where the aftershocks' focal mechanisms display some variations.

Surface Wave Mode Conversion due to Lateral Heterogeneity and its Impact on Waveform Inversions

NASA Astrophysics Data System (ADS)

Datta, A.; Priestley, K. F.; Chapman, C. H.; Roecker, S. W.

2016-12-01

Surface wave tomography based on great circle ray theory has certain limitations which become increasingly significant with increasing frequency. One such limitation is the assumption of different surface wave modes propagating independently from source to receiver, valid only in case of smoothly varying media. In the real Earth, strong lateral gradients can cause significant interconversion among modes, thus potentially wreaking havoc with ray theory based tomographic inversions that make use of multimode information. The issue of mode coupling (with either normal modes or surface wave modes) for accurate modelling and inversion of body wave data has received significant attention in the seismological literature, but its impact on inversion of surface waveforms themselves remains much less understood.We present an empirical study with synthetic data, to investigate this problem with a two-fold approach. In the first part, 2D forward modelling using a new finite difference method that allows modelling a single mode at a time, is used to build a general picture of energy transfer among modes as a function of size, strength and sharpness of lateral heterogeneities. In the second part, we use the example of a multimode waveform inversion technique based on the Cara and Leveque (1987) approach of secondary observables, to invert our synthetic data and assess how mode conversion can affect the process of imaging the Earth. We pay special attention to ensuring that any biases or artefacts in the resulting inversions can be unambiguously attributed to mode conversion effects. This study helps pave the way towards the next generation of (non-numerical) surface wave tomography techniques geared to exploit higher frequencies and mode numbers than are typically used today.

Bessel smoothing filter for spectral-element mesh

NASA Astrophysics Data System (ADS)

Trinh, P. T.; Brossier, R.; Métivier, L.; Virieux, J.; Wellington, P.

2017-06-01

Smoothing filters are extremely important tools in seismic imaging and inversion, such as for traveltime tomography, migration and waveform inversion. For efficiency, and as they can be used a number of times during inversion, it is important that these filters can easily incorporate prior information on the geological structure of the investigated medium, through variable coherent lengths and orientation. In this study, we promote the use of the Bessel filter to achieve these purposes. Instead of considering the direct application of the filter, we demonstrate that we can rely on the equation associated with its inverse filter, which amounts to the solution of an elliptic partial differential equation. This enhances the efficiency of the filter application, and also its flexibility. We apply this strategy within a spectral-element-based elastic full waveform inversion framework. Taking advantage of this formulation, we apply the Bessel filter by solving the associated partial differential equation directly on the spectral-element mesh through the standard weak formulation. This avoids cumbersome projection operators between the spectral-element mesh and a regular Cartesian grid, or expensive explicit windowed convolution on the finite-element mesh, which is often used for applying smoothing operators. The associated linear system is solved efficiently through a parallel conjugate gradient algorithm, in which the matrix vector product is factorized and highly optimized with vectorized computation. Significant scaling behaviour is obtained when comparing this strategy with the explicit convolution method. The theoretical numerical complexity of this approach increases linearly with the coherent length, whereas a sublinear relationship is observed practically. Numerical illustrations are provided here for schematic examples, and for a more realistic elastic full waveform inversion gradient smoothing on the SEAM II benchmark model. These examples illustrate well the efficiency and flexibility of the approach proposed.

Three-Dimensional Anisotropic Acoustic and Elastic Full-Waveform Seismic Inversion

NASA Astrophysics Data System (ADS)

Warner, M.; Morgan, J. V.

2013-12-01

Three-dimensional full-waveform inversion is a high-resolution, high-fidelity, quantitative, seismic imaging technique that has advanced rapidly within the oil and gas industry. The method involves the iterative improvement of a starting model using a series of local linearized updates to solve the full non-linear inversion problem. During the inversion, forward modeling employs the full two-way three-dimensional heterogeneous anisotropic acoustic or elastic wave equation to predict the observed raw field data, wiggle-for-wiggle, trace-by-trace. The method is computationally demanding; it is highly parallelized, and runs on large multi-core multi-node clusters. Here, we demonstrate what can be achieved by applying this newly practical technique to several high-density 3D seismic datasets that were acquired to image four contrasting sedimentary targets: a gas cloud above an oil reservoir, a radially faulted dome, buried fluvial channels, and collapse structures overlying an evaporate sequence. We show that the resulting anisotropic p-wave velocity models match in situ measurements in deep boreholes, reproduce detailed structure observed independently on high-resolution seismic reflection sections, accurately predict the raw seismic data, simplify and sharpen reverse-time-migrated reflection images of deeper horizons, and flatten Kirchhoff-migrated common-image gathers. We also show that full-elastic 3D full-waveform inversion of pure pressure data can generate a reasonable shear-wave velocity model for one of these datasets. For two of the four datasets, the inclusion of significant transversely isotropic anisotropy with a vertical axis of symmetry was necessary in order to fit the kinematics of the field data properly. For the faulted dome, the full-waveform-inversion p-wave velocity model recovers the detailed structure of every fault that can be seen on coincident seismic reflection data. Some of the individual faults represent high-velocity zones, some represent low-velocity zones, some have more-complex internal structure, and some are visible merely as offsets between two regions with contrasting velocity. Although this has not yet been demonstrated quantitatively for this dataset, it seems likely that at least some of this fine structure in the recovered velocity model is related to the detailed lithology, strain history and fluid properties within the individual faults. We have here applied this technique to seismic data that were acquired by the extractive industries, however this inversion scheme is immediately scalable and applicable to a much wider range of problems given sufficient quality and density of observed data. Potential targets range from shallow magma chambers beneath active volcanoes, through whole-crustal sections across plate boundaries, to regional and whole-Earth models.

Inferring global upper-mantle shear attenuation structure by waveform tomography using the spectral element method

NASA Astrophysics Data System (ADS)

Karaoǧlu, Haydar; Romanowicz, Barbara

2018-06-01

We present a global upper-mantle shear wave attenuation model that is built through a hybrid full-waveform inversion algorithm applied to long-period waveforms, using the spectral element method for wavefield computations. Our inversion strategy is based on an iterative approach that involves the inversion for successive updates in the attenuation parameter (δ Q^{-1}_μ) and elastic parameters (isotropic velocity VS, and radial anisotropy parameter ξ) through a Gauss-Newton-type optimization scheme that employs envelope- and waveform-type misfit functionals for the two steps, respectively. We also include source and receiver terms in the inversion steps for attenuation structure. We conducted a total of eight iterations (six for attenuation and two for elastic structure), and one inversion for updates to source parameters. The starting model included the elastic part of the relatively high-resolution 3-D whole mantle seismic velocity model, SEMUCB-WM1, which served to account for elastic focusing effects. The data set is a subset of the three-component surface waveform data set, filtered between 400 and 60 s, that contributed to the construction of the whole-mantle tomographic model SEMUCB-WM1. We applied strict selection criteria to this data set for the attenuation iteration steps, and investigated the effect of attenuation crustal structure on the retrieved mantle attenuation structure. While a constant 1-D Qμ model with a constant value of 165 throughout the upper mantle was used as starting model for attenuation inversion, we were able to recover, in depth extent and strength, the high-attenuation zone present in the depth range 80-200 km. The final 3-D model, SEMUCB-UMQ, shows strong correlation with tectonic features down to 200-250 km depth, with low attenuation beneath the cratons, stable parts of continents and regions of old oceanic crust, and high attenuation along mid-ocean ridges and backarcs. Below 250 km, we observe strong attenuation in the southwestern Pacific and eastern Africa, while low attenuation zones fade beneath most of the cratons. The strong negative correlation of Q^{-1}_μ and VS anomalies at shallow upper-mantle depths points to a common dominant origin for the two, likely due to variations in thermal structure. A comparison with two other global upper-mantle attenuation models shows promising consistency. As we updated the elastic 3-D model in alternate iterations, we found that the VS part of the model was stable, while the ξ structure evolution was more pronounced, indicating that it may be important to include 3-D attenuation effects when inverting for ξ, possibly due to the influence of dispersion corrections on this less well-constrained parameter.

Subsurface Void Characterization with 3-D Time Domain Full Waveform Tomography.

NASA Astrophysics Data System (ADS)

Nguyen, T. D.

2017-12-01

A new three dimensional full waveform inversion (3-D FWI) method is presented for subsurface site characterization at engineering scales (less than 30 m in depth). The method is based on a solution of 3-D elastic wave equations for forward modeling, and a cross-adjoint gradient approach for model updating. The staggered-grid finite-difference technique is used to solve the wave equations, together with implementation of the perfectly matched layer condition for boundary truncation. The gradient is calculated from the forward and backward wavefields. Reversed-in-time displacement residuals are induced as multiple sources at all receiver locations for the backward wavefield. The capability of the presented FWI method is tested on both synthetic and field experimental datasets. The test configuration uses 96 receivers and 117 shots at equal spacing (Fig 1). The inversion results from synthetic data show the ability of characterizing variable low- and high-velocity layers with embedded void (Figs 2-3). The synthetic study shows good potential for detection of voids and abnormalities in the field.

On decomposing stimulus and response waveforms in event-related potentials recordings.

PubMed

Yin, Gang; Zhang, Jun

2011-06-01

Event-related potentials (ERPs) reflect the brain activities related to specific behavioral events, and are obtained by averaging across many trial repetitions with individual trials aligned to the onset of a specific event, e.g., the onset of stimulus (s-aligned) or the onset of the behavioral response (r-aligned). However, the s-aligned and r-aligned ERP waveforms do not purely reflect, respectively, underlying stimulus (S-) or response (R-) component waveform, due to their cross-contaminations in the recorded ERP waveforms. Zhang [J. Neurosci. Methods, 80, pp. 49-63, 1998] proposed an algorithm to recover the pure S-component waveform and the pure R-component waveform from the s-aligned and r-aligned ERP average waveforms-however, due to the nature of this inverse problem, a direct solution is sensitive to noise that disproportionally affects low-frequency components, hindering the practical implementation of this algorithm. Here, we apply the Wiener deconvolution technique to deal with noise in input data, and investigate a Tikhonov regularization approach to obtain a stable solution that is robust against variances in the sampling of reaction-time distribution (when number of trials is low). Our method is demonstrated using data from a Go/NoGo experiment about image classification and recognition.

Advantages of the full-waveform inversion: real data example from the Polish Basin

NASA Astrophysics Data System (ADS)

Malinowski, M.; Operto, S.

2006-12-01

Modern acquisition techniques allow us to gather high-density seismic data even in case of crustal-scale investigations. In combination with increasing availability of computational resources (eg. PC clusters), this allow us to image the Earth's structure on much finer scale than offered by ray-theory based methods (like travel time tomography) by applying the full waveform inversion/tomography method (FWT). Recently, the FWT method was for the first time successfully applied to the real wide-aperture data: 100-km long OBS profile (Operto et al. 2006) and a 15-km long land profile (Operto et al. 2004, Ravaut et al., 2004). We present the results of the application of the FWT method to the GRUNDY 2003 experiment data, which is standing in between the scale of the mentioned datasets. This project was targeted at recognition of the pre-Zechstein strata within the Polish Basin. For a successful investigations relatively low-frequencies and wide-apertures were used. In the 50 by 10 km rectangular area ca. 800 RefTek 125 "Texan" stations with 4.5 Hz geophones were deployed, forming high-density central line (receiver spacing 100 m) and additional 4 parallel profiles. Previously the data were modelled using conventional methods: CDP processing and traveltime tomography. In order to utilise secondary arrivals, we used the frequency-domain FWT method of Pratt et al. (1998). The wide-aperture content of our data leads to a redundant wavenumber coverage which can be partially removed without loss of information by limiting the inversion to few frequencies only. The inversion proceeds by stepping from low to high frequencies and uses the model inferred for one component as the starting one for the next frequency. Before full waveform inversion, the data were preprocessed by QC editing, spectral deconvolution (whitening), bandpass filtering and muting in narrow window around the first arrival. Traveltime tomogram was choosen as the starting model for 2D waveform inversion. The model size was 50x10 km with 25 m FD grid step. We have selected 10 frequencies from 4 to 13 Hz. For each frequency 10 iterations were computed (on a Linux cluster). There is a clear improvement in resolution of the obtained tomographic images by exploiting the full wavefield. The model allows to predict also fairly well the observed seismograms and is consistent with both the geological horizons projected from industrial reflection profiles as well as check-shot velocity log. Benefits of FWT in application to our data seems to be clear: in one step, without the need for performing the forward raytracing modelling, we gained both the quasi-structural image (perturbational model) and the detailed velocity model. In this way we fully exploited the broad range of recorded offsets and reflection angles from pre- to postcritical ones for a successful imaging beneath the Zechstein strata.

Acceleration for 2D time-domain elastic full waveform inversion using a single GPU card

NASA Astrophysics Data System (ADS)

Jiang, Jinpeng; Zhu, Peimin

2018-05-01

Full waveform inversion (FWI) is a challenging procedure due to the high computational cost related to the modeling, especially for the elastic case. The graphics processing unit (GPU) has become a popular device for the high-performance computing (HPC). To reduce the long computation time, we design and implement the GPU-based 2D elastic FWI (EFWI) in time domain using a single GPU card. We parallelize the forward modeling and gradient calculations using the CUDA programming language. To overcome the limitation of relatively small global memory on GPU, the boundary saving strategy is exploited to reconstruct the forward wavefield. Moreover, the L-BFGS optimization method used in the inversion increases the convergence of the misfit function. A multiscale inversion strategy is performed in the workflow to obtain the accurate inversion results. In our tests, the GPU-based implementations using a single GPU device achieve >15 times speedup in forward modeling, and about 12 times speedup in gradient calculation, compared with the eight-core CPU implementations optimized by OpenMP. The test results from the GPU implementations are verified to have enough accuracy by comparing the results obtained from the CPU implementations.

Towards Full-Waveform Ambient Noise Inversion

NASA Astrophysics Data System (ADS)

Sager, K.; Ermert, L. A.; Boehm, C.; Fichtner, A.

2016-12-01

Noise tomography usually works under the assumption that the inter-station ambient noise correlation is equal to a scaled version of the Green function between the two receivers. This assumption, however, is only met under specific conditions, e.g. wavefield diffusivity and equipartitioning, or the isotropic distribution of both mono- and dipolar uncorrelated noise sources. These assumptions are typically not satisfied in the Earth. This inconsistency inhibits the exploitation of the full waveform information contained in noise correlations in order to constrain Earth structure and noise generation. To overcome this limitation, we attempt to develop a method that consistently accounts for the distribution of noise sources, 3D heterogeneous Earth structure and the full seismic wave propagation physics. This is intended to improve the resolution of tomographic images, to refine noise source location, and thereby to contribute to a better understanding of noise generation. We introduce an operator-based formulation for the computation of correlation functions and apply the continuous adjoint method that allows us to compute first and second derivatives of misfit functionals with respect to source distribution and Earth structure efficiently. Based on these developments we design an inversion scheme using a 2D finite-difference code. To enable a joint inversion for noise sources and Earth structure, we investigate the following aspects: The capability of different misfit functionals to image wave speed anomalies and source distribution. Possible source-structure trade-offs, especially to what extent unresolvable structure can be mapped into the inverted noise source distribution and vice versa. In anticipation of real-data applications, we present an extension of the open-source waveform modelling and inversion package Salvus, which allows us to compute correlation functions in 3D media with heterogeneous noise sources at the surface.

Viscoelastic property identification from waveform reconstruction

NASA Astrophysics Data System (ADS)

Leymarie, N.; Aristégui, C.; Audoin, B.; Baste, S.

2002-05-01

An inverse method is proposed for the determination of the viscoelastic properties of material plates from the plane-wave transmitted acoustic field. Innovations lie in a two-step inversion scheme based on the well-known maximum-likelihood principle with an analytic signal formulation. In addition, establishing the analytical formulations of the plate transmission coefficient we implement an efficient and slightly noise-sensitive process suited to both very thin plates and strongly dispersive media.

Validation of the inverse pulse wave transit time series as surrogate of systolic blood pressure in MVAR modeling.

PubMed

Giassi, Pedro; Okida, Sergio; Oliveira, Maurício G; Moraes, Raimes

2013-11-01

Short-term cardiovascular regulation mediated by the sympathetic and parasympathetic branches of the autonomic nervous system has been investigated by multivariate autoregressive (MVAR) modeling, providing insightful analysis. MVAR models employ, as inputs, heart rate (HR), systolic blood pressure (SBP) and respiratory waveforms. ECG (from which HR series is obtained) and respiratory flow waveform (RFW) can be easily sampled from the patients. Nevertheless, the available methods for acquisition of beat-to-beat SBP measurements during exams hamper the wider use of MVAR models in clinical research. Recent studies show an inverse correlation between pulse wave transit time (PWTT) series and SBP fluctuations. PWTT is the time interval between the ECG R-wave peak and photoplethysmography waveform (PPG) base point within the same cardiac cycle. This study investigates the feasibility of using inverse PWTT (IPWTT) series as an alternative input to SBP for MVAR modeling of the cardiovascular regulation. For that, HR, RFW, and IPWTT series acquired from volunteers during postural changes and autonomic blockade were used as input of MVAR models. Obtained results show that IPWTT series can be used as input of MVAR models, replacing SBP measurements in order to overcome practical difficulties related to the continuous sampling of the SBP during clinical exams.

Tsunami waveform inversion of the 2007 Bengkulu, southern Sumatra, earthquake

NASA Astrophysics Data System (ADS)

Fujii, Y.; Satake, K.

2008-09-01

We performed tsunami waveform inversions for the Bengkulu, southern Sumatra, earthquake on September 12, 2007 (Mw 8.4 by USGS). The tsunami was recorded at many tide gauge stations around the Indian Ocean and by a DART system in the deep ocean. The observed tsunami records indicate that the amplitudes were less than several tens of centimeters at most stations, around 1 m at Padang, the nearest station to the source, and a few centimeters at the DART station. For the tsunami waveform inversions, we adopted 20-, 15- and 10-subfault models. The tsunami waveforms computed from the estimated slip distributions explain the observed waveforms at most stations, regardless of the subfault model. We found that large slips were consistently estimated at the deeper part (>24 km) of the fault plane, located more than 100 km from the trench axis. The largest slips of 6-9 m were located about 100-200 km northwest of the epicenter. The deep slips may have contributed to the relatively small tsunami for its earthquake size. The total seismic moment is calculated as 4.7 × 1021 N m (Mw = 8.4) for the 10-subfault model, our preferred model from a comparison of tsunami waveforms at Cocos and the DART station.

Rigorous Approach in Investigation of Seismic Structure and Source Characteristicsin Northeast Asia: Hierarchical and Trans-dimensional Bayesian Inversion

NASA Astrophysics Data System (ADS)

Mustac, M.; Kim, S.; Tkalcic, H.; Rhie, J.; Chen, Y.; Ford, S. R.; Sebastian, N.

2015-12-01

Conventional approaches to inverse problems suffer from non-linearity and non-uniqueness in estimations of seismic structures and source properties. Estimated results and associated uncertainties are often biased by applied regularizations and additional constraints, which are commonly introduced to solve such problems. Bayesian methods, however, provide statistically meaningful estimations of models and their uncertainties constrained by data information. In addition, hierarchical and trans-dimensional (trans-D) techniques are inherently implemented in the Bayesian framework to account for involved error statistics and model parameterizations, and, in turn, allow more rigorous estimations of the same. Here, we apply Bayesian methods throughout the entire inference process to estimate seismic structures and source properties in Northeast Asia including east China, the Korean peninsula, and the Japanese islands. Ambient noise analysis is first performed to obtain a base three-dimensional (3-D) heterogeneity model using continuous broadband waveforms from more than 300 stations. As for the tomography of surface wave group and phase velocities in the 5-70 s band, we adopt a hierarchical and trans-D Bayesian inversion method using Voronoi partition. The 3-D heterogeneity model is further improved by joint inversions of teleseismic receiver functions and dispersion data using a newly developed high-efficiency Bayesian technique. The obtained model is subsequently used to prepare 3-D structural Green's functions for the source characterization. A hierarchical Bayesian method for point source inversion using regional complete waveform data is applied to selected events from the region. The seismic structure and source characteristics with rigorously estimated uncertainties from the novel Bayesian methods provide enhanced monitoring and discrimination of seismic events in northeast Asia.

3-D imaging of large scale buried structure by 1-D inversion of very early time electromagnetic (VETEM) data

USGS Publications Warehouse

Aydmer, A.A.; Chew, W.C.; Cui, T.J.; Wright, D.L.; Smith, D.V.; Abraham, J.D.

2001-01-01

A simple and efficient method for large scale three-dimensional (3-D) subsurface imaging of inhomogeneous background is presented. One-dimensional (1-D) multifrequency distorted Born iterative method (DBIM) is employed in the inversion. Simulation results utilizing synthetic scattering data are given. Calibration of the very early time electromagnetic (VETEM) experimental waveforms is detailed along with major problems encountered in practice and their solutions. This discussion is followed by the results of a large scale application of the method to the experimental data provided by the VETEM system of the U.S. Geological Survey. The method is shown to have a computational complexity that is promising for on-site inversion.

Earthquake source tensor inversion with the gCAP method and 3D Green's functions

NASA Astrophysics Data System (ADS)

Zheng, J.; Ben-Zion, Y.; Zhu, L.; Ross, Z.

2013-12-01

We develop and apply a method to invert earthquake seismograms for source properties using a general tensor representation and 3D Green's functions. The method employs (i) a general representation of earthquake potency/moment tensors with double couple (DC), compensated linear vector dipole (CLVD), and isotropic (ISO) components, and (ii) a corresponding generalized CAP (gCap) scheme where the continuous wave trains are broken into Pnl and surface waves (Zhu & Ben-Zion, 2013). For comparison, we also use the waveform inversion method of Zheng & Chen (2012) and Ammon et al. (1998). Sets of 3D Green's functions are calculated on a grid of 1 km3 using the 3-D community velocity model CVM-4 (Kohler et al. 2003). A bootstrap technique is adopted to establish robustness of the inversion results using the gCap method (Ross & Ben-Zion, 2013). Synthetic tests with 1-D and 3-D waveform calculations show that the source tensor inversion procedure is reasonably reliable and robust. As initial application, the method is used to investigate source properties of the March 11, 2013, Mw=4.7 earthquake on the San Jacinto fault using recordings of ~45 stations up to ~0.2Hz. Both the best fitting and most probable solutions include ISO component of ~1% and CLVD component of ~0%. The obtained ISO component, while small, is found to be a non-negligible positive value that can have significant implications for the physics of the failure process. Work on using higher frequency data for this and other earthquakes is in progress.

Acoustic Full Waveform Inversion to Characterize Near-surface Chemical Explosions

NASA Astrophysics Data System (ADS)

Kim, K.; Rodgers, A. J.

2015-12-01

Recent high-quality, atmospheric overpressure data from chemical high-explosive experiments provide a unique opportunity to characterize near-surface explosions, specifically estimating yield and source time function. Typically, yield is estimated from measured signal features, such as peak pressure, impulse, duration and/or arrival time of acoustic signals. However, the application of full waveform inversion to acoustic signals for yield estimation has not been fully explored. In this study, we apply a full waveform inversion method to local overpressure data to extract accurate pressure-time histories of acoustics sources during chemical explosions. A robust and accurate inversion technique for acoustic source is investigated using numerical Green's functions that take into account atmospheric and topographic propagation effects. The inverted pressure-time history represents the pressure fluctuation at the source region associated with the explosion, and thus, provides a valuable information about acoustic source mechanisms and characteristics in greater detail. We compare acoustic source properties (i.e., peak overpressure, duration, and non-isotropic shape) of a series of explosions having different emplacement conditions and investigate the relationship of the acoustic sources to the yields of explosions. The time histories of acoustic sources may refine our knowledge of sound-generation mechanisms of shallow explosions, and thereby allow for accurate yield estimation based on acoustic measurements. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.

Interpretaion of synthetic seismic time-lapse monitoring data for Korea CCS project based on the acoustic-elastic coupled inversion

NASA Astrophysics Data System (ADS)

Oh, J.; Min, D.; Kim, W.; Huh, C.; Kang, S.

2012-12-01

Recently, the CCS (Carbon Capture and Storage) is one of the promising methods to reduce the CO2 emission. To evaluate the success of the CCS project, various geophysical monitoring techniques have been applied. Among them, the time-lapse seismic monitoring is one of the effective methods to investigate the migration of CO2 plume. To monitor the injected CO2 plume accurately, it is needed to interpret seismic monitoring data using not only the imaging technique but also the full waveform inversion, because subsurface material properties can be estimated through the inversion. However, previous works for interpreting seismic monitoring data are mainly based on the imaging technique. In this study, we perform the frequency-domain full waveform inversion for synthetic data obtained by the acoustic-elastic coupled modeling for the geological model made after Ulleung Basin, which is one of the CO2 storage prospects in Korea. We suppose the injection layer is located in fault-related anticlines in the Dolgorae Deformed Belt and, for more realistic situation, we contaminate the synthetic monitoring data with random noise and outliers. We perform the time-lapse full waveform inversion in two scenarios. One scenario is that the injected CO2 plume migrates within the injection layer and is stably captured. The other scenario is that the injected CO2 plume leaks through the weak part of the cap rock. Using the inverted P- and S-wave velocities and Poisson's ratio, we were able to detect the migration of the injected CO2 plume. Acknowledgment This work was financially supported by the Brain Korea 21 project of Energy Systems Engineering, the "Development of Technology for CO2 Marine Geological Storage" program funded by the Ministry of Land, Transport and Maritime Affairs (MLTM) of Korea and the Korea CCS R&D Center (KCRC) grant funded by the Korea government (Ministry of Education, Science and Technology) (No. 2012-0008926).

Advancements and challenges in crosshole GPR full-waveform inversion for hydrological applications

NASA Astrophysics Data System (ADS)

Klotzsche, A.; Van Der Kruk, J.; Vereecken, H.

2016-12-01

Crosshole ground penetrating radar (GPR) full-waveform inversion (FWI) demonstrated over the last decade a high potential to detect, map, and resolve decimeter-small-scale structures within aquifers. GPR FWI uses Maxwell's equations to find a model that fits the measurements with the entire measured waveform. One big advantage is that by applying one method, we can derive two soil properties: dielectric permittivity and electrical conductivity. Both parameters are sensitive to different soil properties such as soil water content and porosity, or, clay content. Hence, an improved characterization of the critical zone is possible. The application of the FWI to aquifers in Germany, Switzerland, Denmark, and USA showed for all sites improved and higher resolution images than standard ray-based methods and provided new insights in the aquifers' structures. Furthermore, small-scale high contrast layers caused by changes in porosity were characterize and enhanced our understanding of the electromagnetic wave propagation related to these features. However, to obtain reliable and accurate inversion results from experimental data and hence porosity estimates, many detailed steps in acquiring the data, pre-processing and inverting the data need to be carefully followed. Here, we provide an overview of recent developments and advancements of the 2D crosshole GPR FWI that provide improved inversion results for permittivity and electrical conductivity. In addition, we will provide guidelines and point out important challenges and pitfalls that can occur during the inversion of experimental data. We will illustrate the necessary steps that are required to achieve reliable FWI results, which are indicated by e.g. a good fit of the measured and modelled traces, and, absence of a remaining gradient for the final models. Important requirements for a successful application are an accurate time zero correction, good starting models for the FWI, and, a well-estimated source wavelet.

Characterization of moderate ash-and-gas explosions at Santiaguito volcano, Guatemala, from infrasound waveform inversion and thermal infrared measurements

NASA Astrophysics Data System (ADS)

Angelis, S. De; Lamb, O. D.; Lamur, A.; Hornby, A. J.; von Aulock, F. W.; Chigna, G.; Lavallée, Y.; Rietbrock, A.

2016-06-01

The rapid discharge of gas and rock fragments during volcanic eruptions generates acoustic infrasound. Here we present results from the inversion of infrasound signals associated with small and moderate gas-and-ash explosions at Santiaguito volcano, Guatemala, to retrieve the time history of mass eruption rate at the vent. Acoustic waveform inversion is complemented by analyses of thermal infrared imagery to constrain the volume and rise dynamics of the eruption plume. Finally, we combine results from the two methods in order to assess the bulk density of the erupted mixture, constrain the timing of the transition from a momentum-driven jet to a buoyant plume, and to evaluate the relative volume fractions of ash and gas during the initial thrust phase. Our results demonstrate that eruptive plumes associated with small-to-moderate size explosions at Santiaguito only carry minor fractions of ash, suggesting that these events may not involve extensive magma fragmentation in the conduit.

Characterization of moderate ash-and-gas explosions at Santiaguito volcano, Guatemala, from infrasound waveform inversion and thermal infrared measurements.

PubMed

Angelis, S De; Lamb, O D; Lamur, A; Hornby, A J; von Aulock, F W; Chigna, G; Lavallée, Y; Rietbrock, A

2016-06-28

The rapid discharge of gas and rock fragments during volcanic eruptions generates acoustic infrasound. Here we present results from the inversion of infrasound signals associated with small and moderate gas-and-ash explosions at Santiaguito volcano, Guatemala, to retrieve the time history of mass eruption rate at the vent. Acoustic waveform inversion is complemented by analyses of thermal infrared imagery to constrain the volume and rise dynamics of the eruption plume. Finally, we combine results from the two methods in order to assess the bulk density of the erupted mixture, constrain the timing of the transition from a momentum-driven jet to a buoyant plume, and to evaluate the relative volume fractions of ash and gas during the initial thrust phase. Our results demonstrate that eruptive plumes associated with small-to-moderate size explosions at Santiaguito only carry minor fractions of ash, suggesting that these events may not involve extensive magma fragmentation in the conduit.

Resolving the Detailed Spatiotemporal Slip Evolution of Deep Tremor in Western Japan

NASA Astrophysics Data System (ADS)

Ohta, Kazuaki; Ide, Satoshi

2017-12-01

We study the detailed spatiotemporal behavior of deep tremor in western Japan through the development and application of a new slip inversion method. Although many studies now recognize tremor as shear slip along the plate interface manifested in low-frequency earthquake (LFE) swarms, a conventional slip inversion analysis is not available for tremor due to insufficient knowledge of source locations and Green's functions. Here we introduce synthetic template waveforms, which are typical tremor waveforms obtained by stacking LFE seismograms at arranged points along the plate interface. Using these synthetic template waveforms as substitutes for Green's functions, we invert the continuous tremor waveforms using an iterative deconvolution approach with Bayesian constraints. We apply this method to two tremor burst episodes in western and central Shikoku, Japan. The estimated slip distribution from a 12 day tremor burst episode in western Shikoku is heterogeneous, with several patchy areas of slip along the plate interface where rapid moment releases with durations of <100 s regularly occur. We attribute these heterogeneous spatiotemporal slip patterns to heterogeneous material properties along the plate interface. For central Shikoku, where we focus on a tremor burst episode that occurred coincidentally with a very low frequency earthquake (VLF), we observe that the source size of the VLF is much larger than that estimated from tremor activity in western Shikoku. These differences in the size of the slip region may dictate the visibility of VLF signals in observed seismograms, which has implications for the mechanics of slow earthquakes and subduction zone processes.

Resolving the detailed spatiotemporal slip evolution of deep tremor in western Japan

NASA Astrophysics Data System (ADS)

Ohta, K.; Ide, S.

2017-12-01

A quantitative evaluation of the slip evolution of tremor is essential to understand the generation mechanism of slow earthquakes. The recent studies have revealed the most part of tremor signals can be expressed as the superposition of low frequency earthquakes (LFE). However, it is still challenging to explain the entire waveforms of tremor, because a conventional slip inversion analysis is not available for tremor due to insufficient knowledge of source locations and Green's functions. Here we investigate the detailed spatiotemporal behavior of deep tremor in western Japan through the development and application of a new slip inversion method. We introduce synthetic template waveforms, which are typical tremor waveforms obtained by stacking LFE seismograms at arranged points along the plate interface. Using these synthetic template waveforms as substitutes for Green's functions, we invert the continuous tremor waveforms using an iterative deconvolution approach with Bayesian constraints. We apply this method to two tremor burst episodes in western and central Shikoku, Japan. The estimated slip distribution from a 12-day tremor burst episode in western Shikoku is heterogeneous, with several patchy areas of slip along the plate interface where rapid moment releases with durations of <100 s regularly occur. We attribute these heterogeneous spatiotemporal slip patterns to heterogeneous material properties along the plate interface. For central Shikoku, where we focus on a tremor burst episode that occurred coincidentally with a very low frequency earthquake (VLF), we observe that the source size of the VLF is much larger than that estimated from tremor activity in western Shikoku. These differences in the size of the slip region may dictate the visibility of VLF signals in observed seismograms, which has implications for the mechanics of slow earthquakes and subduction zone processes.

Convergence acceleration in scattering series and seismic waveform inversion using nonlinear Shanks transformation

NASA Astrophysics Data System (ADS)

Eftekhar, Roya; Hu, Hao; Zheng, Yingcai

2018-06-01

Iterative solution process is fundamental in seismic inversions, such as in full-waveform inversions and some inverse scattering methods. However, the convergence could be slow or even divergent depending on the initial model used in the iteration. We propose to apply Shanks transformation (ST for short) to accelerate the convergence of the iterative solution. ST is a local nonlinear transformation, which transforms a series locally into another series with an improved convergence property. ST works by separating the series into a smooth background trend called the secular term versus an oscillatory transient term. ST then accelerates the convergence of the secular term. Since the transformation is local, we do not need to know all the terms in the original series which is very important in the numerical implementation. The ST performance was tested numerically for both the forward Born series and the inverse scattering series (ISS). The ST has been shown to accelerate the convergence in several examples, including three examples of forward modeling using the Born series and two examples of velocity inversion based on a particular type of the ISS. We observe that ST is effective in accelerating the convergence and it can also achieve convergence even for a weakly divergent scattering series. As such, it provides a useful technique to invert for a large-contrast medium perturbation in seismic inversion.

Imaging Faults in Carbonate Reservoir using Full Waveform Inversion and Reverse Time Migration of Walkaway VSP Data

NASA Astrophysics Data System (ADS)

Takam Takougang, E. M.; Bouzidi, Y.

2016-12-01

Multi-offset Vertical Seismic Profile (walkaway VSP) data were collected in an oil field located in a shallow water environment dominated by carbonate rocks, offshore the United Arab Emirates. The purpose of the survey was to provide structural information of the reservoir, around and away from the borehole. Five parallel lines were collected using an air gun at 25 m shot interval and 4 m source depth. A typical recording tool with 20 receivers spaced every 15.1 m, and located in a deviated borehole with an angle varying between 0 and 24 degree from the vertical direction, was used to record the data. The recording tool was deployed at different depths for each line, from 521 m to 2742 m depth. Smaller offsets were used for shallow receivers and larger offsets for deeper receivers. The lines merged to form the input dataset for waveform tomography. The total length of the combined lines was 9 km, containing 1344 shots and 100 receivers in the borehole located half-way down. Acoustic full waveform inversion was applied in the frequency domain to derive a high resolution velocity model. The final velocity model derived after the inversion using the frequencies 5-40 Hz, showed good correlation with velocities estimated from vertical incidence VSP and sonic log, confirming the success of the inversion. The velocity model showed anomalous low values in areas that correlate with known location of hydrocarbon reservoir. Pre-stack depth Reverse time migration was then applied using the final velocity model from waveform inversion and the up-going wavefield from the input data. The final estimated source signature from waveform inversion was used as input source for reverse time migration. To save computational memory and time, every 3 shots were used during reverse time migration and the data were low-pass filtered to 30 Hz. Migration artifacts were attenuated using a second order derivative filter. The final migration image shows a good correlation with the waveform tomography velocity model, and highlights a complex network of faults in the reservoir, that could be useful in understanding fluid and hydrocarbon movements. This study shows that the combination of full waveform tomography and reverse time migration can provide high resolution images that can enhance interpretation and characterization of oil reservoirs.

Comparison of weighting techniques for acoustic full waveform inversion

NASA Astrophysics Data System (ADS)

Jeong, Gangwon; Hwang, Jongha; Min, Dong-Joo

2017-12-01

To reconstruct long-wavelength structures in full waveform inversion (FWI), the wavefield-damping and weighting techniques have been used to synthesize and emphasize low-frequency data components in frequency-domain FWI. However, these methods have some weak points. The application of wavefield-damping method on filtered data fails to synthesize reliable low-frequency data; the optimization formula obtained introducing the weighting technique is not theoretically complete, because it is not directly derived from the objective function. In this study, we address these weak points and present how to overcome them. We demonstrate that the source estimation in FWI using damped wavefields fails when the data used in the FWI process does not satisfy the causality condition. This phenomenon occurs when a non-causal filter is applied to data. We overcome this limitation by designing a causal filter. Also we modify the conventional weighting technique so that its optimization formula is directly derived from the objective function, retaining its original characteristic of emphasizing the low-frequency data components. Numerical results show that the newly designed causal filter enables to recover long-wavelength structures using low-frequency data components synthesized by damping wavefields in frequency-domain FWI, and the proposed weighting technique enhances the inversion results.

Waveform-based Bayesian full moment tensor inversion and uncertainty determination for the induced seismicity in an oil/gas field

NASA Astrophysics Data System (ADS)

Gu, Chen; Marzouk, Youssef M.; Toksöz, M. Nafi

2018-03-01

Small earthquakes occur due to natural tectonic motions and are induced by oil and gas production processes. In many oil/gas fields and hydrofracking processes, induced earthquakes result from fluid extraction or injection. The locations and source mechanisms of these earthquakes provide valuable information about the reservoirs. Analysis of induced seismic events has mostly assumed a double-couple source mechanism. However, recent studies have shown a non-negligible percentage of non-double-couple components of source moment tensors in hydraulic fracturing events, assuming a full moment tensor source mechanism. Without uncertainty quantification of the moment tensor solution, it is difficult to determine the reliability of these source models. This study develops a Bayesian method to perform waveform-based full moment tensor inversion and uncertainty quantification for induced seismic events, accounting for both location and velocity model uncertainties. We conduct tests with synthetic events to validate the method, and then apply our newly developed Bayesian inversion approach to real induced seismicity in an oil/gas field in the sultanate of Oman—determining the uncertainties in the source mechanism and in the location of that event.

Density reconstruction in multiparameter elastic full-waveform inversion

NASA Astrophysics Data System (ADS)

Sun, Min'ao; Yang, Jizhong; Dong, Liangguo; Liu, Yuzhu; Huang, Chao

2017-12-01

Elastic full-waveform inversion (EFWI) is a quantitative data fitting procedure that recovers multiple subsurface parameters from multicomponent seismic data. As density is involved in addition to P- and S-wave velocities, the multiparameter EFWI suffers from more serious tradeoffs. In addition, compared with P- and S-wave velocities, the misfit function is less sensitive to density perturbation. Thus, a robust density reconstruction remains a difficult problem in multiparameter EFWI. In this paper, we develop an improved scattering-integral-based truncated Gauss-Newton method to simultaneously recover P- and S-wave velocities and density in EFWI. In this method, the inverse Gauss-Newton Hessian has been estimated by iteratively solving the Gauss-Newton equation with a matrix-free conjugate gradient algorithm. Therefore, it is able to properly handle the parameter tradeoffs. To give a detailed illustration of the tradeoffs between P- and S-wave velocities and density in EFWI, wavefield-separated sensitivity kernels and the Gauss-Newton Hessian are numerically computed, and their distribution characteristics are analyzed. Numerical experiments on a canonical inclusion model and a modified SEG/EAGE Overthrust model have demonstrated that the proposed method can effectively mitigate the tradeoff effects, and improve multiparameter gradients. Thus, a high convergence rate and an accurate density reconstruction can be achieved.

Comparison of seismic waveform inversion results for the rupture history of a finite fault: application to the 1986 North Palm Springs, California, earthquake

USGS Publications Warehouse

Hartzell, S.

1989-01-01

The July 8, 1986, North Palm Strings earthquake is used as a basis for comparison of several different approaches to the solution for the rupture history of a finite fault. The inversion of different waveform data is considered; both teleseismic P waveforms and local strong ground motion records. Linear parametrizations for slip amplitude are compared with nonlinear parametrizations for both slip amplitude and rupture time. Inversions using both synthetic and empirical Green's functions are considered. In general, accurate Green's functions are more readily calculable for the teleseismic problem where simple ray theory and flat-layered velocity structures are usually sufficient. However, uncertainties in the variation in t* with frequency most limit the resolution of teleseismic inversions. A set of empirical Green's functions that are well recorded at teleseismic distances could avoid the uncertainties in attenuation. In the inversion of strong motion data, the accurate calculation of propagation path effects other than attenuation effects is the limiting factor in the resolution of source parameters. -from Author

Full waveform time domain solutions for source and induced magnetotelluric and controlled-source electromagnetic fields using quasi-equivalent time domain decomposition and GPU parallelization

NASA Astrophysics Data System (ADS)

Imamura, N.; Schultz, A.

2015-12-01

Recently, a full waveform time domain solution has been developed for the magnetotelluric (MT) and controlled-source electromagnetic (CSEM) methods. The ultimate goal of this approach is to obtain a computationally tractable direct waveform joint inversion for source fields and earth conductivity structure in three and four dimensions. This is desirable on several grounds, including the improved spatial resolving power expected from use of a multitude of source illuminations of non-zero wavenumber, the ability to operate in areas of high levels of source signal spatial complexity and non-stationarity, etc. This goal would not be obtainable if one were to adopt the finite difference time-domain (FDTD) approach for the forward problem. This is particularly true for the case of MT surveys, since an enormous number of degrees of freedom are required to represent the observed MT waveforms across the large frequency bandwidth. It means that for FDTD simulation, the smallest time steps should be finer than that required to represent the highest frequency, while the number of time steps should also cover the lowest frequency. This leads to a linear system that is computationally burdensome to solve. We have implemented our code that addresses this situation through the use of a fictitious wave domain method and GPUs to speed up the computation time. We also substantially reduce the size of the linear systems by applying concepts from successive cascade decimation, through quasi-equivalent time domain decomposition. By combining these refinements, we have made good progress toward implementing the core of a full waveform joint source field/earth conductivity inverse modeling method. From results, we found the use of previous generation of CPU/GPU speeds computations by an order of magnitude over a parallel CPU only approach. In part, this arises from the use of the quasi-equivalent time domain decomposition, which shrinks the size of the linear system dramatically.

Reappraisal of the 2010 Maule, 2014 Iquique, 2015 Illapel through Inversion of Geodetic Data and Tsunami Waveforms Using the Optimal Time Alignment (OTA) Method

NASA Astrophysics Data System (ADS)

Romano, F.; Lorito, S.; Piatanesi, A.; Volpe, M.; Lay, T.; Tolomei, C.; Murphy, S.; Tonini, R.; Escalante, C.; Castro, M. J.; Gonzalez-Vida, J. M.; Macias, J.

2017-12-01

The Chile subduction zone is one of the most seismically active regions in the world and it hosted a number of great tsunamigenic earthquakes in the past. In particular, during the last 7 years three M8+ earthquakes occurred nearby the Chilean coasts, that is the 2010 M8.8 Maule, the 2014 M8.1 Iquique, and the M8.3 2015 Illapel earthquakes. The rupture process of these earthquakes has been studied by using different kind of geophysical observations such as seismic, geodetic, and tsunami data; in particular, tsunami waveforms are important for constraining the slip on the offshore portion of the fault. However, it has been shown that forward modelling of tsunami data can be affected by unavailability of accurate bathymetric models, especially in the vicinity of the tide-gauges; and in the far field by water density gradients, ocean floor elasticity, or geopotential gravity changes, generally neglected. This could result in a mismatch between observed and predicted tsunami signals thus affecting the retrieved tsunami source image. Recently, a method has been proposed for automatic correction during the nonlinear inversion of the mismatch (optimal time alignment, OTA; Romano et al., GRL, 2016). Here, we present a reappraisal of the joint inversion of tsunami data with OTA procedure and geodetic data, for the Maule, Iquique, and Illapel earthquakes. We compare the results with those obtained by tsunami inversion without using OTA and with other published inversion results.

Resolvability of regional density structure

NASA Astrophysics Data System (ADS)

Plonka, A.; Fichtner, A.

2016-12-01

Lateral density variations are the source of mass transport in the Earth at all scales, acting as drivers of convectivemotion. However, the density structure of the Earth remains largely unknown since classic seismic observables and gravityprovide only weak constraints with strong trade-offs. Current density models are therefore often based on velocity scaling,making strong assumptions on the origin of structural heterogeneities, which may not necessarily be correct. Our goal is to assessif 3D density structure may be resolvable with emerging full-waveform inversion techniques. We have previously quantified the impact of regional-scale crustal density structure on seismic waveforms with the conclusion that reasonably sized density variations within thecrust can leave a strong imprint on both travel times and amplitudes, and, while this can produce significant biases in velocity and Q estimates, the seismic waveform inversion for density may become feasible. In this study we performprincipal component analyses of sensitivity kernels for P velocity, S velocity, and density. This is intended to establish theextent to which these kernels are linearly independent, i.e. the extent to which the different parameters may be constrainedindependently. Since the density imprint we observe is not exclusively linked to travel times and amplitudes of specific phases,we consider waveform differences between complete seismograms. We test the method using a known smooth model of the crust and seismograms with clear Love and Rayleigh waves, showing that - as expected - the first principal kernel maximizes sensitivity to SH and SV velocity structure, respectively, and that the leakage between S velocity, P velocity and density parameter spaces is minimal in the chosen setup. Next, we apply the method to data from 81 events around the Iberian Penninsula, registered in total by 492 stations. The objective is to find a principal kernel which would maximize the sensitivity to density, potentially allowing for independent density resolution, and, as the final goal, for direct density inversion.

Source mechanism of small long-period events at Mount St. Helens in July 2005 using template matching, phase-weighted stacking, and full-waveform inversion

USGS Publications Warehouse

Matoza, Robin S.; Chouet, Bernard A.; Dawson, Phillip B.; Shearer, Peter M.; Haney, Matthew M.; Waite, Gregory P.; Moran, Seth C.; Mikesell, T. Dylan

2015-01-01

Long-period (LP, 0.5-5 Hz) seismicity, observed at volcanoes worldwide, is a recognized signature of unrest and eruption. Cyclic LP “drumbeating” was the characteristic seismicity accompanying the sustained dome-building phase of the 2004–2008 eruption of Mount St. Helens (MSH), WA. However, together with the LP drumbeating was a near-continuous, randomly occurring series of tiny LP seismic events (LP “subevents”), which may hold important additional information on the mechanism of seismogenesis at restless volcanoes. We employ template matching, phase-weighted stacking, and full-waveform inversion to image the source mechanism of one multiplet of these LP subevents at MSH in July 2005. The signal-to-noise ratios of the individual events are too low to produce reliable waveform-inversion results, but the events are repetitive and can be stacked. We apply network-based template matching to 8 days of continuous velocity waveform data from 29 June to 7 July 2005 using a master event to detect 822 network triggers. We stack waveforms for 359 high-quality triggers at each station and component, using a combination of linear and phase-weighted stacking to produce clean stacks for use in waveform inversion. The derived source mechanism pointsto the volumetric oscillation (~10 m3) of a subhorizontal crack located at shallow depth (~30 m) in an area to the south of Crater Glacier in the southern portion of the breached MSH crater. A possible excitation mechanism is the sudden condensation of metastable steam from a shallow pressurized hydrothermal system as it encounters cool meteoric water in the outer parts of the edifice, perhaps supplied from snow melt.

Multi-parameter Full-waveform Inversion for Acoustic VTI Medium with Surface Seismic Data

NASA Astrophysics Data System (ADS)

Cheng, X.; Jiao, K.; Sun, D.; Huang, W.; Vigh, D.

2013-12-01

Full-waveform Inversion (FWI) attracts wide attention recently in oil and gas industry as a new promising tool for high resolution subsurface velocity model building. While the traditional common image point gather based tomography method aims to focus post-migrated data in depth domain, FWI aims to directly fit the observed seismic waveform in either time or frequency domain. The inversion is performed iteratively by updating the velocity fields to reduce the difference between the observed and the simulated data. It has been shown the inversion is very sensitive to the starting velocity fields, and data with long offsets and low frequencies is crucial for the success of FWI to overcome this sensitivity. Considering the importance of data with long offsets and low frequencies, in most geologic environment, anisotropy is an unavoidable topic for FWI especially at long offsets, since anisotropy tends to have more pronounced effects on waves traveled for a great distance. In VTI medium, this means more horizontal velocity will be registered in middle-to-long offset data, while more vertical velocity will be registered in near-to-middle offset data. Up to date, most of real world applications of FWI still remain in isotropic medium, and only a few studies have been shown to account for anisotropy. And most of those studies only account for anisotropy in waveform simulation, but not invert for those anisotropy fields. Multi-parameter inversion for anisotropy fields, even in VTI medium, remains as a hot topic in the field. In this study, we develop a strategy for multi-parameter FWI for acoustic VTI medium with surface seismic data. Because surface seismic data is insensitivity to the delta fields, we decide to hold the delta fields unchanged during our inversion, and invert only for vertical velocity and epsilon fields. Through parameterization analysis and synthetic tests, we find that it is more feasible to invert for the parameterization as vertical and horizontal velocities instead of inverting for the parameterization as vertical velocity and epsilon fields. We develop a hierarchical approach to invert for vertical velocity first but hold epsilon unchanged and only switch to simultaneous inversion when vertical velocity inversion are approaching convergence. During simultaneous inversion, we observe significant acceleration in the convergence when incorporates second order information and preconditioning into inversion. We demonstrate the success of our strategy for VTI FWI using synthetic and real data examples from the Gulf of Mexico. Our results show that incorporation of VTI FWI improves migration of large offset acquisition data, and produces better focused migration images to be used in exploration, production and development of oil fields.

Salvus: A scalable software suite for full-waveform modelling & inversion

NASA Astrophysics Data System (ADS)

Afanasiev, M.; Boehm, C.; van Driel, M.; Krischer, L.; Fichtner, A.

2017-12-01

Full-waveform inversion (FWI), whether at the lab, exploration, or planetary scale, requires the cooperation of five principal components. (1) The geometry of the domain needs to be properly discretized and an initial guess of the model parameters must be projected onto it; (2) Large volumes of recorded waveform data must be collected, organized, and processed; (3) Synthetic waveform data must be efficiently and accurately computed through complex domains; (4) Suitable misfit functions and optimization techniques must be used to relate discrepancies in data space to perturbations in the model; and (5) Some form of workflow management must be employed to schedule and run (1) - (4) in the correct order. Each one of these components can represent a formidable technical challenge which redirects energy from the true task at hand: using FWI to extract new information about some underlying continuum.In this presentation we give an overview of the current status of the Salvus software suite, which was introduced to address the challenges listed above. Specifically, we touch on (1) salvus_mesher, which eases the discretization of complex Earth models into hexahedral meshes; (2) salvus_seismo, which integrates with LASIF and ObsPy to streamline the processing and preparation of seismic data; (3) salvus_wave, a high-performance and scalable spectral-element solver capable of simulating waveforms through general unstructured 2- and 3-D domains, and (4) salvus_opt, an optimization toolbox specifically designed for full-waveform inverse problems. Tying everything together, we also discuss (5) salvus_flow: a workflow package designed to orchestrate and manage the rest of the suite. It is our hope that these developments represent a step towards the automation of large-scale seismic waveform inversion, while also lowering the barrier of entry for new applications. We include several examples of Salvus' use in (extra-) planetary seismology, non-destructive testing, and medical imaging.

Source rupture process of the 2016 Kaikoura, New Zealand earthquake estimated from the kinematic waveform inversion of strong-motion data

NASA Astrophysics Data System (ADS)

Zheng, Ao; Wang, Mingfeng; Yu, Xiangwei; Zhang, Wenbo

2018-03-01

On 2016 November 13, an Mw 7.8 earthquake occurred in the northeast of the South Island of New Zealand near Kaikoura. The earthquake caused severe damages and great impacts on local nature and society. Referring to the tectonic environment and defined active faults, the field investigation and geodetic evidence reveal that at least 12 fault sections ruptured in the earthquake, and the focal mechanism is one of the most complicated in historical earthquakes. On account of the complexity of the source rupture, we propose a multisegment fault model based on the distribution of surface ruptures and active tectonics. We derive the source rupture process of the earthquake using the kinematic waveform inversion method with the multisegment fault model from strong-motion data of 21 stations (0.05-0.35 Hz). The inversion result suggests the rupture initiates in the epicentral area near the Humps fault, and then propagates northeastward along several faults, until the offshore Needles fault. The Mw 7.8 event is a mixture of right-lateral strike and reverse slip, and the maximum slip is approximately 19 m. The synthetic waveforms reproduce the characteristics of the observed ones well. In addition, we synthesize the coseismic offsets distribution of the ruptured region from the slips of upper subfaults in the fault model, which is roughly consistent with the surface breaks observed in the field survey.

A Modified Normalization Technique for Frequency-Domain Full Waveform Inversion

NASA Astrophysics Data System (ADS)

Hwang, J.; Jeong, G.; Min, D. J.; KIM, S.; Heo, J. Y.

2016-12-01

Full waveform inversion (FWI) is a technique to estimate subsurface material properties minimizing the misfit function built with residuals between field and modeled data. To achieve computational efficiency, FWI has been performed in the frequency domain by carrying out modeling in the frequency domain, whereas observed data (time-series data) are Fourier-transformed.One of the main drawbacks of seismic FWI is that it easily gets stuck in local minima because of lacking of low-frequency data. To compensate for this limitation, damped wavefields are used, as in the Laplace-domain waveform inversion. Using damped wavefield in FWI plays a role in generating low-frequency components and help recover long-wavelength structures. With these newly generated low-frequency components, we propose a modified frequency-normalization technique, which has an effect of boosting contribution of low-frequency components to model parameter update.In this study, we introduce the modified frequency-normalization technique which effectively amplifies low-frequency components of damped wavefields. Our method is demonstrated for synthetic data for the SEG/EAGE salt model. AcknowledgementsThis work was supported by the Korea Institute of Energy Technology Evaluation and Planning(KETEP) and the Ministry of Trade, Industry & Energy(MOTIE) of the Republic of Korea (No. 20168510030830) and by the Dual Use Technology Program, granted financial resource from the Ministry of Trade, Industry & Energy, Republic of Korea.

Seismic waveform inversion for core-mantle boundary topography

NASA Astrophysics Data System (ADS)

Colombi, Andrea; Nissen-Meyer, Tarje; Boschi, Lapo; Giardini, Domenico

2014-07-01

The topography of the core-mantle boundary (CMB) is directly linked to the dynamics of both the mantle and the outer core, although it is poorly constrained and understood. Recent studies have produced topography models with mutual agreement up to degree 2. A broad-band waveform inversion strategy is introduced and applied here, with relatively low computational cost and based on a first-order Born approximation. Its performance is validated using synthetic waveforms calculated in theoretical earth models that include different topography patterns with varying lateral wavelengths, from 600 to 2500 km, and magnitudes (˜10 km peak-to-peak). The source-receiver geometry focuses mainly on the Pdiff, PKP, PcP and ScS phases. The results show that PKP branches, PcP and ScS generally perform well and in a similar fashion, while Pdiff yields unsatisfactory results. We investigate also how 3-D mantle correction influences the output models, and find that despite the disturbance introduced, the models recovered do not appear to be biased, provided that the 3-D model is correct. Using cross-correlated traveltimes, we derive new topography models from both P and S waves. The static corrections used to remove the mantle effect are likely to affect the inversion, compromising the agreement between models derived from P and S data. By modelling traveltime residuals starting from sensitivity kernels, we show how the simultaneous use of volumetric and boundary kernels can reduce the bias coming from mantle structures. The joint inversion approach should be the only reliable method to invert for CMB topography using absolute cross-correlation traveltimes.

Waveform inversion of oscillatory signatures in long-period events beneath volcanoes

USGS Publications Warehouse

Kumagai, H.; Chouet, B.A.; Nakano, M.

2002-01-01

The source mechanism of long-period (LP) events is examined using synthetic waveforms generated by the acoustic resonance of a fluid-filled crack. We perform a series of numerical tests in which the oscillatory signatures of synthetic LP waveforms are used to determine the source time functions of the six moment tensor components from waveform inversions assuming a point source. The results indicate that the moment tensor representation is valid for the odd modes of crack resonance with wavelengths 2L/n, 2W/n, n = 3, 5, 7, ..., where L and W are the crack length and width, respectively. For the even modes with wavelengths 2L/n, 2W/n, n = 2, 4, 6,..., a generalized source representation using higher-order tensors is required, although the efficiency of seismic waves radiated by the even modes is expected to be small. We apply the moment tensor inversion to the oscillatory signatures of an LP event observed at Kusatsu-Shirane Volcano, central Japan. Our results point to the resonance of a subhorizontal crack located a few hundred meters beneath the summit crater lakes. The present approach may be useful to quantify the source location, geometry, and force system of LP events, and opens the way for moment tensor inversions of tremor.

Focal mechanisms and moment magnitudes of micro-earthquakes in central Brazil by waveform inversion with quality assessment and inference of the local stress field

NASA Astrophysics Data System (ADS)

Carvalho, Juraci; Barros, Lucas Vieira; Zahradník, Jiří

2016-11-01

This paper documents an investigation on the use of full waveform inversion to retrieve focal mechanisms of 11 micro-earthquakes (Mw 0.8 to 1.4). The events represent aftershocks of a 5.0 mb earthquake that occurred on October 8, 2010 close to the city of Mara Rosa in the state of Goiás, Brazil. The main contribution of the work lies in demonstrating the feasibility of waveform inversion of such weak events. The inversion was made possible thanks to recordings available at 8 temporary seismic stations in epicentral distances of less than 8 km, at which waveforms can be successfully modeled at relatively high frequencies (1.5-2.0 Hz). On average, the fault-plane solutions obtained are in agreement with a composite focal mechanism previously calculated from first-motion polarities. They also agree with the fault geometry inferred from precise relocation of the Mara Rosa aftershock sequence. The focal mechanisms provide an estimate of the local stress field. This paper serves as a pilot study for similar investigations in intraplate regions where the stress-field investigations are difficult due to rare earthquake occurrences, and where weak events must be studied with a detailed quality assessment.

A Global Upper-Mantle Tomographic Model of Shear Attenuation

NASA Astrophysics Data System (ADS)

Karaoglu, H.; Romanowicz, B. A.

2016-12-01

Mapping anelastic 3D structure within the earth's mantle is key to understanding present day mantle dynamics, as it provides complementary constraints to those obtained from elastic structure, with the potential to distinguish between thermal and compositional heterogeneity. For this, we need to measure seismic wave amplitudes, which are sensitive to both elastic (through focusing and scattering) and anelastic structure. The elastic effects are less pronounced at long periods, so previous global upper-mantle attenuation models are based on teleseismic surface wave data, sometimes including overtones. In these studies, elastic effects are considered either indirectly, by eliminating data strongly contaminated by them (e.g. Romanowicz, 1995; Gung and Romanowicz, 2004), or by correcting for elastic focusing effects using an approximate linear approach (Dalton et al., 2008). Additionally, in these studies, the elastic structure is held fixed when inverting for intrinsic attenuation . The importance of (1) having a good starting elastic model, (2) accurate modeling of the seismic wavefield and (3) joint inversion for elastic and anelastic structure, becomes more evident as the targeted resolution level increases. Also, velocity dispersion effects due to anelasticity need to be taken into account. Here, we employ a hybrid full waveform inversion method, inverting jointly for global elastic and anelastic upper mantle structure, starting from the latest global 3D shear velocity model built by our group (French and Romanowicz, 2014), using the spectral element method for the forward waveform modeling (Capdeville et al., 2003), and normal-mode perturbation theory (NACT - Li and Romanowicz, 1995) for kernel computations. We present a 3D upper-mantle anelastic model built by using three component fundamental and overtone surface waveforms down to 60 s as well as long period body waveforms down to 30 s. We also include source and site effects to first order as frequency independent scalar factors. The robustness of the inversion method is assessed through synthetic and resolution tests. We discuss salient features of the resulting anelastic model and in particular the well-resolved strong correlation with tectonics observed in the first 200 km of the mantle.

Fully probabilistic earthquake source inversion on teleseismic scales

NASA Astrophysics Data System (ADS)

Stähler, Simon; Sigloch, Karin

2017-04-01

Seismic source inversion is a non-linear problem in seismology where not just the earthquake parameters but also estimates of their uncertainties are of great practical importance. We have developed a method of fully Bayesian inference for source parameters, based on measurements of waveform cross-correlation between broadband, teleseismic body-wave observations and their modelled counterparts. This approach yields not only depth and moment tensor estimates but also source time functions. These unknowns are parameterised efficiently by harnessing as prior knowledge solutions from a large number of non-Bayesian inversions. The source time function is expressed as a weighted sum of a small number of empirical orthogonal functions, which were derived from a catalogue of >1000 source time functions (STFs) by a principal component analysis. We use a likelihood model based on the cross-correlation misfit between observed and predicted waveforms. The resulting ensemble of solutions provides full uncertainty and covariance information for the source parameters, and permits propagating these source uncertainties into travel time estimates used for seismic tomography. The computational effort is such that routine, global estimation of earthquake mechanisms and source time functions from teleseismic broadband waveforms is feasible. A prerequisite for Bayesian inference is the proper characterisation of the noise afflicting the measurements. We show that, for realistic broadband body-wave seismograms, the systematic error due to an incomplete physical model affects waveform misfits more strongly than random, ambient background noise. In this situation, the waveform cross-correlation coefficient CC, or rather its decorrelation D = 1 - CC, performs more robustly as a misfit criterion than ℓp norms, more commonly used as sample-by-sample measures of misfit based on distances between individual time samples. From a set of over 900 user-supervised, deterministic earthquake source solutions treated as a quality-controlled reference, we derive the noise distribution on signal decorrelation D of the broadband seismogram fits between observed and modelled waveforms. The noise on D is found to approximately follow a log-normal distribution, a fortunate fact that readily accommodates the formulation of an empirical likelihood function for D for our multivariate problem. The first and second moments of this multivariate distribution are shown to depend mostly on the signal-to-noise ratio (SNR) of the CC measurements and on the back-azimuthal distances of seismic stations. References: Stähler, S. C. and Sigloch, K.: Fully probabilistic seismic source inversion - Part 1: Efficient parameterisation, Solid Earth, 5, 1055-1069, doi:10.5194/se-5-1055-2014, 2014. Stähler, S. C. and Sigloch, K.: Fully probabilistic seismic source inversion - Part 2: Modelling errors and station covariances, Solid Earth, 7, 1521-1536, doi:10.5194/se-7-1521-2016, 2016.

Developing a Near Real-time System for Earthquake Slip Distribution Inversion

NASA Astrophysics Data System (ADS)

Zhao, Li; Hsieh, Ming-Che; Luo, Yan; Ji, Chen

2016-04-01

Advances in observational and computational seismology in the past two decades have enabled completely automatic and real-time determinations of the focal mechanisms of earthquake point sources. However, seismic radiations from moderate and large earthquakes often exhibit strong finite-source directivity effect, which is critically important for accurate ground motion estimations and earthquake damage assessments. Therefore, an effective procedure to determine earthquake rupture processes in near real-time is in high demand for hazard mitigation and risk assessment purposes. In this study, we develop an efficient waveform inversion approach for the purpose of solving for finite-fault models in 3D structure. Full slip distribution inversions are carried out based on the identified fault planes in the point-source solutions. To ensure efficiency in calculating 3D synthetics during slip distribution inversions, a database of strain Green tensors (SGT) is established for 3D structural model with realistic surface topography. The SGT database enables rapid calculations of accurate synthetic seismograms for waveform inversion on a regular desktop or even a laptop PC. We demonstrate our source inversion approach using two moderate earthquakes (Mw~6.0) in Taiwan and in mainland China. Our results show that 3D velocity model provides better waveform fitting with more spatially concentrated slip distributions. Our source inversion technique based on the SGT database is effective for semi-automatic, near real-time determinations of finite-source solutions for seismic hazard mitigation purposes.

Frozen Gaussian approximation for 3D seismic tomography

NASA Astrophysics Data System (ADS)

Chai, Lihui; Tong, Ping; Yang, Xu

2018-05-01

Three-dimensional (3D) wave-equation-based seismic tomography is computationally challenging in large scales and high-frequency regime. In this paper, we apply the frozen Gaussian approximation (FGA) method to compute 3D sensitivity kernels and seismic tomography of high-frequency. Rather than standard ray theory used in seismic inversion (e.g. Kirchhoff migration and Gaussian beam migration), FGA is used to compute the 3D high-frequency sensitivity kernels for travel-time or full waveform inversions. Specifically, we reformulate the equations of the forward and adjoint wavefields for the purpose of convenience to apply FGA, and with this reformulation, one can efficiently compute the Green’s functions whose convolutions with source time function produce wavefields needed for the construction of 3D kernels. Moreover, a fast summation method is proposed based on local fast Fourier transform which greatly improves the speed of reconstruction as the last step of FGA algorithm. We apply FGA to both the travel-time adjoint tomography and full waveform inversion (FWI) on synthetic crosswell seismic data with dominant frequencies as high as those of real crosswell data, and confirm again that FWI requires a more sophisticated initial velocity model for the convergence than travel-time adjoint tomography. We also numerically test the accuracy of applying FGA to local earthquake tomography. This study paves the way to directly apply wave-equation-based seismic tomography methods into real data around their dominant frequencies.

A flexible, extendable, modular and computationally efficient approach to scattering-integral-based seismic full waveform inversion

NASA Astrophysics Data System (ADS)

Schumacher, F.; Friederich, W.; Lamara, S.

2016-02-01

We present a new conceptual approach to scattering-integral-based seismic full waveform inversion (FWI) that allows a flexible, extendable, modular and both computationally and storage-efficient numerical implementation. To achieve maximum modularity and extendability, interactions between the three fundamental steps carried out sequentially in each iteration of the inversion procedure, namely, solving the forward problem, computing waveform sensitivity kernels and deriving a model update, are kept at an absolute minimum and are implemented by dedicated interfaces. To realize storage efficiency and maximum flexibility, the spatial discretization of the inverted earth model is allowed to be completely independent of the spatial discretization employed by the forward solver. For computational efficiency reasons, the inversion is done in the frequency domain. The benefits of our approach are as follows: (1) Each of the three stages of an iteration is realized by a stand-alone software program. In this way, we avoid the monolithic, unflexible and hard-to-modify codes that have often been written for solving inverse problems. (2) The solution of the forward problem, required for kernel computation, can be obtained by any wave propagation modelling code giving users maximum flexibility in choosing the forward modelling method. Both time-domain and frequency-domain approaches can be used. (3) Forward solvers typically demand spatial discretizations that are significantly denser than actually desired for the inverted model. Exploiting this fact by pre-integrating the kernels allows a dramatic reduction of disk space and makes kernel storage feasible. No assumptions are made on the spatial discretization scheme employed by the forward solver. (4) In addition, working in the frequency domain effectively reduces the amount of data, the number of kernels to be computed and the number of equations to be solved. (5) Updating the model by solving a large equation system can be done using different mathematical approaches. Since kernels are stored on disk, it can be repeated many times for different regularization parameters without need to solve the forward problem, making the approach accessible to Occam's method. Changes of choice of misfit functional, weighting of data and selection of data subsets are still possible at this stage. We have coded our approach to FWI into a program package called ASKI (Analysis of Sensitivity and Kernel Inversion) which can be applied to inverse problems at various spatial scales in both Cartesian and spherical geometries. It is written in modern FORTRAN language using object-oriented concepts that reflect the modular structure of the inversion procedure. We validate our FWI method by a small-scale synthetic study and present first results of its application to high-quality seismological data acquired in the southern Aegean.

A high-order strong stability preserving Runge-Kutta method for three-dimensional full waveform modeling and inversion of anelastic models

NASA Astrophysics Data System (ADS)

Wang, N.; Shen, Y.; Yang, D.; Bao, X.; Li, J.; Zhang, W.

2017-12-01

Accurate and efficient forward modeling methods are important for high resolution full waveform inversion. Compared with the elastic case, solving anelastic wave equation requires more computational time, because of the need to compute additional material-independent anelastic functions. A numerical scheme with a large Courant-Friedrichs-Lewy (CFL) condition number enables us to use a large time step to simulate wave propagation, which improves computational efficiency. In this work, we apply the fourth-order strong stability preserving Runge-Kutta method with an optimal CFL coeffiecient to solve the anelastic wave equation. We use a fourth order DRP/opt MacCormack scheme for the spatial discretization, and we approximate the rheological behaviors of the Earth by using the generalized Maxwell body model. With a larger CFL condition number, we find that the computational efficient is significantly improved compared with the traditional fourth-order Runge-Kutta method. Then, we apply the scattering-integral method for calculating travel time and amplitude sensitivity kernels with respect to velocity and attenuation structures. For each source, we carry out one forward simulation and save the time-dependent strain tensor. For each station, we carry out three `backward' simulations for the three components and save the corresponding strain tensors. The sensitivity kernels at each point in the medium are the convolution of the two sets of the strain tensors. Finally, we show several synthetic tests to verify the effectiveness of the strong stability preserving Runge-Kutta method in generating accurate synthetics in full waveform modeling, and in generating accurate strain tensors for calculating sensitivity kernels at regional and global scales.

The source parameters of 2013 Mw6.6 Lushan earthquake constrained with the restored local clipped seismic waveforms

NASA Astrophysics Data System (ADS)

Hao, J.; Zhang, J. H.; Yao, Z. X.

2017-12-01

We developed a method to restore the clipped seismic waveforms near epicenter using projection onto convex sets method (Zhang et al, 2016). This method was applied to rescue the local clipped waveforms of 2013 Mw 6.6 Lushan earthquake. We restored 88 out of 93 clipped waveforms of 38 broadband seismic stations of China Earthquake Networks (CEN). The epicenter distance of the nearest station to the epicenter that we can faithfully restore is only about 32 km. In order to investigate if the source parameters of earthquake could be determined exactly with the restored data, restored waveforms are utilized to get the mechanism of Lushan earthquake. We apply the generalized reflection-transmission coefficient matrix method to calculate the synthetic seismic records and simulated annealing method in inversion (Yao and Harkrider, 1983; Hao et al., 2012). We select 5 stations of CEN with the epicenter distance about 200km whose records aren't clipped and three-component velocity records are used. The result shows the strike, dip and rake angles of Lushan earthquake are 200o, 51o and 87o respectively, hereinafter "standard result". Then the clipped and restored seismic waveforms are applied respectively. The strike, dip and rake angles of clipped seismic waveforms are 184o, 53o and 72o respectively. The largest misfit of angle is 16o. In contrast, the strike, dip and rake angles of restored seismic waveforms are 198o, 51o and 87o respectively. It is very close to the "standard result". We also study the rupture history of Lushan earthquake constrained with the restored local broadband and teleseismic waves based on finite fault method (Hao et al., 2013). The result consists with that constrained with the strong motion and teleseismic waves (Hao et al., 2013), especially the location of the patch with larger slip. In real-time seismology, determining the source parameters as soon as possible is important. This method will help us to determine the mechanism of earthquake using the local clipped waveforms. Strong motion stations in China don't have good coverage at present. This method will help us to investigate the rupture history of large earthquake in China using the local clipped data of broadband stations.

Development of FWIGPR, an open-source package for full-waveform inversion of common-offset GPR data

NASA Astrophysics Data System (ADS)

Jazayeri, S.; Kruse, S.

2017-12-01

We introduce a package for full-waveform inversion (FWI) of Ground Penetrating Radar (GPR) data based on a combination of open-source programs. The FWI requires a good starting model, based on direct knowledge of field conditions or on traditional ray-based inversion methods. With a good starting model, the FWI can improve resolution of selected subsurface features. The package will be made available for general use in educational and research activities. The FWIGPR package consists of four main components: 3D to 2D data conversion, source wavelet estimation, forward modeling, and inversion. (These four components additionally require the development, by the user, of a good starting model.) A major challenge with GPR data is the unknown form of the waveform emitted by the transmitter held close to the ground surface. We apply a blind deconvolution method to estimate the source wavelet, based on a sparsity assumption about the reflectivity series of the subsurface model (Gholami and Sacchi 2012). The estimated wavelet is deconvolved from the data and the sparsest reflectivity series with fewest reflectors. The gprMax code (www.gprmax.com) is used as the forward modeling tool and the PEST parameter estimation package (www.pesthomepage.com) for the inversion. To reduce computation time, the field data are converted to an effective 2D equivalent, and the gprMax code can be run in 2D mode. In the first step, the user must create a good starting model of the data, presumably using ray-based methods. This estimated model will be introduced to the FWI process as an initial model. Next, the 3D data is converted to 2D, then the user estimates the source wavelet that best fits the observed data by sparsity assumption of the earth's response. Last, PEST runs gprMax with the initial model and calculates the misfit between the synthetic and observed data, and using an iterative algorithm calling gprMax several times ineach iteration, finds successive models that better fit the data. To gauge whether the iterative process has arrived at a local or global minima, the process can be repeated with a range of starting models. Tests have shown that this package can successfully improve estimates of selected subsurface model parameters for simple synthetic and real data. Ongoing research will focus on FWI of more complex scenarios.

Resolvability of regional density structure and the road to direct density inversion - a principal-component approach to resolution analysis

NASA Astrophysics Data System (ADS)

Płonka, Agnieszka; Fichtner, Andreas

2017-04-01

Lateral density variations are the source of mass transport in the Earth at all scales, acting as drivers of convective motion. However, the density structure of the Earth remains largely unknown since classic seismic observables and gravity provide only weak constraints with strong trade-offs. Current density models are therefore often based on velocity scaling, making strong assumptions on the origin of structural heterogeneities, which may not necessarily be correct. Our goal is to assess if 3D density structure may be resolvable with emerging full-waveform inversion techniques. We have previously quantified the impact of regional-scale crustal density structure on seismic waveforms with the conclusion that reasonably sized density variations within the crust can leave a strong imprint on both travel times and amplitudes, and, while this can produce significant biases in velocity and Q estimates, the seismic waveform inversion for density may become feasible. In this study we perform principal component analyses of sensitivity kernels for P velocity, S velocity, and density. This is intended to establish the extent to which these kernels are linearly independent, i.e. the extent to which the different parameters may be constrained independently. We apply the method to data from 81 events around the Iberian Penninsula, registered in total by 492 stations. The objective is to find a principal kernel which would maximize the sensitivity to density, potentially allowing for as independent as possible density resolution. We find that surface (mosty Rayleigh) waves have significant sensitivity to density, and that the trade-off with velocity is negligible. We also show the preliminary results of the inversion.

Multicomponent pre-stack seismic waveform inversion in transversely isotropic media using a non-dominated sorting genetic algorithm

NASA Astrophysics Data System (ADS)

Padhi, Amit; Mallick, Subhashis

2014-03-01

Inversion of band- and offset-limited single component (P wave) seismic data does not provide robust estimates of subsurface elastic parameters and density. Multicomponent seismic data can, in principle, circumvent this limitation but adds to the complexity of the inversion algorithm because it requires simultaneous optimization of multiple objective functions, one for each data component. In seismology, these multiple objectives are typically handled by constructing a single objective given as a weighted sum of the objectives of individual data components and sometimes with additional regularization terms reflecting their interdependence; which is then followed by a single objective optimization. Multi-objective problems, inclusive of the multicomponent seismic inversion are however non-linear. They have non-unique solutions, known as the Pareto-optimal solutions. Therefore, casting such problems as a single objective optimization provides one out of the entire set of the Pareto-optimal solutions, which in turn, may be biased by the choice of the weights. To handle multiple objectives, it is thus appropriate to treat the objective as a vector and simultaneously optimize each of its components so that the entire Pareto-optimal set of solutions could be estimated. This paper proposes such a novel multi-objective methodology using a non-dominated sorting genetic algorithm for waveform inversion of multicomponent seismic data. The applicability of the method is demonstrated using synthetic data generated from multilayer models based on a real well log. We document that the proposed method can reliably extract subsurface elastic parameters and density from multicomponent seismic data both when the subsurface is considered isotropic and transversely isotropic with a vertical symmetry axis. We also compute approximate uncertainty values in the derived parameters. Although we restrict our inversion applications to horizontally stratified models, we outline a practical procedure of extending the method to approximately include local dips for each source-receiver offset pair. Finally, the applicability of the proposed method is not just limited to seismic inversion but it could be used to invert different data types not only requiring multiple objectives but also multiple physics to describe them.

ASKI: A modular toolbox for scattering-integral-based seismic full waveform inversion and sensitivity analysis utilizing external forward codes

NASA Astrophysics Data System (ADS)

Schumacher, Florian; Friederich, Wolfgang

Due to increasing computational resources, the development of new numerically demanding methods and software for imaging Earth's interior remains of high interest in Earth sciences. Here, we give a description from a user's and programmer's perspective of the highly modular, flexible and extendable software package ASKI-Analysis of Sensitivity and Kernel Inversion-recently developed for iterative scattering-integral-based seismic full waveform inversion. In ASKI, the three fundamental steps of solving the seismic forward problem, computing waveform sensitivity kernels and deriving a model update are solved by independent software programs that interact via file output/input only. Furthermore, the spatial discretizations of the model space used for solving the seismic forward problem and for deriving model updates, respectively, are kept completely independent. For this reason, ASKI does not contain a specific forward solver but instead provides a general interface to established community wave propagation codes. Moreover, the third fundamental step of deriving a model update can be repeated at relatively low costs applying different kinds of model regularization or re-selecting/weighting the inverted dataset without need to re-solve the forward problem or re-compute the kernels. Additionally, ASKI offers the user sensitivity and resolution analysis tools based on the full sensitivity matrix and allows to compose customized workflows in a consistent computational environment. ASKI is written in modern Fortran and Python, it is well documented and freely available under terms of the GNU General Public License (http://www.rub.de/aski).

Global seismic attenuation imaging using full-waveform inversion: a comparative assessment of different choices of misfit functionals

NASA Astrophysics Data System (ADS)

Karaoǧlu, Haydar; Romanowicz, Barbara

2018-02-01

We present the results of synthetic tests that aim at evaluating the relative performance of three different definitions of misfit functionals in the context of 3-D imaging of shear wave attenuation in the earth's upper mantle at the global scale, using long-period full-waveform data. The synthetic tests are conducted with simple hypothetical upper-mantle models that contain Qμ anomalies centred at different depths and locations, with or without additional seismic velocity anomalies. To build synthetic waveform data sets, we performed simulations of 50 events in the hypothetical (target) models, using the spectral element method, filtered in the period range 60-400 s. The selected events are chosen among 273 events used in the development of radially anisotropic model SEMUCB-WM1 and recorded at 495 stations worldwide. The synthetic Z-component waveforms correspond to paths and time intervals (fundamental mode and overtone Rayleigh waves) that exist in the real waveform data set. The inversions for shear attenuation structure are carried out using a Gauss-Newton optimization scheme in which the gradient and Hessian are computed using normal mode perturbation theory. The three different misfit functionals considered are based on time domain waveform (WF) and waveform envelope (E-WF) differences, as well as spectral amplitude ratios (SA), between observed and predicted waveforms. We evaluate the performance of the three misfit functional definitions in the presence of seismic noise and unresolved S-wave velocity heterogeneity and discuss the relative importance of physical dispersion effects due to 3-D Qμ structure. We observed that the performance of WF is poorer than the other two misfit functionals in recovering attenuation structure, unless anelastic dispersion effects are taken into account in the calculation of partial derivatives. WF also turns out to be more sensitive to seismic noise than E-WF and SA. Overall, SA performs best for attenuation imaging. Our tests show that it is important to account for 3-D elastic effects (focusing) before inverting for Qμ. Additionally, we show that including high signal-to-noise ratio overtone wave packets is necessary to resolve Qμ structure at depths greater than 250 km.

Velocity structure of a bottom simulating reflector offshore Peru: Results from full waveform inversion

USGS Publications Warehouse

Pecher, I.A.; Minshull, T.A.; Singh, S.C.; von Huene, Roland E.

1996-01-01

Much of our knowledge of the worldwide distribution of submarine gas hydrates comes from seismic observations of Bottom Simulating Reflectors (BSRs). Full waveform inversion has proven to be a reliable technique for studying the fine structure of BSRs using the compressional wave velocity. We applied a non-linear full waveform inversion technique to a BSR at a location offshore Peru. We first determined the large-scale features of seismic velocity variations using a statistical inversion technique to maximise coherent energy along travel-time curves. These velocities were used for a starting velocity model for the full waveform inversion, which yielded a detailed velocity/depth model in the vicinity of the BSR. We found that the data are best fit by a model in which the BSR consists of a thin, low-velocity layer. The compressional wave velocity drops from 2.15 km/s down to an average of 1.70 km/s in an 18m thick interval, with a minimum velocity of 1.62 km/s in a 6 m interval. The resulting compressional wave velocity was used to estimate gas content in the sediments. Our results suggest that the low velocity layer is a 6-18 m thick zone containing a few percent of free gas in the pore space. The presence of the BSR coincides with a region of vertical uplift. Therefore, we suggest that gas at this BSR is formed by a dissociation of hydrates at the base of the hydrate stability zone due to uplift and subsequently a decrease in pressure.

A long source area of the 1906 Colombia-Ecuador earthquake estimated from observed tsunami waveforms

NASA Astrophysics Data System (ADS)

Yamanaka, Yusuke; Tanioka, Yuichiro; Shiina, Takahiro

2017-12-01

The 1906 Colombia-Ecuador earthquake induced both strong seismic motions and a tsunami, the most destructive earthquake in the history of the Colombia-Ecuador subduction zone. The tsunami propagated across the Pacific Ocean, and its waveforms were observed at tide gauge stations in countries including Panama, Japan, and the USA. This study conducted slip inverse analysis for the 1906 earthquake using these waveforms. A digital dataset of observed tsunami waveforms at the Naos Island (Panama) and Honolulu (USA) tide gauge stations, where the tsunami was clearly observed, was first produced by consulting documents. Next, the two waveforms were applied in an inverse analysis as the target waveform. The results of this analysis indicated that the moment magnitude of the 1906 earthquake ranged from 8.3 to 8.6. Moreover, the dominant slip occurred in the northern part of the assumed source region near the coast of Colombia, where little significant seismicity has occurred, rather than in the southern part. The results also indicated that the source area, with significant slip, covered a long distance, including the southern, central, and northern parts of the region.[Figure not available: see fulltext.

Approximate solutions of acoustic 3D integral equation and their application to seismic modeling and full-waveform inversion

NASA Astrophysics Data System (ADS)

Malovichko, M.; Khokhlov, N.; Yavich, N.; Zhdanov, M.

2017-10-01

Over the recent decades, a number of fast approximate solutions of Lippmann-Schwinger equation, which are more accurate than classic Born and Rytov approximations, were proposed in the field of electromagnetic modeling. Those developments could be naturally extended to acoustic and elastic fields; however, until recently, they were almost unknown in seismology. This paper presents several solutions of this kind applied to acoustic modeling for both lossy and lossless media. We evaluated the numerical merits of those methods and provide an estimation of their numerical complexity. In our numerical realization we use the matrix-free implementation of the corresponding integral operator. We study the accuracy of those approximate solutions and demonstrate, that the quasi-analytical approximation is more accurate, than the Born approximation. Further, we apply the quasi-analytical approximation to the solution of the inverse problem. It is demonstrated that, this approach improves the estimation of the data gradient, comparing to the Born approximation. The developed inversion algorithm is based on the conjugate-gradient type optimization. Numerical model study demonstrates that the quasi-analytical solution significantly reduces computation time of the seismic full-waveform inversion. We also show how the quasi-analytical approximation can be extended to the case of elastic wavefield.

Effect of surface-related Rayleigh and multiple waves on velocity reconstruction with time-domain elastic FWI

NASA Astrophysics Data System (ADS)

Fang, Jinwei; Zhou, Hui; Zhang, Qingchen; Chen, Hanming; Wang, Ning; Sun, Pengyuan; Wang, Shucheng

2018-01-01

It is critically important to assess the effectiveness of elastic full waveform inversion (FWI) algorithms when FWI is applied to real land seismic data including strong surface and multiple waves related to the air-earth boundary. In this paper, we review the realization of the free surface boundary condition in staggered-grid finite-difference (FD) discretization of elastic wave equation, and analyze the impact of the free surface on FWI results. To reduce inputs/outputs (I/O) operations in gradient calculation, we adopt the boundary value reconstruction method to rebuild the source wavefields during the backward propagation of the residual data. A time-domain multiscale inversion strategy is conducted by using a convolutional objective function, and a multi-GPU parallel programming technique is used to accelerate our elastic FWI further. Forward simulation and elastic FWI examples without and with considering the free surface are shown and analyzed, respectively. Numerical results indicate that no free surface incorporated elastic FWI fails to recover a good inversion result from the Rayleigh wave contaminated observed data. By contrast, when the free surface is incorporated into FWI, the inversion results become better. We also discuss the dependency of the Rayleigh waveform incorporated FWI on the accuracy of initial models, especially the accuracy of the shallow part of the initial models.

Anisotropy effects on 3D waveform inversion

NASA Astrophysics Data System (ADS)

Stekl, I.; Warner, M.; Umpleby, A.

2010-12-01

In the recent years 3D waveform inversion has become achievable procedure for seismic data processing. A number of datasets has been inverted and presented (Warner el al 2008, Ben Hadj at all, Sirgue et all 2010) using isotropic 3D waveform inversion. However the question arises will the results be affected by isotropic assumption. Full-wavefield inversion techniques seek to match field data, wiggle-for-wiggle, to synthetic data generated by a high-resolution model of the sub-surface. In this endeavour, correctly matching the travel times of the principal arrivals is a necessary minimal requirement. In many, perhaps most, long-offset and wide-azimuth datasets, it is necessary to introduce some form of p-wave velocity anisotropy to match the travel times successfully. If this anisotropy is not also incorporated into the wavefield inversion, then results from the inversion will necessarily be compromised. We have incorporated anisotropy into our 3D wavefield tomography codes, characterised as spatially varying transverse isotropy with a tilted axis of symmetry - TTI anisotropy. This enhancement approximately doubles both the run time and the memory requirements of the code. We show that neglect of anisotropy can lead to significant artefacts in the recovered velocity models. We will present inversion results of inverting anisotropic 3D dataset by assuming isotropic earth and compare them with anisotropic inversion result. As a test case Marmousi model extended to 3D with no velocity variation in third direction and with added spatially varying anisotropy is used. Acquisition geometry is assumed as OBC with sources and receivers everywhere at the surface. We attempted inversion using both 2D and full 3D acquisition for this dataset. Results show that if no anisotropy is taken into account although image looks plausible most features are miss positioned in depth and space, even for relatively low anisotropy, which leads to incorrect result. This may lead to misinterpretation of results. However if correct physics is used results agree with correct model. Our algorithm is relatively affordable and runs on standard pc clusters in acceptable time. Refferences: H. Ben Hadj Ali, S. Operto and J. Virieux. Velocity model building by 3D frequency-domain full-waveform inversion of wide-aperture seismic data, Geophysics (Special issue: Velocity Model Building), 73(6), P. VE101-VE117 (2008). L. Sirgue, O.I. Barkved, J. Dellinger, J. Etgen, U. Albertin, J.H. Kommedal, Full waveform inversion: the next leap forward in imaging at Valhall, First Brake April 2010 - Issue 4 - Volume 28 M. Warner, I. Stekl, A. Umpleby, Efficient and Effective 3D Wavefield Tomography, 70th EAGE Conference & Exhibition (2008)

Accumulated energy norm for full waveform inversion of marine data

NASA Astrophysics Data System (ADS)

Shin, Changsoo; Ha, Wansoo

2017-12-01

Macro-velocity models are important for imaging the subsurface structure. However, the conventional objective functions of full waveform inversion in the time and the frequency domain have a limited ability to recover the macro-velocity model because of the absence of low-frequency information. In this study, we propose new objective functions that can recover the macro-velocity model by minimizing the difference between the zero-frequency components of the square of seismic traces. Instead of the seismic trace itself, we use the square of the trace, which contains low-frequency information. We apply several time windows to the trace and obtain zero-frequency information of the squared trace for each time window. The shape of the new objective functions shows that they are suitable for local optimization methods. Since we use the acoustic wave equation in this study, this method can be used for deep-sea marine data, in which elastic effects can be ignored. We show that the zero-frequency components of the square of the seismic traces can be used to recover macro-velocities from synthetic and field data.

Lithospheric structure of the Western Alps as seen by full-waveform inversion of CIFALPS teleseismic data

NASA Astrophysics Data System (ADS)

Beller, Stephen; Monteiller, Vadim; Operto, Stéphane; Nolet, Guust; Paul, Anne; Zhao, Liang

2017-04-01

Full-waveform inversion (FWI) is a powerful but constitutionally intensive technique that aims to recover 3D multiparameter images of the subsurface by minimising the waveform difference between the full recorded and modelled seismograms. This method has recently been adapted and successfully applied in lithospheric settings by tackling teleseismic waveform modelling with hybrid methods. For each event, a global scale simulation is performed once and for all to store the wavefield solutions on the edges of the lithospheric target. Then, for each modelling involved in the FWI process, these global scale solutions are injected within the lithospheric medium from the boundaries. We present the results of the application of teleseismic FWI to the data acquired by the CIFALPS experiment that was conducted in the Western Alps to gain new insights its lithospheric structure and geodynamic evolution of the alpine range. Nine teleseismic events were inverted to infer 3D models of density, P-wave velocity and S-wave velocity of the crust and the upper-mantle down to 200 km depth. Our models show clear evidences of continental subduction during the alpine orogeny. They outline a dipping European Moho down to 75 km depth and finely delineate the geometry of the Ivrea body at the suture between European and Adriatic plates. Deeper, in the mantle a slow S-wave velocity anomaly might indicate the location of the European slab detachment. Overall, FWI models give access to new seismic images that fill the resolution gap between smooth tomographic model and sharp receiver function images of the lithosphere and enable integrated interpretations of crustal and upper-mantle structures.

Adjoint tomography and centroid-moment tensor inversion of the Kanto region, Japan

NASA Astrophysics Data System (ADS)

Miyoshi, T.

2017-12-01

A three-dimensional seismic wave speed model in the Kanto region of Japan was developed using adjoint tomography based on large computing. Starting with a model based on previous travel time tomographic results, we inverted the waveforms obtained at seismic broadband stations from 140 local earthquakes in the Kanto region to obtain the P- and S-wave speeds Vp and Vs. The synthetic displacements were calculated using the spectral element method (SEM; e.g. Komatitsch and Tromp 1999; Peter et al. 2011) in which the Kanto region was parameterized using 16 million grid points. The model parameters Vp and Vs were updated iteratively by Newton's method using the misfit and Hessian kernels until the misfit between the observed and synthetic waveforms was minimized. The proposed model reveals several anomalous areas with extremely low Vs values in comparison with those of the initial model. The synthetic waveforms obtained using the newly proposed model for the selected earthquakes show better fit than the initial model to the observed waveforms in different period ranges within 5-30 s. In the present study, all centroid times of the source solutions were determined using time shifts based on cross correlation to prevent high computing resources before the structural inversion. Additionally, parameters of centroid-moment solutions were fully determined using the SEM assuming the 3D structure (e.g. Liu et al. 2004). As a preliminary result, new solutions were basically same as their initial solutions. This may indicate that the 3D structure is not effective for the source estimation. Acknowledgements: This study was supported by JSPS KAKENHI Grant Number 16K21699.

Detection of sinkholes or anomalies using full seismic wave fields.

DOT National Transportation Integrated Search

2013-04-01

This research presents an application of two-dimensional (2-D) time-domain waveform tomography for detection of embedded sinkholes and anomalies. The measured seismic surface wave fields were inverted using a full waveform inversion (FWI) technique, ...

A developed nearly analytic discrete method for forward modeling in the frequency domain

NASA Astrophysics Data System (ADS)

Liu, Shaolin; Lang, Chao; Yang, Hui; Wang, Wenshuai

2018-02-01

High-efficiency forward modeling methods play a fundamental role in full waveform inversion (FWI). In this paper, the developed nearly analytic discrete (DNAD) method is proposed to accelerate frequency-domain forward modeling processes. We first derive the discretization of frequency-domain wave equations via numerical schemes based on the nearly analytic discrete (NAD) method to obtain a linear system. The coefficients of numerical stencils are optimized to make the linear system easier to solve and to minimize computing time. Wavefield simulation and numerical dispersion analysis are performed to compare the numerical behavior of DNAD method with that of the conventional NAD method. The results demonstrate the superiority of our proposed method. Finally, the DNAD method is implemented in frequency-domain FWI, and high-resolution inverse results are obtained.

Simulations of Ground Motion in Southern California based upon the Spectral-Element Method

NASA Astrophysics Data System (ADS)

Tromp, J.; Komatitsch, D.; Liu, Q.

2003-12-01

We use the spectral-element method to simulate ground motion generated by recent well-recorded small earthquakes in Southern California. Simulations are performed using a new sedimentary basin model that is constrained by hundreds of petroleum industry well logs and more than twenty thousand kilometers of seismic reflection profiles. The numerical simulations account for 3D variations of seismic wave speeds and density, topography and bathymetry, and attenuation. Simulations for several small recent events demonstrate that the combination of a detailed sedimentary basin model and an accurate numerical technique facilitates the simulation of ground motion at periods of 2 seconds and longer inside the Los Angeles basin and 6 seconds and longer elsewhere. Peak ground displacement, velocity and acceleration maps illustrate that significant amplification occurs in the basin. Centroid-Moment Tensor mechanisms are obtained based upon Pnl and surface waveforms and numerically calculated 3D Frechet derivatives. We use a combination of waveform and waveform-envelope misfit criteria, and facilitate pure double-couple or zero-trace moment-tensor inversions.

Joint Inversion of 1-Hz GPS Data and Strong Motion Records for the Rupture Process of the 2008 Iwate-Miyagi Nairiku Earthquake: Objectively Determining Relative Weighting

NASA Astrophysics Data System (ADS)

Wang, Z.; Kato, T.; Wang, Y.

2015-12-01

The spatiotemporal fault slip history of the 2008 Iwate-Miyagi Nairiku earthquake, Japan, is obtained by the joint inversion of 1-Hz GPS waveforms and near-field strong motion records. 1-Hz GPS data from GEONET is processed by GAMIT/GLOBK and then a low-pass filter of 0.05 Hz is applied. The ground surface strong motion records from stations of K-NET and Kik-Net are band-pass filtered for the range of 0.05 ~ 0.3 Hz and integrated once to obtain velocity. The joint inversion exploits a broader frequency band for near-field ground motions, which provides excellent constraints for both the detailed slip history and slip distribution. A fully Bayesian inversion method is performed to simultaneously and objectively determine the rupture model, the unknown relative weighting of multiple data sets and the unknown smoothing hyperparameters. The preferred rupture model is stable for different choices of velocity structure model and station distribution, with maximum slip of ~ 8.0 m and seismic moment of 2.9 × 1019 Nm (Mw 6.9). By comparison with the single inversion of strong motion records, the cumulative slip distribution of joint inversion shows sparser slip distribution with two slip asperities. One common slip asperity extends from the hypocenter southeastward to the ground surface of breakage; another slip asperity, which is unique for joint inversion contributed by 1-Hz GPS waveforms, appears in the deep part of fault where very few aftershocks are occurring. The differential moment rate function of joint and single inversions obviously indicates that rich high frequency waves are radiated in the first three seconds but few low frequency waves.

Nature and Role of Subducting Sediments on the Megathrust and Forearc Evolution in the 2004 Great Sumatra Earthquake Rupture Zone: Results from Full Waveform Inversion of Long Offset Seismic Data

NASA Astrophysics Data System (ADS)

Singh, S. C.; Qin, Y.

2015-12-01

On active accretionary margins, the nature of incoming sediments defines the locking mechanism on the megathrust, and the development and evolution of the accretionary wedge. Drilling is the most direct method to characterise the nature of these sediments, but the drilling is very expensive, and provide information at only a few locations. In north Sumatra, an IODP drilling is programmed to take place in July-August 2016. We have performed seismic full waveform inversion of 12 km long offset seismic reflection data acquired by WesternGeco in 2006 over a 35 km zone near the subduction front in the 2004 earthquake rupture zone area that provide detailed quantitative information on the characteristics of the incoming sediments. We first downward continue the surface streamer data to the seafloor, which removes the effect of deep water (~5 km) and brings out the refraction arrivals as the first arrivals. We carry out travel time tomography, and then performed full waveform inversion of seismic refraction data followed by the full waveform inversion of reflection data providing detailed (10-20 m) velocity structure. The sediments in this area are 3-5 km thick where the P-wave velocity increases from 1.6 km/s near the seafloor to more than 4.5 km/s above the oceanic crust. The high velocity of sediments above the basement suggests that the sediments are highly compacted, strengthened the coupling near the subduction front, which might have been responsible for 2004 earthquake rupture propagation up to the subduction front, enhancing the tsunami. We also find several thin velocity layers within the sediments, which might be due to high pore-pressure fluid or free gas. These layers might be responsible for the formation of pseudo-decollement within the forearc sediments that acts as a conveyer belt between highly compacted subducting lower sediments and accreted sediments above. The presence of well intact sediments on the accretionary prism supports this interpretation. Our results provide first hand information about the sediments properties, which will be ground toothed by drilling.

Receiver function HV ratio: a new measurement for reducing non-uniqueness of receiver function waveform inversion

NASA Astrophysics Data System (ADS)

Chong, Jiajun; Chu, Risheng; Ni, Sidao; Meng, Qingjun; Guo, Aizhi

2018-02-01

It is known that a receiver function has relatively weak constraint on absolute seismic wave velocity, and that joint inversion of the receiver function with surface wave dispersion has been widely applied to reduce the trade-off of velocity with interface depth. However, some studies indicate that the receiver function itself is capable for determining the absolute shear-wave velocity. In this study, we propose to measure the receiver function HV ratio which takes advantage of the amplitude information of the receiver function to constrain the shear-wave velocity. Numerical analysis indicates that the receiver function HV ratio is sensitive to the average shear-wave velocity in the depth range it samples, and can help to reduce the non-uniqueness of receiver function waveform inversion. A joint inversion scheme has been developed, and both synthetic tests and real data application proved the feasibility of the joint inversion.

Improved source inversion from joint measurements of translational and rotational ground motions

NASA Astrophysics Data System (ADS)

Donner, S.; Bernauer, M.; Reinwald, M.; Hadziioannou, C.; Igel, H.

2017-12-01

Waveform inversion for seismic point (moment tensor) and kinematic sources is a standard procedure. However, especially in the local and regional distances a lack of appropriate velocity models, the sparsity of station networks, or a low signal-to-noise ratio combined with more complex waveforms hamper the successful retrieval of reliable source solutions. We assess the potential of rotational ground motion recordings to increase the resolution power and reduce non-uniquenesses for point and kinematic source solutions. Based on synthetic waveform data, we perform a Bayesian (i.e. probabilistic) inversion. Thus, we avoid the subjective selection of the most reliable solution according the lowest misfit or other constructed criterion. In addition, we obtain unbiased measures of resolution and possible trade-offs. Testing different earthquake mechanisms and scenarios, we can show that the resolution of the source solutions can be improved significantly. Especially depth dependent components show significant improvement. Next to synthetic data of station networks, we also tested sparse-network and single station cases.

Solving constrained inverse problems for waveform tomography with Salvus

NASA Astrophysics Data System (ADS)

Boehm, C.; Afanasiev, M.; van Driel, M.; Krischer, L.; May, D.; Rietmann, M.; Fichtner, A.

2016-12-01

Finding a good balance between flexibility and performance is often difficult within domain-specific software projects. To achieve this balance, we introduce Salvus: an open-source high-order finite element package built upon PETSc and Eigen, that focuses on large-scale full-waveform modeling and inversion. One of the key features of Salvus is its modular design, based on C++ mixins, that separates the physical equations from the numerical discretization and the mathematical optimization. In this presentation we focus on solving inverse problems with Salvus and discuss (i) dealing with inexact derivatives resulting, e.g., from lossy wavefield compression, (ii) imposing additional constraints on the model parameters, e.g., from effective medium theory, and (iii) integration with a workflow management tool. We present a feasible-point trust-region method for PDE-constrained inverse problems that can handle inexactly computed derivatives. The level of accuracy in the approximate derivatives is controlled by localized error estimates to ensure global convergence of the method. Additional constraints on the model parameters are typically cheap to compute without the need for further simulations. Hence, including them in the trust-region subproblem introduces only a small computational overhead, but ensures feasibility of the model in every iteration. We show examples with homogenization constraints derived from effective medium theory (i.e. all fine-scale updates must upscale to a physically meaningful long-wavelength model). Salvus has a built-in workflow management framework to automate the inversion with interfaces to user-defined misfit functionals and data structures. This significantly reduces the amount of manual user interaction and enhances reproducibility which we demonstrate for several applications from the laboratory to global scale.

Detailed Velocity and Density models of the Cascadia Subduction Zone from Prestack Full-Waveform Inversion

NASA Astrophysics Data System (ADS)

Fortin, W.; Holbrook, W. S.; Mallick, S.; Everson, E. D.; Tobin, H. J.; Keranen, K. M.

2014-12-01

Understanding the geologic composition of the Cascadia Subduction Zone (CSZ) is critically important in assessing seismic hazards in the Pacific Northwest. Despite being a potential earthquake and tsunami threat to millions of people, key details of the structure and fault mechanisms remain poorly understood in the CSZ. In particular, the position and character of the subduction interface remains elusive due to its relative aseismicity and low seismic reflectivity, making imaging difficult for both passive and active source methods. Modern active-source reflection seismic data acquired as part of the COAST project in 2012 provide an opportunity to study the transition from the Cascadia basin, across the deformation front, and into the accretionary prism. Coupled with advances in seismic inversion methods, this new data allow us to produce detailed velocity models of the CSZ and accurate pre-stack depth migrations for studying geologic structure. While still computationally expensive, current computing clusters can perform seismic inversions at resolutions that match that of the seismic image itself. Here we present pre-stack full waveform inversions of the central seismic line of the COAST survey offshore Washington state. The resultant velocity model is produced by inversion at every CMP location, 6.25 m laterally, with vertical resolution of 0.2 times the dominant seismic frequency. We report a good average correlation value above 0.8 across the entire seismic line, determined by comparing synthetic gathers to the real pre-stack gathers. These detailed velocity models, both Vp and Vs, along with the density model, are a necessary step toward a detailed porosity cross section to be used to determine the role of fluids in the CSZ. Additionally, the P-velocity model is used to produce a pre-stack depth migration image of the CSZ.

Trans-dimensional and hierarchical Bayesian approaches toward rigorous estimation of seismic sources and structures in the Northeast Asia

NASA Astrophysics Data System (ADS)

Kim, Seongryong; Tkalčić, Hrvoje; Mustać, Marija; Rhie, Junkee; Ford, Sean

2016-04-01

A framework is presented within which we provide rigorous estimations for seismic sources and structures in the Northeast Asia. We use Bayesian inversion methods, which enable statistical estimations of models and their uncertainties based on data information. Ambiguities in error statistics and model parameterizations are addressed by hierarchical and trans-dimensional (trans-D) techniques, which can be inherently implemented in the Bayesian inversions. Hence reliable estimation of model parameters and their uncertainties is possible, thus avoiding arbitrary regularizations and parameterizations. Hierarchical and trans-D inversions are performed to develop a three-dimensional velocity model using ambient noise data. To further improve the model, we perform joint inversions with receiver function data using a newly developed Bayesian method. For the source estimation, a novel moment tensor inversion method is presented and applied to regional waveform data of the North Korean nuclear explosion tests. By the combination of new Bayesian techniques and the structural model, coupled with meaningful uncertainties related to each of the processes, more quantitative monitoring and discrimination of seismic events is possible.

Joint inversion of regional and teleseismic earthquake waveforms

NASA Astrophysics Data System (ADS)

Baker, Mark R.; Doser, Diane I.

1988-03-01

A least squares joint inversion technique for regional and teleseismic waveforms is presented. The mean square error between seismograms and synthetics is minimized using true amplitudes. Matching true amplitudes in modeling requires meaningful estimates of modeling uncertainties and of seismogram signal-to-noise ratios. This also permits calculating linearized uncertainties on the solution based on accuracy and resolution. We use a priori estimates of earthquake parameters to stabilize unresolved parameters, and for comparison with a posteriori uncertainties. We verify the technique on synthetic data, and on the 1983 Borah Peak, Idaho (M = 7.3), earthquake. We demonstrate the inversion on the August 1954 Rainbow Mountain, Nevada (M = 6.8), earthquake and find parameters consistent with previous studies.

Inversion of ground-motion data from a seismometer array for rotation using a modification of Jaeger's method

USGS Publications Warehouse

Chi, Wu-Cheng; Lee, W.H.K.; Aston, J.A.D.; Lin, C.J.; Liu, C.-C.

2011-01-01

We develop a new way to invert 2D translational waveforms using Jaeger's (1969) formula to derive rotational ground motions about one axis and estimate the errors in them using techniques from statistical multivariate analysis. This procedure can be used to derive rotational ground motions and strains using arrayed translational data, thus providing an efficient way to calibrate the performance of rotational sensors. This approach does not require a priori information about the noise level of the translational data and elastic properties of the media. This new procedure also provides estimates of the standard deviations of the derived rotations and strains. In this study, we validated this code using synthetic translational waveforms from a seismic array. The results after the inversion of the synthetics for rotations were almost identical with the results derived using a well-tested inversion procedure by Spudich and Fletcher (2009). This new 2D procedure can be applied three times to obtain the full, three-component rotations. Additional modifications can be implemented to the code in the future to study different features of the rotational ground motions and strains induced by the passage of seismic waves.

Duration of Tsunami Generation Longer than Duration of Seismic Wave Generation in the 2011 Mw 9.0 Tohoku-Oki Earthquake

NASA Astrophysics Data System (ADS)

Fujihara, S.; Korenaga, M.; Kawaji, K.; Akiyama, S.

2013-12-01

We try to compare and evaluate the nature of tsunami generation and seismic wave generation in occurrence of the 2011 Tohoku-Oki earthquake (hereafter, called as TOH11), in terms of two type of moment rate functions, inferred from finite source imaging of tsunami waveforms and seismic waveforms. Since 1970's, the nature of "tsunami earthquakes" has been discussed in many researches (e.g. Kanamori, 1972; Kanamori and Kikuchi, 1993; Kikuchi and Kanamori, 1995; Ide et al., 1993; Satake, 1994) mostly based on analysis of seismic waveform data , in terms of the "slow" nature of tsunami earthquakes (e.g., the 1992 Nicaragura earthquake). Although TOH11 is not necessarily understood as a tsunami earthquake, TOH11 is one of historical earthquakes that simultaneously generated large seismic waves and tsunami. Also, TOH11 is one of earthquakes which was observed both by seismic observation network and tsunami observation network around the Japanese islands. Therefore, for the purpose of analyzing the nature of tsunami generation, we try to utilize tsunami waveform data as much as possible. In our previous studies of TOH11 (Fujihara et al., 2012a; Fujihara et al., 2012b), we inverted tsunami waveforms at GPS wave gauges of NOWPHAS to image the spatio-temporal slip distribution. The "temporal" nature of our tsunami source model is generally consistent with the other tsunami source models (e.g., Satake et al, 2013). For seismic waveform inversion based on 1-D structure, here we inverted broadband seismograms at GSN stations based on the teleseismic body-wave inversion scheme (Kikuchi and Kanamori, 2003). Also, for seismic waveform inversion considering the inhomogeneous internal structure, we inverted strong motion seismograms at K-NET and KiK-net stations, based on 3-D Green's functions (Fujihara et al., 2013a; Fujihara et al., 2013b). The gross "temporal" nature of our seismic source models are generally consistent with the other seismic source models (e.g., Yoshida et al., 2011; Ide at al., 2011; Yagi and Fukahata, 2011; Suzuki et al., 2011). The comparison of two type of moment rate functions, inferred from finite source imaging of tsunami waveforms and seismic waveforms, suggested that there was the time period common to both seismic wave generation and tsunami generation followed by the time period unique to tsunami generation. At this point, we think that comparison of the absolute values of moment rates is not so meaningful between tsunami waveform inversion and seismic waveform inversion, because of general ambiguity of rigidity values of each subfault in the fault region (assuming the rigidity value of 30 GPa of Yoshida et al (2011)). Considering this, the normalized value of moment rate function was also evaluated and it does not change the general feature of two moment rate functions in terms of duration property. Furthermore, the results suggested that tsunami generation process apparently took more time than seismic wave generation process did. Tsunami can be generated even by "extra" motions resulting from many suggested abnormal mechanisms. These extra motions may be attribute to the relatively larger-scale tsunami generation than expected from the magnitude level from seismic ground motion, and attribute to the longer duration of tsunami generation process.

Intelligent earthquake data processing for global adjoint tomography

NASA Astrophysics Data System (ADS)

Chen, Y.; Hill, J.; Li, T.; Lei, W.; Ruan, Y.; Lefebvre, M. P.; Tromp, J.

2016-12-01

Due to the increased computational capability afforded by modern and future computing architectures, the seismology community is demanding a more comprehensive understanding of the full waveform information from the recorded earthquake seismograms. Global waveform tomography is a complex workflow that matches observed seismic data with synthesized seismograms by iteratively updating the earth model parameters based on the adjoint state method. This methodology allows us to compute a very accurate model of the earth's interior. The synthetic data is simulated by solving the wave equation in the entire globe using a spectral-element method. In order to ensure the inversion accuracy and stability, both the synthesized and observed seismograms must be carefully pre-processed. Because the scale of the inversion problem is extremely large and there is a very large volume of data to both be read and written, an efficient and reliable pre-processing workflow must be developed. We are investigating intelligent algorithms based on a machine-learning (ML) framework that will automatically tune parameters for the data processing chain. One straightforward application of ML in data processing is to classify all possible misfit calculation windows into usable and unusable ones, based on some intelligent ML models such as neural network, support vector machine or principle component analysis. The intelligent earthquake data processing framework will enable the seismology community to compute the global waveform tomography using seismic data from an arbitrarily large number of earthquake events in the fastest, most efficient way.

Transdimensional inversion of scattered body waves for 1D S-wave velocity structure - Application to the Tengchong volcanic area, Southwestern China

NASA Astrophysics Data System (ADS)

Li, Mengkui; Zhang, Shuangxi; Bodin, Thomas; Lin, Xu; Wu, Tengfei

2018-06-01

Inversion of receiver functions is commonly used to recover the S-wave velocity structure beneath seismic stations. Traditional approaches are based on deconvolved waveforms, where the horizontal component of P-wave seismograms is deconvolved by the vertical component. Deconvolution of noisy seismograms is a numerically unstable process that needs to be stabilized by regularization parameters. This biases noise statistics, making it difficult to estimate uncertainties in observed receiver functions for Bayesian inference. This study proposes a method to directly invert observed radial waveforms and to better account for data noise in a Bayesian formulation. We illustrate its feasibility with two synthetic tests having different types of noises added to seismograms. Then, a real site application is performed to obtain the 1-D S-wave velocity structure beneath a seismic station located in the Tengchong volcanic area, Southwestern China. Surface wave dispersion measurements spanning periods from 8 to 65 s are jointly inverted with P waveforms. The results show a complex S-wave velocity structure, as two low velocity zones are observed in the crust and uppermost mantle, suggesting the existence of magma chambers, or zones of partial melt. The upper magma chambers may be the heart source that cause the thermal activity on the surface.

Towards full waveform ambient noise inversion

NASA Astrophysics Data System (ADS)

Sager, Korbinian; Ermert, Laura; Boehm, Christian; Fichtner, Andreas

2018-01-01

In this work we investigate fundamentals of a method—referred to as full waveform ambient noise inversion—that improves the resolution of tomographic images by extracting waveform information from interstation correlation functions that cannot be used without knowing the distribution of noise sources. The fundamental idea is to drop the principle of Green function retrieval and to establish correlation functions as self-consistent observables in seismology. This involves the following steps: (1) We introduce an operator-based formulation of the forward problem of computing correlation functions. It is valid for arbitrary distributions of noise sources in both space and frequency, and for any type of medium, including 3-D elastic, heterogeneous and attenuating media. In addition, the formulation allows us to keep the derivations independent of time and frequency domain and it facilitates the application of adjoint techniques, which we use to derive efficient expressions to compute first and also second derivatives. The latter are essential for a resolution analysis that accounts for intra- and interparameter trade-offs. (2) In a forward modelling study we investigate the effect of noise sources and structure on different observables. Traveltimes are hardly affected by heterogeneous noise source distributions. On the other hand, the amplitude asymmetry of correlations is at least to first order insensitive to unmodelled Earth structure. Energy and waveform differences are sensitive to both structure and the distribution of noise sources. (3) We design and implement an appropriate inversion scheme, where the extraction of waveform information is successively increased. We demonstrate that full waveform ambient noise inversion has the potential to go beyond ambient noise tomography based on Green function retrieval and to refine noise source location, which is essential for a better understanding of noise generation. Inherent trade-offs between source and structure are quantified using Hessian-vector products.

The effects of core-reflected waves on finite fault inversions with teleseismic body wave data

NASA Astrophysics Data System (ADS)

Qian, Yunyi; Ni, Sidao; Wei, Shengji; Almeida, Rafael; Zhang, Han

2017-11-01

Teleseismic body waves are essential for imaging rupture processes of large earthquakes. Earthquake source parameters are usually characterized by waveform analyses such as finite fault inversions using only turning (direct) P and SH waves without considering the reflected phases from the core-mantle boundary (CMB). However, core-reflected waves such as ScS usually have amplitudes comparable to direct S waves due to the total reflection from the CMB and might interfere with the S waves used for inversion, especially at large epicentral distances for long duration earthquakes. In order to understand how core-reflected waves affect teleseismic body wave inversion results, we develop a procedure named Multitel3 to compute Green's functions that contain turning waves (direct P, pP, sP, direct S, sS and reverberations in the crust) and core-reflected waves (PcP, pPcP, sPcP, ScS, sScS and associated reflected phases from the CMB). This ray-based method can efficiently generate synthetic seismograms for turning and core-reflected waves independently, with the flexibility to take into account the 3-D Earth structure effect on the timing between these phases. The performance of this approach is assessed through a series of numerical inversion tests on synthetic waveforms of the 2008 Mw7.9 Wenchuan earthquake and the 2015 Mw7.8 Nepal earthquake. We also compare this improved method with the turning-wave only inversions and explore the stability of the new procedure when there are uncertainties in a priori information (such as fault geometry and epicentre location) or arrival time of core-reflected phases. Finally, a finite fault inversion of the 2005 Mw8.7 Nias-Simeulue earthquake is carried out using the improved Green's functions. Using enhanced Green's functions yields better inversion results as expected. While the finite source inversion with conventional P and SH waves is able to recover large-scale characteristics of the earthquake source, by adding PcP and ScS phases, the inverted slip model and moment rate function better match previous results incorporating field observations, geodetic and seismic data.

Multi-scale Slip Inversion Based on Simultaneous Spatial and Temporal Domain Wavelet Transform

NASA Astrophysics Data System (ADS)

Liu, W.; Yao, H.; Yang, H. Y.

2017-12-01

Finite fault inversion is a widely used method to study earthquake rupture processes. Some previous studies have proposed different methods to implement finite fault inversion, including time-domain, frequency-domain, and wavelet-domain methods. Many previous studies have found that different frequency bands show different characteristics of the seismic rupture (e.g., Wang and Mori, 2011; Yao et al., 2011, 2013; Uchide et al., 2013; Yin et al., 2017). Generally, lower frequency waveforms correspond to larger-scale rupture characteristics while higher frequency data are representative of smaller-scale ones. Therefore, multi-scale analysis can help us understand the earthquake rupture process thoroughly from larger scale to smaller scale. By the use of wavelet transform, the wavelet-domain methods can analyze both the time and frequency information of signals in different scales. Traditional wavelet-domain methods (e.g., Ji et al., 2002) implement finite fault inversion with both lower and higher frequency signals together to recover larger-scale and smaller-scale characteristics of the rupture process simultaneously. Here we propose an alternative strategy with a two-step procedure, i.e., firstly constraining the larger-scale characteristics with lower frequency signals, and then resolving the smaller-scale ones with higher frequency signals. We have designed some synthetic tests to testify our strategy and compare it with the traditional one. We also have applied our strategy to study the 2015 Gorkha Nepal earthquake using tele-seismic waveforms. Both the traditional method and our two-step strategy only analyze the data in different temporal scales (i.e., different frequency bands), while the spatial distribution of model parameters also shows multi-scale characteristics. A more sophisticated strategy is to transfer the slip model into different spatial scales, and then analyze the smooth slip distribution (larger scales) with lower frequency data firstly and more detailed slip distribution (smaller scales) with higher frequency data subsequently. We are now implementing the slip inversion using both spatial and temporal domain wavelets. This multi-scale analysis can help us better understand frequency-dependent rupture characteristics of large earthquakes.

Seismic waveform inversion using neural networks

NASA Astrophysics Data System (ADS)

De Wit, R. W.; Trampert, J.

2012-12-01

Full waveform tomography aims to extract all available information on Earth structure and seismic sources from seismograms. The strongly non-linear nature of this inverse problem is often addressed through simplifying assumptions for the physical theory or data selection, thus potentially neglecting valuable information. Furthermore, the assessment of the quality of the inferred model is often lacking. This calls for the development of methods that fully appreciate the non-linear nature of the inverse problem, whilst providing a quantification of the uncertainties in the final model. We propose to invert seismic waveforms in a fully non-linear way by using artificial neural networks. Neural networks can be viewed as powerful and flexible non-linear filters. They are very common in speech, handwriting and pattern recognition. Mixture Density Networks (MDN) allow us to obtain marginal posterior probability density functions (pdfs) of all model parameters, conditioned on the data. An MDN can approximate an arbitrary conditional pdf as a linear combination of Gaussian kernels. Seismograms serve as input, Earth structure parameters are the so-called targets and network training aims to learn the relationship between input and targets. The network is trained on a large synthetic data set, which we construct by drawing many random Earth models from a prior model pdf and solving the forward problem for each of these models, thus generating synthetic seismograms. As a first step, we aim to construct a 1D Earth model. Training sets are constructed using the Mineos package, which computes synthetic seismograms in a spherically symmetric non-rotating Earth by summing normal modes. We train a network on the body waveforms present in these seismograms. Once the network has been trained, it can be presented with new unseen input data, in our case the body waves in real seismograms. We thus obtain the posterior pdf which represents our final state of knowledge given the information in the training set and the real data.

Seismic tomography of the southern California crust based on spectral-element and adjoint methods

NASA Astrophysics Data System (ADS)

Tape, Carl; Liu, Qinya; Maggi, Alessia; Tromp, Jeroen

2010-01-01

We iteratively improve a 3-D tomographic model of the southern California crust using numerical simulations of seismic wave propagation based on a spectral-element method (SEM) in combination with an adjoint method. The initial 3-D model is provided by the Southern California Earthquake Center. The data set comprises three-component seismic waveforms (i.e. both body and surface waves), filtered over the period range 2-30 s, from 143 local earthquakes recorded by a network of 203 stations. Time windows for measurements are automatically selected by the FLEXWIN algorithm. The misfit function in the tomographic inversion is based on frequency-dependent multitaper traveltime differences. The gradient of the misfit function and related finite-frequency sensitivity kernels for each earthquake are computed using an adjoint technique. The kernels are combined using a source subspace projection method to compute a model update at each iteration of a gradient-based minimization algorithm. The inversion involved 16 iterations, which required 6800 wavefield simulations. The new crustal model, m16, is described in terms of independent shear (VS) and bulk-sound (VB) wave speed variations. It exhibits strong heterogeneity, including local changes of +/-30 per cent with respect to the initial 3-D model. The model reveals several features that relate to geological observations, such as sedimentary basins, exhumed batholiths, and contrasting lithologies across faults. The quality of the new model is validated by quantifying waveform misfits of full-length seismograms from 91 earthquakes that were not used in the tomographic inversion. The new model provides more accurate synthetic seismograms that will benefit seismic hazard assessment.

Improved Phase Corrections for Transoceanic Tsunami Data in Spatial and Temporal Source Estimation: Application to the 2011 Tohoku Earthquake

NASA Astrophysics Data System (ADS)

Ho, Tung-Cheng; Satake, Kenji; Watada, Shingo

2017-12-01

Systemic travel time delays of up to 15 min relative to the linear long waves for transoceanic tsunamis have been reported. A phase correction method, which converts the linear long waves into dispersive waves, was previously proposed to consider seawater compressibility, the elasticity of the Earth, and gravitational potential change associated with tsunami motion. In the present study, we improved this method by incorporating the effects of ocean density stratification, actual tsunami raypath, and actual bathymetry. The previously considered effects amounted to approximately 74% for correction of the travel time delay, while the ocean density stratification, actual raypath, and actual bathymetry, contributed to approximately 13%, 4%, and 9% on average, respectively. The improved phase correction method accounted for almost all the travel time delay at far-field stations. We performed single and multiple time window inversions for the 2011 Tohoku tsunami using the far-field data (>3 h travel time) to investigate the initial sea surface displacement. The inversion result from only far-field data was similar to but smoother than that from near-field data and all stations, including a large sea surface rise increasing toward the trench followed by a migration northward along the trench. For the forward simulation, our results showed good agreement between the observed and computed waveforms at both near-field and far-field tsunami gauges, as well as with satellite altimeter data. The present study demonstrates that the improved method provides a more accurate estimate for the waveform inversion and forward prediction of far-field data.

Period-dependent source rupture behavior of the 2011 Tohoku earthquake estimated by multi period-band Bayesian waveform inversion

NASA Astrophysics Data System (ADS)

Kubo, H.; Asano, K.; Iwata, T.; Aoi, S.

2014-12-01

Previous studies for the period-dependent source characteristics of the 2011 Tohoku earthquake (e.g., Koper et al., 2011; Lay et al., 2012) were based on the short and long period source models using different method. Kubo et al. (2013) obtained source models of the 2011 Tohoku earthquake using multi period-bands waveform data by a common inversion method and discussed its period-dependent source characteristics. In this study, to achieve more in detail spatiotemporal source rupture behavior of this event, we introduce a new fault surface model having finer sub-fault size and estimate the source models in multi period-bands using a Bayesian inversion method combined with a multi-time-window method. Three components of velocity waveforms at 25 stations of K-NET, KiK-net, and F-net of NIED are used in this analysis. The target period band is 10-100 s. We divide this period band into three period bands (10-25 s, 25-50 s, and 50-100 s) and estimate a kinematic source model in each period band using a Bayesian inversion method with MCMC sampling (e.g., Fukuda & Johnson, 2008; Minson et al., 2013, 2014). The parameterization of spatiotemporal slip distribution follows the multi-time-window method (Hartzell & Heaton, 1983). The Green's functions are calculated by the 3D FDM (GMS; Aoi & Fujiwara, 1999) using a 3D velocity structure model (JIVSM; Koketsu et al., 2012). The assumed fault surface model is based on the Pacific plate boundary of JIVSM and is divided into 384 subfaults of about 16 * 16 km^2. The estimated source models in multi period-bands show the following source image: (1) First deep rupture off Miyagi at 0-60 s toward down-dip mostly radiating relatively short period (10-25 s) seismic waves. (2) Shallow rupture off Miyagi at 45-90 s toward up-dip with long duration radiating long period (50-100 s) seismic wave. (3) Second deep rupture off Miyagi at 60-105 s toward down-dip radiating longer period seismic waves then that of the first deep rupture. (4) Deep rupture off Fukushima at 90-135 s. The dominant-period difference of the seismic-wave radiation between two deep ruptures off Miyagi may result from the mechanism that small-scale heterogeneities on the fault are removed by the first rupture. This difference can be also interpreted by the concept of multi-scale dynamic rupture (Ide & Aochi, 2005).

Full waveform inversion using envelope-based global correlation norm

NASA Astrophysics Data System (ADS)

Oh, Ju-Won; Alkhalifah, Tariq

2018-05-01

To increase the feasibility of full waveform inversion on real data, we suggest a new objective function, which is defined as the global correlation of the envelopes of modelled and observed data. The envelope-based global correlation norm has the advantage of the envelope inversion that generates artificial low-frequency information, which provides the possibility to recover long-wavelength structure in an early stage. In addition, the envelope-based global correlation norm maintains the advantage of the global correlation norm, which reduces the sensitivity of the misfit to amplitude errors so that the performance of inversion on real data can be enhanced when the exact source wavelet is not available and more complex physics are ignored. Through the synthetic example for 2-D SEG/EAGE overthrust model with inaccurate source wavelet, we compare the performance of four different approaches, which are the least-squares waveform inversion, least-squares envelope inversion, global correlation norm and envelope-based global correlation norm. Finally, we apply the envelope-based global correlation norm on the 3-D Ocean Bottom Cable (OBC) data from the North Sea. The envelope-based global correlation norm captures the strong reflections from the high-velocity caprock and generates artificial low-frequency reflection energy that helps us recover long-wavelength structure of the model domain in the early stages. From this long-wavelength model, the conventional global correlation norm is sequentially applied to invert for higher-resolution features of the model.

The Earthquake‐Source Inversion Validation (SIV) Project

USGS Publications Warehouse

Mai, P. Martin; Schorlemmer, Danijel; Page, Morgan T.; Ampuero, Jean-Paul; Asano, Kimiyuki; Causse, Mathieu; Custodio, Susana; Fan, Wenyuan; Festa, Gaetano; Galis, Martin; Gallovic, Frantisek; Imperatori, Walter; Käser, Martin; Malytskyy, Dmytro; Okuwaki, Ryo; Pollitz, Fred; Passone, Luca; Razafindrakoto, Hoby N. T.; Sekiguchi, Haruko; Song, Seok Goo; Somala, Surendra N.; Thingbaijam, Kiran K. S.; Twardzik, Cedric; van Driel, Martin; Vyas, Jagdish C.; Wang, Rongjiang; Yagi, Yuji; Zielke, Olaf

2016-01-01

Finite‐fault earthquake source inversions infer the (time‐dependent) displacement on the rupture surface from geophysical data. The resulting earthquake source models document the complexity of the rupture process. However, multiple source models for the same earthquake, obtained by different research teams, often exhibit remarkable dissimilarities. To address the uncertainties in earthquake‐source inversion methods and to understand strengths and weaknesses of the various approaches used, the Source Inversion Validation (SIV) project conducts a set of forward‐modeling exercises and inversion benchmarks. In this article, we describe the SIV strategy, the initial benchmarks, and current SIV results. Furthermore, we apply statistical tools for quantitative waveform comparison and for investigating source‐model (dis)similarities that enable us to rank the solutions, and to identify particularly promising source inversion approaches. All SIV exercises (with related data and descriptions) and statistical comparison tools are available via an online collaboration platform, and we encourage source modelers to use the SIV benchmarks for developing and testing new methods. We envision that the SIV efforts will lead to new developments for tackling the earthquake‐source imaging problem.

A parallel algorithm for 2D visco-acoustic frequency-domain full-waveform inversion: application to a dense OBS data set

NASA Astrophysics Data System (ADS)

Sourbier, F.; Operto, S.; Virieux, J.

2006-12-01

We present a distributed-memory parallel algorithm for 2D visco-acoustic full-waveform inversion of wide-angle seismic data. Our code is written in fortran90 and use MPI for parallelism. The algorithm was applied to real wide-angle data set recorded by 100 OBSs with a 1-km spacing in the eastern-Nankai trough (Japan) to image the deep structure of the subduction zone. Full-waveform inversion is applied sequentially to discrete frequencies by proceeding from the low to the high frequencies. The inverse problem is solved with a classic gradient method. Full-waveform modeling is performed with a frequency-domain finite-difference method. In the frequency-domain, solving the wave equation requires resolution of a large unsymmetric system of linear equations. We use the massively parallel direct solver MUMPS (http://www.enseeiht.fr/irit/apo/MUMPS) for distributed-memory computer to solve this system. The MUMPS solver is based on a multifrontal method for the parallel factorization. The MUMPS algorithm is subdivided in 3 main steps: a symbolic analysis step that performs re-ordering of the matrix coefficients to minimize the fill-in of the matrix during the subsequent factorization and an estimation of the assembly tree of the matrix. Second, the factorization is performed with dynamic scheduling to accomodate numerical pivoting and provides the LU factors distributed over all the processors. Third, the resolution is performed for multiple sources. To compute the gradient of the cost function, 2 simulations per shot are required (one to compute the forward wavefield and one to back-propagate residuals). The multi-source resolutions can be performed in parallel with MUMPS. In the end, each processor stores in core a sub-domain of all the solutions. These distributed solutions can be exploited to compute in parallel the gradient of the cost function. Since the gradient of the cost function is a weighted stack of the shot and residual solutions of MUMPS, each processor computes the corresponding sub-domain of the gradient. In the end, the gradient is centralized on the master processor using a collective communation. The gradient is scaled by the diagonal elements of the Hessian matrix. This scaling is computed only once per frequency before the first iteration of the inversion. Estimation of the diagonal terms of the Hessian requires performing one simulation per non redondant shot and receiver position. The same strategy that the one used for the gradient is used to compute the diagonal Hessian in parallel. This algorithm was applied to a dense wide-angle data set recorded by 100 OBSs in the eastern Nankai trough, offshore Japan. Thirteen frequencies ranging from 3 and 15 Hz were inverted. Tweny iterations per frequency were computed leading to 260 tomographic velocity models of increasing resolution. The velocity model dimensions are 105 km x 25 km corresponding to a finite-difference grid of 4201 x 1001 grid with a 25-m grid interval. The number of shot was 1005 and the number of inverted OBS gathers was 93. The inversion requires 20 days on 6 32-bits bi-processor nodes with 4 Gbytes of RAM memory per node when only the LU factorization is performed in parallel. Preliminary estimations of the time required to perform the inversion with the fully-parallelized code is 6 and 4 days using 20 and 50 processors respectively.

Rapid kinematic finite source inversion for Tsunamic Early Warning using high rate GNSS data

NASA Astrophysics Data System (ADS)

Chen, K.; Liu, Z.; Song, Y. T.

2017-12-01

Recently, Global Navigation Satellite System (GNSS) has been used for rapid earthquake source inversion towards tsunami early warning. In practice, two approaches, i.e., static finite source inversion based on permanent co-seismic offsets and kinematic finite source inversion using high-rate (>= 1 Hz) co-seismic displacement waveforms, are often employed to fulfill the task. The static inversion is relatively easy to be implemented and does not require additional constraints on rupture velocity, duration, and temporal variation. However, since most GNSS receivers are deployed onshore locating on one side of the subduction fault, there is very limited resolution on near-trench fault slip using GNSS in static finite source inversion. On the other hand, the high-rate GNSS displacement waveforms, which contain the timing information of earthquake rupture explicitly and static offsets implicitly, have the potential to improve near-trench resolution by reconciling with the depth-dependent megathrust rupture behaviors. In this contribution, we assess the performance of rapid kinematic finite source inversion using high-rate GNSS by three selected historical tsunamigenic cases: the 2010 Mentawai, 2011 Tohoku and 2015 Illapel events. With respect to the 2010 Mentawai case, it is a typical tsunami earthquake with most slip concentrating near the trench. The static inversion has little resolution there and incorrectly puts slip at greater depth (>10km). In contrast, the recorded GNSS displacement waveforms are deficit in high-frequency energy, the kinematic source inversion recovers a shallow slip patch (depth less than 6 km) and tsunami runups are predicted quite reasonably. For the other two events, slip from kinematic and static inversion show similar characteristics and comparable tsunami scenarios, which may be related to dense GNSS network and behavior of the rupture. Acknowledging the complexity of kinematic source inversion in real-time, we adopt the back-projection approach to provide constraint on rupture velocity.

Tsunami waveform inversion of the 2007 Bengkulu, southern Sumatra earthquake

NASA Astrophysics Data System (ADS)

Fujii, Y.; Satake, K.

2007-12-01

We have performed tsunami waveform inversion for the 2007 Bengkulu, southern Sumatra earthquake on September 12, 2007 (4.520°S, 101.374°E, Mw=8.4 at 11:10:26 UTC according to USGS), and found that the large slips were located on deeper part (> 20 km) of the fault plane, more than 100 km from the trench axis. The deep slip might have contributed the relatively small tsunami for its earthquake size. The largest slips more than 6 m were located beneath Pagais Islands, about 100-200 km northwest of the epicenter. The obtained slip distribution yields a total seismic moment of 3.6 × 1021 Nm (Mw = 8.3). The tsunami generated by this earthquake was recorded at many tide gauge stations located in and around the Indian Ocean. The DART system installed in deep ocean and maintained by Thai Meteorological Department (TMD) also captured this tsunami. We have downloaded the tsunami waveforms at 16 stations from University of Hawaii Sea Level Center's (UHSLC) and National Oceanic & Atmospheric Administration's (NOAA) web sites. The observed tsunami records indicate that the tsunami amplitudes were less than several tens of cm at most stations, around 1 m at Padang, nearest station to the source, and a few cm at DART station. For the tsunami waveforms inversion, we divided the source area (length: 250 km, width: 200 km) into 20 subfaults. Tsunami waveforms from each subfault (50 km × 50 km) or Greens functions were calculated by numerically solving the linear shallow-water long-wave equations. We adopted the focal mechanism of Global CMT solution (strike: 327°, dip: 12°, rake: 114°) for each subfault, and assumed a rise time of 1 min. The computed tsunami waveforms from the estimated slip distribution explain the observed waveforms at most of tide gauges and DART station.

Double-Difference Global Adjoint Tomography

NASA Astrophysics Data System (ADS)

Orsvuran, R.; Bozdag, E.; Lei, W.; Tromp, J.

2017-12-01

The adjoint method allows us to incorporate full waveform simulations in inverse problems. Misfit functions play an important role in extracting the relevant information from seismic waveforms. In this study, our goal is to apply the Double-Difference (DD) methodology proposed by Yuan et al. (2016) to global adjoint tomography. Dense seismic networks, such as USArray, lead to higher-resolution seismic images underneath continents. However, the imbalanced distribution of stations and sources poses challenges in global ray coverage. We adapt double-difference multitaper measurements to global adjoint tomography. We normalize each DD measurement by its number of pairs, and if a measurement has no pair, as may frequently happen for data recorded at oceanic stations, classical multitaper measurements are used. As a result, the differential measurements and pair-wise weighting strategy help balance uneven global kernel coverage. Our initial experiments with minor- and major-arc surface waves show promising results, revealing more pronounced structure near dense networks while reducing the prominence of paths towards cluster of stations. We have started using this new measurement in global adjoint inversions, addressing azimuthal anisotropy in upper mantle. Meanwhile, we are working on combining the double-difference approach with instantaneous phase measurements to emphasize contributions of scattered waves in global inversions and extending it to body waves. We will present our results and discuss challenges and future directions in the context of global tomographic inversions.

Full Waveform Inversion Using Student's t Distribution: a Numerical Study for Elastic Waveform Inversion and Simultaneous-Source Method

NASA Astrophysics Data System (ADS)

Jeong, Woodon; Kang, Minji; Kim, Shinwoong; Min, Dong-Joo; Kim, Won-Ki

2015-06-01

Seismic full waveform inversion (FWI) has primarily been based on a least-squares optimization problem for data residuals. However, the least-squares objective function can suffer from its weakness and sensitivity to noise. There have been numerous studies to enhance the robustness of FWI by using robust objective functions, such as l 1-norm-based objective functions. However, the l 1-norm can suffer from a singularity problem when the residual wavefield is very close to zero. Recently, Student's t distribution has been applied to acoustic FWI to give reasonable results for noisy data. Student's t distribution has an overdispersed density function compared with the normal distribution, and is thus useful for data with outliers. In this study, we investigate the feasibility of Student's t distribution for elastic FWI by comparing its basic properties with those of the l 2-norm and l 1-norm objective functions and by applying the three methods to noisy data. Our experiments show that the l 2-norm is sensitive to noise, whereas the l 1-norm and Student's t distribution objective functions give relatively stable and reasonable results for noisy data. When noise patterns are complicated, i.e., due to a combination of missing traces, unexpected outliers, and random noise, FWI based on Student's t distribution gives better results than l 1- and l 2-norm FWI. We also examine the application of simultaneous-source methods to acoustic FWI based on Student's t distribution. Computing the expectation of the coefficients of gradient and crosstalk noise terms and plotting the signal-to-noise ratio with iteration, we were able to confirm that crosstalk noise is suppressed as the iteration progresses, even when simultaneous-source FWI is combined with Student's t distribution. From our experiments, we conclude that FWI based on Student's t distribution can retrieve subsurface material properties with less distortion from noise than l 1- and l 2-norm FWI, and the simultaneous-source method can be adopted to improve the computational efficiency of FWI based on Student's t distribution.

DOE Office of Scientific and Technical Information (OSTI.GOV)

bedle, H; Matzel, E; Flanagan, M

This report summarizes the data analysis achieved during Heather Bedle's eleven-week Technical Scholar internship at Lawrence Livermore National Labs during the early summer 2006. The work completed during this internship resulted in constraints on the crustal and upper mantle S-velocity structure in Northern Africa, the Mediterranean, the Middle East, and Europe, through the fitting of regional waveform data. This data extends current raypath coverage and will be included in a joint inversion along with data from surface wave group velocity measurements, S and P teleseismic arrival time data, and receiver function data to create an improved velocity model of themore » upper mantle in this region. The tectonic structure of the North African/Mediterranean/Europe/Middle Eastern study region is extremely heterogeneous. This region consists of, among others, stable cratons and platforms such as the West Africa Craton, and Baltica in Northern Europe; oceanic subduction zones throughout the Mediterranean Sea where the African and Eurasian plate collide; regions of continental collision as the Arabian Plate moves northward into the Turkish Plate; and rifting in the Red Sea, separating the Arabian and Nubian shields. With such diverse tectonic structures, many of the waveforms were difficult to fit. This is not unexpected as the waveforms are fit using an averaged structure. In many cases the raypaths encounter several tectonic features, complicating the waveform, and making it hard for the software to converge on a 1D average structure. Overall, the quality of the waveform data was average, with roughly 30% of the waveforms being discarded due to excessive noise that interfered with the frequency ranges of interest. An inversion for the 3D S-velocity structure of this region was also performed following the methodology of Partitioned Waveform Inversion (Nolet, 1990; Van der Lee and Nolet, 1997). The addition of the newly fit waveforms drastically extends the range of the model. The model now extends as far east in Africa to cover Chad and Niger, and reaches south to cover Zambia. The model is also stretched eastward to cover the eastern half of India, and northward to cover the southern portion of Scandinavia.« less

Global and local waveform simulations using the VERCE platform

NASA Astrophysics Data System (ADS)

Garth, Thomas; Saleh, Rafiq; Spinuso, Alessandro; Gemund, Andre; Casarotti, Emanuele; Magnoni, Federica; Krischner, Lion; Igel, Heiner; Schlichtweg, Horst; Frank, Anton; Michelini, Alberto; Vilotte, Jean-Pierre; Rietbrock, Andreas

2017-04-01

In recent years the potential to increase resolution of seismic imaging by full waveform inversion has been demonstrated on a range of scales from basin to continental scales. These techniques rely on harnessing the computational power of large supercomputers, and running large parallel codes to simulate the seismic wave field in a three-dimensional geological setting. The VERCE platform is designed to make these full waveform techniques accessible to a far wider spectrum of the seismological community. The platform supports the two widely used spectral element simulation programs SPECFEM3D Cartesian, and SPECFEM3D globe, allowing users to run a wide range of simulations. In the SPECFEM3D Cartesian implementation the user can run waveform simulations on a range of pre-loaded meshes and velocity models for specific areas, or upload their own velocity model and mesh. In the new SPECFEM3D globe implementation, the user will be able to select from a number of continent scale model regions, or perform waveform simulations for the whole earth. Earthquake focal mechanisms can be downloaded within the platform, for example from the GCMT catalogue, or users can upload their own focal mechanism catalogue through the platform. The simulations can be run on a range of European supercomputers in the PRACE network. Once a job has been submitted and run through the platform, the simulated waveforms can be manipulated or downloaded for further analysis. The misfit between the simulated and recorded waveforms can then be calculated through the platform through three interoperable workflows, for raw-data access (FDSN) and caching, pre-processing and finally misfit. The last workflow makes use of the Pyflex analysis software. In addition, the VERCE platform can be used to produce animations of waveform propagation through the velocity model, and synthetic shakemaps. All these data-products are made discoverable and re-usable thanks to the VERCE data and metadata management layer. We demonstrate the functionality of the VERCE platform with two use cases, one using the pre-loaded velocity model and mesh for the Maule area of Chile using the SPECFEM3D Cartesian workflow, and one showing the output of a global simulation using the SPECFEM3D globe workflow. It is envisioned that this tool will allow a much greater range of seismologists to access these full waveform inversion tools, and aid full waveform tomographic and source inversion, synthetic shakemap production and other full waveform applications, in a wide range of tectonic settings.

Some practical aspects of prestack waveform inversion using a genetic algorithm: An example from the east Texas Woodbine gas sand

DOE Office of Scientific and Technical Information (OSTI.GOV)

Mallick, S.

1999-03-01

In this paper, a prestack inversion method using a genetic algorithm (GA) is presented, and issues relating to the implementation of prestack GA inversion in practice are discussed. GA is a Monte-Carlo type inversion, using a natural analogy to the biological evolution process. When GA is cast into a Bayesian framework, a priori information of the model parameters and the physics of the forward problem are used to compute synthetic data. These synthetic data can then be matched with observations to obtain approximate estimates of the marginal a posteriori probability density (PPD) functions in the model space. Plots of thesemore » PPD functions allow an interpreter to choose models which best describe the specific geologic setting and lead to an accurate prediction of seismic lithology. Poststack inversion and prestack GA inversion were applied to a Woodbine gas sand data set from East Texas. A comparison of prestack inversion with poststack inversion demonstrates that prestack inversion shows detailed stratigraphic features of the subsurface which are not visible on the poststack inversion.« less

Estimation of Dynamic Friction Process of the Akatani Landslide Based on the Waveform Inversion and Numerical Simulation

NASA Astrophysics Data System (ADS)

Yamada, M.; Mangeney, A.; Moretti, L.; Matsushi, Y.

2014-12-01

Understanding physical parameters, such as frictional coefficients, velocity change, and dynamic history, is important issue for assessing and managing the risks posed by deep-seated catastrophic landslides. Previously, landslide motion has been inferred qualitatively from topographic changes caused by the event, and occasionally from eyewitness reports. However, these conventional approaches are unable to evaluate source processes and dynamic parameters. In this study, we use broadband seismic recordings to trace the dynamic process of the deep-seated Akatani landslide that occurred on the Kii Peninsula, Japan, which is one of the best recorded large slope failures. Based on the previous results of waveform inversions and precise topographic surveys done before and after the event, we applied numerical simulations using the SHALTOP numerical model (Mangeney et al., 2007). This model describes homogeneous continuous granular flows on a 3D topography based on a depth averaged thin layer approximation. We assume a Coulomb's friction law with a constant friction coefficient, i. e. the friction is independent of the sliding velocity. We varied the friction coefficients in the simulation so that the resulting force acting on the surface agrees with the single force estimated from the seismic waveform inversion. Figure shows the force history of the east-west components after the band-pass filtering between 10-100 seconds. The force history of the simulation with frictional coefficient 0.27 (thin red line) the best agrees with the result of seismic waveform inversion (thick gray line). Although the amplitude is slightly different, phases are coherent for the main three pulses. This is an evidence that the point-source approximation works reasonably well for this particular event. The friction coefficient during the sliding was estimated to be 0.38 based on the seismic waveform inversion performed by the previous study and on the sliding block model (Yamada et al., 2013), whereas the frictional coefficient estimated from the numerical simulation was about 0.27. This discrepancy may be due to the digital elevation model, to the other forces such as pressure gradients and centrifugal acceleration included in the model. However, quantitative interpretation of this difference requires further investigation.

Forward and Inverse Modeling of Near-Field Seismic Waveforms from Underground Nuclear Explosions for Effective Source Functions and Structure Parameters.

DTIC Science & Technology

1987-04-05

IP o , I-S " M4.7 :" * AMIWILTON & U, .-- EALY(I969) : o H CARROLL(1966) HADLEY (19811 C . Figure 2. P and S-wave velocity structure for Pahute Mesa...8217; 0 .02 s wh ilIe S -. cI by C ) >, s) thIe kta i Is o f t he wav e for:7s are quite well modeled bot h ir tr~~e inversion nd in tefrad mod e Iin~ indi...ESTIMATION 7-Te source parameters determined through waveform inversion for the fo: s o r c ri i c e h v h s~ ahute Mesa events studied are sum.:rarited in

Joint design of large-tip-angle parallel RF pulses and blipped gradient trajectories.

PubMed

Cao, Zhipeng; Donahue, Manus J; Ma, Jun; Grissom, William A

2016-03-01

To design multichannel large-tip-angle kT-points and spokes radiofrequency (RF) pulses and gradient waveforms for transmit field inhomogeneity compensation in high field magnetic resonance imaging. An algorithm to design RF subpulse weights and gradient blip areas is proposed to minimize a magnitude least-squares cost function that measures the difference between realized and desired state parameters in the spin domain, and penalizes integrated RF power. The minimization problem is solved iteratively with interleaved target phase updates, RF subpulse weights updates using the conjugate gradient method with optimal control-based derivatives, and gradient blip area updates using the conjugate gradient method. Two-channel parallel transmit simulations and experiments were conducted in phantoms and human subjects at 7 T to demonstrate the method and compare it to small-tip-angle-designed pulses and circularly polarized excitations. The proposed algorithm designed more homogeneous and accurate 180° inversion and refocusing pulses than other methods. It also designed large-tip-angle pulses on multiple frequency bands with independent and joint phase relaxation. Pulses designed by the method improved specificity and contrast-to-noise ratio in a finger-tapping spin echo blood oxygen level dependent functional magnetic resonance imaging study, compared with circularly polarized mode refocusing. A joint RF and gradient waveform design algorithm was proposed and validated to improve large-tip-angle inversion and refocusing at ultrahigh field. © 2015 Wiley Periodicals, Inc.

Point spread functions for earthquake source imaging: An interpretation based on seismic interferometry

USGS Publications Warehouse

Nakahara, Hisashi; Haney, Matt

2015-01-01

Recently, various methods have been proposed and applied for earthquake source imaging, and theoretical relationships among the methods have been studied. In this study, we make a follow-up theoretical study to better understand the meanings of earthquake source imaging. For imaging problems, the point spread function (PSF) is used to describe the degree of blurring and degradation in an obtained image of a target object as a response of an imaging system. In this study, we formulate PSFs for earthquake source imaging. By calculating the PSFs, we find that waveform source inversion methods remove the effect of the PSF and are free from artifacts. However, the other source imaging methods are affected by the PSF and suffer from the effect of blurring and degradation due to the restricted distribution of receivers. Consequently, careful treatment of the effect is necessary when using the source imaging methods other than waveform inversions. Moreover, the PSF for source imaging is found to have a link with seismic interferometry with the help of the source-receiver reciprocity of Green’s functions. In particular, the PSF can be related to Green’s function for cases in which receivers are distributed so as to completely surround the sources. Furthermore, the PSF acts as a low-pass filter. Given these considerations, the PSF is quite useful for understanding the physical meaning of earthquake source imaging.

Parameterization of the Voice Source by Combining Spectral Decay and Amplitude Features of the Glottal Flow.

ERIC Educational Resources Information Center

Alku, Paavo; Vilkman, Erkki; Laukkanen, Anne-Maria

1998-01-01

A new method is presented for the parameterization of glottal volume velocity waveforms that have been estimated by inverse filtering acoustic speech pressure signals. The new technique combines two features of voice production: the AC value and the spectral decay of the glottal flow. Testing found the new parameter correlates strongly with the…

Analysis of impulse signals with Hylaty ELF station

NASA Astrophysics Data System (ADS)

Kulak, A.; Mlynarczyk, J.; Ostrowski, M.; Kubisz, J.; Michalec, A.

2012-04-01

Lighting discharges generate electromagnetic field pulses that propagate in the Earth-ionosphere waveguide. The attenuation in the ELF range is so small that the pulses originating from strong atmospheric discharges can be observed even several thousand kilometers away from the individual discharge. The recorded waveform depends on the discharge process, the Earth-ionosphere waveguide properties on the source-receiver path, and the transfer function of the receiver. If the distance from the source is known, an inverse method can be used for reconstructing the current moment waveform and the charge moment of the discharge. In order to reconstruct the source parameters from the recorded signal a reliable model of the radio wave propagation in the Earth-ionosphere waveguide as well as practical signal processing techniques are necessary. We present two methods, both based on analytical formulas. The first method allows for fast calculation of the charge moment of relatively short atmospheric discharges. It is based on peak amplitude measurement of the recorded magnetic component of the ELF EM field and it takes into account the receiver characteristics. The second method, called "inverse channel method" allows reconstructing the complete current moment waveform of strong atmospheric discharges that exhibit the continuing current phase, such as Gigantic Jets and Sprites. The method makes it possible to fully remove from the observed waveform the distortions related to the receiver's impulse response as well as the influence of the Earth-ionosphere propagation channel. Our ELF station is equipped with two magnetic antennas for Bx and By components measurement in the 0.03 to 55 Hz frequency range. ELF Data recording is carried out since 1993, with continuous data acquisition since 2005. The station features low noise level and precise timing. It is battery powered and located in the sparsely populated area, far from major electric power lines, which results in high quality signal recordings and allows for precise calculations of the charge moments of upward discharges and strong cloud-to-ground discharges originating from distant sources. The same data is used for Schumann resonance observation. We demonstrate the use of our methods based on recent recordings from the Hylaty ELF station. We include examples of GJ (Gigantic Jet) and TGF (Terrestrial Gamma-ray Flash) related discharges.

Velocity and Attenuation Structure of the Earth's Inner Core Boundary From Semi-Automatic Waveform Modeling

NASA Astrophysics Data System (ADS)

Jin, J.; Song, X.; Sun, D.; Helmberger, D. V.

2013-12-01

The structure of the Earth's inner core boundary (ICB) is complex. Hemispherical differences and local variations of velocity and attenuation structures, as well as the ICB topography have been reported in previous studies. We are using an automatic waveform modeling method to improve the resolution of the ICB structures. The full waveforms of triplicated PKP phases at distance ranges from 120 to 165 degrees are used to model the lowermost 200 km of the outer core and the uppermost 600km of the inner core. Given a 1D velocity and attenuation model, synthetic seismograms are generated by Generalized Ray Theory. We are also experimenting 2D synthetic methods (WKM, AXISEM, and 2D FD) for 2D models (in the mantle and the inner core). The source time function is determined by observed seismic data. We use neighborhood algorithm to search for a group of models that minimize the misfit between predictions and observations. Tests on synthetic data show the efficiency of this method in resolving detailed velocity and attenuation structures of the ICB simultaneously. We are analyzing seismic record sections at dense arrays along different paths and will report our modeling and inversion results in the meeting.

Tomographic inversion of P-wave velocity and Q structures beneath the Kirishima volcanic complex, Southern Japan, based on finite difference calculations of complex traveltimes

USGS Publications Warehouse

Tomatsu, T.; Kumagai, H.; Dawson, P.B.

2001-01-01

We estimate the P-wave velocity and attenuation structures beneath the Kirishima volcanic complex, southern Japan, by inverting the complex traveltimes (arrival times and pulse widths) of waveform data obtained during an active seismic experiment conducted in 1994. In this experiment, six 200-250 kg shots were recorded at 163 temporary seismic stations deployed on the volcanic complex. We use first-arrival times for the shots, which were hand-measured interactively. The waveform data are Fourier transformed into the frequency domain and analysed using a new method based on autoregressive modelling of complex decaying oscillations in the frequency domain to determine pulse widths for the first-arrival phases. A non-linear inversion method is used to invert 893 first-arrival times and 325 pulse widths to estimate the velocity and attenuation structures of the volcanic complex. Wavefronts for the inversion are calculated with a finite difference method based on the Eikonal equation, which is well suited to estimating the complex traveltimes for the structures of the Kirishima volcano complex, where large structural heterogeneities are expected. The attenuation structure is derived using ray paths derived from the velocity structure. We obtain 3-D velocity and attenuation structures down to 1.5 and 0.5 km below sea level, respectively. High-velocity pipe-like structures with correspondingly low attenuation are found under the summit craters. These pipe-like structures are interpreted as remnant conduits of solidified magma. No evidence of a shallow magma chamber is visible in the tomographic images.

Adjoint Tomography of the Southern California Crust (Invited) (Invited)

NASA Astrophysics Data System (ADS)

Tape, C.; Liu, Q.; Maggi, A.; Tromp, J.

2009-12-01

We iteratively improve a three-dimensional tomographic model of the southern California crust using numerical simulations of seismic wave propagation based on a spectral-element method (SEM) in combination with an adjoint method. The initial 3D model is provided by the Southern California Earthquake Center. The dataset comprises three-component seismic waveforms (i.e. both body and surface waves), filtered over the period range 2-30 s, from 143 local earthquakes recorded by a network of 203 stations. Time windows for measurements are automatically selected by the FLEXWIN algorithm. The misfit function in the tomographic inversion is based on frequency-dependent multitaper traveltime differences. The gradient of the misfit function and related finite-frequency sensitivity kernels for each earthquake are computed using an adjoint technique. The kernels are combined using a source subspace projection method to compute a model update at each iteration of a gradient-based minimization algorithm. The inversion involved 16 iterations, which required 6800 wavefield simulations and a total of 0.8 million CPU hours. The new crustal model, m16, is described in terms of independent shear (Vs) and bulk-sound (Vb) wavespeed variations. It exhibits strong heterogeneity, including local changes of ±30% with respect to the initial 3D model. The model reveals several features that relate to geologic observations, such as sedimentary basins, exhumed batholiths, and contrasting lithologies across faults. The quality of the new model is validated by quantifying waveform misfits of full-length seismograms from 91 earthquakes that were not used in the tomographic inversion. The new model provides more accurate synthetic seismograms that will benefit seismic hazard assessment.

Time-domain wavefield reconstruction inversion

NASA Astrophysics Data System (ADS)

Li, Zhen-Chun; Lin, Yu-Zhao; Zhang, Kai; Li, Yuan-Yuan; Yu, Zhen-Nan

2017-12-01

Wavefield reconstruction inversion (WRI) is an improved full waveform inversion theory that has been proposed in recent years. WRI method expands the searching space by introducing the wave equation into the objective function and reconstructing the wavefield to update model parameters, thereby improving the computing efficiency and mitigating the influence of the local minimum. However, frequency-domain WRI is difficult to apply to real seismic data because of the high computational memory demand and requirement of time-frequency transformation with additional computational costs. In this paper, wavefield reconstruction inversion theory is extended into the time domain, the augmented wave equation of WRI is derived in the time domain, and the model gradient is modified according to the numerical test with anomalies. The examples of synthetic data illustrate the accuracy of time-domain WRI and the low dependency of WRI on low-frequency information.

Application of an iterative least-squares waveform inversion of strong-motion and teleseismic records to the 1978 Tabas, Iran, earthquake

USGS Publications Warehouse

Hartzell, S.; Mendoza, C.

1991-01-01

An iterative least-squares technique is used to simultaneously invert the strong-motion records and teleseismic P waveforms for the 1978 Tabas, Iran, earthquake to deduce the rupture history. The effects of using different data sets and different parametrizations of the problem (linear versus nonlinear) are considered. A consensus of all the inversion runs indicates a complex, multiple source for the Tabas earthquake, with four main source regions over a fault length of 90 km and an average rupture velocity of 2.5 km/sec. -from Authors

Stochastic Seismic Inversion and Migration for Offshore Site Investigation in the Northern Gulf of Mexico

NASA Astrophysics Data System (ADS)

Son, J.; Medina-Cetina, Z.

2017-12-01

We discuss the comparison between deterministic and stochastic optimization approaches to the nonlinear geophysical full-waveform inverse problem, based on the seismic survey data from Mississippi Canyon in the Northern Gulf of Mexico. Since the subsea engineering and offshore construction projects actively require reliable ground models from various site investigations, the primary goal of this study is to reconstruct the accurate subsurface information of the soil and rock material profiles under the seafloor. The shallow sediment layers have naturally formed heterogeneous formations which may cause unwanted marine landslides or foundation failures of underwater infrastructure. We chose the quasi-Newton and simulated annealing as deterministic and stochastic optimization algorithms respectively. Seismic forward modeling based on finite difference method with absorbing boundary condition implements the iterative simulations in the inverse modeling. We briefly report on numerical experiments using a synthetic data as an offshore ground model which contains shallow artificial target profiles of geomaterials under the seafloor. We apply the seismic migration processing and generate Voronoi tessellation on two-dimensional space-domain to improve the computational efficiency of the imaging stratigraphical velocity model reconstruction. We then report on the detail of a field data implementation, which shows the complex geologic structures in the Northern Gulf of Mexico. Lastly, we compare the new inverted image of subsurface site profiles in the space-domain with the previously processed seismic image in the time-domain at the same location. Overall, stochastic optimization for seismic inversion with migration and Voronoi tessellation show significant promise to improve the subsurface imaging of ground models and improve the computational efficiency required for the full waveform inversion. We anticipate that by improving the inversion process of shallow layers from geophysical data will better support the offshore site investigation.

Surface wave tomography of the Ontong Java Plateau: Seismic probing of the largest igneous province

NASA Astrophysics Data System (ADS)

Richardson, William Philip

1998-12-01

Large igneous provinces (LIP), such as the gigantic Cretaceous oceanic plateaus, the Ontong-Java, the Manihiki and the Kerguelen, are part of a globally distributed diverse suite of massive crustal features considered to be episodic representations of mantle dynamics (Coffin and Eldholm, 1994). The Ontong Java Plateau in the central western Pacific is by far the largest (and presumably thickest) of these provinces and is believed to have been emplaced rapidly in the Aptian, ˜122 Ma (Tarduno et al., 1991). From 1994 to 1996 four PASSCAL broadband seismic stations were deployed in an array north of the OJP. Analysis was conducted on vertical component broadband seismograms from events recorded on the Micronesian Seismic Experiment array between January 1994 and March 1996. The purpose of this experiment is to investigate the crustal and upper mantle structure of the Ontong Java Plateau (OJP) employing surface wave tomographic methods. Using the partitioned waveform inversion method (Nolet, 1990) and earthquakes with published Centroid Moment Tensor (Dziewonski et al., 1981) solutions, we produce waveform fits from source-to-receiver paths that primarily sample the OJP. From these waveform fits, linearized constraints on shear velocity suggest: (1) a massively thickened crust over the center of the OJP-greater than 35km over central areas of the plateau while thinning off-center; (2) a pronounced low-velocity zone down to ˜300km depth-a robust result in agreement with recent geochemical predictions (Neal et al., 1997); (3) the probability of lateral heterogeneity across the OJP. Finally, by combining many single waveform inversions (van der Lee and Nolet, 1997b) a 3-D shear velocity model can be computed for the Ontong Java Plateau and the nearby Caroline Basin. New constraints on the crustal thickness (and hence the volume extruded) are presented, thereby adding to the understanding of the overall tectonic setting and possible emplacement mechanism of the structure.

Mapping the rupture process of moderate earthquakes by inverting accelerograms

USGS Publications Warehouse

Hellweg, M.; Boatwright, J.

1999-01-01

We present a waveform inversion method that uses recordings of small events as Green's functions to map the rupture growth of moderate earthquakes. The method fits P and S waveforms from many stations simultaneously in an iterative procedure to estimate the subevent rupture time and amplitude relative to the Green's function event. We invert the accelerograms written by two moderate Parkfield earthquakes using smaller events as Green's functions. The first earthquake (M = 4.6) occurred on November 14, 1993, at a depth of 11 km under Middle Mountain, in the assumed preparation zone for the next Parkfield main shock. The second earthquake (M = 4.7) occurred on December 20, 1994, some 6 km to the southeast, at a depth of 9 km on a section of the San Andreas fault with no previous microseismicity and little inferred coseismic slip in the 1966 Parkfield earthquake. The inversion results are strikingly different for the two events. The average stress release in the 1993 event was 50 bars, distributed over a geometrically complex area of 0.9 km2. The average stress release in the 1994 event was only 6 bars, distributed over a roughly elliptical area of 20 km2. The ruptures of both events appear to grow spasmodically into relatively complex shapes: the inversion only constrains the ruptures to grow more slowly than the S wave velocity but does not use smoothness constraints. Copyright 1999 by the American Geophysical Union.

Crustal Stress and Strain Distribution in Sicily (Southern Italy) from Joint Analysis of Seismicity and Geodetic Data

NASA Astrophysics Data System (ADS)

Presti, D.; Neri, G.; Aloisi, M.; Cannavo, F.; Orecchio, B.; Palano, M.; Siligato, G.; Totaro, C.

2014-12-01

An updated database of earthquake focal mechanisms is compiled for the Sicilian region (southern Italy) and surrounding off-shore areas where the Nubia-Eurasia convergence coexists with the very-slow residual rollback of the Ionian subducting slab. High-quality solutions selected from literature and catalogs have been integrated with new solutions estimated in the present work using the Cut And Paste (CAP) waveform inversion method. In the CAP algorithm (Zhao and Helmberger, 1994; Zhu and Helmberger, 1996), each waveform is broken up into Pnl and surface wave segments, which are weighted differently during the inversion procedure. Integration of the new solutions with the ones selected from literature and official catalogs led us to collect a database consisting exclusively of waveform inversion data relative to earthquakes with minimum magnitude 2.6. The seismicity and focal mechanism distributions have been compared with crustal motion and strain data coming from GNSS analyses. For this purpose GNSS-based observations collected over the investigated area by episodic measurements (1994-2013) as well as continuous monitoring (since 2006) were processed by the GAMIT/GLOBK software packages (Herring et al., 2010) following the approach described in Palano et al. (2011). To adequately investigate the crustal deformation pattern, the estimated GNSS velocities were aligned to a fixed Eurasian reference frame. The good agreement found between seismic and geodetic information contributes to better define seismotectonic domains characterized by different kinematics. Moving from the available geophysical information and from an early application of FEM algorithms, we have also started to investigate stress/strain fields in the crust of the study area including depth dependence and relationships with rupture of the main seismogenic structures.

Resolving the fine-scale velocity structure of continental hyperextension at the Deep Galicia Margin using full-waveform inversion

NASA Astrophysics Data System (ADS)

Davy, R. G.; Morgan, J. V.; Minshull, T. A.; Bayrakci, G.; Bull, J. M.; Klaeschen, D.; Reston, T. J.; Sawyer, D. S.; Lymer, G.; Cresswell, D.

2018-01-01

Continental hyperextension during magma-poor rifting at the Deep Galicia Margin is characterized by a complex pattern of faulting, thin continental fault blocks and the serpentinization, with local exhumation, of mantle peridotites along the S-reflector, interpreted as a detachment surface. In order to understand fully the evolution of these features, it is important to image seismically the structure and to model the velocity structure to the greatest resolution possible. Traveltime tomography models have revealed the long-wavelength velocity structure of this hyperextended domain, but are often insufficient to match accurately the short-wavelength structure observed in reflection seismic imaging. Here, we demonstrate the application of 2-D time-domain acoustic full-waveform inversion (FWI) to deep-water seismic data collected at the Deep Galicia Margin, in order to attain a high-resolution velocity model of continental hyperextension. We have used several quality assurance procedures to assess the velocity model, including comparison of the observed and modeled waveforms, checkerboard tests, testing of parameter and inversion strategy and comparison with the migrated reflection image. Our final model exhibits an increase in the resolution of subsurface velocities, with particular improvement observed in the westernmost continental fault blocks, with a clear rotation of the velocity field to match steeply dipping reflectors. Across the S-reflector, there is a sharpening in the velocity contrast, with lower velocities beneath S indicative of preferential mantle serpentinization. This study supports the hypothesis that normal faulting acts to hydrate the upper-mantle peridotite, observed as a systematic decrease in seismic velocities, consistent with increased serpentinization. Our results confirm the feasibility of applying the FWI method to sparse, deep-water crustal data sets.

First results from a full-waveform inversion of the African continent using Salvus

NASA Astrophysics Data System (ADS)

van Herwaarden, D. P.; Afanasiev, M.; Krischer, L.; Trampert, J.; Fichtner, A.

2017-12-01

We present the initial results from an elastic full-waveform inversion (FWI) of the African continent which is melded together within the framework of the Collaborative Seismic Earth Model (CSEM) project. The continent of Africa is one of the most geophysically interesting regions on the planet. More specifically, Africa contains the Afar Depression, which is the only place on Earth where incipient seafloor spreading is sub-aerially exposed, along with other anomalous features such as the topography in the south, and several smaller surface expressions such as the Cameroon Volcanic Line and Congo Basin. Despite its significance, relatively few tomographic images exist of Africa, and, as a result, the debate on the geophysical origins of Africa's anomalies is rich and ongoing. Tomographic images of Africa present unique challenges due to uneven station coverage: while tectonically active areas such as the Afar rift are well sampled, much of the continent exhibits a severe lack of seismic stations. And, while Africa is mostly surrounded by tectonically active spreading plate boundaries, the interior of the continent is seismically quiet. To mitigate such issues, our simulation domain is extended to include earthquakes occurring in the South Atlantic and along the western edge of South America. Waveform modelling and inversion is performed using Salvus, a flexible and high-performance software suite based on the spectral-element method. Recently acquired recordings from the AfricaArray and NARS seismic networks are used to complement data obtained from global networks. We hope that this new model presents a fresh high-resolution image of African geodynamic structure, and helps advance the debate regarding the causative mechanisms of its surface anomalies.

Resolving Isotropic Components from Regional Waves using Grid Search and Moment Tensor Inversion Methods

NASA Astrophysics Data System (ADS)

Ichinose, G. A.; Saikia, C. K.

2007-12-01

We applied the moment tensor (MT) analysis scheme to identify seismic sources using regional seismograms based on the representation theorem for the elastic wave displacement field. This method is applied to estimate the isotropic (ISO) and deviatoric MT components of earthquake, volcanic, and isotropic sources within the Basin and Range Province (BRP) and western US. The ISO components from Hoya, Bexar, Montello and Junction were compared to recently well recorded recent earthquakes near Little Skull Mountain, Scotty's Junction, Eureka Valley, and Fish Lake Valley within southern Nevada. We also examined "dilatational" sources near Mammoth Lakes Caldera and two mine collapses including the August 2007 event in Utah recorded by US Array. Using our formulation we have first implemented the full MT inversion method on long period filtered regional data. We also applied a grid-search technique to solve for the percent deviatoric and %ISO moments. By using the grid-search technique, high-frequency waveforms are used with calibrated velocity models. We modeled the ISO and deviatoric components (spall and tectonic release) as separate events delayed in time or offset in space. Calibrated velocity models helped the resolution of the ISO components and decrease the variance over the average, initial or background velocity models. The centroid location and time shifts are velocity model dependent. Models can be improved as was done in previously published work in which we used an iterative waveform inversion method with regional seismograms from four well recorded and constrained earthquakes. The resulting velocity models reduced the variance between predicted synthetics by about 50 to 80% for frequencies up to 0.5 Hz. Tests indicate that the individual path-specific models perform better at recovering the earthquake MT solutions even after using a sparser distribution of stations than the average or initial models.

High-resolution near-surface velocity model building using full-waveform inversion—a case study from southwest Sweden

NASA Astrophysics Data System (ADS)

Adamczyk, A.; Malinowski, M.; Malehmir, A.

2014-06-01

Full-waveform inversion (FWI) is an iterative optimization technique that provides high-resolution models of subsurface properties. Frequency-domain, acoustic FWI was applied to seismic data acquired over a known quick-clay landslide scar in southwest Sweden. We inverted data from three 2-D seismic profiles, 261-572 m long, two of them shot with small charges of dynamite and one with a sledgehammer. To our best knowledge this is the first published application of FWI to sledgehammer data. Both sources provided data suitable for waveform inversion, the sledgehammer data containing even wider frequency spectrum. Inversion was performed for frequency groups between 27.5 and 43.1 Hz for the explosive data and 27.5-51.0 Hz for the sledgehammer. The lowest inverted frequency was limited by the resonance frequency of the standard 28-Hz geophones used in the survey. High-velocity granitic bedrock in the area is undulated and very shallow (15-100 m below the surface), and exhibits a large P-wave velocity contrast to the overlying normally consolidated sediments. In order to mitigate the non-linearity of the inverse problem we designed a multiscale layer-stripping inversion strategy. Obtained P-wave velocity models allowed to delineate the top of the bedrock and revealed distinct layers within the overlying sediments of clays and coarse-grained materials. Models were verified in an extensive set of validating procedures and used for pre-stack depth migration, which confirmed their robustness.

Long-period GPS waveforms. What can GPS bring to Earth seismic velocity models?

NASA Astrophysics Data System (ADS)

Kelevitz, Krisztina; Houlié, Nicolas; Boschi, Lapo; Nissen-Meyer, Tarje; Giardini, Domenico

2014-05-01

It is now commonly admitted that high rate GPS observations can provide reliable surface displacement waveforms (Cervelli, et al., 2001; Langbein, et al., 2006; Houlié, et al., 2006; Houlié et al., 2011). For long-period (T>5s) transients, it was shown that GPS and seismometer (STS-1) displacements are in agreement at least for vertical component (Houlié, et al., Sci. Rep. 2011). We propose here to supplement existing long-period seismic networks with high rate (>= 1Hz) GPS data in order to improve the resolution of global seismic velocity models. GPS measurements are providing a wide range of frequencies, going beyond the range of STS-1 in the low frequency end. Nowadays, almost 10.000 GPS receivers would be able to record data at 1 Hz with 3000+ stations already streaming data in Real-Time (RT). The reasons for this quick expansion are the price of receivers, their low maintenance, and the wide range of activities they can be used for (transport, science, public apps, navigation, etc.). We are presenting work completed on the 1Hz GPS records of the Hokkaido earthquake (25th of September, 2003, Mw=8.3). 3D Waveforms have been computed with an improved, stabilised inversion algorithm in order to constrain the ground motion history. Through the better resolution of inversion of the GPS phase observations, we determine displacement waveforms of frequencies ranging from 0.77 mHz to 330 mHz for a selection of sites. We compare inverted GPS waveforms with STS-1 waveforms and synthetic waveforms computed using 3D global wave propagation with SPECFEM. At co-located sites (STS-1 and GPS located within 10km) the agreement is good for the vertical component between seismic (both real and synthetic) and GPS waveforms.

Broadband waveform inversion of moderate earthquakes in the Messina Straits, southern Italy

NASA Astrophysics Data System (ADS)

D'Amico, Sebastiano; Orecchio, Barbara; Presti, Debora; Zhu, Lupei; Herrmann, Robert B.; Neri, Giancarlo

2010-04-01

We report the first application of the Cut and Paste (CAP) method to compute earthquake focal mechanisms in the Messina Straits area by waveform inversion of Pnl and surface wave segments. This application of CAP has furnished new knowledge about low-magnitude earthquake mechanics that will be useful for improved understanding of the local geodynamics. This is possible because the CAP inversion technique can be applied to small earthquakes, for which traditional moment tensor inversion methods are not appropriate and P-onset focal mechanisms in the study area fail because of a lack of sufficient observations. We estimate the focal mechanisms of 23 earthquakes with local magnitudes in the range of 3-4 occurring in the 2004-2008 time period, and recorded by the broadband stations of the Italian National Seismic Network and the Mediterranean Very Broadband Seismographic Network (MedNet) run by the Istituto Nazionale di Geofisica e Vulcanologia (INGV). The solutions show that normal faulting is the prevailing style of seismic deformation in the northern part of the study area while co-existence of normal faulting and strike-slip has been detected in the southern part. In the whole area of investigation the T-axes of focal mechanisms display a preferential northwest-southeast direction of extension. Combined with the findings of previous investigations, this improved database of focal mechanisms allows us to better detail the transitional area between the extensional domain related to subduction trench retreat (southern Calabria) and the compressional one associated with continental collision (western-central Sicily). The observed spatial change of seismic deformation regime offers new data to current seismotectonic and seismic hazard investigations in the area of Messina Straits where a magnitude 7.2 earthquake caused more than 60,000 casualties on 28 December 1908.

Waveform inversion of very long period impulsive signals associated with magmatic injection beneath Kilauea Volcano, Hawaii

USGS Publications Warehouse

Ohminato, T.; Chouet, B.A.; Dawson, P.; Kedar, S.

1998-01-01

We use data from broadband seismometers deployed around the summit of Kilauea Volcano to quantify the mechanism associated with a transient in the flow of magma feeding the east rift eruption of the volcano. The transient is marked by rapid inflation of the Kilauea summit peaking at 22 ??rad 4.5 hours after the event onset, followed by slow deflation over a period of 3 days. Superimposed on the summit inflation is a series of sawtooth displacement pulses, each characterized by a sudden drop in amplitude lasting 5-10 s followed by an exponential recovery lasting 1-3 min. The sawtooth waveforms display almost identical shapes, suggesting a process involving the repeated activation of a fixed source. The particle motion associated with each sawtooth is almost linear, and its major swing shows compressional motion at all stations. Analyses of semblance and particle motion are consistent with a point source located 1 km beneath the northeast edge of the Halemaumau pit crater. To estimate the source mechanism, we apply a moment tensor inversion to the waveform data, assuming a point source embedded in a homogeneous half-space with compressional and shear wave velocities representative of the average medium properties at shallow depth under Kilauea. Synthetic waveforms are constructed by a superposition of impulse responses for six moment tensor components and three single force components. The origin times of individual impulses are distributed along the time axis at appropriately small, equal intervals, and their amplitudes are determined by least squares. In this inversion, the source time functions of the six tensor and three force components are determined simultaneously. We confirm the accuracy of the inversion method through a series of numerical tests. The results from the inversion show that the waveform data are well explained by a pulsating transport mechanism operating on a subhorizontal crack linking the summit reservoir to the east rift of Kilauea. The crack acts like a buffer in which a batch of fluid (magma and/or gas) accumulates over a period of 1-3 min before being rapidly injected into a larger reservoir (possibly the east rift) over a timescale of 5-10 s. The seismic moment and volume change associated with a typical batch of fluid are approximately 1014 N m and 3000 m3, respectively. Our results also point to the existence of a single force component with amplitude of 109 N, which may be explained as the drag force generated by the flow of viscous magma through a narrow constriction in the flow path. The total volume of magma associated with the 4.5-hour-long activation of the pulsating source is roughly 500,000 m3 in good agreement with the integrated volume flow rate of magma estimated near the eruptive site.

The shift-invariant discrete wavelet transform and application to speech waveform analysis.

PubMed

Enders, Jörg; Geng, Weihua; Li, Peijun; Frazier, Michael W; Scholl, David J

2005-04-01

The discrete wavelet transform may be used as a signal-processing tool for visualization and analysis of nonstationary, time-sampled waveforms. The highly desirable property of shift invariance can be obtained at the cost of a moderate increase in computational complexity, and accepting a least-squares inverse (pseudoinverse) in place of a true inverse. A new algorithm for the pseudoinverse of the shift-invariant transform that is easier to implement in array-oriented scripting languages than existing algorithms is presented together with self-contained proofs. Representing only one of the many and varied potential applications, a recorded speech waveform illustrates the benefits of shift invariance with pseudoinvertibility. Visualization shows the glottal modulation of vowel formants and frication noise, revealing secondary glottal pulses and other waveform irregularities. Additionally, performing sound waveform editing operations (i.e., cutting and pasting sections) on the shift-invariant wavelet representation automatically produces quiet, click-free section boundaries in the resulting sound. The capabilities of this wavelet-domain editing technique are demonstrated by changing the rate of a recorded spoken word. Individual pitch periods are repeated to obtain a half-speed result, and alternate individual pitch periods are removed to obtain a double-speed result. The original pitch and formant frequencies are preserved. In informal listening tests, the results are clear and understandable.

The shift-invariant discrete wavelet transform and application to speech waveform analysis

NASA Astrophysics Data System (ADS)

Enders, Jörg; Geng, Weihua; Li, Peijun; Frazier, Michael W.; Scholl, David J.

2005-04-01

The discrete wavelet transform may be used as a signal-processing tool for visualization and analysis of nonstationary, time-sampled waveforms. The highly desirable property of shift invariance can be obtained at the cost of a moderate increase in computational complexity, and accepting a least-squares inverse (pseudoinverse) in place of a true inverse. A new algorithm for the pseudoinverse of the shift-invariant transform that is easier to implement in array-oriented scripting languages than existing algorithms is presented together with self-contained proofs. Representing only one of the many and varied potential applications, a recorded speech waveform illustrates the benefits of shift invariance with pseudoinvertibility. Visualization shows the glottal modulation of vowel formants and frication noise, revealing secondary glottal pulses and other waveform irregularities. Additionally, performing sound waveform editing operations (i.e., cutting and pasting sections) on the shift-invariant wavelet representation automatically produces quiet, click-free section boundaries in the resulting sound. The capabilities of this wavelet-domain editing technique are demonstrated by changing the rate of a recorded spoken word. Individual pitch periods are repeated to obtain a half-speed result, and alternate individual pitch periods are removed to obtain a double-speed result. The original pitch and formant frequencies are preserved. In informal listening tests, the results are clear and understandable. .

Constructing a starting 3D shear velocity model with sharp interfaces for SEM-based upper mantle tomography in North America

NASA Astrophysics Data System (ADS)

Calo, M.; Bodin, T.; Yuan, H.; Romanowicz, B. A.; Larmat, C. S.; Maceira, M.

2013-12-01

Seismic tomography is currently evolving towards 3D earth models that satisfy full seismic waveforms at increasingly high frequencies. This evolution is possible thanks to the advent of powerful numerical methods such as the Spectral Element Method (SEM) that allow accurate computation of the seismic wavefield in complex media, and the drastic increase of computational resources. However, the production of such models requires handling complex misfit functions with more than one local minimum. Standard linearized inversion methods (such as gradient methods) have two main drawbacks: 1) they produce solution models highly dependent on the starting model; 2) they do not provide a means of estimating true model uncertainties. However, these issues can be addressed with stochastic methods that can sample the space of possible solutions efficiently. Such methods are prohibitively challenging computationally in 3D, but increasingly accessible in 1D. In previous work (Yuan and Romanowicz, 2010; Yuan et al., 2011) we developed a continental scale anisotropic upper mantle model of north America based on a combination of long period seismic waveforms and SKS splitting measurements, showing the pervasive presence of layering of anisotropy in the cratonic lithosphere with significant variations in depth of the mid-lithospheric boundary. The radial anisotropy part of the model has been recently updated using the spectral element method for forward wavefield computations and waveform data from the latest deployments of USarray (Yuan and Romanowicz, 2013). However, the long period waveforms (periods > 40s) themselves only provide a relatively smooth view of the mantle if the starting model is smooth, and the mantle discontinuities necessary for geodynamical interpretation are not imaged. Increasing the frequency of the computations to constrain smaller scale features is possible, but challenging computationally, and at the risk of falling in local minima of the misfit function. In this work we propose instead to directly tackle the non-linearity of the inverse problem by using stochastic methods to construct a 3D starting model with a good estimate of the depths of the main layering interfaces. We present preliminary results of the construction of such a starting 3D model based on: (1) Regionalizing the study area to define provinces within which lateral variations are smooth; (2) Applying trans-dimensional stochastic inversion (Bodin et al., 2012) to obtain accurate 1D models in each province as well as the corresponding error distribution, constrained by receiver function and surface wave dispersion data as well as the previously constructed 3D model (name), and (3) connecting these models laterally using data-driven smoothing operators to obtain a starting 3D model with errors. References Bodin, T.,et al. 2012, Transdimensional inversion of receiver functions and surface wave dispersion, J. Geophys. Res., 117, B02301, doi:10.1029/2011JB008560. Yuan and Romanowicz, 2013, in revison. Yuan, H., et al. 2011, 3-D shear wave radially and azimuthally anisotropic velocity model of the North American upper mantle. Geophysical Journal International, 184: 1237-1260. doi: 10.1111/j.1365-246X.2010.04901.x Yuan, H. & Romanowicz, B., 2010. Lithospheric layering in the North American Craton, Nature, 466, 1063-1068.

Elastic and anelastic structure of the lowermost mantle beneath the Western Pacific using waveform inversion

NASA Astrophysics Data System (ADS)

Konishi, K.; Deschamps, F.; Fuji, N.

2015-12-01

We investigate quasi-2D elastic and anelastic structure of the lowermost mantle beneath the Western Pacific by inverting S and ScS waveforms. The transverse component data were obtained from F-net for 32 deep sources beneath Tonga and Fiji, filtered between 12.5 and 200 s. We observe a regional variation of S and ScS arrival times and amplitude ratio, according to which we divide our region of interest into four sub-regions and perform 1D waveform inversion for S-wave velocity and Qμ value simultaneously. We find S-shaped structure of S-wave velocity beneath the whole region with sub-regional variation of S-wave velocity peak depths, which can explain regional difference in travel times. Qμ structure varies with sub-regions as well, but the physical interpretation has not yet done.

Moment Inversion of the DPRK Nuclear Tests Using Finite-Difference Three-dimensional Strain Green's Tensors

NASA Astrophysics Data System (ADS)

Bao, X.; Shen, Y.; Wang, N.

2017-12-01

Accurate estimation of the source moment is important for discriminating underground explosions from earthquakes and other seismic sources. In this study, we invert for the full moment tensors of the recent seismic events (since 2016) at the Democratic People's Republic of Korea (PRRK) Punggye-ri test site. We use waveform data from broadband seismic stations located in China, Korea, and Japan in the inversion. Using a non-staggered-grid, finite-difference algorithm, we calculate the strain Green's tensors (SGT) based on one-dimensional (1D) and three-dimensional (3D) Earth models. Taking advantage of the source-receiver reciprocity, a SGT database pre-calculated and stored for the Punggye-ri test site is used in inversion for the source mechanism of each event. With the source locations estimated from cross-correlation using regional Pn and Pn-coda waveforms, we obtain the optimal source mechanism that best fits synthetics to the observed waveforms of both body and surface waves. The moment solutions of the first three events (2016-01-06, 2016-09-09, and 2017-09-03) show dominant isotropic components, as expected from explosions, though there are also notable non-isotropic components. The last event ( 8 minutes after the mb6.3 explosion in 2017) contained mainly implosive component, suggesting a collapse following the explosion. The solutions from the 3D model can better fit observed waveforms than the corresponding solutions from the 1D model. The uncertainty in the resulting moment solution is influenced by heterogeneities not resolved by the Earth model according to the waveform misfit. Using the moment solutions, we predict the peak ground acceleration at the Punggye-ri test site and compare the prediction with corresponding InSAR and other satellite images.

Multiparameter elastic full waveform inversion with facies-based constraints

NASA Astrophysics Data System (ADS)

Zhang, Zhen-dong; Alkhalifah, Tariq; Naeini, Ehsan Zabihi; Sun, Bingbing

2018-06-01

Full waveform inversion (FWI) incorporates all the data characteristics to estimate the parameters described by the assumed physics of the subsurface. However, current efforts to utilize FWI beyond improved acoustic imaging, like in reservoir delineation, faces inherent challenges related to the limited resolution and the potential trade-off between the elastic model parameters. Some anisotropic parameters are insufficiently updated because of their minor contributions to the surface collected data. Adding rock physics constraints to the inversion helps mitigate such limited sensitivity, but current approaches to add such constraints are based on including them as a priori knowledge mostly valid around the well or as a global constraint for the whole area. Since similar rock formations inside the Earth admit consistent elastic properties and relative values of elasticity and anisotropy parameters (this enables us to define them as a seismic facies), utilizing such localized facies information in FWI can improve the resolution of inverted parameters. We propose a novel approach to use facies-based constraints in both isotropic and anisotropic elastic FWI. We invert for such facies using Bayesian theory and update them at each iteration of the inversion using both the inverted models and a priori information. We take the uncertainties of the estimated parameters (approximated by radiation patterns) into consideration and improve the quality of estimated facies maps. Four numerical examples corresponding to different acquisition, physical assumptions and model circumstances are used to verify the effectiveness of the proposed method.

Adjoint tomography of the crust and upper mantle structure beneath the Kanto region using broadband seismograms

NASA Astrophysics Data System (ADS)

Miyoshi, Takayuki; Obayashi, Masayuki; Peter, Daniel; Tono, Yoko; Tsuboi, Seiji

2017-12-01

A three-dimensional seismic wave speed model in the Kanto region of Japan was developed using adjoint tomography for application in the effective reproduction of observed waveforms. Starting with a model based on previous travel time tomographic results, we inverted the waveforms obtained at seismic broadband stations from 140 local earthquakes in the Kanto region to obtain the P- and S-wave speeds V p and V s . Additionally, all centroid times of the source solutions were determined before the structural inversion. The synthetic displacements were calculated using the spectral-element method (SEM) in which the Kanto region was parameterized using 16 million grid points. The model parameters V p and V s were updated iteratively by Newton's method using the misfit and Hessian kernels until the misfit between the observed and synthetic waveforms was minimized. Computations of the forward and adjoint simulations were conducted on the K computer in Japan. The optimized SEM code required a total of 6720 simulations using approximately 62,000 node hours to obtain the final model after 16 iterations. The proposed model reveals several anomalous areas with extremely low- V s values in comparison with those of the initial model. These anomalies were found to correspond to geological features, earthquake sources, and volcanic regions with good data coverage and resolution. The synthetic waveforms obtained using the newly proposed model for the selected earthquakes showed better fit than the initial model to the observed waveforms in different period ranges within 5-30 s. This result indicates that the model can accurately predict actual waveforms. [Figure not available: see fulltext.

Real-time Tsunami Inundation Prediction Using High Performance Computers

NASA Astrophysics Data System (ADS)

Oishi, Y.; Imamura, F.; Sugawara, D.

2014-12-01

Recently off-shore tsunami observation stations based on cabled ocean bottom pressure gauges are actively being deployed especially in Japan. These cabled systems are designed to provide real-time tsunami data before tsunamis reach coastlines for disaster mitigation purposes. To receive real benefits of these observations, real-time analysis techniques to make an effective use of these data are necessary. A representative study was made by Tsushima et al. (2009) that proposed a method to provide instant tsunami source prediction based on achieving tsunami waveform data. As time passes, the prediction is improved by using updated waveform data. After a tsunami source is predicted, tsunami waveforms are synthesized from pre-computed tsunami Green functions of linear long wave equations. Tsushima et al. (2014) updated the method by combining the tsunami waveform inversion with an instant inversion of coseismic crustal deformation and improved the prediction accuracy and speed in the early stages. For disaster mitigation purposes, real-time predictions of tsunami inundation are also important. In this study, we discuss the possibility of real-time tsunami inundation predictions, which require faster-than-real-time tsunami inundation simulation in addition to instant tsunami source analysis. Although the computational amount is large to solve non-linear shallow water equations for inundation predictions, it has become executable through the recent developments of high performance computing technologies. We conducted parallel computations of tsunami inundation and achieved 6.0 TFLOPS by using 19,000 CPU cores. We employed a leap-frog finite difference method with nested staggered grids of which resolution range from 405 m to 5 m. The resolution ratio of each nested domain was 1/3. Total number of grid points were 13 million, and the time step was 0.1 seconds. Tsunami sources of 2011 Tohoku-oki earthquake were tested. The inundation prediction up to 2 hours after the earthquake occurs took about 2 minutes, which would be sufficient for a practical tsunami inundation predictions. In the presentation, the computational performance of our faster-than-real-time tsunami inundation model will be shown, and preferable tsunami wave source analysis for an accurate inundation prediction will also be discussed.

A strategy for the application of frequency domain acoustic waveform tomography to marine Walkaway VSP data

NASA Astrophysics Data System (ADS)

Bouzidi, Y.; Takam Takougang, E. M.

2016-12-01

Two dimensional frequency domain acoustic waveform tomography was applied to walkaway VSP data from an oil field in a shallow water environment, offshore the United Arab Emirates, to form a high resolution velocity model of the subsurface around and away from the borehole. Five close parallel walkaway VSP lines were merged to form a 9 km line, with 1344 shots at 25 m shot interval and 4 m shot depth. Each line was recorded using a typical recording tool with 20 receivers at 15.1 m receiver intervals. The recording tool was deployed in a deviated borehole at different depths for each line (521-2742 m depth). Waveform tomography was performed following a specific inversion strategy to mitigate non-linearity. Three parameters were critical for the success of the inversion: the starting model obtained from traveltime tomography, the preconditioning of the input data used for amplitudes correction to remove of shear waves and noise, and a judicious selection of the time damping constant τ to suppress late arrivals in the Laplace-Fourier domain. Several values of the time damping constant were tested, and 2 values, 0.5 s and 0.8 s that suppress waveforms arriving after 1.2 s and 2 s respectively, were retained. The inversion was performed in 2 stages, with frequencies ranging from 5 Hz to 40 Hz. The values of the time damping term τ = 0.5 s and τ = 0.8 s were used in sequence for the frequencies 5-25 Hz, and τ = 0.8 s was used for the frequencies 25-40 Hz. A group of 5 frequencies at 0.5 Hz intervals were used and 6 iterations were performed. A velocity model that generally correlates well with the sonic log and estimated velocities from normal incidence VSP was obtained. The results confirmed the success of the inversion strategy. The velocity model shows zones with anomalous low velocities below 2000 m depth that correlate with known locations of hydrocarbons reservoirs. with known locations of hydrocarbon reservoirs. However, between 500 m and 1200 m depth, the velocity model appears to be slightly underestimated, which can be explained by possible elastic effects and out-of-plane structures not considered during the inversion. This result shows that acoustic waveform tomography can be successfully applied to walkaway VSP data when a good preconditioning of the input data and inversion strategy are used.

Quantification of Uncertainty in Full-Waveform Moment Tensor Inversion for Regional Seismicity

NASA Astrophysics Data System (ADS)

Jian, P.; Hung, S.; Tseng, T.

2013-12-01

Routinely and instantaneously determined moment tensor solutions deliver basic information for investigating faulting nature of earthquakes and regional tectonic structure. The accuracy of full-waveform moment tensor inversion mostly relies on azimuthal coverage of stations, data quality and previously known earth's structure (i.e., impulse responses or Green's functions). However, intrinsically imperfect station distribution, noise-contaminated waveform records and uncertain earth structure can often result in large deviations of the retrieved source parameters from the true ones, which prohibits the use of routinely reported earthquake catalogs for further structural and tectonic interferences. Duputel et al. (2012) first systematically addressed the significance of statistical uncertainty estimation in earthquake source inversion and exemplified that the data covariance matrix, if prescribed properly to account for data dependence and uncertainty due to incomplete and erroneous data and hypocenter mislocation, cannot only be mapped onto the uncertainty estimate of resulting source parameters, but it also aids obtaining more stable and reliable results. Over the past decade, BATS (Broadband Array in Taiwan for Seismology) has steadily devoted to building up a database of good-quality centroid moment tensor (CMT) solutions for moderate to large magnitude earthquakes that occurred in Taiwan area. Because of the lack of the uncertainty quantification and reliability analysis, it remains controversial to use the reported CMT catalog directly for further investigation of regional tectonics, near-source strong ground motions, and seismic hazard assessment. In this study, we develop a statistical procedure to make quantitative and reliable estimates of uncertainty in regional full-waveform CMT inversion. The linearized inversion scheme adapting efficient estimation of the covariance matrices associated with oversampled noisy waveform data and errors of biased centroid positions is implemented and inspected for improving source parameter determination of regional seismicity in Taiwan. Synthetic inversion tests demonstrate the resolved moment tensors would better match the hypothetical CMT solutions, and tend to suppress unreal non-double-couple components and reduce the trade-off between focal mechanism and centroid depth if individual signal-to-noise ratios and correlation lengths for 3-component seismograms at each station and mislocation uncertainties are properly taken into account. We further testify the capability of our scheme in retrieving the robust CMT information for mid-sized (Mw~3.5) and offshore earthquakes in Taiwan, which offers immediate and broad applications in detailed modelling of regional stress field and deformation pattern and mapping of subsurface velocity structures.

Real-time monitoring and massive inversion of source parameters of very long period seismic signals: An application to Stromboli Volcano, Italy

USGS Publications Warehouse

Auger, E.; D'Auria, L.; Martini, M.; Chouet, B.; Dawson, P.

2006-01-01

We present a comprehensive processing tool for the real-time analysis of the source mechanism of very long period (VLP) seismic data based on waveform inversions performed in the frequency domain for a point source. A search for the source providing the best-fitting solution is conducted over a three-dimensional grid of assumed source locations, in which the Green's functions associated with each point source are calculated by finite differences using the reciprocal relation between source and receiver. Tests performed on 62 nodes of a Linux cluster indicate that the waveform inversion and search for the best-fitting signal over 100,000 point sources require roughly 30 s of processing time for a 2-min-long record. The procedure is applied to post-processing of a data archive and to continuous automatic inversion of real-time data at Stromboli, providing insights into different modes of degassing at this volcano. Copyright 2006 by the American Geophysical Union.

Determination of source parameters of the 2017 Mount Agung volcanic earthquake from moment-tensor inversion method using local broadband seismic waveforms

NASA Astrophysics Data System (ADS)

Madlazim; Prastowo, T.; Supardiyono; Hardy, T.

2018-03-01

Monitoring of volcanoes has been an important issue for many purposes, particularly hazard mitigation. With regard to this, the aims of the present work are to estimate and analyse source parameters of a volcanic earthquake driven by recent magmatic events of Mount Agung in Bali island that occurred on September 28, 2017. The broadband seismogram data consisting of 3 local component waveforms were recorded by the IA network of 5 seismic stations: SRBI, DNP, BYJI, JAGI, and TWSI (managed by BMKG). These land-based observatories covered a full 4-quadrant region surrounding the epicenter. The methods used in the present study were seismic moment-tensor inversions, where the data were all analyzed to extract the parameters, namely moment magnitude, type of a volcanic earthquake indicated by percentages of seismic components: compensated linear vector dipole (CLVD), isotropic (ISO), double-couple (DC), and source depth. The results are given in the forms of variance reduction of 65%, a magnitude of M W 3.6, a CLVD of 40%, an ISO of 33%, a DC of 27% and a centroid-depth of 9.7 km. These suggest that the unusual earthquake was dominated by a vertical CLVD component, implying the dominance of uplift motion of magmatic fluid flow inside the volcano.

Sensitivity Kernels for the Cross-Convolution Measure: Eliminate the Source in Waveform Tomography

NASA Astrophysics Data System (ADS)

Menke, W. H.

2017-12-01

We use the adjoint method to derive sensitivity kernels for the cross-convolution measure, a goodness-of-fit criterion that is applicable to seismic data containing closely-spaced multiple arrivals, such as reverberating compressional waves and split shear waves. In addition to a general formulation, specific expressions for sensitivity with respect to density, Lamé parameter and shear modulus are derived for a isotropic elastic solid. As is typical of adjoint methods, the kernels depend upon an adjoint field, the source of which, in this case, is the reference displacement field, pre-multiplied by a matrix of cross-correlations of components of the observed field. We use a numerical simulation to evaluate the resolving power of a topographic inversion that employs the cross-convolution measure. The estimated resolving kernel shows is point-like, indicating that the cross-convolution measure will perform well in waveform tomography settings.

Tsunami Source Estimate for the 1960 Chilean Earthquake from Near- and Far-Field Observations

NASA Astrophysics Data System (ADS)

Ho, T.; Satake, K.; Watada, S.; Fujii, Y.

2017-12-01

The tsunami source of the 1960 Chilean earthquake was estimated from the near- and far-field tsunami data. The 1960 Chilean earthquake is known as the greatest earthquake instrumentally ever recorded. This earthquake caused a large tsunami which was recorded by 13 near-field tidal gauges in South America, and 84 far-field stations around the Pacific Ocean at the coasts of North America, Asia, and Oceania. The near-field stations had been used for estimating the tsunami source [Fujii and Satake, Pageoph, 2013]. However, far-field tsunami waveforms have not been utilized because of the discrepancy between observed and simulated waveforms. The observed waveforms at the far-field stations are found systematically arrived later than the simulated waveforms. This phenomenon has been also observed in the tsunami of the 2004 Sumatra earthquake, the 2010 Chilean earthquake, and the 2011 Tohoku earthquake. Recently, the factors for the travel time delay have been explained [Watada et al., JGR, 2014; Allgeyer and Cummins, GRL, 2014], so the far-field data are usable for tsunami source estimation. The phase correction method [Watada et al., JGR, 2014] converts the tsunami waveforms computed by the linear long wave into the dispersive waveform which accounts for the effects of elasticity of the Earth and ocean, ocean density stratification, and gravitational potential change associated with tsunami propagation. We apply the method to correct the computed waveforms. For the preliminary initial sea surface height inversion, we use 12 near-field stations and 63 far-field stations, located in the South and North America, islands in the Pacific Ocean, and the Oceania. The estimated tsunami source from near-field stations is compared with the result from both near- and far-field stations. Two estimated sources show a similar pattern: a large sea surface displacement concentrated at the south of the epicenter close to the coast and extended to south. However, the source estimated from near-field stations shows larger displacement than one from both dataset.

Slope tomography based on eikonal solvers and the adjoint-state method

NASA Astrophysics Data System (ADS)

Tavakoli F., B.; Operto, S.; Ribodetti, A.; Virieux, J.

2017-06-01

Velocity macromodel building is a crucial step in the seismic imaging workflow as it provides the necessary background model for migration or full waveform inversion. In this study, we present a new formulation of stereotomography that can handle more efficiently long-offset acquisition, complex geological structures and large-scale data sets. Stereotomography is a slope tomographic method based upon a semi-automatic picking of local coherent events. Each local coherent event, characterized by its two-way traveltime and two slopes in common-shot and common-receiver gathers, is tied to a scatterer or a reflector segment in the subsurface. Ray tracing provides a natural forward engine to compute traveltime and slopes but can suffer from non-uniform ray sampling in presence of complex media and long-offset acquisitions. Moreover, most implementations of stereotomography explicitly build a sensitivity matrix, leading to the resolution of large systems of linear equations, which can be cumbersome when large-scale data sets are considered. Overcoming these issues comes with a new matrix-free formulation of stereotomography: a factored eikonal solver based on the fast sweeping method to compute first-arrival traveltimes and an adjoint-state formulation to compute the gradient of the misfit function. By solving eikonal equation from sources and receivers, we make the computational cost proportional to the number of sources and receivers while it is independent of picked events density in each shot and receiver gather. The model space involves the subsurface velocities and the scatterer coordinates, while the dips of the reflector segments are implicitly represented by the spatial support of the adjoint sources and are updated through the joint localization of nearby scatterers. We present an application on the complex Marmousi model for a towed-streamer acquisition and a realistic distribution of local events. We show that the estimated model, built without any prior knowledge of the velocities, provides a reliable initial model for frequency-domain FWI of long-offset data for a starting frequency of 4 Hz, although some artefacts at the reservoir level result from a deficit of illumination. This formulation of slope tomography provides a computationally efficient alternative to waveform inversion method such as reflection waveform inversion or differential-semblance optimization to build an initial model for pre-stack depth migration and conventional FWI.

Full-waveform inversion for the Iranian plateau

NASA Astrophysics Data System (ADS)

Masouminia, N.; Fichtner, A.; Rahimi, H.

2017-12-01

We aim to obtain a detailed tomographic model for the Iranian plateau facilitated by full-waveform inversion. By using this method, we intend to better constrain the 3-D structure of the crust and the upper mantle in the region. The Iranian plateau is a complex tectonic area resulting from the collision of the Arabian and Eurasian tectonic plates. This region is subject to complex tectonic processes such as Makran subduction zone, which runs along the southeastern coast of Iran, and the convergence of the Arabian and- Eurasian plates, which itself led to another subduction under Central Iran. This continent-continent collision has also caused shortening and crustal thickening, which can be seen today as Zagros mountain range in the south and Kopeh Dagh mountain range in the northeast. As a result of such a tectonic activity, the crust and the mantle beneath the region are expected to be highly heterogeneous. To further our understanding of the region and its tectonic history, a detailed 3-D velocity model is required.To construct a 3-D model, we propose to use full-waveform inversion, which allows us to incorporate all types of waves recorded in the seismogram, including body waves as well as fundamental- and higher-mode surface waves. Exploiting more information from the observed data using this approach is likely to constrain features which have not been found by classical tomography studies so far. We address the forward problem using Salvus - a numerical wave propagation solver, based on spectral-element method and run on high-performance computers. The solver allows us to simulate wave field propagating in highly heterogeneous, attenuating and anisotropic media, respecting the surface topography. To improve the model, we solve the optimization problem. Solution of this optimization problem is based on an iterative approach which employs adjoint methods to calculate the gradient and uses steepest descent and conjugate-gradient methods to minimize the objective function. Each iteration of such an approach is expected to bring the model closer to the true model.Our model domain extends between 25°N and 40°N in latitude and 42°E and 63°E in longitude. To constrain the 3-D structure of the area we use 83 broadband seismic stations and 146 earthquakes with magnitude Mw>4.5 -that occurred in the region between 2012 and 2017.

Downscaling Smooth Tomographic Models: Separating Intrinsic and Apparent Anisotropy

NASA Astrophysics Data System (ADS)

Bodin, Thomas; Capdeville, Yann; Romanowicz, Barbara

2016-04-01

In recent years, a number of tomographic models based on full waveform inversion have been published. Due to computational constraints, the fitted waveforms are low pass filtered, which results in an inability to map features smaller than half the shortest wavelength. However, these tomographic images are not a simple spatial average of the true model, but rather an effective, apparent, or equivalent model that provides a similar 'long-wave' data fit. For example, it can be shown that a series of horizontal isotropic layers will be seen by a 'long wave' as a smooth anisotropic medium. In this way, the observed anisotropy in tomographic models is a combination of intrinsic anisotropy produced by lattice-preferred orientation (LPO) of minerals, and apparent anisotropy resulting from the incapacity of mapping discontinuities. Interpretations of observed anisotropy (e.g. in terms of mantle flow) requires therefore the separation of its intrinsic and apparent components. The "up-scaling" relations that link elastic properties of a rapidly varying medium to elastic properties of the effective medium as seen by long waves are strongly non-linear and their inverse highly non-unique. That is, a smooth homogenized effective model is equivalent to a large number of models with discontinuities. In the 1D case, Capdeville et al (GJI, 2013) recently showed that a tomographic model which results from the inversion of low pass filtered waveforms is an homogenized model, i.e. the same as the model computed by upscaling the true model. Here we propose a stochastic method to sample the ensemble of layered models equivalent to a given tomographic profile. We use a transdimensional formulation where the number of layers is variable. Furthermore, each layer may be either isotropic (1 parameter) or intrinsically anisotropic (2 parameters). The parsimonious character of the Bayesian inversion gives preference to models with the least number of parameters (i.e. least number of layers, and maximum number of isotropic layers). The non-uniqueness of the problem can be addressed by adding high frequency data such as receiver functions, able to map first order discontinuities. We show with synthetic tests that this method enables us to distinguish between intrinsic and apparent anisotropy in tomographic models, as layers with intrinsic anisotropy are only present when required by the data. A real data example is presented based on the latest global model produced at Berkeley.

Fine-scale thermohaline ocean structure retrieved with 2-D prestack full-waveform inversion of multichannel seismic data: Application to the Gulf of Cadiz (SW Iberia)

NASA Astrophysics Data System (ADS)

Dagnino, D.; Sallarès, V.; Biescas, B.; Ranero, C. R.

2016-08-01

This work demonstrates the feasibility of 2-D time-domain, adjoint-state acoustic full-waveform inversion (FWI) to retrieve high-resolution models of ocean physical parameters such as sound speed, temperature and salinity. The proposed method is first described and then applied to prestack multichannel seismic (MCS) data acquired in the Gulf of Cadiz (SW Iberia) in 2007 in the framework of the Geophysical Oceanography project. The inversion strategy flow includes specifically designed data preconditioning for acoustic noise reduction, followed by the inversion of sound speed in the shotgather domain. We show that the final sound speed model has a horizontal resolution of ˜ 70 m, which is two orders of magnitude better than that of the initial model constructed with coincident eXpendable Bathy Thermograph (XBT) data, and close to the theoretical resolution of O(λ). Temperature (T) and salinity (S) are retrieved with the same lateral resolution as sound speed by combining the inverted sound speed model with the thermodynamic equation of seawater and a local, depth-dependent T-S relation derived from regional conductivity-temperature-depth (CTD) measurements of the National Oceanic and Atmospheric Administration (NOAA) database. The comparison of the inverted T and S models with XBT and CTD casts deployed simultaneously to the MCS acquisition shows that the thermohaline contrasts are resolved with an accuracy of 0.18oC for temperature and 0.08 PSU for salinity. The combination of oceanographic and MCS data into a common, pseudo-automatic inversion scheme allows to quantitatively resolve submeso-scale features that ought to be incorporated into larger-scale ocean models of oceans structure and circulation.

Arbitrary waveform modulated pulse EPR at 200 GHz

NASA Astrophysics Data System (ADS)

Kaminker, Ilia; Barnes, Ryan; Han, Songi

2017-06-01

We report here on the implementation of arbitrary waveform generation (AWG) capabilities at ∼200 GHz into an Electron Paramagnetic Resonance (EPR) and Dynamic Nuclear Polarization (DNP) instrument platform operating at 7 T. This is achieved with the integration of a 1 GHz, 2 channel, digital to analog converter (DAC) board that enables the generation of coherent arbitrary waveforms at Ku-band frequencies with 1 ns resolution into an existing architecture of a solid state amplifier multiplier chain (AMC). This allows for the generation of arbitrary phase- and amplitude-modulated waveforms at 200 GHz with >150 mW power. We find that the non-linearity of the AMC poses significant difficulties in generating amplitude-modulated pulses at 200 GHz. We demonstrate that in the power-limited regime of ω1 < 1 MHz phase-modulated pulses were sufficient to achieve significant improvements in broadband (>10 MHz) spin manipulation in incoherent (inversion), as well as coherent (echo formation) experiments. Highlights include the improvement by one order of magnitude in inversion bandwidth compared to that of conventional rectangular pulses, as well as a factor of two in improvement in the refocused echo intensity at 200 GHz.

Optimal Inversion Parameters for Full Waveform Inversion using OBS Data Set

NASA Astrophysics Data System (ADS)

Kim, S.; Chung, W.; Shin, S.; Kim, D.; Lee, D.

2017-12-01

In recent years, full Waveform Inversion (FWI) has been the most researched technique in seismic data processing. It uses the residuals between observed and modeled data as an objective function; thereafter, the final subsurface velocity model is generated through a series of iterations meant to minimize the residuals.Research on FWI has expanded from acoustic media to elastic media. In acoustic media, the subsurface property is defined by P-velocity; however, in elastic media, properties are defined by multiple parameters, such as P-velocity, S-velocity, and density. Further, the elastic media can also be defined by Lamé constants, density or impedance PI, SI; consequently, research is being carried out to ascertain the optimal parameters.From results of advanced exploration equipment and Ocean Bottom Seismic (OBS) survey, it is now possible to obtain multi-component seismic data. However, to perform FWI on these data and generate an accurate subsurface model, it is important to determine optimal inversion parameters among (Vp, Vs, ρ), (λ, μ, ρ), and (PI, SI) in elastic media. In this study, staggered grid finite difference method was applied to simulate OBS survey. As in inversion, l2-norm was set as objective function. Further, the accurate computation of gradient direction was performed using the back-propagation technique and its scaling was done using the Pseudo-hessian matrix.In acoustic media, only Vp is used as the inversion parameter. In contrast, various sets of parameters, such as (Vp, Vs, ρ) and (λ, μ, ρ) can be used to define inversion in elastic media. Therefore, it is important to ascertain the parameter that gives the most accurate result for inversion with OBS data set.In this study, we generated Vp and Vs subsurface models by using (λ, μ, ρ) and (Vp, Vs, ρ) as inversion parameters in every iteration, and compared the final two FWI results.This research was supported by the Basic Research Project(17-3312) of the Korea Institute of Geoscience and Mineral Resources(KIGAM) funded by the Ministry of Science, ICT and Future Planning of Korea.

Studies of earthquakes and microearthquakes using near-field seismic and geodetic observations

NASA Astrophysics Data System (ADS)

O'Toole, Thomas Bartholomew

The Centroid-Moment Tensor (CMT) method allows an optimal point-source description of an earthquake to be recovered from a set of seismic observations, and, for over 30 years, has been routinely applied to determine the location and source mechanism of teleseismically recorded earthquakes. The CMT approach is, however, entirely general: any measurements of seismic displacement fields could, in theory, be used within the CMT inversion formulation, so long as the treatment of the earthquake as a point source is valid for that data. We modify the CMT algorithm to enable a variety of near-field seismic observables to be inverted for the source parameters of an earthquake. The first two data types that we implement are provided by Global Positioning System receivers operating at sampling frequencies of 1,Hz and above. When deployed in the seismic near field, these instruments may be used as long-period-strong-motion seismometers, recording displacement time series that include the static offset. We show that both the displacement waveforms, and static displacements alone, can be used to obtain CMT solutions for moderate-magnitude earthquakes, and that performing analyses using these data may be useful for earthquake early warning. We also investigate using waveform recordings - made by conventional seismometers deployed at the surface, or by geophone arrays placed in boreholes - to determine CMT solutions, and their uncertainties, for microearthquakes induced by hydraulic fracturing. A similar waveform inversion approach could be applied in many other settings where induced seismicity and microseismicity occurs..

Influence of global heterogeneities on regional imaging based upon full waveform inversion of teleseismic wavefield

NASA Astrophysics Data System (ADS)

Monteiller, Vadim; Beller, Stephen; Operto, Stephane; Virieux, Jean

2015-04-01

The current development of dense seismic arrays and high performance computing make feasible today application of full-waveform inversion (FWI) on teleseismic data for high-resolution lithospheric imaging. In teleseismic configuration, the source is often considered to first order as a planar wave that impinges the base of the lithospheric target located below the receiver array. Recently, injection methods coupling global propagation in 1D or axisymmetric earth model with regional 3D methods (Discontinuous Galerkin finite element methods, Spectral elements methods or finite differences) allow us to consider more realistic teleseismic phases. Those teleseismic phases can be propagated inside 3D regional model in order to exploit not only the forward-scattered waves propagating up to the receiver but also second-order arrivals that are back-scattered from the free-surface and the reflectors before their recordings on the surface. However, those computation are performed assuming simple global model. In this presentation, we review some key specifications that might be considered for mitigating the effect on FWI of heterogeneities situated outside the regional domain. We consider synthetic models and data computed using our recently developed hybrid method AxiSEM/SEM. The global simulation is done by AxiSEM code which allows us to consider axisymmetric anomalies. The 3D regional computation is performed by Spectral Element Method. We investigate the effect of external anomalies on the regional model obtained by FWI when one neglects them by considering only 1D global propagation. We also investigate the effect of the source time function and the focal mechanism on results of the FWI approach.

Inverts permittivity and conductivity with structural constraint in GPR FWI based on truncated Newton method

NASA Astrophysics Data System (ADS)

Ren, Qianci

2018-04-01

Full waveform inversion (FWI) of ground penetrating radar (GPR) is a promising technique to quantitatively evaluate the permittivity and conductivity of near subsurface. However, these two parameters are simultaneously inverted in the GPR FWI, increasing the difficulty to obtain accurate inversion results for both parameters. In this study, I present a structural constrained GPR FWI procedure to jointly invert the two parameters, aiming to force a structural relationship between permittivity and conductivity in the process of model reconstruction. The structural constraint is enforced by a cross-gradient function. In this procedure, the permittivity and conductivity models are inverted alternately at each iteration and updated with hierarchical frequency components in the frequency domain. The joint inverse problem is solved by the truncated Newton method which considering the effect of Hessian operator and using the approximated solution of Newton equation to be the perturbation model in the updating process. The joint inversion procedure is tested by three synthetic examples. The results show that jointly inverting permittivity and conductivity in GPR FWI effectively increases the structural similarities between the two parameters, corrects the structures of parameter models, and significantly improves the accuracy of conductivity model, resulting in a better inversion result than the individual inversion.

Rapid finite-fault inversions in Southern California using Cybershake Green's functions

NASA Astrophysics Data System (ADS)

Thio, H. K.; Polet, J.

2017-12-01

We have developed a system for rapid finite fault inversion for intermediate and large Southern California earthquakes using local, regional and teleseismic seismic waveforms as well as geodetic data. For modeling the local seismic data, we use 3D Green's functions from the Cybershake project, which were made available to us courtesy of the Southern California Earthquake Center (SCEC). The use of 3D Green's functions allows us to extend the inversion to higher frequency waveform data and smaller magnitude earthquakes, in addition to achieving improved solutions in general. The ultimate aim of this work is to develop the ability to provide high quality finite fault models within a few hours after any damaging earthquake in Southern California, so that they may be used as input to various post-earthquake assessment tools such as ShakeMap, as well as by the scientific community and other interested parties. Additionally, a systematic determination of finite fault models has value as a resource for scientific studies on detailed earthquake processes, such as rupture dynamics and scaling relations. We are using an established least-squares finite fault inversion method that has been applied extensively both on large as well as smaller regional earthquakes, in conjunction with the 3D Green's functions, where available, as well as 1D Green's functions for areas for which the Cybershake library has not yet been developed. We are carrying out validation and calibration of this system using significant earthquakes that have occurred in the region over the last two decades, spanning a range of locations and magnitudes (5.4 and higher).

Expanding the frontiers of waveform imaging with Salvus

NASA Astrophysics Data System (ADS)

Afanasiev, M.; Boehm, C.; van Driel, M.; Krischer, L.; Fichtner, A.

2017-12-01

Mechanical waves are natural harbingers of information. From medical ultrasound to the normal modes of Sun, wave motion is often our best window into the character of some underlying continuum. For over a century, geophysicists have been using this window to peer deep into the Earth, developing techniques that have gone on to underlie much of world's energy economy. As computers and numerical techniques have become more powerful over the last several decades, seismologists have begun to scale back classical simplifying approximations of wave propagation physics. As a result, we are now approaching the ideal of `full-waveform inversion'; maximizing the aperture of our window by taking the full complexity of wave motion into account.Salvus is a modern high-performance software suite which aims to bring recent developments in geophysical waveform inversion to new and exciting domains. In this short presentation we will look at the connections between these applications, with examples from non-destructive testing, medical imaging, seismic exploration, and (extra-) planetary seismology.

A Sensitivity Analysis of Tsunami Inversions on the Number of Stations

NASA Astrophysics Data System (ADS)

An, Chao; Liu, Philip L.-F.; Meng, Lingsen

2018-05-01

Current finite-fault inversions of tsunami recordings generally adopt as many tsunami stations as possible to better constrain earthquake source parameters. In this study, inversions are evaluated by the waveform residual that measures the difference between model predictions and recordings, and the dependence of the quality of inversions on the number tsunami stations is derived. Results for the 2011 Tohoku event show that, if the tsunami stations are optimally located, the waveform residual decreases significantly with the number of stations when the number is 1 ˜ 4 and remains almost constant when the number is larger than 4, indicating that 2 ˜ 4 stations are able to recover the main characteristics of the earthquake source. The optimal location of tsunami stations is explained in the text. Similar analysis is applied to the Manila Trench in the South China Sea using artificially generated earthquakes and virtual tsunami stations. Results confirm that 2 ˜ 4 stations are necessary and sufficient to constrain the earthquake source parameters, and the optimal sites of stations are recommended in the text. The conclusion is useful for the design of new tsunami warning systems. Current strategies of tsunameter network design mainly focus on the early detection of tsunami waves from potential sources to coastal regions. We therefore recommend that, in addition to the current strategies, the waveform residual could also be taken into consideration so as to minimize the error of tsunami wave prediction for warning purposes.

The Variability and Interpretation of Earthquake Source Mechanisms in The Geysers Geothermal Field From a Bayesian Standpoint Based on the Choice of a Noise Model

NASA Astrophysics Data System (ADS)

Mustać, Marija; Tkalčić, Hrvoje; Burky, Alexander L.

2018-01-01

Moment tensor (MT) inversion studies of events in The Geysers geothermal field mostly focused on microseismicity and found a large number of earthquakes with significant non-double-couple (non-DC) seismic radiation. Here we concentrate on the largest events in the area in recent years using a hierarchical Bayesian MT inversion. Initially, we show that the non-DC components of the MT can be reliably retrieved using regional waveform data from a small number of stations. Subsequently, we present results for a number of events and show that accounting for noise correlations can lead to retrieval of a lower isotropic (ISO) component and significantly different focal mechanisms. We compute the Bayesian evidence to compare solutions obtained with different assumptions of the noise covariance matrix. Although a diagonal covariance matrix produces a better waveform fit, inversions that account for noise correlations via an empirically estimated noise covariance matrix account for interdependences of data errors and are preferred from a Bayesian point of view. This implies that improper treatment of data noise in waveform inversions can result in fitting the noise and misinterpreting the non-DC components. Finally, one of the analyzed events is characterized as predominantly DC, while the others still have significant non-DC components, probably as a result of crack opening, which is a reasonable hypothesis for The Geysers geothermal field geological setting.

Line-source simulation for shallow-seismic data. Part 2: full-waveform inversion—a synthetic 2-D case study

NASA Astrophysics Data System (ADS)

Schäfer, M.; Groos, L.; Forbriger, T.; Bohlen, T.

2014-09-01

Full-waveform inversion (FWI) of shallow-seismic surface waves is able to reconstruct lateral variations of subsurface elastic properties. Line-source simulation for point-source data is required when applying algorithms of 2-D adjoint FWI to recorded shallow-seismic field data. The equivalent line-source response for point-source data can be obtained by convolving the waveforms with √{t^{-1}} (t: traveltime), which produces a phase shift of π/4. Subsequently an amplitude correction must be applied. In this work we recommend to scale the seismograms with √{2 r v_ph} at small receiver offsets r, where vph is the phase velocity, and gradually shift to applying a √{t^{-1}} time-domain taper and scaling the waveforms with r√{2} for larger receiver offsets r. We call this the hybrid transformation which is adapted for direct body and Rayleigh waves and demonstrate its outstanding performance on a 2-D heterogeneous structure. The fit of the phases as well as the amplitudes for all shot locations and components (vertical and radial) is excellent with respect to the reference line-source data. An approach for 1-D media based on Fourier-Bessel integral transformation generates strong artefacts for waves produced by 2-D structures. The theoretical background for both approaches is presented in a companion contribution. In the current contribution we study their performance when applied to waves propagating in a significantly 2-D-heterogeneous structure. We calculate synthetic seismograms for 2-D structure for line sources as well as point sources. Line-source simulations obtained from the point-source seismograms through different approaches are then compared to the corresponding line-source reference waveforms. Although being derived by approximation the hybrid transformation performs excellently except for explicitly back-scattered waves. In reconstruction tests we further invert point-source synthetic seismograms by a 2-D FWI to subsurface structure and evaluate its ability to reproduce the original structural model in comparison to the inversion of line-source synthetic data. Even when applying no explicit correction to the point-source waveforms prior to inversion only moderate artefacts appear in the results. However, the overall performance is best in terms of model reproduction and ability to reproduce the original data in a 3-D simulation if inverted waveforms are obtained by the hybrid transformation.

Ultra-low velocity zones beneath the Philippine and Tasman Seas revealed by a trans-dimensional Bayesian waveform inversion

NASA Astrophysics Data System (ADS)

Pachhai, Surya; Dettmer, Jan; Tkalčić, Hrvoje

2015-11-01

Ultra-low velocity zones (ULVZs) are small-scale structures in the Earth's lowermost mantle inferred from the analysis of seismological observations. These structures exhibit a strong decrease in compressional (P)-wave velocity, shear (S)-wave velocity, and an increase in density. Quantifying the elastic properties of ULVZs is crucial for understanding their physical origin, which has been hypothesized either as partial melting, iron enrichment, or a combination of the two. Possible disambiguation of these hypotheses can lead to a better understanding of the dynamic processes of the lowermost mantle, such as, percolation, stirring and thermochemical convection. To date, ULVZs have been predominantly studied by forward waveform modelling of seismic waves that sample the core-mantle boundary region. However, ULVZ parameters (i.e. velocity, density, and vertical and lateral extent) obtained through forward modelling are poorly constrained because inferring Earth structure from seismic observations is a non-linear inverse problem with inherent non-uniqueness. To address these issues, we developed a trans-dimensional hierarchical Bayesian inversion that enables rigorous estimation of ULVZ parameter values and their uncertainties, including the effects of model selection. The model selection includes treating the number of layers and the vertical extent of the ULVZ as unknowns. The posterior probability density (solution to the inverse problem) of the ULVZ parameters is estimated by reversible jump Markov chain Monte Carlo sampling that employs parallel tempering to improve efficiency/convergence. First, we apply our method to study the resolution of complex ULVZ structure (including gradually varying structure) by probabilistically inverting simulated noisy waveforms. Then, two data sets sampling the CMB beneath the Philippine and Tasman Seas are considered in the inversion. Our results indicate that both ULVZs are more complex than previously suggested. For the Philippine Sea data, we find a strong decrease in S-wave velocity, which indicates the presence of iron-rich material, albeit this result is accompanied with larger parameter uncertainties than in a previous study. For the Tasman Sea data, our analysis yields a well-constrained S-wave velocity that gradually decreases with depth. We conclude that this ULVZ represents a partial melt of iron-enriched material with higher melt content near its bottom.

Bayesian ISOLA: new tool for automated centroid moment tensor inversion

NASA Astrophysics Data System (ADS)

Vackář, Jiří; Burjánek, Jan; Gallovič, František; Zahradník, Jiří; Clinton, John

2017-08-01

We have developed a new, fully automated tool for the centroid moment tensor (CMT) inversion in a Bayesian framework. It includes automated data retrieval, data selection where station components with various instrumental disturbances are rejected and full-waveform inversion in a space-time grid around a provided hypocentre. A data covariance matrix calculated from pre-event noise yields an automated weighting of the station recordings according to their noise levels and also serves as an automated frequency filter suppressing noisy frequency ranges. The method is tested on synthetic and observed data. It is applied on a data set from the Swiss seismic network and the results are compared with the existing high-quality MT catalogue. The software package programmed in Python is designed to be as versatile as possible in order to be applicable in various networks ranging from local to regional. The method can be applied either to the everyday network data flow, or to process large pre-existing earthquake catalogues and data sets.

Joint inversion of time-lapse VSP data for monitoring CO2 injection at the Farnsworth EOR field in Texas

NASA Astrophysics Data System (ADS)

Zhang, M.; Gao, K.; Balch, R. S.; Huang, L.

2016-12-01

During the Development Phase (Phase III) of the U.S. Southwest Regional Partnership on Carbon Sequestration (SWP), time-lapse 3D vertical seismic profiling (VSP) data were acquired to monitor CO2 injection/migration at the Farnsworth Enhanced Oil Recovery (EOR) field, in partnership with the industrial partner Chaparral Energy. The project is to inject a million tons of carbon dioxide into the target formation, the deep oil-bearing Morrow Formation in the Farnsworth Unit EOR field. Quantitative time-lapse seismic monitoring has the potential to track CO2 movement in geologic carbon storage sites. Los Alamos National Laboratory (LANL) has recently developed new full-waveform inversion methods to jointly invert time-lapse seismic data for changes in elastic and anisotropic parameters in target monitoring regions such as a CO2 reservoir. We apply our new joint inversion methods to time-lapse VSP data acquired at the Farnsworth EOR filed, and present some preliminary results showing geophysical properties changes in the reservoir.

Rupture process of the 2016 Mw 7.8 Ecuador earthquake from joint inversion of InSAR data and teleseismic P waveforms

NASA Astrophysics Data System (ADS)

Yi, Lei; Xu, Caijun; Wen, Yangmao; Zhang, Xu; Jiang, Guoyan

2018-01-01

The 2016 Ecuador earthquake ruptured the Ecuador-Colombia subduction interface where several historic megathrust earthquakes had occurred. In order to determine a detailed rupture model, Interferometric Synthetic Aperture Radar (InSAR) images and teleseismic data sets were objectively weighted by using a modified Akaika's Bayesian Information Criterion (ABIC) method to jointly invert for the rupture process of the earthquake. In modeling the rupture process, a constrained waveform length method, unlike the traditional subjective selected waveform length method, was used since the lengths of inverted waveforms were strictly constrained by the rupture velocity and rise time (the slip duration time). The optimal rupture velocity and rise time of the earthquake were estimated from grid search, which were determined to be 2.0 km/s and 20 s, respectively. The inverted model shows that the event is dominated by thrust movement and the released moment is 5.75 × 1020 Nm (Mw 7.77). The slip distribution extends southward along the Ecuador coast line in an elongated stripe at a depth between 10 and 25 km. The slip model is composed of two asperities and slipped over 4 m. The source time function is approximate 80 s that separated into two segments corresponding to the two asperities. The small magnitude of the slip occurred in the updip section of the fault plane resulted in small tsunami waves that were verified by observations near the coast. We suggest a possible situation that the rupture zone of the 2016 earthquake is likely not overlapped with that of the 1942 earthquake.

Platform for Postprocessing Waveform-Based NDE

NASA Technical Reports Server (NTRS)

Roth, Don

2008-01-01

Taking advantage of the similarities that exist among all waveform-based non-destructive evaluation (NDE) methods, a common software platform has been developed containing multiple- signal and image-processing techniques for waveforms and images. The NASA NDE Signal and Image Processing software has been developed using the latest versions of LabVIEW, and its associated Advanced Signal Processing and Vision Toolkits. The software is useable on a PC with Windows XP and Windows Vista. The software has been designed with a commercial grade interface in which two main windows, Waveform Window and Image Window, are displayed if the user chooses a waveform file to display. Within these two main windows, most actions are chosen through logically conceived run-time menus. The Waveform Window has plots for both the raw time-domain waves and their frequency- domain transformations (fast Fourier transform and power spectral density). The Image Window shows the C-scan image formed from information of the time-domain waveform (such as peak amplitude) or its frequency-domain transformation at each scan location. The user also has the ability to open an image, or series of images, or a simple set of X-Y paired data set in text format. Each of the Waveform and Image Windows contains menus from which to perform many user actions. An option exists to use raw waves obtained directly from scan, or waves after deconvolution if system wave response is provided. Two types of deconvolution, time-based subtraction or inverse-filter, can be performed to arrive at a deconvolved wave set. Additionally, the menu on the Waveform Window allows preprocessing of waveforms prior to image formation, scaling and display of waveforms, formation of different types of images (including non-standard types such as velocity), gating of portions of waves prior to image formation, and several other miscellaneous and specialized operations. The menu available on the Image Window allows many further image processing and analysis operations, some of which are found in commercially-available image-processing software programs (such as Adobe Photoshop), and some that are not (removing outliers, Bscan information, region-of-interest analysis, line profiles, and precision feature measurements).

Resolution of VTI anisotropy with elastic full-waveform inversion: theory and basic numerical examples

NASA Astrophysics Data System (ADS)

Podgornova, O.; Leaney, S.; Liang, L.

2018-07-01

Extracting medium properties from seismic data faces some limitations due to the finite frequency content of the data and restricted spatial positions of the sources and receivers. Some distributions of the medium properties make low impact on the data (including none). If these properties are used as the inversion parameters, then the inverse problem becomes overparametrized, leading to ambiguous results. We present an analysis of multiparameter resolution for the linearized inverse problem in the framework of elastic full-waveform inversion. We show that the spatial and multiparameter sensitivities are intertwined and non-sensitive properties are spatial distributions of some non-trivial combinations of the conventional elastic parameters. The analysis accounts for the Hessian information and frequency content of the data; it is semi-analytical (in some scenarios analytical), easy to interpret and enhances results of the widely used radiation pattern analysis. Single-type scattering is shown to have limited sensitivity, even for full-aperture data. Finite-frequency data lose multiparameter sensitivity at smooth and fine spatial scales. Also, we establish ways to quantify a spatial-multiparameter coupling and demonstrate that the theoretical predictions agree well with the numerical results.

The Ellipticity Filter-A Proposed Solution to the Mixed Event Problem in Nuclear Seismic Discrimination

DTIC Science & Technology

1974-09-07

ellipticity filter. The source waveforms are recreated by an inverse transform of those complex ampli- tudes associated with the same azimuth...terms of the three complex data points and the ellipticity. Having solved the equations for all frequency bins, the inverse transform of...Transform of those complex amplitudes associated with Source 1, yielding the signal a (t). Similarly, take the inverse Transform of all

Seismic wave-speed structure beneath the metropolitan area of Japan based on adjoint tomography

NASA Astrophysics Data System (ADS)

Miyoshi, T.; Obayashi, M.; Tono, Y.; Tsuboi, S.

2015-12-01

We have obtained a three-dimensional (3D) model of seismic wave-speed structure beneath the metropolitan area of Japan. We applied the spectral-element method (e.g. Komatitsch and Tromp 1999) and adjoint method (Liu and Tromp 2006) to the broadband seismograms in order to infer the 3D model. We used the travel-time tomography result (Matsubara and Obara 2011) as an initial 3D model and used broadband waveforms recorded at the NIED F-net stations. We selected 147 earthquakes with magnitude of larger than 4.5 from the F-net earthquake catalog and used their bandpass filtered seismograms between 5 and 20 second with a high S/N ratio. The 3D model used for the forward and adjoint simulations is represented as a region of approximately 500 by 450 km in horizontal and 120 km in depth. Minimum period of theoretical waveforms was 4.35 second. For the adjoint inversion, we picked up the windows of the body waves from the observed and theoretical seismograms. We used SPECFEM3D_Cartesian code (e.g. Peter et al. 2011) for the forward and adjoint simulations, and their simulations were implemented by K-computer in RIKEN. Each iteration required about 0.1 million CPU hours at least. The model parameters of Vp and Vs were updated by using the steepest descent method. We obtained the fourth iterative model (M04), which reproduced observed waveforms better than the initial model. The shear wave-speed of M04 was significantly smaller than the initial model at any depth. The model of compressional wave-speed was not improved by inversion because of small alpha kernel values. Acknowledgements: This research was partly supported by MEXT Strategic Program for Innovative Research. We thank to the NIED for providing seismological data.

Measurement of leaky Lamb wave dispersion curves with application on coating characterization

NASA Astrophysics Data System (ADS)

Lee, Yung-Chun; Cheng, Sheng Wen

2001-04-01

This paper describes a new measurement system for measuring dispersion curves of leaky Lamb waves. The measurement system is based on a focusing PVDF transducer, the defocusing measurement, the V(f,z) waveform processing method, and an image displaying technique. The measurement system is applied for the determination of thin-film elastic properties, namely Young's modulus and shear modulus, by the inversion of dispersion curves measured from a thin-film/plate configuration. Elastic constants of electro-deposited nickel layers are determined with this method.

3-D acoustic waveform simulation and inversion at Yasur Volcano, Vanuatu

NASA Astrophysics Data System (ADS)

Iezzi, A. M.; Fee, D.; Matoza, R. S.; Austin, A.; Jolly, A. D.; Kim, K.; Christenson, B. W.; Johnson, R.; Kilgour, G.; Garaebiti, E.; Kennedy, B.; Fitzgerald, R.; Key, N.

2016-12-01

Acoustic waveform inversion shows promise for improved eruption characterization that may inform volcano monitoring. Well-constrained inversion can provide robust estimates of volume and mass flux, increasing our ability to monitor volcanic emissions (potentially in real-time). Previous studies have made assumptions about the multipole source mechanism, which can be thought of as the combination of pressure fluctuations from a volume change, directionality, and turbulence. This infrasound source could not be well constrained up to this time due to infrasound sensors only being deployed on Earth's surface, so the assumption of no vertical dipole component has been made. In this study we deploy a high-density seismo-acoustic network, including multiple acoustic sensors along a tethered balloon around Yasur Volcano, Vanuatu. Yasur has frequent strombolian eruptions from any one of its three active vents within a 400 m diameter crater. The third dimension (vertical) of pressure sensor coverage allows us to begin to constrain the acoustic source components in a profound way, primarily the horizontal and vertical components and their previously uncharted contributions to volcano infrasound. The deployment also has a geochemical and visual component, including FLIR, FTIR, two scanning FLYSPECs, and a variety of visual imagery. Our analysis employs Finite-Difference Time-Domain (FDTD) modeling to obtain the full 3D Green's functions for each propagation path. This method, following Kim et al. (2015), takes into account realistic topographic scattering based on a digital elevation model created using structure-from-motion techniques. We then invert for the source location and source-time function, constraining the contribution of the vertical sound radiation to the source. The final outcome of this inversion is an infrasound-derived volume flux as a function of time, which we then compare to those derived independently from geochemical techniques as well as the inversion of seismic data. Kim, K., Fee, D., Yokoo, A., & Lees, J. M. (2015). Acoustic source inversion to estimate volume flux from volcanic explosions. Geophysical Research Letters, 42(13), 5243-5249

Toward regional corrections of long period CMT inversions using InSAR

NASA Astrophysics Data System (ADS)

Shakibay Senobari, N.; Funning, G.; Ferreira, A. M.

2017-12-01

One of InSAR's main strengths, with respect to other methods of studying earthquakes, is finding the accurate location of the best point source (or `centroid') for an earthquake. While InSAR data have great advantages for study of shallow earthquakes, the number of earthquakes for which we have InSAR data is low, compared with the number of earthquakes recorded seismically. And though improvements to SAR satellite constellations have enhanced the use of InSAR data during earthquake response, post-event data still have a latency on the order of days. On the other hand, earthquake centroid inversion methods using long period seismic data (e.g. the Global CMT method) are fast but include errors caused by inaccuracies in both the Earth velocity model and in wave propagation assumptions (e.g. Hjörleifsdóttir and Ekström, 2010; Ferreira and Woodhouse, 2006). Here we demonstrate a method that combines the strengths of both methods, calculating regional travel-time corrections for long-period waveforms using accurate centroid locations from InSAR, then applying these to other events that occur in the same region. Our method is based on the observation that synthetic seismograms produced from InSAR source models and locations match the data very well except for some phase shifts (travel time biases) between the two waveforms, likely corresponding to inaccuracies in Earth velocity models (Weston et al., 2014). Our previous work shows that adding such phase shifts to the Green's functions can improve the accuracy of long period seismic CMT inversions by reducing tradeoffs between the moment tensor components and centroid location (e.g. Shakibay Senobari et al., AGU Fall Meeting 2015). Preliminary work on several pairs of neighboring events (e.g. Landers-Hector Mine, the 2000 South Iceland earthquake sequences) shows consistent azimuthal patterns of these phase shifts for nearby events at common stations. These phase shift patterns strongly suggest that it is possible to determine regional corrections for the source regions of these events. The aim of this project is to perform a full CMT inversion using the phase shift corrections, calculated for nearby events, to observe improvement in CMT locations and solutions. We will demonstrate our method on the five M 6 events that occurred in central Italy between 1997 and 2016.

Time-domain full-waveform inversion of Rayleigh and Love waves in presence of free-surface topography

NASA Astrophysics Data System (ADS)

Pan, Yudi; Gao, Lingli; Bohlen, Thomas

2018-05-01

Correct estimation of near-surface seismic-wave velocity when encountering lateral heterogeneity and free surface topography is one of the challenges to current shallow seismic. We propose to use time-domain full-waveform inversion (FWI) of surface waves, including both Rayleigh and Love waves, to solve this problem. We adopt a 2D time-domain finite-difference method with an improved vacuum formulation (IVF) to simulate shallow-seismic Rayleigh wave in presence of free-surface topography. We modify the IVF for SH-wave equation for the simulation of Love wave in presence of topographic free surface and prove its accuracy by benchmark tests. Checkboard model tests are performed in both cases when free-surface topography is included or neglected in FWI. Synthetic model containing a dipping planar free surface and lateral heterogeneity was then tested, in both cases of considering and neglecting free-surface topography. Both checkerboard and synthetic models show that Rayleigh- and Love-wave FWI have similar ability of reconstructing near-surface structures when free-surface topography is considered, while Love-wave FWI could reconstruct near-surface structures better than Rayleigh-wave when free-surface topography is neglected.

Waveform inversion of mantle Love waves: The born seismogram approach

NASA Technical Reports Server (NTRS)

Tanimoto, T.

1983-01-01

Normal mode theory, extended to the slightly laterally heterogeneous Earth by the first-order Born approximation, is applied to the waveform inversion of mantle Love waves (200-500 sec) for the Earth's lateral heterogeneity at l=2 and a spherically symmetric anelasticity (Q sub mu) structure. The data are from the Global Digital Seismograph Network (GDSN). The l=2 pattern is very similar to the results of other studies that used either different methods, such as phase velocity measurements and multiplet location measurements, or a different data set, such as mantle Rayleigh waves from different instruments. The results are carefully analyzed for variance reduction and are most naturally explained by heterogeneity in the upper 420 km. Because of the poor resolution of the data set for the deep interior, however, a fairly large heterogeneity in the transition zones, of the order of up to 3.5% in shear wave velocity, is allowed. It is noteworthy that Love waves of this period range can not constrain the structure below 420 km and thus any model presented by similar studies below this depth are likely to be constrained by Rayleigh waves (spheroidal modes) only.

Waveform inversion of mantle Love waves - The Born seismogram approach

NASA Technical Reports Server (NTRS)

Tanimoto, T.

1984-01-01

Normal mode theory, extended to the slightly laterally heterogeneous earth by the first-order Born approximation, is applied to the waveform inversion of mantle Love waves (200-500 sec) for the earth's lateral heterogeneity at l = 2 and a spherically symmetric anelasticity (Q sub mu) structure. The data are from the Global Digital Seismograph Network (GDSN). The l = 2 pattern is very similar to the results of other studies that used either different methods, such as phase velocity measurements and multiplet location measurements, or a different data set, such as mantle Rayleigh waves from different instruments. The results are carefully analyzed for variance reduction and are most naturally explained by heterogeneity in the upper 420 km. Because of the poor resolution of the data set for the deep interior, however, a fairly large heterogeneity in the transition zones, of the order of up to 3.5 percent in shear wave velocity, is allowed. It is noteworthy that Love waves of this period range can not constrain the structure below 420 km and thus any model presented by similar studies below this depth are likely to be constrained by Rayleigh waves (spheroidal modes) only.

Wavefield complexity and stealth structures: Resolution constraints by wave physics

NASA Astrophysics Data System (ADS)

Nissen-Meyer, T.; Leng, K.

2017-12-01

Imaging the Earth's interior relies on understanding how waveforms encode information from heterogeneous multi-scale structure. This relation is given by elastodynamics, but forward modeling in the context of tomography primarily serves to deliver synthetic waveforms and gradients for the inversion procedure. While this is entirely appropriate, it depreciates a wealth of complementary inference that can be obtained from the complexity of the wavefield. Here, we are concerned with the imprint of realistic multi-scale Earth structure on the wavefield, and the question on the inherent physical resolution limit of structures encoded in seismograms. We identify parameter and scattering regimes where structures remain invisible as a function of seismic wavelength, structural multi-scale geometry, scattering strength, and propagation path. Ultimately, this will aid in interpreting tomographic images by acknowledging the scope of "forgotten" structures, and shall offer guidance for optimising the selection of seismic data for tomography. To do so, we use our novel 3D modeling method AxiSEM3D which tackles global wave propagation in visco-elastic, anisotropic 3D structures with undulating boundaries at unprecedented resolution and efficiency by exploiting the inherent azimuthal smoothness of wavefields via a coupled Fourier expansion-spectral-element approach. The method links computational cost to wavefield complexity and thereby lends itself well to exploring the relation between waveforms and structures. We will show various examples of multi-scale heterogeneities which appear or disappear in the waveform, and argue that the nature of the structural power spectrum plays a central role in this. We introduce the concept of wavefield learning to examine the true wavefield complexity for a complexity-dependent modeling framework and discriminate which scattering structures can be retrieved by surface measurements. This leads to the question of physical invisibility and the tomographic resolution limit, and offers insight as to why tomographic images still show stark differences for smaller-scale heterogeneities despite progress in modeling and data resolution. Finally, we give an outlook on how we expand this modeling framework towards an inversion procedure guided by wavefield complexity.

Crustal Structure Beneath Taiwan Using Frequency-band Inversion of Receiver Function Waveforms

NASA Astrophysics Data System (ADS)

Tomfohrde, D. A.; Nowack, R. L.

Receiver function analysis is used to determine local crustal structure beneath Taiwan. We have performed preliminary data processing and polarization analysis for the selection of stations and events and to increase overall data quality. Receiver function analysis is then applied to data from the Taiwan Seismic Network to obtain radial and transverse receiver functions. Due to the limited azimuthal coverage, only the radial receiver functions are analyzed in terms of horizontally layered crustal structure for each station. In order to improve convergence of the receiver function inversion, frequency-band inversion (FBI) is implemented, in which an iterative inversion procedure with sequentially higher low-pass corner frequencies is used to stabilize the waveform inversion. Frequency-band inversion is applied to receiver functions at six stations of the Taiwan Seismic Network. Initial 20-layer crustal models are inverted for using prior tomographic results for the initial models. The resulting 20-1ayer models are then simplified to 4 to 5 layer models and input into an alternating depth and velocity frequency-band inversion. For the six stations investigated, the resulting simplified models provide an average estimate of 38 km for the Moho thickness surrounding the Central Range of Taiwan. Also, the individual station estimates compare well with the recent tomographic model of and the refraction results of Rau and Wu (1995) and the refraction results of Ma and Song (1997).

Frequency domain and full waveform time domain inversion of ground based magnetometer, electrometer and incoherent scattering radar arrays to image strongly heterogenous 3-D Earth structure, ionospheric structure, and to predict the intensity of GICs in the power grid

NASA Astrophysics Data System (ADS)

Schultz, A.; Imamura, N.; Bonner, L. R., IV; Cosgrove, R. B.

2016-12-01

Ground-based magnetometer and electrometer arrays provide the means to probe the structure of the Earth's interior, the interactions of space weather with the ionosphere, and to anticipate the intensity of geomagnetically induced currents (GICs) in power grids. We present a local-to-continental scale view of a heterogeneous 3-D crust and mantle as determined from magnetotelluric (MT) observations across arrays of ground-based electric and magnetic field sensors. MT impedance tensors describe the relationship between electric and magnetic fields at a given site, thus implicitly they contain all known information on the 3-D electrical resistivity structure beneath and surrounding that site. By using multivariate transfer functions to project real-time magnetic observatory network data to areas surrounding electric power grids, and by projecting those magnetic fields through MT impedance tensors, the projected magnetic field can be transformed into predictions of electric fields along the path of the transmission lines, an essential element of predicting the intensity of GICs in the grid. Finally, we explore GICs, i.e. Earth-ionosphere coupling directly in the time-domain. We consider the fully coupled EM system, where we allow for a non-stationary ionospheric source field of arbitrary complexity above a 3-D Earth. We solve the simultaneous inverse problem for 3-D Earth conductivity and source field structure directly in the time domain. In the present work, we apply this method to magnetotelluric data obtained from a synchronously operating array of 25 MT stations that collected continuous MT waveform data in the interior of Alaska during the autumn and winter of 2015 under the footprint of the Poker Flat (Alaska) Incoherent Scattering Radar (PFISR). PFISR data yield functionals of the ionospheric electric field and ionospheric conductivity that constrain the MT source field. We show that in this region conventional robust MT processing methods struggle to produce reliable MT response functions at periods much greater than about 2,000 s, a consequence, we believe, of the complexity of the ionospheric source fields in this high latitude setting. This provides impetus for direct waveform inversion methods that dispense with typical parametric assumptions made about the MT source fields.

Recovering Long-wavelength Velocity Models using Spectrogram Inversion with Single- and Multi-frequency Components

NASA Astrophysics Data System (ADS)

Ha, J.; Chung, W.; Shin, S.

2015-12-01

Many waveform inversion algorithms have been proposed in order to construct subsurface velocity structures from seismic data sets. These algorithms have suffered from computational burden, local minima problems, and the lack of low-frequency components. Computational efficiency can be improved by the application of back-propagation techniques and advances in computing hardware. In addition, waveform inversion algorithms, for obtaining long-wavelength velocity models, could avoid both the local minima problem and the effect of the lack of low-frequency components in seismic data. In this study, we proposed spectrogram inversion as a technique for recovering long-wavelength velocity models. In spectrogram inversion, decomposed frequency components from spectrograms of traces, in the observed and calculated data, are utilized to generate traces with reproduced low-frequency components. Moreover, since each decomposed component can reveal the different characteristics of a subsurface structure, several frequency components were utilized to analyze the velocity features in the subsurface. We performed the spectrogram inversion using a modified SEG/SEGE salt A-A' line. Numerical results demonstrate that spectrogram inversion could also recover the long-wavelength velocity features. However, inversion results varied according to the frequency components utilized. Based on the results of inversion using a decomposed single-frequency component, we noticed that robust inversion results are obtained when a dominant frequency component of the spectrogram was utilized. In addition, detailed information on recovered long-wavelength velocity models was obtained using a multi-frequency component combined with single-frequency components. Numerical examples indicate that various detailed analyses of long-wavelength velocity models can be carried out utilizing several frequency components.

A Comparison between Predicted and Observed Atmospheric States and their Effects on Infrasonic Source Time Function Inversion at Source Physics Experiment 6

NASA Astrophysics Data System (ADS)

Aur, K. A.; Poppeliers, C.; Preston, L. A.

2017-12-01

The Source Physics Experiment (SPE) consists of a series of underground chemical explosions at the Nevada National Security Site (NNSS) designed to gain an improved understanding of the generation and propagation of physical signals in the near and far field. Characterizing the acoustic and infrasound source mechanism from underground explosions is of great importance to underground explosion monitoring. To this end we perform full waveform source inversion of infrasound data collected from the SPE-6 experiment at distances from 300 m to 6 km and frequencies up to 20 Hz. Our method requires estimating the state of the atmosphere at the time of each experiment, computing Green's functions through these atmospheric models, and subsequently inverting the observed data in the frequency domain to obtain a source time function. To estimate the state of the atmosphere at the time of the experiment, we utilize the Weather Research and Forecasting - Data Assimilation (WRF-DA) modeling system to derive a unified atmospheric state model by combining Global Energy and Water Cycle Experiment (GEWEX) Continental-scale International Project (GCIP) data and locally obtained sonde and surface weather observations collected at the time of the experiment. We synthesize Green's functions through these atmospheric models using Sandia's moving media acoustic propagation simulation suite (TDAAPS). These models include 3-D variations in topography, temperature, pressure, and wind. We compare inversion results using the atmospheric models derived from the unified weather models versus previous modeling results and discuss how these differences affect computed source waveforms with respect to observed waveforms at various distances. Sandia National Laboratories is a multi-mission laboratory managed and operated by National Technology and Engineering Solutions of Sandia LLC, a wholly owned subsidiary of Honeywell International Inc. for the U.S. Department of Energy's National Nuclear Security Administration under contract DE-NA0003525.

An Oscillation of the Crack-like Conduit at Nevado del Ruiz Volcano, Colombia, Inferred from Multi-band Analyses of Very Long Period Seismic Events

NASA Astrophysics Data System (ADS)

Maeda, Y.; Kumagai, H.; Londono, J. M.; Lopez, C. M.; Castaño, L. M.; Beatriz, B.; García, L.

2017-12-01

Nevado del Ruiz is an active volcano in Colombia, which continues eruption activity and has been monitored by 13 broadband and 3 short-period seismic stations. In 2015-2016, a joint Japan-Colombia team installed an automatic event detection and location system based on the amplitude source location (ASL) method. Kumagai et al. (IAVCEI, 2017) indicated the existence of a magma conduit extending from the NW flank to the summit based on ASL analyses of various seismic signals including long-period (LP) and very long period (VLP) events and tremors in a 5-10 Hz frequency band. In this study, we analyzed the VLP events by waveform inversion using eight summit stations in a frequency band of 0.3-0.7 Hz. We selected 14 VLP events from May to December 2016 based on signal-to-noise ratios and simplicity of the waveforms. We assumed a homogeneous P-wave velocity of 3.5 km/s with topography in the calculation of the Green functions. We conducted frequency-domain waveform inversion assuming a tensile crack source and investigated the best location and orientation of the crack by a grid search. The inversion results pointed to a low-angle ( 30°) NW-dipping crack near the top of the conduit (approximately 1 km below the summit). The estimated source time functions displayed two or three cycles of oscillations with the seismic moment of order of 1010-1011 N m. For these 14 events, the ASLs from the 5-10 Hz frequency band were also near the top of the conduit. These results suggest the VLP and high-frequency signals are generated by an oscillation of the crack-like conduit near the summit, which may be triggered by a volume change of magma ascending in the conduit.

Effects of dwell time of excitation waveform on meniscus movements for a tubular piezoelectric print-head: experiments and model

NASA Astrophysics Data System (ADS)

Chang, Jiaqing; Liu, Yaxin; Huang, Bo

2017-07-01

In inkjet applications, it is normal to search for an optimal drive waveform when dispensing a fresh fluid or adjusting a newly fabricated print-head. To test trial waveforms with different dwell times, a camera and a strobe light were used to image the protruding or retracting liquid tongues without ejecting any droplets. An edge detection method was used to calculate the lengths of the liquid tongues to draw the meniscus movement curves. The meniscus movement is determined by the time-domain response of the acoustic pressure at the nozzle of the print-head. Starting at the inverse piezoelectric effect, a mathematical model which considers the liquid viscosity in acoustic propagation is constructed to study the acoustic pressure response at the nozzle of the print-head. The liquid viscosity retards the propagation speed and dampens the harmonic amplitude. The pressure response, which is the combined effect of the acoustic pressures generated during the rising time and the falling time and after their propagations and reflections, explains the meniscus movements well. Finally, the optimal dwell time for droplet ejections is discussed.

Simultaneous inversion of intrinsic and scattering attenuation parameters incorporating multiple scattering effect

NASA Astrophysics Data System (ADS)

Ogiso, M.

2017-12-01

Heterogeneous attenuation structure is important for not only understanding the earth structure and seismotectonics, but also ground motion prediction. Attenuation of ground motion in high frequency range is often characterized by the distribution of intrinsic and scattering attenuation parameters (intrinsic Q and scattering coefficient). From the viewpoint of ground motion prediction, both intrinsic and scattering attenuation affect the maximum amplitude of ground motion while scattering attenuation also affect the duration time of ground motion. Hence, estimation of both attenuation parameters will lead to sophisticate the ground motion prediction. In this study, we try to estimate both parameters in southwestern Japan in a tomographic manner. We will conduct envelope fitting of seismic coda since coda has sensitivity to both intrinsic attenuation and scattering coefficients. Recently, Takeuchi (2016) successfully calculated differential envelope when these parameters have fluctuations. We adopted his equations to calculate partial derivatives of these parameters since we did not need to assume homogeneous velocity structure. Matrix for inversion of structural parameters would become too huge to solve in a straightforward manner. Hence, we adopted ART-type Bayesian Reconstruction Method (Hirahara, 1998) to project the difference of envelopes to structural parameters iteratively. We conducted checkerboard reconstruction test. We assumed checkerboard pattern of 0.4 degree interval in horizontal direction and 20 km in depth direction. Reconstructed structures well reproduced the assumed pattern in shallower part while not in deeper part. Since the inversion kernel has large sensitivity around source and stations, resolution in deeper part would be limited due to the sparse distribution of earthquakes. To apply the inversion method which described above to actual waveforms, we have to correct the effects of source and site amplification term. We consider these issues to estimate the actual intrinsic and scattering structures of the target region.Acknowledgment We used the waveforms of Hi-net, NIED. This study was supported by the Earthquake Research Institute of the University of Tokyo cooperative research program.

Source process of a long-period event at Kilauea volcano, Hawaii

USGS Publications Warehouse

Kumagai, H.; Chouet, B.A.; Dawson, P.B.

2005-01-01

We analyse a long-period (LP) event observed by a dense seismic network temporarily operated at Kilauea volcano, Hawaii, in 1996. We systematically perform spectral analyses, waveform inversions and forward modeling of the LP event to quantify its source process. Spectral analyses identify two dominant spectral frequencies at 0.6 and 1.3 Hz with associated Q values in the range 10-20. Results from waveform inversions assuming six moment-tensor and three single-force components point to the resonance of a horizontal crack located at a depth of approximately 150 m near the northeastern rim of the Halemaumau pit crater. Waveform simulations based on a fluid-filled crack model suggest that the observed frequencies and Q values can be explained by a crack filled with a hydrothermal fluid in the form of either bubbly water or steam. The shallow hydrothermal crack located directly above the magma conduit may have been heated by volcanic gases leaking from the conduit. The enhanced flux of heat raised the overall pressure of the hydrothermal fluid in the crack and induced a rapid discharge of fluid from the crack, which triggered the acoustic vibrations of the resonator generating the LP waveform. The present study provides further support to the idea that LP events originate in the resonance of a crack. ?? 2005 RAS.

Effects of Conjugate Gradient Methods and Step-Length Formulas on the Multiscale Full Waveform Inversion in Time Domain: Numerical Experiments

NASA Astrophysics Data System (ADS)

Liu, Youshan; Teng, Jiwen; Xu, Tao; Badal, José; Liu, Qinya; Zhou, Bing

2017-05-01

We carry out full waveform inversion (FWI) in time domain based on an alternative frequency-band selection strategy that allows us to implement the method with success. This strategy aims at decomposing the seismic data within partially overlapped frequency intervals by carrying out a concatenated treatment of the wavelet to largely avoid redundant frequency information to adapt to wavelength or wavenumber coverage. A pertinent numerical test proves the effectiveness of this strategy. Based on this strategy, we comparatively analyze the effects of update parameters for the nonlinear conjugate gradient (CG) method and step-length formulas on the multiscale FWI through several numerical tests. The investigations of up to eight versions of the nonlinear CG method with and without Gaussian white noise make clear that the HS (Hestenes and Stiefel in J Res Natl Bur Stand Sect 5:409-436, 1952), CD (Fletcher in Practical methods of optimization vol. 1: unconstrained optimization, Wiley, New York, 1987), and PRP (Polak and Ribière in Revue Francaise Informat Recherche Opertionelle, 3e Année 16:35-43, 1969; Polyak in USSR Comput Math Math Phys 9:94-112, 1969) versions are more efficient among the eight versions, while the DY (Dai and Yuan in SIAM J Optim 10:177-182, 1999) version always yields inaccurate result, because it overestimates the deeper parts of the model. The application of FWI algorithms using distinct step-length formulas, such as the direct method ( Direct), the parabolic search method ( Search), and the two-point quadratic interpolation method ( Interp), proves that the Interp is more efficient for noise-free data, while the Direct is more efficient for Gaussian white noise data. In contrast, the Search is less efficient because of its slow convergence. In general, the three step-length formulas are robust or partly insensitive to Gaussian white noise and the complexity of the model. When the initial velocity model deviates far from the real model or the data are contaminated by noise, the objective function values of the Direct and Interp are oscillating at the beginning of the inversion, whereas that of the Search decreases consistently.

Unity power factor converter

NASA Technical Reports Server (NTRS)

Wester, Gene W. (Inventor)

1980-01-01

A unity power factor converter capable of effecting either inversion (dc-to-dc) or rectification (ac-to-dc), and capable of providing bilateral power control from a DC source (or load) through an AC transmission line to a DC load (or source) for power flow in either direction, is comprised of comparators for comparing the AC current i with an AC signal i.sub.ref (or its phase inversion) derived from the AC ports to generate control signals to operate a switch control circuit for high speed switching to shape the AC current waveform to a sine waveform, and synchronize it in phase and frequency with the AC voltage at the AC ports, by selectively switching the connections to a series inductor as required to increase or decrease the current i.

NEW APPLICATIONS IN THE INVERSION OF ACOUSTIC FULL WAVEFORM LOGS - RELATING MODE EXCITATION TO LITHOLOGY.

USGS Publications Warehouse

Paillet, Frederick L.; Cheng, C.H.; Meredith, J.A.

1987-01-01

Existing techniques for the quantitative interpretation of waveform data have been based on one of two fundamental approaches: (1) simultaneous identification of compressional and shear velocities; and (2) least-squares minimization of the difference between experimental waveforms and synthetic seismograms. Techniques based on the first approach do not always work, and those based on the second seem too numerically cumbersome for routine application during data processing. An alternative approach is tested here, in which synthetic waveforms are used to predict relative mode excitation in the composite waveform. Synthetic waveforms are generated for a series of lithologies ranging from hard, crystalline rocks (Vp equals 6. 0 km/sec. and Poisson's ratio equals 0. 20) to soft, argillaceous sediments (Vp equals 1. 8 km/sec. and Poisson's ratio equals 0. 40). The series of waveforms illustrates a continuous change within this range of rock properties. Mode energy within characteristic velocity windows is computed for each of the modes in the set of synthetic waveforms. The results indicate that there is a consistent variation in mode excitation in lithology space that can be used to construct a unique relationship between relative mode excitation and lithology.

Application of genetic algorithms to focal mechanism determination

NASA Astrophysics Data System (ADS)

Kobayashi, Reiji; Nakanishi, Ichiro

1994-04-01

Genetic algorithms are a new class of methods for global optimization. They resemble Monte Carlo techniques, but search for solutions more efficiently than uniform Monte Carlo sampling. In the field of geophysics, genetic algorithms have recently been used to solve some non-linear inverse problems (e.g., earthquake location, waveform inversion, migration velocity estimation). We present an application of genetic algorithms to focal mechanism determination from first-motion polarities of P-waves and apply our method to two recent large events, the Kushiro-oki earthquake of January 15, 1993 and the SW Hokkaido (Japan Sea) earthquake of July 12, 1993. Initial solution and curvature information of the objective function that gradient methods need are not required in our approach. Moreover globally optimal solutions can be efficiently obtained. Calculation of polarities based on double-couple models is the most time-consuming part of the source mechanism determination. The amount of calculations required by the method designed in this study is much less than that of previous grid search methods.

Fast in-memory elastic full-waveform inversion using consumer-grade GPUs

NASA Astrophysics Data System (ADS)

Sivertsen Bergslid, Tore; Birger Raknes, Espen; Arntsen, Børge

2017-04-01

Full-waveform inversion (FWI) is a technique to estimate subsurface properties by using the recorded waveform produced by a seismic source and applying inverse theory. This is done through an iterative optimization procedure, where each iteration requires solving the wave equation many times, then trying to minimize the difference between the modeled and the measured seismic data. Having to model many of these seismic sources per iteration means that this is a highly computationally demanding procedure, which usually involves writing a lot of data to disk. We have written code that does forward modeling and inversion entirely in memory. A typical HPC cluster has many more CPUs than GPUs. Since FWI involves modeling many seismic sources per iteration, the obvious approach is to parallelize the code on a source-by-source basis, where each core of the CPU performs one modeling, and do all modelings simultaneously. With this approach, the GPU is already at a major disadvantage in pure numbers. Fortunately, GPUs can more than make up for this hardware disadvantage by performing each modeling much faster than a CPU. Another benefit of parallelizing each individual modeling is that it lets each modeling use a lot more RAM. If one node has 128 GB of RAM and 20 CPU cores, each modeling can use only 6.4 GB RAM if one is running the node at full capacity with source-by-source parallelization on the CPU. A parallelized per-source code using GPUs can use 64 GB RAM per modeling. Whenever a modeling uses more RAM than is available and has to start using regular disk space the runtime increases dramatically, due to slow file I/O. The extremely high computational speed of the GPUs combined with the large amount of RAM available for each modeling lets us do high frequency FWI for fairly large models very quickly. For a single modeling, our GPU code outperforms the single-threaded CPU-code by a factor of about 75. Successful inversions have been run on data with frequencies up to 40 Hz for a model of 2001 by 600 grid points with 5 m grid spacing and 5000 time steps, in less than 2.5 minutes per source. In practice, using 15 nodes (30 GPUs) to model 101 sources, each iteration took approximately 9 minutes. For reference, the same inversion run with our CPU code uses two hours per iteration. This was done using only a very simple wavefield interpolation technique, saving every second timestep. Using a more sophisticated checkpointing or wavefield reconstruction method would allow us to increase this model size significantly. Our results show that ordinary gaming GPUs are a viable alternative to the expensive professional GPUs often used today, when performing large scale modeling and inversion in geophysics.

Microseismic imaging using a source function independent full waveform inversion method

NASA Astrophysics Data System (ADS)

Wang, Hanchen; Alkhalifah, Tariq

2018-07-01

At the heart of microseismic event measurements is the task to estimate the location of the source microseismic events, as well as their ignition times. The accuracy of locating the sources is highly dependent on the velocity model. On the other hand, the conventional microseismic source locating methods require, in many cases, manual picking of traveltime arrivals, which do not only lead to manual effort and human interaction, but also prone to errors. Using full waveform inversion (FWI) to locate and image microseismic events allows for an automatic process (free of picking) that utilizes the full wavefield. However, FWI of microseismic events faces incredible nonlinearity due to the unknown source locations (space) and functions (time). We developed a source function independent FWI of microseismic events to invert for the source image, source function and the velocity model. It is based on convolving reference traces with these observed and modelled to mitigate the effect of an unknown source ignition time. The adjoint-state method is used to derive the gradient for the source image, source function and velocity updates. The extended image for the source wavelet in Z axis is extracted to check the accuracy of the inverted source image and velocity model. Also, angle gathers are calculated to assess the quality of the long wavelength component of the velocity model. By inverting for the source image, source wavelet and the velocity model simultaneously, the proposed method produces good estimates of the source location, ignition time and the background velocity for synthetic examples used here, like those corresponding to the Marmousi model and the SEG/EAGE overthrust model.

Tsunami Waves Joint Inversion Using Tsunami Inundation, Tsunami Deposits Distribution and Marine-Terrestrial Sediment Signal in Tsunami Deposit

NASA Astrophysics Data System (ADS)

Tang, H.; WANG, J.

2017-12-01

Population living close to coastlines is increasing, which creates higher risks due to coastal hazards, such as the tsunami. However, the generation of a tsunami is not fully understood yet, especially for paleo-tsunami. Tsunami deposits are one of the concrete evidence in the geological record which we can apply for studying paleo-tsunami. The understanding of tsunami deposits has significantly improved over the last decades. There are many inversion models (e.g. TsuSedMod, TSUFLIND, and TSUFLIND-EnKF) to study the overland-flow characteristics based on tsunami deposits. However, none of them tries to reconstruct offshore tsunami wave characteristics (wave form, wave height, and length) based on tsunami deposits. Here we present a state-of-the-art inverse approach to reconstruct offshore tsunami wave based on the tsunami inundation data, the spatial distribution of tsunami deposits and Marine-terrestrial sediment signal in the tsunami deposits. Ensemble Kalman Filter (EnKF) Method is used for assimilating both sediment transport simulations and the field observation data. While more computationally expensive, the EnKF approach potentially provides more accurate reconstructions for tsunami waveform. In addition to the improvement of inversion results, the ensemble-based method can also quantify the uncertainties of the results. Meanwhile, joint inversion improves the resolution of tsunami waves compared with inversions using any single data type. The method will be tested by field survey data and gauge data from the 2011 Tohoku tsunami on Sendai plain area.

Data inversion immune to cycle-skipping using AWI

NASA Astrophysics Data System (ADS)

Guasch, L.; Warner, M.; Umpleby, A.; Yao, G.; Morgan, J. V.

2014-12-01

Over the last decade, 3D Full Waveform Inversion (FWI) has become a standard model-building tool in exploration seismology, especially in oil and gas applications -thanks to the high quality (spatial density of sources and receivers) datasets acquired by the industry. FWI provides superior quantitative images than its travel-time counterparts (travel-time based inversion methods) because it aims to match all the information in the observations instead of a severely restricted subset of them, namely picked arrivals.The downside is that the solution space explored by FWI has a high number of local minima, and since the solution is restricted to local optimization methods (due to the objective function evaluation cost), the success of the inversion is subject to starting within the basin of attraction of the global minimum.Local minima can exist for a wide variety of reasons, and it seems unlikely that a formulation of the problem that can eliminate all of them -by defining the optimization problem in a form that results in a monotonic objective function- exist. However, a significant amount of local minima are created by the definition of data misfit. In its standard formulation FWI compares observed data (field data) with predicted data (generated with a synthetic model) by subtracting one from the other, and the objective function is defined as some norm of this difference. The combination of this criteria and the fact that seismic data is oscillatory produces the well-known phenomenon of cycle-skipping, where model updates try to match nearest cycles from one dataset to the other.In order to avoid cycle-skipping we propose a different comparison between observed and predicted data, based on Wiener filters, which exploits the fact that the "identity" Wiener filter is a spike at zero lag. This gives rise to a new objective function without cycle-skipped related local minima, and therefore suppress the need of accurate starting models or low frequencies in the data. This new technique, called Adaptive Waveform Inversion (AWI) appears always superior to conventional FWI.

3D frequency-domain ultrasound waveform tomography breast imaging

NASA Astrophysics Data System (ADS)

Sandhu, Gursharan Yash; West, Erik; Li, Cuiping; Roy, Olivier; Duric, Neb

2017-03-01

Frequency-domain ultrasound waveform tomography is a promising method for the visualization and characterization of breast disease. It has previously been shown to accurately reconstruct the sound speed distributions of breasts of varying densities. The reconstructed images show detailed morphological and quantitative information that can help differentiate different types of breast disease including benign and malignant lesions. The attenuation properties of an ex vivo phantom have also been assessed. However, the reconstruction algorithms assumed a 2D geometry while the actual data acquisition process was not. Although clinically useful sound speed images can be reconstructed assuming this mismatched geometry, artifacts from the reconstruction process exist within the reconstructed images. This is especially true for registration across different modalities and when the 2D assumption is violated. For example, this happens when a patient's breast is rapidly sloping. It is also true for attenuation imaging where energy lost or gained out of the plane gets transformed into artifacts within the image space. In this paper, we will briefly review ultrasound waveform tomography techniques, give motivation for pursuing the 3D method, discuss the 3D reconstruction algorithm, present the results of 3D forward modeling, show the mismatch that is induced by the violation of 3D modeling via numerical simulations, and present a 3D inversion of a numerical phantom.

A Bootstrap-Based Probabilistic Optimization Method to Explore and Efficiently Converge in Solution Spaces of Earthquake Source Parameter Estimation Problems: Application to Volcanic and Tectonic Earthquakes

NASA Astrophysics Data System (ADS)

Dahm, T.; Heimann, S.; Isken, M.; Vasyura-Bathke, H.; Kühn, D.; Sudhaus, H.; Kriegerowski, M.; Daout, S.; Steinberg, A.; Cesca, S.

2017-12-01

Seismic source and moment tensor waveform inversion is often ill-posed or non-unique if station coverage is poor or signals are weak. Therefore, the interpretation of moment tensors can become difficult, if not the full model space is explored, including all its trade-offs and uncertainties. This is especially true for non-double couple components of weak or shallow earthquakes, as for instance found in volcanic, geothermal or mining environments.We developed a bootstrap-based probabilistic optimization scheme (Grond), which is based on pre-calculated Greens function full waveform databases (e.g. fomosto tool, doi.org/10.5880/GFZ.2.1.2017.001). Grond is able to efficiently explore the full model space, the trade-offs and the uncertainties of source parameters. The program is highly flexible with respect to the adaption to specific problems, the design of objective functions, and the diversity of empirical datasets.It uses an integrated, robust waveform data processing based on a newly developed Python toolbox for seismology (Pyrocko, see Heimann et al., 2017, http://doi.org/10.5880/GFZ.2.1.2017.001), and allows for visual inspection of many aspects of the optimization problem. Grond has been applied to the CMT moment tensor inversion using W-phases, to nuclear explosions in Korea, to meteorite atmospheric explosions, to volcano-tectonic events during caldera collapse and to intra-plate volcanic and tectonic crustal events.Grond can be used to optimize simultaneously seismological waveforms, amplitude spectra and static displacements of geodetic data as InSAR and GPS (e.g. KITE, Isken et al., 2017, http://doi.org/10.5880/GFZ.2.1.2017.002). We present examples of Grond optimizations to demonstrate the advantage of a full exploration of source parameter uncertainties for interpretation.

Reverse-Time Imaging Based on Full-Waveform Inverted Velocity Model for Nondestructive Testing of Heterogeneous Engineered Structures

NASA Astrophysics Data System (ADS)

Nguyen, L. T.; Modrak, R. T.; Saenger, E. H.; Tromp, J.

2017-12-01

Reverse-time migration (RTM) can reconstruct reflectors and scatterers by cross-correlating the source wavefield and the receiver wavefield given a known velocity model of the background. In nondestructive testing, however, the engineered structure under inspection is often composed of layers of various materials and the background material has been degraded non-uniformly because of environmental or operational effects. On the other hand, ultrasonic waveform tomography based on the principles of full-waveform inversion (FWI) has succeeded in detecting anomalous features in engineered structures. But the building of the wave velocity model of the comprehensive small-size and high-contrast defect(s) is difficult because it requires computationally expensive high-frequency numerical wave simulations and an accurate understanding of large-scale background variations of the engineered structure.To reduce computational cost and improve detection of small defects, a useful approach is to divide the waveform tomography procedure into two steps: first, a low-frequency model-building step aimed at recovering background structure using FWI, and second, a high-frequency imaging step targeting defects using RTM. Through synthetic test cases, we show that the two-step procedure appears more promising in most cases than a single-step inversion. In particular, we find that the new workflow succeeds in the challenging scenario where the defect lies along preexisting layer interface in a composite bridge deck and in related experiments involving noisy data or inaccurate source parameters. The results reveal the potential of the new wavefield imaging method and encourage further developments in data processing, enhancing computation power, and optimizing the imaging workflow itself so that the procedure can efficiently be applied to geometrically complex 3D solids and waveguides. Lastly, owing to the scale invariance of the elastic wave equation, this imaging procedure can be transferred to applications in regional scales as well.

Models of brachial to finger pulse wave distortion and pressure decrement.

PubMed

Gizdulich, P; Prentza, A; Wesseling, K H

1997-03-01

To model the pulse wave distortion and pressure decrement occurring between brachial and finger arteries. Distortion reversion and decrement correction were also our aims. Brachial artery pressure was recorded intra-arterially and finger pressure was recorded non-invasively by the Finapres technique in 53 adult human subjects. Mean pressure was subtracted from each pressure waveform and Fourier analysis applied to the pulsations. A distortion model was estimated for each subject and averaged over the group. The average inverse model was applied to the full finger pressure waveform. The pressure decrement was modelled by multiple regression on finger systolic and diastolic levels. Waveform distortion could be described by a general, frequency dependent model having a resonance at 7.3 Hz. The general inverse model has an anti-resonance at this frequency. It converts finger to brachial pulsations thereby reducing average waveform distortion from 9.7 (s.d. 3.2) mmHg per sample for the finger pulse to 3.7 (1.7) mmHg for the converted pulse. Systolic and diastolic level differences between finger and brachial arterial pressures changed from -4 (15) and -8 (11) to +8 (14) and +8 (12) mmHg, respectively, after inverse modelling, with pulse pressures correct on average. The pressure decrement model reduced both the mean and the standard deviation of systolic and diastolic level differences to 0 (13) and 0 (8) mmHg. Diastolic differences were thus reduced most. Brachial to finger pulse wave distortion due to wave reflection in arteries is almost identical in all subjects and can be modelled by a single resonance. The pressure decrement due to flow in arteries is greatest for high pulse pressures superimposed on low means.

A high-order 3-D spectral-element method for the forward modelling and inversion of gravimetric data—Application to the western Pyrenees

NASA Astrophysics Data System (ADS)

Martin, Roland; Chevrot, Sébastien; Komatitsch, Dimitri; Seoane, Lucia; Spangenberg, Hannah; Wang, Yi; Dufréchou, Grégory; Bonvalot, Sylvain; Bruinsma, Sean

2017-04-01

We image the internal density structure of the Pyrenees by inverting gravity data using an a priori density model derived by scaling a Vp model obtained by full waveform inversion of teleseismic P-waves. Gravity anomalies are computed via a 3-D high-order finite-element integration in the same high-order spectral-element grid as the one used to solve the wave equation and thus to obtain the velocity model. The curvature of the Earth and surface topography are taken into account in order to obtain a density model as accurate as possible. The method is validated through comparisons with exact semi-analytical solutions. We show that the spectral-element method drastically accelerates the computations when compared to other more classical methods. Different scaling relations between compressional velocity and density are tested, and the Nafe-Drake relation is the one that leads to the best agreement between computed and observed gravity anomalies. Gravity data inversion is then performed and the results allow us to put more constraints on the density structure of the shallow crust and on the deep architecture of the mountain range.

Efficient Monte Carlo sampling of inverse problems using a neural network-based forward—applied to GPR crosshole traveltime inversion

NASA Astrophysics Data System (ADS)

Hansen, T. M.; Cordua, K. S.

2017-12-01

Probabilistically formulated inverse problems can be solved using Monte Carlo-based sampling methods. In principle, both advanced prior information, based on for example, complex geostatistical models and non-linear forward models can be considered using such methods. However, Monte Carlo methods may be associated with huge computational costs that, in practice, limit their application. This is not least due to the computational requirements related to solving the forward problem, where the physical forward response of some earth model has to be evaluated. Here, it is suggested to replace a numerical complex evaluation of the forward problem, with a trained neural network that can be evaluated very fast. This will introduce a modeling error that is quantified probabilistically such that it can be accounted for during inversion. This allows a very fast and efficient Monte Carlo sampling of the solution to an inverse problem. We demonstrate the methodology for first arrival traveltime inversion of crosshole ground penetrating radar data. An accurate forward model, based on 2-D full-waveform modeling followed by automatic traveltime picking, is replaced by a fast neural network. This provides a sampling algorithm three orders of magnitude faster than using the accurate and computationally expensive forward model, and also considerably faster and more accurate (i.e. with better resolution), than commonly used approximate forward models. The methodology has the potential to dramatically change the complexity of non-linear and non-Gaussian inverse problems that have to be solved using Monte Carlo sampling techniques.

The Earthquake Source Inversion Validation (SIV) - Project: Summary, Status, Outlook

NASA Astrophysics Data System (ADS)

Mai, P. M.

2017-12-01

Finite-fault earthquake source inversions infer the (time-dependent) displacement on the rupture surface from geophysical data. The resulting earthquake source models document the complexity of the rupture process. However, this kinematic source inversion is ill-posed and returns non-unique solutions, as seen for instance in multiple source models for the same earthquake, obtained by different research teams, that often exhibit remarkable dissimilarities. To address the uncertainties in earthquake-source inversions and to understand strengths and weaknesses of various methods, the Source Inversion Validation (SIV) project developed a set of forward-modeling exercises and inversion benchmarks. Several research teams then use these validation exercises to test their codes and methods, but also to develop and benchmark new approaches. In this presentation I will summarize the SIV strategy, the existing benchmark exercises and corresponding results. Using various waveform-misfit criteria and newly developed statistical comparison tools to quantify source-model (dis)similarities, the SIV platforms is able to rank solutions and identify particularly promising source inversion approaches. Existing SIV exercises (with related data and descriptions) and all computational tools remain available via the open online collaboration platform; additional exercises and benchmark tests will be uploaded once they are fully developed. I encourage source modelers to use the SIV benchmarks for developing and testing new methods. The SIV efforts have already led to several promising new techniques for tackling the earthquake-source imaging problem. I expect that future SIV benchmarks will provide further innovations and insights into earthquake source kinematics that will ultimately help to better understand the dynamics of the rupture process.

Using 3D Simulation of Elastic Wave Propagation in Laplace Domain for Electromagnetic-Seismic Inverse Modeling

NASA Astrophysics Data System (ADS)

Petrov, P.; Newman, G. A.

2010-12-01

Quantitative imaging of the subsurface objects is essential part of modern geophysical technology important in oil and gas exploration and wide-range engineering applications. A significant advancement in developing a robust, high resolution imaging technology is concerned with using the different geophysical measurements (gravity, EM and seismic) sense the subsurface structure. A joint image of the subsurface geophysical attributes (velocity, electrical conductivity and density) requires the consistent treatment of the different geophysical data (electromagnetic and seismic) due to their differing physical nature - diffusive and attenuated propagation of electromagnetic energy and nonlinear, multiple scattering wave propagation of seismic energy. Recent progress has been reported in the solution of this problem by reducing the complexity of seismic wave field. Works formed by Shin and Cha (2009 and 2008) suggests that low-pass filtering the seismic trace via Laplace-Fourier transformation can be an effective approach for obtaining seismic data that has similar spatial resolution to EM data. The effect of Laplace- Fourier transformation on the low-pass filtered trace changes the modeling of the seismic wave field from multi-wave propagation to diffusion. The key benefit of transformation is that diffusive wave-field inversion works well for both data sets seismic (Shin and Cha, 2008) and electromagnetic (Commer and Newman 2008, Newman et al., 2010). Moreover the different data sets can also be matched for similar and consistent resolution. Finally, the low pass seismic image is also an excellent choice for a starting model when analyzing the entire seismic waveform to recover the high spatial frequency components of the seismic image; its reflectivity (Shin and Cha, 2009). Without a good starting model full waveform seismic imaging and migration can encounter serious difficulties. To produce seismic wave fields consistent for joint imaging in the Laplace-Fourier domain we had developed 3D code for full-wave field simulation in the elastic media which take into account nonlinearity introduced by free-surface effects. Our approach is based on the velocity-stress formulation. In the contrast to conventional formulation we defined the material properties such as density and Lame constants not at nodal points but within cells. This second order finite differences method formulated in the cell-based grid, generate numerical solutions compatible with analytical ones within the range errors determinate by dispersion analysis. Our simulator will be embedded in an inversion scheme for joint seismic- electromagnetic imaging. It also offers possibilities for preconditioning the seismic wave propagation problems in the frequency domain. References. Shin, C. & Cha, Y. (2009), Waveform inversion in the Laplace-Fourier domain, Geophys. J. Int. 177(3), 1067- 1079. Shin, C. & Cha, Y. H. (2008), Waveform inversion in the Laplace domain, Geophys. J. Int. 173(3), 922-931. Commer, M. & Newman, G. (2008), New advances in three-dimensional controlled-source electromagnetic inversion, Geophys. J. Int. 172(2), 513-535. Newman, G. A., Commer, M. & Carazzone, J. J. (2010), Imaging CSEM data in the presence of electrical anisotropy, Geophysics, in press.

Seismotectonics at the junction of the Philippine Sea plate and the Eurasian plate, in light of the 1990 Hualien earthquake and the near-field waveform inversion

NASA Astrophysics Data System (ADS)

Cheng, Hou-Sheng; Mozziconacci, Laetitia; Chang, Emmy T. Y.; Huang, Bor-Shouh

2016-04-01

In eastern Taiwan, the Longitudinal Valley (LV) is the suture zone separating the Eurasian plate (EUP) to the West from the Philippine Sea plate (PSP) to the East. The northern tip of the LV (near Hualien city) is the junction point where the collision evolve northward to a subduction of the PSP under the EUP. As a result, a high seismic activity is observed. Based on the CWB (Central Weather Bureau, Taiwan) earthquake catalog, four distinct seismic clusters can be observed in this area since 1990. We restrict our effort to the cluster caused by a doublet events of 1990 with two moderate-large earthquakes. The first shock of these doublet occurred on 13rd December with ML 6.5. Seventeen hours later and 15 km to the southeast occurred the second shock of ML 6.7. A campaign seismic network of 15 short-period stations - the Hualien Temporary Seismic Network (HTSN) was deployed during 2 months to detect the aftershocks of the doublet. By applying the near-field waveform inversion to the HTSN records, we can retrieve the focal mechanism solutions (FMS) from 50 aftershocks of local magnitude ranging from 2.5 to 5.0. A modified version of the program "FMNEAR" is adopted in this study, which has been proven to be efficient to retrieve FMS for small-to-moderate earthquakes with a limited number of stations. In practice, the near-field waveforms, were processed by band-pass filter between 0.52 and 1.2 Hz. Synthetic waveforms are built from the discrete wave number method of Bouchon (1981). The inversion is done by grid searches on the FMS parameters while the rake is inverted, the best result gives the lowest waveform misfit. The waveform adjustment are improved by depth optimization and a specific 1D velocity model for each station. Focal depths of events are in average 10km deeper than the depth determined by the island-wide seismic stations that suffered from the lack of stations to the east due to the ocean. The FMS of the 50 aftershocks can be classified into three groups according to their mechanisms and the P- and T-axes. The 3 groups distribute from north to south. The northern one is the largest one and is located along the northern and middle part of the northern segment of the LVF (NLVF). It is mainly reverse in type and display homogeneous FMS. Our hypothesis is that the fault generated the doublet is related to the structure activated by this first group. In the middle part, the second group is dominantly normal while the last group spreads in the southern portion of the NLVF with more strike-slip events.

Adjoint Sensitivity Method to Determine Optimal Set of Stations for Tsunami Source Inversion

NASA Astrophysics Data System (ADS)

Gusman, A. R.; Hossen, M. J.; Cummins, P. R.; Satake, K.

2017-12-01

We applied the adjoint sensitivity technique in tsunami science for the first time to determine an optimal set of stations for a tsunami source inversion. The adjoint sensitivity (AS) method has been used in numerical weather prediction to find optimal locations for adaptive observations. We implemented this technique to Green's Function based Time Reverse Imaging (GFTRI), which is recently used in tsunami source inversion in order to reconstruct the initial sea surface displacement, known as tsunami source model. This method has the same source representation as the traditional least square (LSQ) source inversion method where a tsunami source is represented by dividing the source region into a regular grid of "point" sources. For each of these, Green's function (GF) is computed using a basis function for initial sea surface displacement whose amplitude is concentrated near the grid point. We applied the AS method to the 2009 Samoa earthquake tsunami that occurred on 29 September 2009 in the southwest Pacific, near the Tonga trench. Many studies show that this earthquake is a doublet associated with both normal faulting in the outer-rise region and thrust faulting in the subduction interface. To estimate the tsunami source model for this complex event, we initially considered 11 observations consisting of 5 tide gauges and 6 DART bouys. After implementing AS method, we found the optimal set of observations consisting with 8 stations. Inversion with this optimal set provides better result in terms of waveform fitting and source model that shows both sub-events associated with normal and thrust faulting.

Structures of Xishan village landslide in Li County, Sichuan, China, inferred from high-frequency receiver functions of local earthquakes

NASA Astrophysics Data System (ADS)

Wei, Z.; Chu, R.

2017-12-01

Teleseismic receiver function methods are widely used to study the deep structural information beneath the seismic station. However, teleseismic waveforms are difficult to extract the high-frequency receiver function, which are insufficient to constrain the shallow structure because of the inelastic attenuation effect of the earth. In this study, using the local earthquake waveforms collected from 3 broadband stations deployed on the Xishan village landslide in Li County in Sichuan Province, we used the high-frequency receiver function method to study the shallow structure beneath the landslide. We developed the Vp-k (Vp/Vs) staking method of receiver functions, and combined with the H-k stacking and waveform inversion methods of receiver functions to invert the landslide's thickness, S-wave velocity and average Vp/Vs ratio beneath these stations, and compared the thickness with the borehole results. Our results show small-scale lateral variety of velocity structure, a 78-143m/s lower S-wave velocity in the bottom layer and 2.4-3.1 Vp/Vs ratio in the landslide. The observed high Vp/Vs ratio and low S-wave velocity in the bottom layer of the landslide are consistent with low electrical resistivity and water-rich in the bottom layer, suggesting a weak shear strength and potential danger zone in landslide h1. Our study suggest that the local earthquake receiver function can obtain the shallow velocity structural information and supply some seismic constrains for the landslide catastrophe mitigation.

Characterization of a viscoelastic heterogeneous object with an effective model by nonlinear full waveform inversion

NASA Astrophysics Data System (ADS)

Mesgouez, A.

2018-05-01

The determination of equivalent viscoelastic properties of heterogeneous objects remains challenging in various scientific fields such as (geo)mechanics, geophysics or biomechanics. The present investigation addresses the issue of the identification of effective constitutive properties of a binary object by using a nonlinear and full waveform inversion scheme. The inversion process, without any regularization technique or a priori information, aims at minimizing directly the discrepancy between the full waveform responses of a bi-material viscoelastic cylindrical object and its corresponding effective homogeneous object. It involves the retrieval of five constitutive equivalent parameters. Numerical simulations are performed in a laboratory-scale two-dimensional configuration: a transient acoustic plane wave impacts the object and the diffracted fluid pressure, solid stress or velocity component fields are determined using a semi-analytical approach. Results show that the retrieval of the density and of the real parts of both the compressional and the shear wave velocities have been carried out successfully regarding the number and location of sensors, the type of sensors, the size of the searching space, the frequency range of the incident plane pressure wave, and the change in the geometric or mechanical constitution of the bi-material object. The retrieval of the imaginary parts of the wave velocities can reveal in some cases the limitations of the proposed approach.

Two-dimensional frequency-domain acoustic full-waveform inversion with rugged topography

NASA Astrophysics Data System (ADS)

Zhang, Qian-Jiang; Dai, Shi-Kun; Chen, Long-Wei; Li, Kun; Zhao, Dong-Dong; Huang, Xing-Xing

2015-09-01

We studied finite-element-method-based two-dimensional frequency-domain acoustic FWI under rugged topography conditions. The exponential attenuation boundary condition suitable for rugged topography is proposed to solve the cutoff boundary problem as well as to consider the requirement of using the same subdivision grid in joint multifrequency inversion. The proposed method introduces the attenuation factor, and by adjusting it, acoustic waves are sufficiently attenuated in the attenuation layer to minimize the cutoff boundary effect. Based on the law of exponential attenuation, expressions for computing the attenuation factor and the thickness of attenuation layers are derived for different frequencies. In multifrequency-domain FWI, the conjugate gradient method is used to solve equations in the Gauss-Newton algorithm and thus minimize the computation cost in calculating the Hessian matrix. In addition, the effect of initial model selection and frequency combination on FWI is analyzed. Examples using numerical simulations and FWI calculations are used to verify the efficiency of the proposed method.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Eken Tuna, Kevin Mayeda, Abraham Hofstetter, Rengin Gok, Gonca Orgulu, Niyazi Turkelli

A recently developed coda magnitude methodology was applied to selected broadband stations in Turkey for the purpose of testing the coda method in a large, laterally complex region. As found in other, albeit smaller regions, coda envelope amplitude measurements are significantly less variable than distance-corrected direct wave measurements (i.e., L{sub g} and surface waves) by roughly a factor 3-to-4. Despite strong lateral crustal heterogeneity in Turkey, they found that the region could be adequately modeled assuming a simple 1-D, radially symmetric path correction. After calibrating the stations ISP, ISKB and MALT for local and regional distances, single-station moment-magnitude estimates (M{submore » W}) derived from the coda spectra were in excellent agreement with those determined from multistation waveform modeling inversions, exhibiting a data standard deviation of 0.17. Though the calibration was validated using large events, the results of the calibration will extend M{sub W} estimates to significantly smaller events which could not otherwise be waveform modeled. The successful application of the method is remarkable considering the significant lateral complexity in Turkey and the simple assumptions used in the coda method.« less

TDR water content inverse profiling in layered soils during infiltration and evaporation

NASA Astrophysics Data System (ADS)

Greco, R.; Guida, A.

2009-04-01

During the last three decades, time domain reflectometry (TDR) has become one of the most commonly used tools for soil water content measurements either in laboratory or in the field. Indeed, TDR provides easy and cheap water content estimations with relatively small disturbance to the investigated soil. TDR measurements of soil water content are based on the strong correlation between relative dielectric permittivity of wet soil and its volumetric water content. Several expressions of the relationship between relative dielectric permittivity and volumetric water content have been proposed, empirically stated (Topp et al., 1980) as well as based on semi-analytical approach to dielectric mixing models (Roth et al., 1990; Whalley, 1993). So far, TDR field applications suffered the limitation due to the capability of the technique of estimating only the mean water content in the volume investigated by the probe. Whereas the knowledge of non homogeneous vertical water content profiles was needed, it was necessary to install either several vertical probes of different length or several horizontal probes placed in the soil at different depths, in both cases strongly increasing soil disturbance as well as the complexity of the measurements. Several studies have been recently dedicated to the development of inversion methods aimed to extract more information from TDR waveforms, in order to estimate non homogeneous moisture profiles along the axis of the metallic probe used for TDR measurements. A common feature of all these methods is that electromagnetic transient through the wet soil along the probe is mathematically modelled, assuming that the unknown soil water content distribution corresponds to the best agreement between simulated and measured waveforms. In some cases the soil is modelled as a series of small layers with different dielectric properties, and the waveform is obtained as the result of the superposition of multiple reflections arising from impedance discontinuities between the layers (Nguyen et al., 1997; Todoroff et al., 1998; Heimovaara, 2001; Moret et al., 2006). Other methods consider the dielectric properties of the soil as smoothly variable along probe axis (Greco, 1999; Oswald et al., 2003; Greco, 2006). Aim of the study is testing the applicability to layered soils of the inverse method for the estimation of water content profiles along vertical TDR waveguides, originally applied in laboratory to homogeneous soil samples with monotonic moisture distributions (Greco, 2006), and recently extended to field measurements with more general water content profiles (Greco and Guida, 2008). Influence of soil electrical conductivity, uniqueness of solution, choices of parametrization, parameters identifiabilty, sensitivity of the method to chosen parameters variations are discussed. Finally, the results of the application of the inverse method to a series of infiltration and evaporation experiments carried out in a flume filled with three soil layers of different physical characteristics are presented. ACKNOWLEDGEMENTS The research was co-financed by the Italian Ministry of University, by means of the PRIN 2006 PRIN program, within the research project entitled ‘Definition of critical rainfall thresholds for destructive landslides for civil protection purposes'. REFERENCES Greco, R., 1999. Measurement of water content profiles by single TDR experiments. In: Feyen, J., Wiyo, K. (Eds.), Modelling of Transport Processes in Soils. Wageningen Pers, Wageningen, the Netherlands, pp. 276-283. Greco, R., 2006. Soil water content inverse profiling from single TDR waveforms. J. Hydrol. 317, 325-339. Greco R., Guida A., 2008. Field measurements of topsoil moisture profiles by vertical TDR probes. J. Hydrol. 348, 442- 451. Heimovaara, T.J., 2001. Frequency domain modelling of TDR waveforms in order to obtain frequency dependent dielectric properties of soil samples: a theoretical approach. In: TDR 2001 - Second International Symposium on Time Domain Reflectometry for Innovative Geotechnical Applications. Northwestern University, Evanston, Illinois, pp. 19-21. Moret, D., Arrue, J.L., Lopez, M.V., Gracia, R., 2006. A new TDR waveform analysis approach for soil moisture profiling using a single probe. J. Hydrol. 321, 163-172. Nguyen, B.L., Bruining, J., Slob, E.C., 1997. Saturation profiles from dielectric (frequency domain reflectometry) measurements in porous media. In: Proceedings of International Workshop on characterization and Measurements of the Hydraulic Properties of Unsaturated Porous Media, Riverside, California, pp. 363-375. Oswald, B., Benedickter, H.R., Ba¨chtold, W., Flu¨hler, H., 2003. Spatially resolved water content profiles from inverted time domain reflectometry signals. Water Resour. Res. 39 (12), 1357. Todoroff, P., Lorion, R., Lan Sun Luk, J.-D., 1998. L'utilisation des génétiques pour l'identification de profils hydriques de sol a` partir de courbes réflectométriques. CR Acad. Sci. Paris, Sciences de la terre et des plane`tes 327, 607-610. Topp, G.C., Davis, J.L., Annan, A.P., 1980. Electromagnetic determination of soil water content: measurement in coaxial transmission lines. Water Resour. Res. 16, 574-582. Roth, K., Schulin, R., Fluhler, H., Attinger, W., 1990. Calibration of time domain reflectometry for water content measurement using a composite dielectric approach. Water Resour. Res. 26, 2267-2273. Whalley, W.R., 1993. Considerations on the use of time domain reflectometry (TDR) for measuring soil water content. J. Soil Sci. 44, 1-9.

The Linearized Bregman Method for Frugal Full-waveform Inversion with Compressive Sensing and Sparsity-promoting

NASA Astrophysics Data System (ADS)

Chai, Xintao; Tang, Genyang; Peng, Ronghua; Liu, Shaoyong

2018-03-01

Full-waveform inversion (FWI) reconstructs the subsurface properties from acquired seismic data via minimization of the misfit between observed and simulated data. However, FWI suffers from considerable computational costs resulting from the numerical solution of the wave equation for each source at each iteration. To reduce the computational burden, constructing supershots by combining several sources (aka source encoding) allows mitigation of the number of simulations at each iteration, but it gives rise to crosstalk artifacts because of interference between the individual sources of the supershot. A modified Gauss-Newton FWI (MGNFWI) approach showed that as long as the difference between the initial and true models permits a sparse representation, the ℓ _1-norm constrained model updates suppress subsampling-related artifacts. However, the spectral-projected gradient ℓ _1 (SPGℓ _1) algorithm employed by MGNFWI is rather complicated that makes its implementation difficult. To facilitate realistic applications, we adapt a linearized Bregman (LB) method to sparsity-promoting FWI (SPFWI) because of the efficiency and simplicity of LB in the framework of ℓ _1-norm constrained optimization problem and compressive sensing. Numerical experiments performed with the BP Salt model, the Marmousi model and the BG Compass model verify the following points. The FWI result with LB solving ℓ _1-norm sparsity-promoting problem for the model update outperforms that generated by solving ℓ _2-norm problem in terms of crosstalk elimination and high-fidelity results. The simpler LB method performs comparably and even superiorly to the complicated SPGℓ _1 method in terms of computational efficiency and model quality, making the LB method a viable alternative for realistic implementations of SPFWI.

GPR impedance inversion for imaging and characterization of buried archaeological remains: A case study at Mudu city cite in Suzhou, China

NASA Astrophysics Data System (ADS)

Liu, Yu; Shi, Zhanjie; Wang, Bangbing; Yu, Tianxiang

2018-01-01

As a method with high resolution, GPR has been extensively used in archaeological surveys. However, conventional GPR profile can only provide limited geometry information, such as the shape or location of the interface, but can't give the distribution of physical properties which could help identify the historical remains more directly. A common way for GPR to map parameter distribution is the common-midpoint velocity analysis, but it provides limited resolution. Another research hotspot, the full-waveform inversion, is unstable and relatively dependent on the initial model. Coring method could give direct information in drilling site, while the accurate result is only limited in several boreholes. In this paper, we propose a new scheme to enhance imaging and characterization of archaeological targets by fusion of GPR and coring data. The scheme mainly involves the impedance inversion of conventional common-offset GPR data, which uses well log to compensate GPR data and finally obtains a high-resolution estimation of permittivity. The core analysis result also contributes to interpretation of the inversion result. To test this method, we did a case study at Mudu city site in Suzhou, China. The results provide clear images of the ancient city's moat and wall subsurface and improve the characterization of archaeological targets. It is shown that this method is effective and feasible for archaeological exploration.

Hamiltonian Monte Carlo Inversion of Seismic Sources in Complex Media

NASA Astrophysics Data System (ADS)

Fichtner, A.; Simutė, S.

2017-12-01

We present a probabilistic seismic source inversion method that properly accounts for 3D heterogeneous Earth structure and provides full uncertainty information on the timing, location and mechanism of the event. Our method rests on two essential elements: (1) reciprocity and spectral-element simulations in complex media, and (2) Hamiltonian Monte Carlo sampling that requires only a small amount of test models. Using spectral-element simulations of 3D, visco-elastic, anisotropic wave propagation, we precompute a data base of the strain tensor in time and space by placing sources at the positions of receivers. Exploiting reciprocity, this receiver-side strain data base can be used to promptly compute synthetic seismograms at the receiver locations for any hypothetical source within the volume of interest. The rapid solution of the forward problem enables a Bayesian solution of the inverse problem. For this, we developed a variant of Hamiltonian Monte Carlo (HMC) sampling. Taking advantage of easily computable derivatives, HMC converges to the posterior probability density with orders of magnitude less samples than derivative-free Monte Carlo methods. (Exact numbers depend on observational errors and the quality of the prior). We apply our method to the Japanese Islands region where we previously constrained 3D structure of the crust and upper mantle using full-waveform inversion with a minimum period of around 15 s.

A study of high density bit transition requirements versus the effects on BCH error correcting coding

NASA Technical Reports Server (NTRS)

Ingels, F.; Schoggen, W. O.

1981-01-01

Several methods for increasing bit transition densities in a data stream are summarized, discussed in detail, and compared against constraints imposed by the 2 MHz data link of the space shuttle high rate multiplexer unit. These methods include use of alternate pulse code modulation waveforms, data stream modification by insertion, alternate bit inversion, differential encoding, error encoding, and use of bit scramblers. The psuedo-random cover sequence generator was chosen for application to the 2 MHz data link of the space shuttle high rate multiplexer unit. This method is fully analyzed and a design implementation proposed.

Monitoring of Volcanic Activity by Sub-mm Geodetic Analyses

NASA Astrophysics Data System (ADS)

Miura, S.; Mare, Y.; Ichiki, M.; Demachi, T.; Tachibana, K.; Nishimura, T.

2017-12-01

Volcanic earthquakes have been occurring beneath Zao volcano in northern Honshu, Japan since 2013, following the increase of deep low frequency earthquakes from 2012. On account of a burst of seismicity initiated in April 2015, the JMA announced a warning of eruption, however, the seismicity gradually decreased for the next two months and the warning was canceled in June. In the same time period, minor expansive deformation was observed by GNSS. Small earthquakes are still occurring, and low-freq. earthquakes (LPE) occur sometimes accompanied by static tilt changes. In this study, we try to extract the sub-mm displacements from the LPE waveforms observed by broadband seismometers (BBS) and utilize them for geodetic inversion to monitor volcanic activities. Thun et al. (2015, 2016) devised an efficient method using a running median filter (RMF) to remove LP noises, which contaminate displacement waveforms. They demonstrated the reproducibility of the waveforms corresponding to the experimentally given sub-mm displacements in the laboratory. They also apply the method to the field LPE data obtained from several volcanoes to show static displacements. The procedure is outlined as follows: (1) Unfiltered removal of the instrument response, (2) LP noise estimate by LPF with a corner frequency of 5/M, where M (seconds) is the time window of the RMF and should be at least three times the length of the rise time. (3) Subtract the noise estimated from step (2). (4) Integrate to obtain displacement waveforms. We apply the method to the BBS waveform at a distance of about 1.5 km ESE from the summit crater of Zao Volcano associated with a LPE on April 1, 2017. Assuming the time window M as 300 seconds, we successfully obtained the displacement history: taking the rise time of about 2 minutes, the site was gradually uplifted with the amount of about 50-60 µm and then subsided with HF displacements in the next 2 minutes resulting about 20-30 µm static upheaval. Comparing the waveforms obtained by the RMF and the conventional methods, the onset of the upheaval is delayed a little in the former one, while the latter one is consistent with the inclination change observed by the tiltmeter installed at the same site. It may be caused by the characteristics of the RMF and attention should be paid to detailed discussions of the source time histories.

Zephyr: Open-source Parallel Seismic Waveform Inversion in an Integrated Python-based Framework

NASA Astrophysics Data System (ADS)

Smithyman, B. R.; Pratt, R. G.; Hadden, S. M.

2015-12-01

Seismic Full-Waveform Inversion (FWI) is an advanced method to reconstruct wave properties of materials in the Earth from a series of seismic measurements. These methods have been developed by researchers since the late 1980s, and now see significant interest from the seismic exploration industry. As researchers move towards implementing advanced numerical modelling (e.g., 3D, multi-component, anisotropic and visco-elastic physics), it is desirable to make use of a modular approach, minimizing the effort developing a new set of tools for each new numerical problem. SimPEG (http://simpeg.xyz) is an open source project aimed at constructing a general framework to enable geophysical inversion in various domains. In this abstract we describe Zephyr (https://github.com/bsmithyman/zephyr), which is a coupled research project focused on parallel FWI in the seismic context. The software is built on top of Python, Numpy and IPython, which enables very flexible testing and implementation of new features. Zephyr is an open source project, and is released freely to enable reproducible research. We currently implement a parallel, distributed seismic forward modelling approach that solves the 2.5D (two-and-one-half dimensional) viscoacoustic Helmholtz equation at a range modelling frequencies, generating forward solutions for a given source behaviour, and gradient solutions for a given set of observed data. Solutions are computed in a distributed manner on a set of heterogeneous workers. The researcher's frontend computer may be separated from the worker cluster by a network link to enable full support for computation on remote clusters from individual workstations or laptops. The present codebase introduces a numerical discretization equivalent to that used by FULLWV, a well-known seismic FWI research codebase. This makes it straightforward to compare results from Zephyr directly with FULLWV. The flexibility introduced by the use of a Python programming environment makes extension of the codebase with new methods much more straightforward. This enables comparison and integration of new efforts with existing results.

Use of the ventricular propagated excitation model in the magnetocardiographic inverse problem for reconstruction of electrophysiological properties.

PubMed

Ohyu, Shigeharu; Okamoto, Yoshiwo; Kuriki, Shinya

2002-06-01

A novel magnetocardiographic inverse method for reconstructing the action potential amplitude (APA) and the activation time (AT) on the ventricular myocardium is proposed. This method is based on the propagated excitation model, in which the excitation is propagated through the ventricle with nonuniform height of action potential. Assumption of stepwise waveform on the transmembrane potential was introduced in the model. Spatial gradient of transmembrane potential, which is defined by APA and AT distributed in the ventricular wall, is used for the computation of a current source distribution. Based on this source model, the distributions of APA and AT are inversely reconstructed from the QRS interval of magnetocardiogram (MCG) utilizing a maximum a posteriori approach. The proposed reconstruction method was tested through computer simulations. Stability of the methods with respect to measurement noise was demonstrated. When reference APA was provided as a uniform distribution, root-mean-square errors of estimated APA were below 10 mV for MCG signal-to-noise ratios greater than, or equal to, 20 dB. Low-amplitude regions located at several sites in reference APA distributions were correctly reproduced in reconstructed APA distributions. The goal of our study is to develop a method for detecting myocardial ischemia through the depression of reconstructed APA distributions.

Separation of Migration and Tomography Modes of Full-Waveform Inversion in the Plane Wave Domain

NASA Astrophysics Data System (ADS)

Yao, Gang; da Silva, Nuno V.; Warner, Michael; Kalinicheva, Tatiana

2018-02-01

Full-waveform inversion (FWI) includes both migration and tomography modes. The migration mode acts like a nonlinear least squares migration to map model interfaces with reflections, while the tomography mode behaves as tomography to build a background velocity model. The migration mode is the main response of inverting reflections, while the tomography mode exists in response to inverting both the reflections and refractions. To emphasize one of the two modes in FWI, especially for inverting reflections, the separation of the two modes in the gradient of FWI is required. Here we present a new method to achieve this separation with an angle-dependent filtering technique in the plane wave domain. We first transform the source and residual wavefields into the plane wave domain with the Fourier transform and then decompose them into the migration and tomography components using the opening angles between the transformed source and residual plane waves. The opening angles close to 180° contribute to the tomography component, while the others correspond to the migration component. We find that this approach is very effective and robust even when the medium is relatively complicated with strong lateral heterogeneities, highly dipping reflectors, and strong anisotropy. This is well demonstrated by theoretical analysis and numerical tests with a synthetic data set and a field data set.

Applicability of source scaling relations for crustal earthquakes to estimation of the ground motions of the 2016 Kumamoto earthquake

NASA Astrophysics Data System (ADS)

Irikura, Kojiro; Miyakoshi, Ken; Kamae, Katsuhiro; Yoshida, Kunikazu; Somei, Kazuhiro; Kurahashi, Susumu; Miyake, Hiroe

2017-01-01

A two-stage scaling relationship of the source parameters for crustal earthquakes in Japan has previously been constructed, in which source parameters obtained from the results of waveform inversion of strong motion data are combined with parameters estimated based on geological and geomorphological surveys. A three-stage scaling relationship was subsequently developed to extend scaling to crustal earthquakes with magnitudes greater than M w 7.4. The effectiveness of these scaling relationships was then examined based on the results of waveform inversion of 18 recent crustal earthquakes ( M w 5.4-6.9) that occurred in Japan since the 1995 Hyogo-ken Nanbu earthquake. The 2016 Kumamoto earthquake, with M w 7.0, was one of the largest earthquakes to occur since dense and accurate strong motion observation networks, such as K-NET and KiK-net, were deployed after the 1995 Hyogo-ken Nanbu earthquake. We examined the applicability of the scaling relationships of the source parameters of crustal earthquakes in Japan to the 2016 Kumamoto earthquake. The rupture area and asperity area were determined based on slip distributions obtained from waveform inversion of the 2016 Kumamoto earthquake observations. We found that the relationship between the rupture area and the seismic moment for the 2016 Kumamoto earthquake follows the second-stage scaling within one standard deviation ( σ = 0.14). The ratio of the asperity area to the rupture area for the 2016 Kumamoto earthquake is nearly the same as ratios previously obtained for crustal earthquakes. Furthermore, we simulated the ground motions of this earthquake using a characterized source model consisting of strong motion generation areas (SMGAs) based on the empirical Green's function (EGF) method. The locations and areas of the SMGAs were determined through comparison between the synthetic ground motions and observed motions. The sizes of the SMGAs were nearly coincident with the asperities with large slip. The synthetic ground motions obtained using the EGF method agree well with the observed motions in terms of acceleration, velocity, and displacement within the frequency range of 0.3-10 Hz. These findings indicate that the 2016 Kumamoto earthquake is a standard event that follows the scaling relationship of crustal earthquakes in Japan.

Isotropic source terms of San Jacinto fault zone earthquakes based on waveform inversions with a generalized CAP method

NASA Astrophysics Data System (ADS)

Ross, Z. E.; Ben-Zion, Y.; Zhu, L.

2015-02-01

We analyse source tensor properties of seven Mw > 4.2 earthquakes in the complex trifurcation area of the San Jacinto Fault Zone, CA, with a focus on isotropic radiation that may be produced by rock damage in the source volumes. The earthquake mechanisms are derived with generalized `Cut and Paste' (gCAP) inversions of three-component waveforms typically recorded by >70 stations at regional distances. The gCAP method includes parameters ζ and χ representing, respectively, the relative strength of the isotropic and CLVD source terms. The possible errors in the isotropic and CLVD components due to station variability is quantified with bootstrap resampling for each event. The results indicate statistically significant explosive isotropic components for at least six of the events, corresponding to ˜0.4-8 per cent of the total potency/moment of the sources. In contrast, the CLVD components for most events are not found to be statistically significant. Trade-off and correlation between the isotropic and CLVD components are studied using synthetic tests with realistic station configurations. The associated uncertainties are found to be generally smaller than the observed isotropic components. Two different tests with velocity model perturbation are conducted to quantify the uncertainty due to inaccuracies in the Green's functions. Applications of the Mann-Whitney U test indicate statistically significant explosive isotropic terms for most events consistent with brittle damage production at the source.

Characterizing crustal and uppermost mantle anisotropy with a depth-dependent tilted hexagonally symmetric elastic tensor: theory and examples

NASA Astrophysics Data System (ADS)

Feng, L.; Xie, J.; Ritzwoller, M. H.

2017-12-01

Two major types of surface wave anisotropy are commonly observed by seismologists but are only rarely interpreted jointly: apparent radial anisotropy, which is the difference in propagation speed between horizontally and vertically polarized waves inferred from Love and Rayleigh waves, and apparent azimuthal anisotropy, which is the directional dependence of surface wave speeds (usually Rayleigh waves). We describe a method of inversion that interprets simultaneous observations of radial and azimuthal anisotropy under the assumption of a hexagonally symmetric elastic tensor with a tilted symmetry axis defined by dip and strike angles. With a full-waveform numerical solver based on the spectral element method (SEM), we verify the validity of the forward theory used for the inversion. We also present two examples, in the US and Tibet, in which we have successfully applied the tomographic method to demonstrate that the two types of apparent anisotropy can be interpreted jointly as a tilted hexagonally symmetric medium.

Surface wave tomography of North America and the Caribbean using global and regional broad-band networks: Phase velocity maps and limitations of ray theory

USGS Publications Warehouse

Godey, S.; Snieder, R.; Villasenor, A.; Benz, H.M.

2003-01-01

We present phase velocity maps of fundamental mode Rayleigh waves across the North American and Caribbean plates. Our data set consists of 1846 waveforms from 172 events recorded at 91 broad-band stations operating in North America. We compute phase velocity maps in four narrow period bands between 50 and 150 s using a non-linear waveform inversion method that solves for phase velocity perturbations relative to a reference Earth model (PREM). Our results show a strong velocity contrast between high velocities beneath the stable North American craton, and lower velocities in the tectonically active western margin, in agreement with other regional and global surface wave tomography studies. We perform detailed comparisons with global model results, which display good agreement between phase velocity maps in the location and amplitude of the anomalies. However, forward modelling shows that regional maps are more accurate for predicting waveforms. In addition, at long periods, the amplitude of the velocity anomalies imaged in our regional phase velocity maps is three time larger than in global phase velocity models. This amplitude factor is necessary to explain the data accurately, showing that regional models provide a better image of velocity structures. Synthetic tests show that the raypath coverage used in this study enables one to resolve velocity features of the order of 800-1000 km. However, only larger length-scale features are observed in the phase velocity maps. The limitation in resolution of our maps can be attributed to the wave propagation theory used in the inversion. Ray theory does not account for off-great-circle ray propagation effects, such as ray bending or scattering. For wavelengths less than 1000 km, scattering effects are significant and may need to be considered.

Rupture process of the September 12, 2007 Southern Sumatra earthquake from tsunami waveform inversion

NASA Astrophysics Data System (ADS)

Lorito, S.; Romano, F.; Piatanesi, A.

2007-12-01

The aim of this work is to infer the slip distribution and mean rupture velocity along the rupture zone of the 12 September 2007 Southern Sumatra, Indonesia from available tide-gauge records of the tsunami. We select waveforms from 12 stations, distributed along the west coast of Sumatra and in the whole Indian Ocean (11 GLOSS stations and 1 DART buoy). We assume the fault plane and the slip direction to be consistent with both the geometry of the subducting plate and the early focal mechanism solutions. Then we subdivide the fault plane into several subfaults (both along strike and down dip) and compute the corresponding Green's functions by numerical solution of the shallow water equations through a finite difference method. The slip distribution and rupture velocity are determined simultaneously by means of a simulated annealing technique. We compare the recorded and synthetic waveforms in the time domain, using a cost function that is a trade-off between the L1 and L2 norms. Preliminary synthetic checkerboard tests, using the station coverage and the sampling interval of the available data, indicate that the main features of the rupture process may be robustly inverted.

Digital Oblique Remote Ionospheric Sensing (DORIS) Program Development

DTIC Science & Technology

1992-04-01

waveforms. A new with the ARTIST software (Reinisch and Iluang. autoscaling technique for oblique ionograms 1983, Gamache et al., 1985) which is...development and performance of a complete oblique ionogram autoscaling and inversion algorithm is presented. The inver.i-,n algorithm uses a three...OTIH radar. 14. SUBJECT TERMS 15. NUMBER OF PAGES Oblique Propagation; Oblique lonogram Autoscaling ; i Electron Density Profile Inversion; Simulated 16

3-D S-velocity structure in the lowermost mantle beneath the Northern Pacific

NASA Astrophysics Data System (ADS)

Suzuki, Y.; Kawai, K.; Geller, R. J.; Borgeaud, A. F. E.; Konishi, K.

2017-12-01

We previously (Suzuki et al., EPS, 2016) reported the results of waveform inversion to infer the three-dimensional (3-D) S-velocity structure in the lowermost 400 km of the mantle (the Dʺ region) beneath the Northern Pacific region. Our dataset consists of about 20,000 transverse component broadband body-wave seismograms observed at North American stations (mainly USArray) for 131 intermediate and deep earthquakes which occurred beneath the western Pacific subduction region. Synthetic resolution tests indicate that our methods and dataset can resolve the velocity structure in the target region with a horizontal scale of about 150 km and a vertical scale of about 50 km. The 3-D S-velocity model obtained in that study shows three prominent features: (i) horizontal high-velocity anomalies up to about 3 per cent faster than the Preliminary Reference Earth Model (PREM) with a thickness of a few hundred km and a lower boundary which is at most about 150 km above the core-mantle boundary (CMB), (ii) low-velocity anomalies about 2.5 per cent slower than PREM beneath the high-velocity anomalies at the base of the lower mantle, (iii) a thin (about 150 km) low-velocity structure continuous from the base of the low-velocity zone to at least 400 km above the CMB. We interpret these features respectively as: (i) remnants of slab material where the Mg-perovskite to Mg-post-perovskite phase transition could have occurred within the slab, (ii, iii) large amounts of hot and less dense materials beneath the cold Kula or Pacific slab remnants immediately above the CMB which ascend and form a passive plume upwelling at the edge of the slab remnants. Since our initial work we subsequently conducted waveform inversion using both the transverse- and radial-component horizontal waveform data to infer the isotropic shear velocity structure in the lowermost mantle beneath the Northern Pacific in more detail. We also compute partial derivatives with respect to the 5 independent elastic constants (A, C, F, L, N) of a transversely isotropy (TI) medium, and conduct a synthetic resolution test to examine the ability of our methods and dataset to resolve the anisotropic structure in this region using two-component waveform data.

Time domain viscoelastic full waveform inversion

NASA Astrophysics Data System (ADS)

Fabien-Ouellet, Gabriel; Gloaguen, Erwan; Giroux, Bernard

2017-06-01

Viscous attenuation can have a strong impact on seismic wave propagation, but it is rarely taken into account in full waveform inversion (FWI). When viscoelasticity is considered in time domain FWI, the displacement formulation of the wave equation is usually used instead of the popular velocity-stress formulation. However, inversion schemes rely on the adjoint equations, which are quite different for the velocity-stress formulation than for the displacement formulation. In this paper, we apply the adjoint state method to the isotropic viscoelastic wave equation in the velocity-stress formulation based on the generalized standard linear solid rheology. By applying linear transformations to the wave equation before deriving the adjoint state equations, we obtain two symmetric sets of partial differential equations for the forward and adjoint variables. The resulting sets of equations only differ by a sign change and can be solved by the same numerical implementation. We also investigate the crosstalk between parameter classes (velocity and attenuation) of the viscoelastic equation. More specifically, we show that the attenuation levels can be used to recover the quality factors of P and S waves, but that they are very sensitive to velocity errors. Finally, we present a synthetic example of viscoelastic FWI in the context of monitoring CO2 geological sequestration. We show that FWI based on our formulation can indeed recover P- and S-wave velocities and their attenuation levels when attenuation is high enough. Both changes in velocity and attenuation levels recovered with FWI can be used to track the CO2 plume during and after injection. Further studies are required to evaluate the performance of viscoelastic FWI on real data.

Long-period noise source inversion in a 3-D heterogeneous Earth

NASA Astrophysics Data System (ADS)

Sager, K.; Ermert, L. A.; Afanasiev, M.; Boehm, C.; Fichtner, A.

2017-12-01

We have implemented a new method for ambient noise source inversion that fully honors finite-frequency wave propagation and 3-D heterogeneous Earth structure.Here, we present results of its first application to the Earth's long-period background signal, the hum, in a period band of around 120 - 300 s. In addition to being a computationally convenient test case, the hum is also the topic of ongoing research in its own right, because different physical mechanisms have been proposed for its excitation. The broad patterns of this model for South and North hemisphere winter are qualitatively consistent with previous long-term studies of the hum sources; however, thanks to methodological improvements, the iterative refinement, and the use of a comparatively extensive dataset, we retrieve a more detailed model in certain locations. In particular, our results support findings that the dominant hum sources are focused along coasts and shelf areas, particularly in the North hemisphere winter, with a possible though not well-constrained contribution of pelagic sources. Additionally, our findings indicate that hum source locations in the ocean, tentatively linked to locally high bathymetry, are important contributors particularly during South hemisphere winter. These results, in conjunction with synthetic recovery tests and observed cross-correlation waveforms, suggest that hum sources are rather narrowly concentrated in space, with length scales on the order of few hundred kilometers. Future work includes the extension of the model to spring and fall season and to shorter periods, as well as its use in full-waveform ambient noise inversion for 3-D Earth structure.

Fault Slip Distribution and Optimum Sea Surface Displacement of the 2017 Tehuantepec Earthquake in Mexico (Mw 8.2) Estimated from Tsunami Waveforms

NASA Astrophysics Data System (ADS)

Gusman, A. R.; Satake, K.; Mulia, I. E.

2017-12-01

An intraplate normal fault earthquake (Mw 8.2) occurred on 8 September 2017 in the Tehuantepec seismic gap of the Middle America Trench. The submarine earthquake generated a tsunami which was recorded by coastal tide gauges and offshore DART buoys. We used the tsunami waveforms recorded at 16 stations to estimate the fault slip distribution and an optimum sea surface displacement of the earthquake. A steep fault dipping to the northeast with strike of 315°, dip of 73°and rake of -96° based on the USGS W-phase moment tensor solution was assumed for the slip inversion. To independently estimate the sea surface displacement without assuming earthquake fault parameters, we used the B-spline function for the unit sources. The distribution of the unit sources was optimized by a Genetic Algorithm - Pattern Search (GA-PS) method. Tsunami waveform inversion resolves a spatially compact region of large slip (4-10 m) with a dimension of 100 km along the strike and 80 km along the dip in the depth range between 40 km and 110 km. The seismic moment calculated from the fault slip distribution with assumed rigidity of 6 × 1010 Nm-2 is 2.46 × 1021 Nm (Mw 8.2). The optimum displacement model suggests that the sea surface was uplifted up to 0.5 m and subsided down to -0.8 m. The deep location of large fault slip may be the cause of such small sea surface displacements. The simulated tsunami waveforms from the optimum sea surface displacement can reproduce the observations better than those from fault slip distribution; the normalized root mean square misfit for the sea surface displacement is 0.89, while that for the fault slip distribution is 1.04. We simulated the tsunami propagation using the optimum sea surface displacement model. Large tsunami amplitudes up to 2.5 m were predicted to occur inside and around a lagoon located between Salina Cruz and Puerto Chiapas. Figure 1. a) Sea surface displacement for the 2017 Tehuantepec earthquake estimated by tsunami waveforms. b) Map of simulated maximum tsunami amplitude and comparison between observed (blue circles) and simulated (red circles) tsunami maximum amplitude along the coast.

Cross hole GPR traveltime inversion using a fast and accurate neural network as a forward model

NASA Astrophysics Data System (ADS)

Mejer Hansen, Thomas

2017-04-01

Probabilistic formulated inverse problems can be solved using Monte Carlo based sampling methods. In principle both advanced prior information, such as based on geostatistics, and complex non-linear forward physical models can be considered. However, in practice these methods can be associated with huge computational costs that in practice limit their application. This is not least due to the computational requirements related to solving the forward problem, where the physical response of some earth model has to be evaluated. Here, it is suggested to replace a numerical complex evaluation of the forward problem, with a trained neural network that can be evaluated very fast. This will introduce a modeling error, that is quantified probabilistically such that it can be accounted for during inversion. This allows a very fast and efficient Monte Carlo sampling of the solution to an inverse problem. We demonstrate the methodology for first arrival travel time inversion of cross hole ground-penetrating radar (GPR) data. An accurate forward model, based on 2D full-waveform modeling followed by automatic travel time picking, is replaced by a fast neural network. This provides a sampling algorithm three orders of magnitude faster than using the full forward model, and considerably faster, and more accurate, than commonly used approximate forward models. The methodology has the potential to dramatically change the complexity of the types of inverse problems that can be solved using non-linear Monte Carlo sampling techniques.

Moment Tensor Analysis of Shallow Sources

NASA Astrophysics Data System (ADS)

Chiang, A.; Dreger, D. S.; Ford, S. R.; Walter, W. R.; Yoo, S. H.

2015-12-01

A potential issue for moment tensor inversion of shallow seismic sources is that some moment tensor components have vanishing amplitudes at the free surface, which can result in bias in the moment tensor solution. The effects of the free-surface on the stability of the moment tensor method becomes important as we continue to investigate and improve the capabilities of regional full moment tensor inversion for source-type identification and discrimination. It is important to understand these free surface effects on discriminating shallow explosive sources for nuclear monitoring purposes. It may also be important in natural systems that have shallow seismicity such as volcanoes and geothermal systems. In this study, we apply the moment tensor based discrimination method to the HUMMING ALBATROSS quarry blasts. These shallow chemical explosions at approximately 10 m depth and recorded up to several kilometers distance represent rather severe source-station geometry in terms of vanishing traction issues. We show that the method is capable of recovering a predominantly explosive source mechanism, and the combined waveform and first motion method enables the unique discrimination of these events. Recovering the correct yield using seismic moment estimates from moment tensor inversion remains challenging but we can begin to put error bounds on our moment estimates using the NSS technique.

Source Process of the 2007 Niigata-ken Chuetsu-oki Earthquake Derived from Near-fault Strong Motion Data

NASA Astrophysics Data System (ADS)

Aoi, S.; Sekiguchi, H.; Morikawa, N.; Ozawa, T.; Kunugi, T.; Shirasaka, M.

2007-12-01

The 2007 Niigata-ken Chuetsu-oki earthquake occurred on July 16th, 2007, 10:13 JST. We performed a multi- time window linear waveform inversion analysis (Hartzell and Heaton, 1983) to estimate the rupture process from the near fault strong motion data of 14 stations from K-NET, KiK-net, F-net, JMA, and Niigata prefecture. The fault plane for the mainshock has not been clearly determined yet from the aftershock distribution, so that we performed two waveform inversions for north-west dipping fault (Model A) and south-east dipping fault (Model B). Their strike, dip, and rake are set to those of the moment tensor solutions by F-net. Fault plane model of 30 km length by 24 km width is set to cover aftershock distribution within 24 hours after the mainshock. Theoretical Green's functions were calculated by the discrete wavenumber method (Bouchon, 1981) and the R/T matrix method (Kennett, 1983) with the different stratified medium for each station based on the velocity structure including the information form the reflection survey and borehole logging data. Convolution of moving dislocation was introduced to represent the rupture propagation in an each subfault (Sekiguchi et al., 2002). The observed acceleration records were integrated into velocity except of F-net velocity data, and bandpass filtered between 0.1 and 1.0 Hz. We solved least-squared equation to obtain slip amount of each time window on each subfault to minimize squared residual of the waveform fitting between observed and synthetic waveforms. Both models provide moment magnitudes of 6.7. Regarding Model A, we obtained large slip in the south-west deeper part of the rupture starting point, which is close to Kashiwazaki-city. The second or third velocity pulses of observed velocity waveforms seem to be composed of slip from the asperity. Regarding Model B, we obtained large slip in the southwest shallower part of the rupture starting point, which is also close to Kashiwazaki-city. In both models, we found small slip near the rupture starting point, and largest slip at about ten kilometer in the south-west of the rupture starting point with the maximum slip of 2.3 and 2.5 m for Models A and B, respectively. The difference of the residual between observed and synthetic waveforms for both models is not significant, therefore it is difficult to conclude which fault plane is appropriate to explain. The estimated large-slip regions in the inverted source models with the Models A and B are located near the cross point of the two fault plane models, which should have similar radiation pattern. This situation may be one of the reasons why judgment of the fault plane orientation is such difficult. We need careful examinations not only strong motion data but also geodetic data to further explore the fault orientation and the source process of this earthquake.

Final Project Report: Imaging Fault Zones Using a Novel Elastic Reverse-Time Migration Imaging Technique

DOE Office of Scientific and Technical Information (OSTI.GOV)

Huang, Lianjie; Chen, Ting; Tan, Sirui

Imaging fault zones and fractures is crucial for geothermal operators, providing important information for reservoir evaluation and management strategies. However, there are no existing techniques available for directly and clearly imaging fault zones, particularly for steeply dipping faults and fracture zones. In this project, we developed novel acoustic- and elastic-waveform inversion methods for high-resolution velocity model building. In addition, we developed acoustic and elastic reverse-time migration methods for high-resolution subsurface imaging of complex subsurface structures and steeply-dipping fault/fracture zones. We first evaluated and verified the improved capabilities of our newly developed seismic inversion and migration imaging methods using synthetic seismicmore » data. Our numerical tests verified that our new methods directly image subsurface fracture/fault zones using surface seismic reflection data. We then applied our novel seismic inversion and migration imaging methods to a field 3D surface seismic dataset acquired at the Soda Lake geothermal field using Vibroseis sources. Our migration images of the Soda Lake geothermal field obtained using our seismic inversion and migration imaging algorithms revealed several possible fault/fracture zones. AltaRock Energy, Inc. is working with Cyrq Energy, Inc. to refine the geologic interpretation at the Soda Lake geothermal field. Trenton Cladouhos, Senior Vice President R&D of AltaRock, was very interested in our imaging results of 3D surface seismic data from the Soda Lake geothermal field. He planed to perform detailed interpretation of our images in collaboration with James Faulds and Holly McLachlan of University of Nevada at Reno. Using our high-resolution seismic inversion and migration imaging results can help determine the optimal locations to drill wells for geothermal energy production and reduce the risk of geothermal exploration.« less

W17_geowave “3D full waveform geophysical models”

DOE Office of Scientific and Technical Information (OSTI.GOV)

Larmat, Carene; Maceira, Monica; Roy, Corinna

2018-02-12

Performance of the MCMC inversion according to the number of cores for the computation. A) 64 cores. B) 480 cores. C) 816 cores. The true model is represented by the black line. Vsv is the wave speed of S waves polarized in the vertical plane, ξ is an anisotropy parameter. The Earth is highly anisotropics; the wavespeed of seismic waves depends on the polarization of the wave. Seismic inversion of the elastic structure is usually limited to isotropic information such as Vsv. Our research looked at the inversion of Earth anisotropy.

An Improved Method for Seismic Event Depth and Moment Tensor Determination: CTBT Related Application

NASA Astrophysics Data System (ADS)

Stachnik, J.; Rozhkov, M.; Baker, B.

2016-12-01

According to the Protocol to CTBT, International Data Center is required to conduct expert technical analysis and special studies to improve event parameters and assist State Parties in identifying the source of specific event. Determination of seismic event source mechanism and its depth is a part of these tasks. It is typically done through a strategic linearized inversion of the waveforms for a complete or subset of source parameters, or similarly defined grid search through precomputed Greens Functions created for particular source models. We show preliminary results using the latter approach from an improved software design and applied on a moderately powered computer. In this development we tried to be compliant with different modes of CTBT monitoring regime and cover wide range of source-receiver distances (regional to teleseismic), resolve shallow source depths, provide full moment tensor solution based on body and surface waves recordings, be fast to satisfy both on-demand studies and automatic processing and properly incorporate observed waveforms and any uncertainties a priori as well as accurately estimate posteriori uncertainties. Implemented HDF5 based Green's Functions pre-packaging allows much greater flexibility in utilizing different software packages and methods for computation. Further additions will have the rapid use of Instaseis/AXISEM full waveform synthetics added to a pre-computed GF archive. Along with traditional post processing analysis of waveform misfits through several objective functions and variance reduction, we follow a probabilistic approach to assess the robustness of moment tensor solution. In a course of this project full moment tensor and depth estimates are determined for DPRK 2009, 2013 and 2016 events and shallow earthquakes using a new implementation of waveform fitting of teleseismic P waves. A full grid search over the entire moment tensor space is used to appropriately sample all possible solutions. A recent method by Tape & Tape (2012) to discretize the complete moment tensor space from a geometric perspective is used. Moment tensors for DPRK events show isotropic percentages greater than 50%. Depth estimates for the DPRK events range from 1.0-1.4 km. Probabilistic uncertainty estimates on the moment tensor parameters provide robustness to solution.

Moment tensor solutions estimated using optimal filter theory for 51 selected earthquakes, 1980-1984

USGS Publications Warehouse

Sipkin, S.A.

1987-01-01

The 51 global events that occurred from January 1980 to March 1984, which were chosen by the convenors of the Symposium on Seismological Theory and Practice, have been analyzed using a moment tensor inversion algorithm (Sipkin). Many of the events were routinely analyzed as part of the National Earthquake Information Center's (NEIC) efforts to publish moment tensor and first-motion fault-plane solutions for all moderate- to large-sized (mb>5.7) earthquakes. In routine use only long-period P-waves are used and the source-time function is constrained to be a step-function at the source (??-function in the far-field). Four of the events were of special interest, and long-period P, SH-wave solutions were obtained. For three of these events, an unconstrained inversion was performed. The resulting time-dependent solutions indicated that, for many cases, departures of the solutions from pure double-couples are caused by source complexity that has not been adequately modeled. These solutions also indicate that source complexity of moderate-sized events can be determined from long-period data. Finally, for one of the events of special interest, an inversion of the broadband P-waveforms was also performed, demonstrating the potential for using broadband waveform data in inversion procedures. ?? 1987.

High-resolution moisture profiles from full-waveform probabilistic inversion of TDR signals

NASA Astrophysics Data System (ADS)

Laloy, Eric; Huisman, Johan Alexander; Jacques, Diederik

2014-11-01

This study presents an novel Bayesian inversion scheme for high-dimensional undetermined TDR waveform inversion. The methodology quantifies uncertainty in the moisture content distribution, using a Gaussian Markov random field (GMRF) prior as regularization operator. A spatial resolution of 1 cm along a 70-cm long TDR probe is considered for the inferred moisture content. Numerical testing shows that the proposed inversion approach works very well in case of a perfect model and Gaussian measurement errors. Real-world application results are generally satisfying. For a series of TDR measurements made during imbibition and evaporation from a laboratory soil column, the average root-mean-square error (RMSE) between maximum a posteriori (MAP) moisture distribution and reference TDR measurements is 0.04 cm3 cm-3. This RMSE value reduces to less than 0.02 cm3 cm-3 for a field application in a podzol soil. The observed model-data discrepancies are primarily due to model inadequacy, such as our simplified modeling of the bulk soil electrical conductivity profile. Among the important issues that should be addressed in future work are the explicit inference of the soil electrical conductivity profile along with the other sampled variables, the modeling of the temperature-dependence of the coaxial cable properties and the definition of an appropriate statistical model of the residual errors.

Direct Iterative Nonlinear Inversion by Multi-frequency T-matrix Completion

NASA Astrophysics Data System (ADS)

Jakobsen, M.; Wu, R. S.

2016-12-01

Researchers in the mathematical physics community have recently proposed a conceptually new method for solving nonlinear inverse scattering problems (like FWI) which is inspired by the theory of nonlocality of physical interactions. The conceptually new method, which may be referred to as the T-matrix completion method, is very interesting since it is not based on linearization at any stage. Also, there are no gradient vectors or (inverse) Hessian matrices to calculate. However, the convergence radius of this promising T-matrix completion method is seriously restricted by it's use of single-frequency scattering data only. In this study, we have developed a modified version of the T-matrix completion method which we believe is more suitable for applications to nonlinear inverse scattering problems in (exploration) seismology, because it makes use of multi-frequency data. Essentially, we have simplified the single-frequency T-matrix completion method of Levinson and Markel and combined it with the standard sequential frequency inversion (multi-scale regularization) method. For each frequency, we first estimate the experimental T-matrix by using the Moore-Penrose pseudo inverse concept. Then this experimental T-matrix is used to initiate an iterative procedure for successive estimation of the scattering potential and the T-matrix using the Lippmann-Schwinger for the nonlinear relation between these two quantities. The main physical requirements in the basic iterative cycle is that the T-matrix should be data-compatible and the scattering potential operator should be dominantly local; although a non-local scattering potential operator is allowed in the intermediate iterations. In our simplified T-matrix completion strategy, we ensure that the T-matrix updates are always data compatible simply by adding a suitable correction term in the real space coordinate representation. The use of singular-value decomposition representations are not required in our formulation since we have developed an efficient domain decomposition method. The results of several numerical experiments for the SEG/EAGE salt model illustrate the importance of using multi-frequency data when performing frequency domain full waveform inversion in strongly scattering media via the new concept of T-matrix completion.

Accurate step-FMCW ultrasound ranging and comparison with pulse-echo signaling methods

NASA Astrophysics Data System (ADS)

Natarajan, Shyam; Singh, Rahul S.; Lee, Michael; Cox, Brian P.; Culjat, Martin O.; Grundfest, Warren S.; Lee, Hua

2010-03-01

This paper presents a method setup for high-frequency ultrasound ranging based on stepped frequency-modulated continuous waves (FMCW), potentially capable of producing a higher signal-to-noise ratio (SNR) compared to traditional pulse-echo signaling. In current ultrasound systems, the use of higher frequencies (10-20 MHz) to enhance resolution lowers signal quality due to frequency-dependent attenuation. The proposed ultrasound signaling format, step-FMCW, is well-known in the radar community, and features lower peak power, wider dynamic range, lower noise figure and simpler electronics in comparison to pulse-echo systems. In pulse-echo ultrasound ranging, distances are calculated using the transmit times between a pulse and its subsequent echoes. In step-FMCW ultrasonic ranging, the phase and magnitude differences at stepped frequencies are used to sample the frequency domain. Thus, by taking the inverse Fourier transform, a comprehensive range profile is recovered that has increased immunity to noise over conventional ranging methods. Step-FMCW and pulse-echo waveforms were created using custom-built hardware consisting of an arbitrary waveform generator and dual-channel super heterodyne receiver, providing high SNR and in turn, accuracy in detection.

Source Mechanisms of Destructive Tsunamigenic Earthquakes occurred along the Major Subduction Zones

NASA Astrophysics Data System (ADS)

Yolsal-Çevikbilen, Seda; Taymaz, Tuncay; Ulutaş, Ergin

2016-04-01

Subduction zones, where an oceanic plate is subducted down into the mantle by tectonic forces, are potential tsunami locations. Many big, destructive and tsunamigenic earthquakes (Mw > 7.5) and high amplitude tsunami waves are observed along the major subduction zones particularly near Indonesia, Japan, Kuril and Aleutan Islands, Gulf of Alaska, Southern America. Not all earthquakes are tsunamigenic; in order to generate a tsunami, the earthquake must occur under or near the ocean, be large, and create significant vertical movements of the seafloor. It is also known that tsunamigenic earthquakes release their energy over a couple of minutes, have long source time functions and slow-smooth ruptures. In this study, we performed point-source inversions by using teleseismic long-period P- and SH- and broad-band P-waveforms recorded by the Federation of Digital Seismograph Networks (FDSN) and the Global Digital Seismograph Network (GDSN) stations. We obtained source mechanism parameters and finite-fault slip distributions of recent destructive ten earthquakes (Mw ≥ 7.5) by comparing the shapes and amplitudes of long period P- and SH-waveforms, recorded in the distance range of 30° - 90°, with synthetic waveforms. We further obtained finite-fault rupture histories of those earthquakes to determine the faulting area (fault length and width), maximum displacement, rupture duration and stress drop. We applied a new back-projection method that uses teleseismic P-waveforms to integrate the direct P-phase with reflected phases from structural discontinuities near the source, and customized it to estimate the spatio-temporal distribution of the seismic energy release of earthquakes. Inversion results exhibit that recent tsunamigenic earthquakes show dominantly thrust faulting mechanisms with small amount of strike-slip components. Their focal depths are also relatively shallow (h < 40 km). As an example, the September 16, 2015 Illapel (Chile) earthquake (Mw: 8.3; h: 26 km) reflects the major characteristics of the Peru-Chile subduction zone between the Nazca and South America Plates. The size, location, depth and focal mechanism of this earthquake are consistent with its occurrence on the megathrust interface in this region. This study is supported by the Scientific and Technological Research Council of Turkey (TUBITAK, Project No: CAYDAG - 114Y066).

Medium change based image estimation from application of inverse algorithms to coda wave measurements

NASA Astrophysics Data System (ADS)

Zhan, Hanyu; Jiang, Hanwan; Jiang, Ruinian

2018-03-01

Perturbations worked as extra scatters will cause coda waveform distortions; thus, coda wave with long propagation time and traveling path are sensitive to micro-defects in strongly heterogeneous media such as concretes. In this paper, we conduct varied external loads on a life-size concrete slab which contains multiple existing micro-cracks, and a couple of sources and receivers are installed to collect coda wave signals. The waveform decorrelation coefficients (DC) at different loads are calculated for all available source-receiver pair measurements. Then inversions of the DC results are applied to estimate the associated distribution density values in three-dimensional regions through kernel sensitivity model and least-square algorithms, which leads to the images indicating the micro-cracks positions. This work provides an efficiently non-destructive approach to detect internal defects and damages of large-size concrete structures.

Estimation of the gravitational wave polarizations from a nontemplate search

NASA Astrophysics Data System (ADS)

Di Palma, Irene; Drago, Marco

2018-01-01

Gravitational wave astronomy is just beginning, after the recent success of the four direct detections of binary black hole (BBH) mergers and the first observation from a binary neutron star inspiral, with the expectation of many more events to come. Given the possibility to detect waves from not exactly modeled astrophysical processes, it is fundamental to be ready to calculate the polarization waveforms in the case of searches using nontemplate algorithms. In such a case, the waveform polarizations are the only quantities that contain direct information about the generating process. We present the performance of a new valuable tool to estimate the inverse solution of gravitational wave transient signals, starting from the analysis of the signal properties of a nontemplate algorithm that is open to a wider class of gravitational signals not covered by template algorithms. We highlight the contributions to the wave polarization associated with the detector response, the sky localization, and the polarization angle of the source. In this paper we present the performances of such a method and its implications by using two main classes of transient signals, resembling the limiting case for most simple and complicated morphologies. The performances are encouraging for the tested waveforms: the correlation between the original and the reconstructed waveforms spans from better than 80% for simple morphologies to better than 50% for complicated ones. For a nontemplate search these results can be considered satisfactory to reconstruct the astrophysical progenitor.

Seismic source inversion using Green's reciprocity and a 3-D structural model for the Japanese Islands

NASA Astrophysics Data System (ADS)

Simutė, S.; Fichtner, A.

2015-12-01

We present a feasibility study for seismic source inversions using a 3-D velocity model for the Japanese Islands. The approach involves numerically calculating 3-D Green's tensors, which is made efficient by exploiting Green's reciprocity. The rationale for 3-D seismic source inversion has several aspects. For structurally complex regions, such as the Japan area, it is necessary to account for 3-D Earth heterogeneities to prevent unknown structure polluting source solutions. In addition, earthquake source characterisation can serve as a means to delineate existing faults. Source parameters obtained for more realistic Earth models can then facilitate improvements in seismic tomography and early warning systems, which are particularly important for seismically active areas, such as Japan. We have created a database of numerically computed 3-D Green's reciprocals for a 40°× 40°× 600 km size area around the Japanese Archipelago for >150 broadband stations. For this we used a regional 3-D velocity model, recently obtained from full waveform inversion. The model includes attenuation and radial anisotropy and explains seismic waveform data for periods between 10 - 80 s generally well. The aim is to perform source inversions using the database of 3-D Green's tensors. As preliminary steps, we present initial concepts to address issues that are at the basis of our approach. We first investigate to which extent Green's reciprocity works in a discrete domain. Considering substantial amounts of computed Green's tensors we address storage requirements and file formatting. We discuss the importance of the initial source model, as an intelligent choice can substantially reduce the search volume. Possibilities to perform a Bayesian inversion and ways to move to finite source inversion are also explored.

Full moment tensor and source location inversion based on full waveform adjoint method

NASA Astrophysics Data System (ADS)

Morency, C.

2012-12-01

The development of high-performance computing and numerical techniques enabled global and regional tomography to reach high levels of precision, and seismic adjoint tomography has become a state-of-the-art tomographic technique. The method was successfully used for crustal tomography of Southern California (Tape et al., 2009) and Europe (Zhu et al., 2012). Here, I will focus on the determination of source parameters (full moment tensor and location) based on the same approach (Kim et al, 2011). The method relies on full wave simulations and takes advantage of the misfit between observed and synthetic seismograms. An adjoint wavefield is calculated by back-propagating the difference between observed and synthetics from the receivers to the source. The interaction between this adjoint wavefield and the regular forward wavefield helps define Frechet derivatives of the source parameters, that is, the sensitivity of the misfit with respect to the source parameters. Source parameters are then recovered by minimizing the misfit based on a conjugate gradient algorithm using the Frechet derivatives. First, I will demonstrate the method on synthetic cases before tackling events recorded at the Geysers. The velocity model used at the Geysers is based on the USGS 3D velocity model. Waveform datasets come from the Northern California Earthquake Data Center. Finally, I will discuss strategies to ultimately use this method to characterize smaller events for microseismic and induced seismicity monitoring. References: - Tape, C., Q. Liu, A. Maggi, and J. Tromp, 2009, Adjoint tomography of the Southern California crust: Science, 325, 988992. - Zhu, H., Bozdag, E., Peter, D., and Tromp, J., 2012, Structure of the European upper mantle revealed by adjoint method: Nature Geoscience, 5, 493-498. - Kim, Y., Q. Liu, and J. Tromp, 2011, Adjoint centroid-moment tensor inversions: Geophys. J. Int., 186, 264278. Prepared by LLNL under Contract DE-AC52-07NA27344.

A new algorithm for three-dimensional joint inversion of body wave and surface wave data and its application to the Southern California plate boundary region

NASA Astrophysics Data System (ADS)

Fang, Hongjian; Zhang, Haijiang; Yao, Huajian; Allam, Amir; Zigone, Dimitri; Ben-Zion, Yehuda; Thurber, Clifford; van der Hilst, Robert D.

2016-05-01

We introduce a new algorithm for joint inversion of body wave and surface wave data to get better 3-D P wave (Vp) and S wave (Vs) velocity models by taking advantage of the complementary strengths of each data set. Our joint inversion algorithm uses a one-step inversion of surface wave traveltime measurements at different periods for 3-D Vs and Vp models without constructing the intermediate phase or group velocity maps. This allows a more straightforward modeling of surface wave traveltime data with the body wave arrival times. We take into consideration the sensitivity of surface wave data with respect to Vp in addition to its large sensitivity to Vs, which means both models are constrained by two different data types. The method is applied to determine 3-D crustal Vp and Vs models using body wave and Rayleigh wave data in the Southern California plate boundary region, which has previously been studied with both double-difference tomography method using body wave arrival times and ambient noise tomography method with Rayleigh and Love wave group velocity dispersion measurements. Our approach creates self-consistent and unique models with no prominent gaps, with Rayleigh wave data resolving shallow and large-scale features and body wave data constraining relatively deeper structures where their ray coverage is good. The velocity model from the joint inversion is consistent with local geological structures and produces better fits to observed seismic waveforms than the current Southern California Earthquake Center (SCEC) model.

Automated rapid finite fault inversion for megathrust earthquakes: Application to the Maule (2010), Iquique (2014) and Illapel (2015) great earthquakes

NASA Astrophysics Data System (ADS)

Benavente, Roberto; Cummins, Phil; Dettmer, Jan

2016-04-01

Rapid estimation of the spatial and temporal rupture characteristics of large megathrust earthquakes by finite fault inversion is important for disaster mitigation. For example, estimates of the spatio-temporal evolution of rupture can be used to evaluate population exposure to tsunami waves and ground shaking soon after the event by providing more accurate predictions than possible with point source approximations. In addition, rapid inversion results can reveal seismic source complexity to guide additional, more detailed subsequent studies. This work develops a method to rapidly estimate the slip distribution of megathrust events while reducing subjective parameter choices by automation. The method is simple yet robust and we show that it provides excellent preliminary rupture models as soon as 30 minutes for three great earthquakes in the South-American subduction zone. This may slightly change for other regions depending on seismic station coverage but method can be applied to any subduction region. The inversion is based on W-phase data since it is rapidly and widely available and of low amplitude which avoids clipping at close stations for large events. In addition, prior knowledge of the slab geometry (e.g. SLAB 1.0) is applied and rapid W-phase point source information (time delay and centroid location) is used to constrain the fault geometry and extent. Since the linearization by multiple time window (MTW) parametrization requires regularization, objective smoothing is achieved by the discrepancy principle in two fully automated steps. First, the residuals are estimated assuming unknown noise levels, and second, seeking a subsequent solution which fits the data to noise level. The MTW scheme is applied with positivity constraints and a solution is obtained by an efficient non-negative least squares solver. Systematic application of the algorithm to the Maule (2010), Iquique (2014) and Illapel (2015) events illustrates that rapid finite fault inversion with teleseismic data is feasible and provides meaningful results. The results for the three events show excellent data fits and are consistent with other solutions showing most of the slip occurring close to the trench for the Maule an Illapel events and some deeper slip for the Iquique event. Importantly, the Illapel source model predicts tsunami waveforms of close agreement with observed waveforms. Finally, we develop a new Bayesian approach to approximate uncertainties as part of the rapid inversion scheme with positivity constraints. Uncertainties are estimated by approximating the posterior distribution as a multivariate log-normal distribution. While solving for the posterior adds some additional computational cost, we illustrate that uncertainty estimation is important for meaningful interpretation of finite fault models.

Imaging the density distributions at the regional scale using full waveform and gravity data inversion - Application to the Pyrenees

NASA Astrophysics Data System (ADS)

Martin, Roland; Chevrot, Sébastien; Wang, Yi; Spangenberg, Hannah; Goubet, Marie; Monteiller, Vadim; Komatitsch, Dimitri; Seoane, Lucia; Dufréchou, Grégory

2017-04-01

We present a hybrid inversion method that allows us to image density distributions at the regional scale using both seismic and gravity data. One main goal is to obtain densities and seismic wave velocities (P and S) in the lithosphere with a fine resolution to get important constraints on the mineralogic composition and thermal state of the lithosphere. In the context of the Pyrenees (located between Spain and France), accurate Vp and Vs seismic velocity models are computed first on a 3D spectral element grid at the scale of the Pyrenees by inverting teleseismic full waveforms. In a second step, Vp velocities are mapped to densities using empirical relations to build an a priori density model. BGI and BRGM Bouguer gravity anomaly data sets are then inverted on the same 3D spectral element grid as the Vp model at a resolution of 1-2 km by using high-order numerical integration formulae. Solutions are compared to those obtained using classical semi-analytical techniques. This procedure opens the possibility to invert both teleseismic and gravity data on the same finite-element grid. It can handle topography of the free surface in the same spectral-element distorted mesh that is used to solve the wave equation, without performing extra interpolations between different grids and models. WGS84 curvature, SRTM or ETOPO1 topographies are used.

Seismic Evidence for Fluid/Gas Beneath the Mentawai Fore-Arc Basin, Central Sumatra

NASA Astrophysics Data System (ADS)

Huot, Gabriel; Singh, Satish C.

2018-02-01

Since 2004, there have been three great interplate earthquakes (Mw > 8.0) offshore Sumatra. In addition to rupturing the megathrust, these earthquakes might also have ruptured the backthrusts that bound the Andaman Islands to the Mentawai Islands toward the forearc basins. Here we apply a combination of traveltime tomography and seismic full waveform inversion to an ultralong offset seismic reflection profile from the Mentawai forearc basin, in the region of the 2007 Mw 8.4 Bengkulu earthquake. We perform a waveform inversion of far-offset data followed by a waveform inversion of near-offset data using the starting model derived from the traveltime tomography. Our results show the presence of a large, low-velocity anomaly above the backthrust. The seismic reflection image indicates that this low-velocity anomaly lies either within highly compacted sediments from the accretionary wedge or within highly deformed sediments from the forearc basin. The porosity estimation, using the effective medium theory, suggests that a small amount of gas (from 2 to 13%) or a significant amount of fluid (from 17 to 40%) could generate this low-velocity zone. The presence of fluids and the observation of bottom simulating reflector below a push-up ridge might be associated with mud diapirism. The fluids could originate locally from the dewatering of the sediments from the accretionary wedge or forearc basin. The high reflectivity of the backthrust in this region might also indicate deeper fluid origin, either from underplated sediments on the subduction interface or from the serpentinized mantle wedge.

Tsunami Source Inversion Using Tide Gauge and DART Tsunami Waveforms of the 2017 Mw8.2 Mexico Earthquake

NASA Astrophysics Data System (ADS)

Adriano, Bruno; Fujii, Yushiro; Koshimura, Shunichi; Mas, Erick; Ruiz-Angulo, Angel; Estrada, Miguel

2018-01-01

On September 8, 2017 (UTC), a normal-fault earthquake occurred 87 km off the southeast coast of Mexico. This earthquake generated a tsunami that was recorded at coastal tide gauge and offshore buoy stations. First, we conducted a numerical tsunami simulation using a single-fault model to understand the tsunami characteristics near the rupture area, focusing on the nearby tide gauge stations. Second, the tsunami source of this event was estimated from inversion of tsunami waveforms recorded at six coastal stations and three buoys located in the deep ocean. Using the aftershock distribution within 1 day following the main shock, the fault plane orientation had a northeast dip direction (strike = 320°, dip = 77°, and rake =-92°). The results of the tsunami waveform inversion revealed that the fault area was 240 km × 90 km in size with most of the largest slip occurring on the middle and deepest segments of the fault. The maximum slip was 6.03 m from a 30 × 30 km2 segment that was 64.82 km deep at the center of the fault area. The estimated slip distribution showed that the main asperity was at the center of the fault area. The second asperity with an average slip of 5.5 m was found on the northwest-most segments. The estimated slip distribution yielded a seismic moment of 2.9 × 10^{21} Nm (Mw = 8.24), which was calculated assuming an average rigidity of 7× 10^{10} N/m2.

Analysis of Waveform Retracking Methods in Antarctic Ice Sheet Based on CRYOSAT-2 Data

NASA Astrophysics Data System (ADS)

Xiao, F.; Li, F.; Zhang, S.; Hao, W.; Yuan, L.; Zhu, T.; Zhang, Y.; Zhu, C.

2017-09-01

Satellite altimetry plays an important role in many geoscientific and environmental studies of Antarctic ice sheet. The ranging accuracy is degenerated near coasts or over nonocean surfaces, due to waveform contamination. A postprocess technique, known as waveform retracking, can be used to retrack the corrupt waveform and in turn improve the ranging accuracy. In 2010, the CryoSat-2 satellite was launched with the Synthetic aperture Interferometric Radar ALtimeter (SIRAL) onboard. Satellite altimetry waveform retracking methods are discussed in the paper. Six retracking methods including the OCOG method, the threshold method with 10 %, 25 % and 50 % threshold level, the linear and exponential 5-β parametric methods are used to retrack CryoSat-2 waveform over the transect from Zhongshan Station to Dome A. The results show that the threshold retracker performs best with the consideration of waveform retracking success rate and RMS of retracking distance corrections. The linear 5-β parametric retracker gives best waveform retracking precision, but cannot make full use of the waveform data.

3-D acoustic waveform simulation and inversion supplemented by infrasound sensors on a tethered weather balloon at Yasur Volcano, Vanuatu

NASA Astrophysics Data System (ADS)

Iezzi, A. M.; Fee, D.; Matoza, R. S.; Jolly, A. D.; Kim, K.; Christenson, B. W.; Johnson, R.; Kilgour, G.; Garaebiti, E.; Austin, A.; Kennedy, B.; Fitzgerald, R.; Gomez, C.; Key, N.

2017-12-01

Well-constrained acoustic waveform inversion can provide robust estimates of erupted volume and mass flux, increasing our ability to monitor volcanic emissions (potentially in real-time). Previous studies have made assumptions about the multipole source mechanism, which can be represented as the combination of pressure fluctuations from a volume change, directionality, and turbulence. The vertical dipole has not been addressed due to ground-based recording limitations. In this study we deployed a high-density seismo-acoustic network around Yasur Volcano, Vanuatu, including multiple acoustic sensors along a tethered balloon that was moved every 15-60 minutes. Yasur has frequent strombolian eruptions every 1-4 minutes from any one of three active vents within a 400 m diameter crater. Our experiment captured several explosions from each vent at 38 tether locations covering 200 in azimuth and a take-off range of 50 (Jolly et. al., in review). Additionally, FLIR, FTIR, and a variety of visual imagery were collected during the deployment to aid in the seismo-acoustic interpretations. The third dimension (vertical) of pressure sensor coverage allows us to more completely constrain the acoustic source. Our analysis employs Finite-Difference Time-Domain (FDTD) modeling to obtain the full 3-D Green's functions for each propagation path. This method, following Kim et al. (2015), takes into account realistic topographic scattering based on a high-resolution digital elevation model created using structure-from-motion techniques. We then invert for the source location and multipole source-time function using a grid-search approach. We perform this inversion for multiple events from vents A and C to examine the source characteristics of the vents, including an infrasound-derived volume flux as a function of time. These volumes fluxes are then compared to those derived independently from geochemical and seismic inversion techniques. Jolly, A., Matoza, R., Fee, D., Kennedy, B., Iezzi, A., Fitzgerald, R., Austin, A., & Johnson, R. (in review). Kim, K., Fee, D., Yokoo, A., & Lees, J. M. (2015). Acoustic source inversion to estimate volume flux from volcanic explosions. Geophysical Research Letters, 42(13), 5243-5249.

Time-reversal and Bayesian inversion

NASA Astrophysics Data System (ADS)

Debski, Wojciech

2017-04-01

Probabilistic inversion technique is superior to the classical optimization-based approach in all but one aspects. It requires quite exhaustive computations which prohibit its use in huge size inverse problems like global seismic tomography or waveform inversion to name a few. The advantages of the approach are, however, so appealing that there is an ongoing continuous afford to make the large inverse task as mentioned above manageable with the probabilistic inverse approach. One of the perspective possibility to achieve this goal relays on exploring the internal symmetry of the seismological modeling problems in hand - a time reversal and reciprocity invariance. This two basic properties of the elastic wave equation when incorporating into the probabilistic inversion schemata open a new horizons for Bayesian inversion. In this presentation we discuss the time reversal symmetry property, its mathematical aspects and propose how to combine it with the probabilistic inverse theory into a compact, fast inversion algorithm. We illustrate the proposed idea with the newly developed location algorithm TRMLOC and discuss its efficiency when applied to mining induced seismic data.

Numerical results for near surface time domain electromagnetic exploration: a full waveform approach

NASA Astrophysics Data System (ADS)

Sun, H.; Li, K.; Li, X., Sr.; Liu, Y., Sr.; Wen, J., Sr.

2015-12-01

Time domain or Transient electromagnetic (TEM) survey including types with airborne, semi-airborne and ground play important roles in applicants such as geological surveys, ground water/aquifer assess [Meju et al., 2000; Cox et al., 2010], metal ore exploration [Yang and Oldenburg, 2012], prediction of water bearing structures in tunnels [Xue et al., 2007; Sun et al., 2012], UXO exploration [Pasion et al., 2007; Gasperikova et al., 2009] etc. The common practice is introducing a current into a transmitting (Tx) loop and acquire the induced electromagnetic field after the current is cut off [Zhdanov and Keller, 1994]. The current waveforms are different depending on instruments. Rectangle is the most widely used excitation current source especially in ground TEM. Triangle and half sine are commonly used in airborne and semi-airborne TEM investigation. In most instruments, only the off time responses are acquired and used in later analysis and data inversion. Very few airborne instruments acquire the on time and off time responses together. Although these systems acquire the on time data, they usually do not use them in the interpretation.This abstract shows a novel full waveform time domain electromagnetic method and our recent modeling results. The benefits comes from our new algorithm in modeling full waveform time domain electromagnetic problems. We introduced the current density into the Maxwell's equation as the transmitting source. This approach allows arbitrary waveforms, such as triangle, half-sine, trapezoidal waves or scatter record from equipment, being used in modeling. Here, we simulate the establishing and induced diffusion process of the electromagnetic field in the earth. The traditional time domain electromagnetic with pure secondary fields can also be extracted from our modeling results. The real time responses excited by a loop source can be calculated using the algorithm. We analyze the full time gates responses of homogeneous half space and two layered models with half sine current waveform as examples. We find the on time responses are quite sensitive to resistivity or depth changes. The results show the potential use of full waveform responses in time domain electromagnetic surveys.

Pacific slab beneath northeast China revealed by regional and teleseismic waveform modeling

NASA Astrophysics Data System (ADS)

WANG, X.; Chen, Q. F.; Wei, S.

2015-12-01

Accurate velocity and geometry of the slab is essential for better understanding of the thermal, chemical structure of the mantle earth, as well as geodynamics. Recent tomography studies show similar morphology of the subducting Pacific slab beneath northeast China, which was stagnant in the mantle transition zone with thickness of more than 200km and an average velocity perturbation of ~1.5% [Fukao and Obayashi, 2013]. Meanwhile, waveform-modeling studies reveal that the Pacific slab beneath Japan and Kuril Island has velocity perturbation up to 5% and thickness up to 90km [Chen et al., 2007; Zhan et al., 2014]. These discrepancies are probably caused by the smoothing and limited data coverage in the tomographic inversions. Here we adopted 1D and 2D waveform modeling methods to study the fine structure of Pacific slab beneath northeast China using dense regional permanent and temporary broadband seismic records. The residual S- and P-wave travel time, difference between data and 1D synthetics, shows significant difference between the eastern and western stations. S-wave travel time residuals indicate 5-10s earlier arrivals for stations whose ray path lies within the slab, compared with those out of the slab. Teleseimic waveforms were used to rule out the major contribution of the possible low velocity structure above 200km. Furthermore, we use 2D finite-difference waveform modeling to confirm the velocity perturbation and geometry of the slab. Our result shows that the velocity perturbation in the slab is significantly higher than those reported in travel-time tomography studies. ReferencesChen, M., J. Tromp, D. Helmberger, and H. Kanamori (2007), Waveform modeling of the slab beneath Japan, J. Geophys. Res.-Solid Earth, 112(B2), 19, doi:10.1029/2006jb004394.Fukao, Y., and M. Obayashi (2013), Subducted slabs stagnant above, penetrating through, and trapped below the 660 km discontinuity, J. Geophys. Res.-Solid Earth, 118(11), 5920-5938, doi:10.1002/2013jb010466.Zhan, Z. W., D. V. Helmberger, and D. Z. Li (2014), Imaging subducted slab structure beneath the Sea of Okhotsk with teleseismic waveforms, Phys. Earth Planet. Inter., 232, 30-35, doi:10.1016/j.pepi.2014.03.008.

Tying the spectral ends of the M9.0 Tohoku-Oki earthquake sequence: High-frequency radiators and early quasi-static afterslip

NASA Astrophysics Data System (ADS)

Inbal, A.; Ampuero, J. P.; Avouac, J.; Lengliné, O.; Helmberger, D. V.

2012-12-01

The March 11, 2011 M9.0 Tohoku-Oki earthquake was recorded by dense seismological and geodetical networks deployed in Japan, as well as by a vast number of seismic stations worldwide. These observations allow us to study the properties of the subduction interface with unprecedented accuracy and resolution. Here we examine the spectral tails of the co- and post-seismic stages using local geodetic and seismological recordings. First, we study the details of high-frequency (HF) energy radiation during the rupture by using strong-motion recordings. Second, we jointly invert 1Hz GPS, ocean-bottom GPS and aftershock data for the spatio-temporal distribution of early afterslip. In order to constrain the spatial distribution of HF radiators we model waveform envelopes recorded by Kik-net borehole accelerometers located in northeastern Japan. We compute theoretical envelopes for waves traveling in a heterogeneous scattering medium, and invert for the location and amplitude of energy radiators for frequencies ranging from 1 to 16 Hz. Because the inversion is extremely sensitive to the response of individual sites, we adopt an empirical approach and iteratively separate the source and site terms from the stacked spectra of numerous events recorded by the network. The output response functions for each site are used to stabilize the inversion. Preliminary results are consistent with far-field observations and suggest that the HF energy emitted during the M9.0 event originated at the down-dip limit of the rupture zone. We apply waveform cross-correlation to identify repeating events within the aftershock sequence, and locate them by match-filtering their waveforms with known templates. Many of these events occur on seismic asperities loaded by the surrounding creep. We jointly invert the slip histories on these fault patches and the available GPS data for the spatio-temporal distribution of afterslip during the first few hours following the mainshock. We use the Principal Component Analysis Inversion Method to determine the time history of slip on the megathrust during seismic slip and aseismic afterslip. The eigenfunctions are constrained in an iterative process that incorporates the slip histories of seismic asperities. This approach allows documenting the seismic and aseismic phases in a self-consistent manner. The GPS-only inversion places most of the early afterslip east of the hypocenter up to the trench, an area that seemed to have undergone dynamic overshoot.

Salvus: A flexible open-source package for waveform modelling and inversion from laboratory to global scales

NASA Astrophysics Data System (ADS)

Afanasiev, M.; Boehm, C.; van Driel, M.; Krischer, L.; May, D.; Rietmann, M.; Fichtner, A.

2016-12-01

Recent years have been witness to the application of waveform inversion to new and exciting domains, ranging from non-destructive testing to global seismology. Often, each new application brings with it novel wave propagation physics, spatial and temporal discretizations, and models of variable complexity. Adapting existing software to these novel applications often requires a significant investment of time, and acts as a barrier to progress. To combat these problems we introduce Salvus, a software package designed to solve large-scale full-waveform inverse problems, with a focus on both flexibility and performance. Based on a high order finite (spectral) element discretization, we have built Salvus to work on unstructured quad/hex meshes in both 2 or 3 dimensions, with support for P1-P3 bases on triangles and tetrahedra. A diverse (and expanding) collection of wave propagation physics are supported (i.e. coupled solid-fluid). With a focus on the inverse problem, functionality is provided to ease integration with internal and external optimization libraries. Additionally, a python-based meshing package is included to simplify the generation and manipulation of regional to global scale Earth models (quad/hex), with interfaces available to external mesh generators for complex engineering-scale applications (quad/hex/tri/tet). Finally, to ensure that the code remains accurate and maintainable, we build upon software libraries such as PETSc and Eigen, and follow modern software design and testing protocols. Salvus bridges the gap between research and production codes with a design based on C++ mixins and Python wrappers that separates the physical equations from the numerical core. This allows domain scientists to add new equations using a high-level interface, without having to worry about optimized implementation details. Our goal in this presentation is to introduce the code, show several examples across the scales, and discuss some of the extensible design points.

Determination of source process and the tsunami simulation of the 2013 Santa Cruz earthquake

NASA Astrophysics Data System (ADS)

Park, S. C.; Lee, J. W.; Park, E.; Kim, S.

2014-12-01

In order to understand the characteristics of large tsunamigenic earthquakes, we analyzed the earthquake source process of the 2013 Santa Cruz earthquake and simulated the following tsunami. We first estimated the fault length of about 200 km using 3-day aftershock distribution and the source duration of about 110 seconds using the duration of high-frequency energy radiation (Hara, 2007). Moment magnitude was estimated to be 8.0 using the formula of Hara (2007). From the results of 200 km of fault length and 110 seconds of source duration, we used the initial value of rupture velocity as 1.8 km/s for teleseismic waveform inversions. Teleseismic body wave inversion was carried out using the inversion package by Kikuchi and Kanamori (1991). Teleseismic P waveform data from 14 stations were used and band-pass filter of 0.005 ~ 1 Hz was applied. Our best-fit solution indicated that the earthquake occurred on the northwesterly striking (strike = 305) and shallowly dipping (dip = 13) fault plane. Focal depth was determined to be 23 km indicating shallow event. Moment magnitude of 7.8 was obtained showing somewhat smaller than the result obtained above and that of previous study (Lay et al., 2013). Large slip area was seen around the hypocenter. Using the slip distribution obtained by teleseismic waveform inversion, we calculated the surface deformations using formulas of Okada (1985) assuming as the initial change of sea water by tsunami. Then tsunami simulation was carred out using Conell Multi-grid Coupled Tsunami Model (COMCOT) code and 1 min-grid topographic data for water depth from the General Bathymetric Chart of the Ocenas (GEBCO). According to the tsunami simulation, most of tsunami waves propagated to the directions of southwest and northeast which are perpendicular to the fault strike. DART buoy data were used to verify our simulation. In the presentation, we will discuss more details on the results of source process and tsunami simulation and compare them with the previous study.

Salvus: A flexible high-performance and open-source package for waveform modelling and inversion from laboratory to global scales

NASA Astrophysics Data System (ADS)

Afanasiev, Michael; Boehm, Christian; van Driel, Martin; Krischer, Lion; May, Dave; Rietmann, Max; Fichtner, Andreas

2017-04-01

Recent years have been witness to the application of waveform inversion to new and exciting domains, ranging from non-destructive testing to global seismology. Often, each new application brings with it novel wave propagation physics, spatial and temporal discretizations, and models of variable complexity. Adapting existing software to these novel applications often requires a significant investment of time, and acts as a barrier to progress. To combat these problems we introduce Salvus, a software package designed to solve large-scale full-waveform inverse problems, with a focus on both flexibility and performance. Currently based on an abstract implementation of high order finite (spectral) elements, we have built Salvus to work on unstructured quad/hex meshes in both 2 or 3 dimensions, with support for P1-P3 bases on triangles and tetrahedra. A diverse (and expanding) collection of wave propagation physics are supported (i.e. viscoelastic, coupled solid-fluid). With a focus on the inverse problem, functionality is provided to ease integration with internal and external optimization libraries. Additionally, a python-based meshing package is included to simplify the generation and manipulation of regional to global scale Earth models (quad/hex), with interfaces available to external mesh generators for complex engineering-scale applications (quad/hex/tri/tet). Finally, to ensure that the code remains accurate and maintainable, we build upon software libraries such as PETSc and Eigen, and follow modern software design and testing protocols. Salvus bridges the gap between research and production codes with a design based on C++ template mixins and Python wrappers that separates the physical equations from the numerical core. This allows domain scientists to add new equations using a high-level interface, without having to worry about optimized implementation details. Our goal in this presentation is to introduce the code, show several examples across the scales, and discuss some of the extensible design points.

P- and S-wave Receiver Function Imaging with Scattering Kernels

NASA Astrophysics Data System (ADS)

Hansen, S. M.; Schmandt, B.

2017-12-01

Full waveform inversion provides a flexible approach to the seismic parameter estimation problem and can account for the full physics of wave propagation using numeric simulations. However, this approach requires significant computational resources due to the demanding nature of solving the forward and adjoint problems. This issue is particularly acute for temporary passive-source seismic experiments (e.g. PASSCAL) that have traditionally relied on teleseismic earthquakes as sources resulting in a global scale forward problem. Various approximation strategies have been proposed to reduce the computational burden such as hybrid methods that embed a heterogeneous regional scale model in a 1D global model. In this study, we focus specifically on the problem of scattered wave imaging (migration) using both P- and S-wave receiver function data. The proposed method relies on body-wave scattering kernels that are derived from the adjoint data sensitivity kernels which are typically used for full waveform inversion. The forward problem is approximated using ray theory yielding a computationally efficient imaging algorithm that can resolve dipping and discontinuous velocity interfaces in 3D. From the imaging perspective, this approach is closely related to elastic reverse time migration. An energy stable finite-difference method is used to simulate elastic wave propagation in a 2D hypothetical subduction zone model. The resulting synthetic P- and S-wave receiver function datasets are used to validate the imaging method. The kernel images are compared with those generated by the Generalized Radon Transform (GRT) and Common Conversion Point stacking (CCP) methods. These results demonstrate the potential of the kernel imaging approach to constrain lithospheric structure in complex geologic environments with sufficiently dense recordings of teleseismic data. This is demonstrated using a receiver function dataset from the Central California Seismic Experiment which shows several dipping interfaces related to the tectonic assembly of this region. Figure 1. Scattering kernel examples for three receiver function phases. A) direct P-to-s (Ps), B) direct S-to-p and C) free-surface PP-to-s (PPs).

Full-waveform inversion of surface waves in exploration geophysics

NASA Astrophysics Data System (ADS)

Borisov, D.; Gao, F.; Williamson, P.; Tromp, J.

2017-12-01

Full-waveform inversion (FWI) is a data fitting approach to estimate high-resolution properties of the Earth from seismic data by minimizing the misfit between observed and calculated seismograms. In land seismics, the source on the ground generates high-amplitude surface waves, which generally represent most of the energy recorded by ground sensors. Although surface waves are widely used in global seismology and engineering studies, they are typically treated as noise within the seismic exploration community since they mask deeper reflections from the intervals of exploration interest. This is mainly due to the fact that surface waves decay exponentially with depth and for a typical frequency range (≈[5-50] Hz) sample only the very shallow part of the subsurface, but also because they are much more sensitive to S-wave than P-wave velocities. In this study, we invert surface waves in the hope of using them as additional information for updating the near surface. In a heterogeneous medium, the main challenge of surface wave inversion is associated with their dispersive character, which makes it difficult to define a starting model for conventional FWI which can avoid cycle-skipping. The standard approach to dealing with this is by inverting the dispersion curves in the Fourier (f-k) domain to generate locally 1-D models, typically for the shear wavespeeds only. However this requires that the near-surface zone be more or less horizontally invariant over a sufficient distance for the spatial Fourier transform to be applicable. In regions with significant topography, such as foothills, this is not the case, so we revert to the time-space domain, but aim to minimize the differences of envelopes in the early stages of the inversion to resolve the cycle-skipping issue. Once the model is good enough, we revert to the classic waveform-difference inversion. We first present a few synthetic examples. We show that classical FWI might be trapped in a local minimum even for relatively simple scenario, while FWI with envelopes is stable and can converge using an inaccurate starting model. We also perform resolution analysis using a checkerboard test. We then present a field example. The final shear wavespeed model is compared to the results from the inversion of dispersion curves.

Ambient Seismic Source Inversion in a Heterogeneous Earth: Theory and Application to the Earth's Hum

NASA Astrophysics Data System (ADS)

Ermert, Laura; Sager, Korbinian; Afanasiev, Michael; Boehm, Christian; Fichtner, Andreas

2017-11-01

The sources of ambient seismic noise are extensively studied both to better understand their influence on ambient noise tomography and related techniques, and to infer constraints on their excitation mechanisms. Here we develop a gradient-based inversion method to infer the space-dependent and time-varying source power spectral density of the Earth's hum from cross correlations of continuous seismic data. The precomputation of wavefields using spectral elements allows us to account for both finite-frequency sensitivity and for three-dimensional Earth structure. Although similar methods have been proposed previously, they have not yet been applied to data to the best of our knowledge. We apply this method to image the seasonally varying sources of Earth's hum during North and South Hemisphere winter. The resulting models suggest that hum sources are localized, persistent features that occur at Pacific coasts or shelves and in the North Atlantic during North Hemisphere winter, as well as South Pacific coasts and several distinct locations in the Southern Ocean in South Hemisphere winter. The contribution of pelagic sources from the central North Pacific cannot be constrained. Besides improving the accuracy of noise source locations through the incorporation of finite-frequency effects and 3-D Earth structure, this method may be used in future cross-correlation waveform inversion studies to provide initial source models and source model updates.

Moment tensor analysis of very shallow sources

DOE PAGES

Chiang, Andrea; Dreger, Douglas S.; Ford, Sean R.; ...

2016-10-11

An issue for moment tensor (MT) inversion of shallow seismic sources is that some components of the Green’s functions have vanishing amplitudes at the free surface, which can result in bias in the MT solution. The effects of the free surface on the stability of the MT method become important as we continue to investigate and improve the capabilities of regional full MT inversion for source–type identification and discrimination. It is important to understand free–surface effects on discriminating shallow explosive sources for nuclear monitoring purposes. It may also be important in natural systems that have very shallow seismicity, such asmore » volcanic and geothermal systems. We examine the effects of the free surface on the MT via synthetic testing and apply the MT–based discrimination method to three quarry blasts from the HUMMING ALBATROSS experiment. These shallow chemical explosions at ~10 m depth and recorded up to several kilometers distance represent rather severe source–station geometry in terms of free–surface effects. We show that the method is capable of recovering a predominantly explosive source mechanism, and the combined waveform and first–motion method enables the unique discrimination of these events. Furthermore, recovering the design yield using seismic moment estimates from MT inversion remains challenging, but we can begin to put error bounds on our moment estimates using the network sensitivity solution technique.« less

Time-domain least-squares migration using the Gaussian beam summation method

NASA Astrophysics Data System (ADS)

Yang, Jidong; Zhu, Hejun; McMechan, George; Yue, Yubo

2018-04-01

With a finite recording aperture, a limited source spectrum and unbalanced illumination, traditional imaging methods are insufficient to generate satisfactory depth profiles with high resolution and high amplitude fidelity. This is because traditional migration uses the adjoint operator of the forward modeling rather than the inverse operator. We propose a least-squares migration approach based on the time-domain Gaussian beam summation, which helps to balance subsurface illumination and improve image resolution. Based on the Born approximation for the isotropic acoustic wave equation, we derive a linear time-domain Gaussian beam modeling operator, which significantly reduces computational costs in comparison with the spectral method. Then, we formulate the corresponding adjoint Gaussian beam migration, as the gradient of an L2-norm waveform misfit function. An L1-norm regularization is introduced to the inversion to enhance the robustness of least-squares migration, and an approximated diagonal Hessian is used as a preconditioner to speed convergence. Synthetic and field data examples demonstrate that the proposed approach improves imaging resolution and amplitude fidelity in comparison with traditional Gaussian beam migration.

Time-domain least-squares migration using the Gaussian beam summation method

NASA Astrophysics Data System (ADS)

Yang, Jidong; Zhu, Hejun; McMechan, George; Yue, Yubo

2018-07-01

With a finite recording aperture, a limited source spectrum and unbalanced illumination, traditional imaging methods are insufficient to generate satisfactory depth profiles with high resolution and high amplitude fidelity. This is because traditional migration uses the adjoint operator of the forward modelling rather than the inverse operator. We propose a least-squares migration approach based on the time-domain Gaussian beam summation, which helps to balance subsurface illumination and improve image resolution. Based on the Born approximation for the isotropic acoustic wave equation, we derive a linear time-domain Gaussian beam modelling operator, which significantly reduces computational costs in comparison with the spectral method. Then, we formulate the corresponding adjoint Gaussian beam migration, as the gradient of an L2-norm waveform misfit function. An L1-norm regularization is introduced to the inversion to enhance the robustness of least-squares migration, and an approximated diagonal Hessian is used as a pre-conditioner to speed convergence. Synthetic and field data examples demonstrate that the proposed approach improves imaging resolution and amplitude fidelity in comparison with traditional Gaussian beam migration.

Sensitivity analysis of seismic waveforms to upper-mantle discontinuities using the adjoint method

NASA Astrophysics Data System (ADS)

Koroni, Maria; Bozdağ, Ebru; Paulssen, Hanneke; Trampert, Jeannot

2017-09-01

Using spectral-element simulations of wave propagation, we investigated the sensitivity of seismic waveforms, recorded on transverse components, to upper-mantle discontinuities in 1-D and 3-D background models. These sensitivity kernels, or Fréchet derivatives, illustrate the spatial sensitivity to model parameters, of which those for shear wave speed and the surface topography of internal boundaries are discussed in this paper. We focus on the boundaries at 400 and 670 km depth of the mantle transition zone. SS precursors have frequently been used to infer the topography of upper-mantle discontinuities. These seismic phases are underside reflections off these boundaries and are usually analysed in the distance range of 110°-160°. This distance range is chosen to minimize the interference from other waves. We show sensitivity kernels for consecutive time windows at three characteristic epicentral distances within the 110°-160° range. The sensitivity kernels are computed with the adjoint method using synthetic data. From our simulations we can draw three main conclusions: (i) The exact Fréchet derivatives show that in all time windows, and also in those centred on the SS precursors, there is interference from other waves. This explains the difficulty reported in the literature to correct for 3-D shear wave speed perturbations, even if the 3-D structure is perfectly known. (ii) All studies attempting to map the topography of the 400 and 670 km discontinuities to date assume that the traveltimes of SS precursors can be linearly decomposed into a 3-D elastic structure and a topography part. We recently showed that such a linear decomposition is not possible for SS precursors, and the sensitivity kernels presented in this paper explain why. (iii) In agreement with previous work, we show that other parts of the seismograms have greater sensitivity to upper-mantle discontinuities than SS precursors, especially multiply bouncing S waves exploiting the S-wave triplications due to the mantle transition zone. These phases can potentially improve the inference of global topographic variations of the upper-mantle discontinuities in the context of full waveform inversion in a joint inversion for (an)elastic parameters and topography.

a method of gravity and seismic sequential inversion and its GPU implementation

NASA Astrophysics Data System (ADS)

Liu, G.; Meng, X.

2011-12-01

In this abstract, we introduce a gravity and seismic sequential inversion method to invert for density and velocity together. For the gravity inversion, we use an iterative method based on correlation imaging algorithm; for the seismic inversion, we use the full waveform inversion. The link between the density and velocity is an empirical formula called Gardner equation, for large volumes of data, we use the GPU to accelerate the computation. For the gravity inversion method , we introduce a method based on correlation imaging algorithm,it is also a interative method, first we calculate the correlation imaging of the observed gravity anomaly, it is some value between -1 and +1, then we multiply this value with a little density ,this value become the initial density model. We get a forward reuslt with this initial model and also calculate the correaltion imaging of the misfit of observed data and the forward data, also multiply the correaltion imaging result a little density and add it to the initial model, then do the same procedure above , at last ,we can get a inversion density model. For the seismic inveron method ,we use a mothod base on the linearity of acoustic wave equation written in the frequency domain,with a intial velociy model, we can get a good velocity result. In the sequential inversion of gravity and seismic , we need a link formula to convert between density and velocity ,in our method , we use the Gardner equation. Driven by the insatiable market demand for real time, high-definition 3D images, the programmable NVIDIA Graphic Processing Unit (GPU) as co-processor of CPU has been developed for high performance computing. Compute Unified Device Architecture (CUDA) is a parallel programming model and software environment provided by NVIDIA designed to overcome the challenge of using traditional general purpose GPU while maintaining a low learn curve for programmers familiar with standard programming languages such as C. In our inversion processing, we use the GPU to accelerate our gravity and seismic inversion. Taking the gravity inversion as an example, its kernels are gravity forward simulation and correlation imaging, after the parallelization in GPU, in 3D case,the inversion module, the original five CPU loops are reduced to three,the forward module the original five CPU loops are reduced to two. Acknowledgments We acknowledge the financial support of Sinoprobe project (201011039 and 201011049-03), the Fundamental Research Funds for the Central Universities (2010ZY26 and 2011PY0183), the National Natural Science Foundation of China (41074095) and the Open Project of State Key Laboratory of Geological Processes and Mineral Resources (GPMR0945).

Source process of the 2016 Kumamoto earthquake (Mj7.3) inferred from kinematic inversion of strong-motion records

NASA Astrophysics Data System (ADS)

Yoshida, Kunikazu; Miyakoshi, Ken; Somei, Kazuhiro; Irikura, Kojiro

2017-05-01

In this study, we estimated source process of the 2016 Kumamoto earthquake from strong-motion data by using the multiple-time window linear kinematic waveform inversion method to discuss generation of strong motions and to explain crustal deformation pattern with a seismic source inversion model. A four-segment fault model was assumed based on the aftershock distribution, active fault traces, and interferometric synthetic aperture radar data. Three western segments were set to be northwest-dipping planes, and the most eastern segment under the Aso caldera was examined to be a southeast-dipping plane. The velocity structure models used in this study were estimated by using waveform modeling of moderate earthquakes that occurred in the source region. We applied a two-step approach of the inversions of 20 strong-motion datasets observed by K-NET and KiK-net by using band-pass-filtered strong-motion data at 0.05-0.5 Hz and then at 0.05-1.0 Hz. The rupture area of the fault plane was determined by applying the criterion of Somerville et al. (Seismol Res Lett 70:59-80, 1999) to the inverted slip distribution. From the first-step inversion, the fault length was trimmed from 52 to 44 km, whereas the fault width was kept at 18 km. The trimmed rupture area was not changed in the second-step inversion. The source model obtained from the two-step approach indicated 4.7 × 1019 Nm of the total moment release and 1.8 m average slip of the entire fault with a rupture area of 792 km2. Large slip areas were estimated in the seismogenic zone and in the shallow part corresponding to the surface rupture that occurred during the Mj7.3 mainshock. The areas of the high peak moment rate correlated roughly with those of large slip; however, the moment rate functions near the Earth surface have low peak, bell shape, and long duration. These subfaults with long-duration moment release are expected to cause weak short-period ground motions. We confirmed that the southeast dipping of the most eastern segment is more plausible rather than northwest-dipping from the observed subsidence around the central cones of the Aso volcano.[Figure not available: see fulltext.

Resolution analysis of finite fault source inversion using one- and three-dimensional Green's functions 2. Combining seismic and geodetic data

USGS Publications Warehouse

Wald, D.J.; Graves, R.W.

2001-01-01

Using numerical tests for a prescribed heterogeneous earthquake slip distribution, we examine the importance of accurate Green's functions (GF) for finite fault source inversions which rely on coseismic GPS displacements and leveling line uplift alone and in combination with near-source strong ground motions. The static displacements, while sensitive to the three-dimensional (3-D) structure, are less so than seismic waveforms and thus are an important contribution, particularly when used in conjunction with waveform inversions. For numerical tests of an earthquake source and data distribution modeled after the 1994 Northridge earthquake, a joint geodetic and seismic inversion allows for reasonable recovery of the heterogeneous slip distribution on the fault. In contrast, inaccurate 3-D GFs or multiple 1-D GFs allow only partial recovery of the slip distribution given strong motion data alone. Likewise, using just the GPS and leveling line data requires significant smoothing for inversion stability, and hence, only a blurred vision of the prescribed slip is recovered. Although the half-space approximation for computing the surface static deformation field is no longer justifiable based on the high level of accuracy for current GPS data acquisition and the computed differences between 3-D and half-space surface displacements, a layered 1-D approximation to 3-D Earth structure provides adequate representation of the surface displacement field. However, even with the half-space approximation, geodetic data can provide additional slip resolution in the joint seismic and geodetic inversion provided a priori fault location and geometry are correct. Nevertheless, the sensitivity of the static displacements to the Earth structure begs caution for interpretation of surface displacements, particularly those recorded at monuments located in or near basin environments. Copyright 2001 by the American Geophysical Union.

Decomposition Techniques for Icesat/glas Full-Waveform Data

NASA Astrophysics Data System (ADS)

Liu, Z.; Gao, X.; Li, G.; Chen, J.

2018-04-01

The geoscience laser altimeter system (GLAS) on the board Ice, Cloud, and land Elevation Satellite (ICESat), is the first long-duration space borne full-waveform LiDAR for measuring the topography of the ice shelf and temporal variation, cloud and atmospheric characteristics. In order to extract the characteristic parameters of the waveform, the key step is to process the full waveform data. In this paper, the modified waveform decomposition method is proposed to extract the echo components from full-waveform. First, the initial parameter estimation is implemented through data preprocessing and waveform detection. Next, the waveform fitting is demonstrated using the Levenberg-Marquard (LM) optimization method. The results show that the modified waveform decomposition method can effectively extract the overlapped echo components and missing echo components compared with the results from GLA14 product. The echo components can also be extracted from the complex waveforms.

Modeling the 16 September 2015 Chile tsunami source with the inversion of deep-ocean tsunami records by means of the r - solution method

NASA Astrophysics Data System (ADS)

Voronina, Tatyana; Romanenko, Alexey; Loskutov, Artem

2017-04-01

The key point in the state-of-the-art in the tsunami forecasting is constructing a reliable tsunami source. In this study, we present an application of the original numerical inversion technique to modeling the tsunami sources of the 16 September 2015 Chile tsunami. The problem of recovering a tsunami source from remote measurements of the incoming wave in the deep-water tsunameters is considered as an inverse problem of mathematical physics in the class of ill-posed problems. This approach is based on the least squares and the truncated singular value decomposition techniques. The tsunami wave propagation is considered within the scope of the linear shallow-water theory. As in inverse seismic problem, the numerical solutions obtained by mathematical methods become unstable due to the presence of noise in real data. A method of r-solutions makes it possible to avoid instability in the solution to the ill-posed problem under study. This method seems to be attractive from the computational point of view since the main efforts are required only once for calculating the matrix whose columns consist of computed waveforms for each harmonic as a source (an unknown tsunami source is represented as a part of a spatial harmonics series in the source area). Furthermore, analyzing the singular spectra of the matrix obtained in the course of numerical calculations one can estimate the future inversion by a certain observational system that will allow offering a more effective disposition for the tsunameters with the help of precomputations. In other words, the results obtained allow finding a way to improve the inversion by selecting the most informative set of available recording stations. The case study of the 6 February 2013 Solomon Islands tsunami highlights a critical role of arranging deep-water tsunameters for obtaining the inversion results. Implementation of the proposed methodology to the 16 September 2015 Chile tsunami has successfully produced tsunami source model. The function recovered by the method proposed can find practical applications both as an initial condition for various optimization approaches and for computer calculation of the tsunami wave propagation.

INVERSION OF SOURCE TIME FUNCTION USING BOREHOLE ARRAY SONIC WAVEFORMS. (R825225)

EPA Science Inventory

The perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Concl...

Accurate calibration of waveform data measured by the Plasma Wave Experiment on board the ARASE satellite

NASA Astrophysics Data System (ADS)

Kitahara, M.; Katoh, Y.; Hikishima, M.; Kasahara, Y.; Matsuda, S.; Kojima, H.; Ozaki, M.; Yagitani, S.

2017-12-01

The Plasma Wave Experiment (PWE) is installed on board the ARASE satellite to measure the electric field in the frequency range from DC to 10 MHz, and the magnetic field in the frequency range from a few Hz to 100 kHz using two dipole wire-probe antennas (WPT) and three magnetic search coils (MSC), respectively. In particular, the Waveform Capture (WFC), one of the receivers of the PWE, can detect electromagnetic field waveform in the frequency range from a few Hz to 20 kHz. The Software-type Wave Particle Interaction Analyzer (S-WPIA) is installed on the ARASE satellite to measure the energy exchange between plasma waves and particles. Since S-WPIA uses the waveform data measured by WFC to calculate the relative phase angle between the wave magnetic field and velocity of energetic electrons, the high-accuracy is required to calibration of both amplitude and phase of the waveform data. Generally, the calibration procedure of the signal passed through a receiver consists of three steps; the transformation into spectra, the calibration by the transfer function of a receiver, and the inverse transformation of the calibrated spectra into the time domain. Practically, in order to reduce the side robe effect, a raw data is filtered by a window function in the time domain before applying Fourier transform. However, for the case that a first order differential coefficient of the phase transfer function of the system is not negligible, the phase of the window function convoluted into the calibrated spectra is shifted differently at each frequency, resulting in a discontinuity in the time domain of the calibrated waveform data. To eliminate the effect of the phase shift of a window function, we suggest several methods to calibrate a waveform data accurately and carry out simulations assuming simple sinusoidal waves as an input signal and using transfer functions of WPT, MSC, and WFC obtained in pre-flight tests. In consequence, we conclude that the following two methods can reduce an error contaminated through the calibration to less than 0.1 % of amplitude of input waves; (1) a Turkey-type window function with a flat top region of one-third of the window length and (2) modification of the window function for each frequency by referring the estimation of the phase shift due to the first order differential coefficient from the transfer functions.

Full waveform seismic AVAZ signatures of anisotropic shales by integrated rock physics and the reflectivity method

NASA Astrophysics Data System (ADS)

Liu, Xiwu; Guo, Zhiqi; Han, Xu

2018-06-01

A set of parallel vertical fractures embedded in a vertically transverse isotropy (VTI) background leads to orthorhombic anisotropy and corresponding azimuthal seismic responses. We conducted seismic modeling of full waveform amplitude variations versus azimuth (AVAZ) responses of anisotropic shale by integrating a rock physics model and a reflectivity method. The results indicate that the azimuthal variation of P-wave velocity tends to be more complicated for orthorhombic medium compared to the horizontally transverse isotropy (HTI) case, especially at high polar angles. Correspondingly, for the HTI layer in the theoretical model, the short axis of the azimuthal PP amplitudes at the top interface is parallel to the fracture strike, while the long axis at the bottom reflection directs the fracture strike. In contrast, the orthorhombic layer in the theoretical model shows distinct AVAZ responses in terms of PP reflections. Nevertheless, the azimuthal signatures of the R- and T-components of the mode-converted PS reflections show similar AVAZ features for the HTI and orthorhombic layers, which may imply that the PS responses are dominated by fractures. For the application to real data, a seismic-well tie based on upscaled data and a reflectivity method illustrate good agreement between the reference layers and the corresponding reflected events. Finally, the full waveform seismic AVAZ responses of the Longmaxi shale formation are computed for the cases of HTI and orthorhombic anisotropy for comparison. For the two cases, the azimuthal features represent differences mainly in amplitudes, while slightly in the phases of the reflected waveforms. Azimuth variations in the PP reflections from the reference layers show distinct behaviors for the HTI and orthorhombic cases, while the mode-converted PS reflections in terms of the R- and T-components show little differences in azimuthal features. It may suggest that the behaviors of the PS waves are dominated by vertically aligned fractures. This work provides further insight into the azimuthal seismic response of orthorhombic shales. The proposed method may help to improve the seismic-well tie, seismic interpretation, and inversion results using an azimuth anisotropy dataset.

Full Seismic Waveform Tomography of the Japan region using Adjoint Methods

NASA Astrophysics Data System (ADS)

Steptoe, Hamish; Fichtner, Andreas; Rickers, Florian; Trampert, Jeannot

2013-04-01

We present a full-waveform tomographic model of the Japan region based on spectral-element wave propagation, adjoint techniques and seismic data from dense station networks. This model is intended to further our understanding of both the complex regional tectonics and the finite rupture processes of large earthquakes. The shallow Earth structure of the Japan region has been the subject of considerable tomographic investigation. The islands of Japan exist in an area of significant plate complexity: subduction related to the Pacific and Philippine Sea plates is responsible for the majority of seismicity and volcanism of Japan, whilst smaller micro-plates in the region, including the Okhotsk, and Okinawa and Amur, part of the larger North America and Eurasia plates respectively, contribute significant local intricacy. In response to the need to monitor and understand the motion of these plates and their associated faults, numerous seismograph networks have been established, including the 768 station high-sensitivity Hi-net network, 84 station broadband F-net and the strong-motion seismograph networks K-net and KiK-net in Japan. We also include the 55 station BATS network of Taiwan. We use this exceptional coverage to construct a high-resolution model of the Japan region from the full-waveform inversion of over 15,000 individual component seismograms from 53 events that occurred between 1997 and 2012. We model these data using spectral-element simulations of seismic wave propagation at a regional scale over an area from 120°-150°E and 20°-50°N to a depth of around 500 km. We quantify differences between observed and synthetic waveforms using time-frequency misfits allowing us to separate both phase and amplitude measurements whilst exploiting the complete waveform at periods of 15-60 seconds. Fréchet kernels for these misfits are calculated via the adjoint method and subsequently used in an iterative non-linear conjugate-gradient optimization. Finally, we employ custom smoothing algorithms to remove the singularities of the Fréchet kernels and artifacts introduced by the heterogeneous coverage in oceanic regions of the model.

Source analysis of the August 11, 2012 Varzaghan twin earthquakes in NW Iran

NASA Astrophysics Data System (ADS)

Amini, Samar; Zarifi, Zoya; Roberts, Roland

2013-04-01

On 11th of August 2012, Varzaghan city in northwest of Iran experienced two earthquakes within a short time interval. The first earthquake with magnitude 6.4(Mw) at 12:23 GMT was followed by another earthquake with magnitude 6.2 (Mw) 11 minutes later, just 10 km from the location of the first one. These two earthquakes were followed by numerous aftershocks with magnitude up to 5.3. The official reports suggest a death toll of 300 and more than 3000 injuries for these twin events. Though the earthquakes were moderate size, they were felt in Azarbaijan and Armenia with no major damage. The quiescence of seismicity in the close vicinity of the recent intraplate events make the precise identification of the causative fault(s) difficult, though the Ahar fault is reported to be the ruptured fault for these events, which had not been recognized properly before. Westward drift of the Caspian Sea in the NW of Iran and the NNE direction of collision between the Arabian and Eurasian plates control the stress regime in this area. The Global CMT solution reported a pure strike slip fault for the first earthquake and an oblique thrust fault for the second one. We used broadband data (between 20 to 80°) obtained from IRIS to invert for the slip distribution of these events using the teleseismic body waveform inversion method of Kikuchi and Kanamori. More than 100 waveforms (P, vertical component) have been used for the first earthquake however our choice for the analysis of the second earthquake was limited to just 26 waveforms due to wave interference of the first and second events. Using the obtained results in inversion, we have calculated the coulomb stress transfer to study the possible triggering effect of the first earthquake on the second one and the correlation between the area of stress shadow and/or excitation with the aftershocks distribution.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Nurhandoko, Bagus Endar B.; Wely, Woen; Setiadi, Herlan

It is already known that tomography has a great impact for analyzing and mapping unknown objects based on inversion, travel time as well as waveform inversion. Therefore, tomography has used in wide area, not only in medical but also in petroleum as well as mining. Recently, tomography method is being applied in several mining industries. A case study of tomography imaging has been carried out in DOZ ( Deep Ore Zone ) block caving mine, Tembagapura, Papua. Many researchers are undergoing to investigate the properties of DOZ cave not only outside but also inside which is unknown. Tomography takes amore » part for determining this objective.The sources are natural from the seismic events that caused by mining induced seismicity and rocks deformation activity, therefore it is called as passive seismic. These microseismic travel time data are processed by Simultaneous Iterative Reconstruction Technique (SIRT). The result of the inversion can be used for DOZ cave monitoring. These information must be used for identifying weak zone inside the cave. In addition, these results of tomography can be used to determine DOZ and cave information to support mine activity in PT. Freeport Indonesia.« less

Modern Workflow Full Waveform Inversion Applied to North America and the Northern Atlantic

NASA Astrophysics Data System (ADS)

Krischer, Lion; Fichtner, Andreas; Igel, Heiner

2015-04-01

We present the current state of a new seismic tomography model obtained using full waveform inversion of the crustal and upper mantle structure beneath North America and the Northern Atlantic, including the westernmost part of Europe. Parts of the eastern portion of the initial model consists of previous models by Fichtner et al. (2013) and Rickers et al. (2013). The final results of this study will contribute to the 'Comprehensive Earth Model' being developed by the Computational Seismology group at ETH Zurich. Significant challenges include the size of the domain, the uneven event and station coverage, and the strong east-west alignment of seismic ray paths across the North Atlantic. We use as much data as feasible, resulting in several thousand recordings per event depending on the receivers deployed at the earthquakes' origin times. To manage such projects in a reproducible and collaborative manner, we, as tomographers, should abandon ad-hoc scripts and one-time programs, and adopt sustainable and reusable solutions. Therefore we developed the LArge-scale Seismic Inversion Framework (LASIF - http://lasif.net), an open-source toolbox for managing seismic data in the context of non-linear iterative inversions that greatly reduces the time to research. Information on the applied processing, modelling, iterative model updating, what happened during each iteration, and so on are systematically archived. This results in a provenance record of the final model which in the end significantly enhances the reproducibility of iterative inversions. Additionally, tools for automated data download across different data centers, window selection, misfit measurements, parallel data processing, and input file generation for various forward solvers are provided.

Advanced analysis of complex seismic waveforms to characterize the subsurface Earth structure

NASA Astrophysics Data System (ADS)

Jia, Tianxia

2011-12-01

This thesis includes three major parts, (1) Body wave analysis of mantle structure under the Calabria slab, (2) Spatial Average Coherency (SPAC) analysis of microtremor to characterize the subsurface structure in urban areas, and (3) Surface wave dispersion inversion for shear wave velocity structure. Although these three projects apply different techniques and investigate different parts of the Earth, their aims are the same, which is to better understand and characterize the subsurface Earth structure by analyzing complex seismic waveforms that are recorded on the Earth surface. My first project is body wave analysis of mantle structure under the Calabria slab. Its aim is to better understand the subduction structure of the Calabria slab by analyzing seismograms generated by natural earthquakes. The rollback and subduction of the Calabrian Arc beneath the southern Tyrrhenian Sea is a case study of slab morphology and slab-mantle interactions at short spatial scale. I analyzed the seismograms traversing the Calabrian slab and upper mantle wedge under the southern Tyrrhenian Sea through body wave dispersion, scattering and attenuation, which are recorded during the PASSCAL CAT/SCAN experiment. Compressional body waves exhibit dispersion correlating with slab paths, which is high-frequency components arrivals being delayed relative to low-frequency components. Body wave scattering and attenuation are also spatially correlated with slab paths. I used this correlation to estimate the positions of slab boundaries, and further suggested that the observed spatial variation in near-slab attenuation could be ascribed to mantle flow patterns around the slab. My second project is Spatial Average Coherency (SPAC) analysis of microtremors for subsurface structure characterization. Shear-wave velocity (Vs) information in soil and rock has been recognized as a critical parameter for site-specific ground motion prediction study, which is highly necessary for urban areas located in seismic active zones. SPAC analysis of microtremors provides an efficient way to estimate Vs structure. Compared with other Vs estimating methods, SPAC is noninvasive and does not require any active sources, and therefore, it is especially useful in big cities. I applied SPAC method in two urban areas. The first is the historic city, Charleston, South Carolina, where high levels of seismic hazard lead to great public concern. Accurate Vs information, therefore, is critical for seismic site classification and site response studies. The second SPAC study is in Manhattan, New York City, where depths of high velocity contrast and soil-to-bedrock are different along the island. The two experiments show that Vs structure could be estimated with good accuracy using SPAC method compared with borehole and other techniques. SPAC is proved to be an effective technique for Vs estimation in urban areas. One important issue in seismology is the inversion of subsurface structures from surface recordings of seismograms. My third project focuses on solving this complex geophysical inverse problems, specifically, surface wave phase velocity dispersion curve inversion for shear wave velocity. In addition to standard linear inversion, I developed advanced inversion techniques including joint inversion using borehole data as constrains, nonlinear inversion using Monte Carlo, and Simulated Annealing algorithms. One innovative way of solving the inverse problem is to make inference from the ensemble of all acceptable models. The statistical features of the ensemble provide a better way to characterize the Earth model.

Seismotectonics of the Eastern Himalayan System and Indo-Burman Convergence Zone Using Seismic Waveform Inversion

NASA Astrophysics Data System (ADS)

Kumar, A.; Mitra, S.; Suresh, G.

2014-12-01

The Eastern Himalayan System (east of 88°E) is distinct from the rest of the India-Eurasia continental collision, due to a wider zone of distributed deformation, oblique convergence across two orthogonal plate boundaries and near absence of foreland basin sedimentary strata. To understand the seismotectonics of this region we study the spatial distribution and source mechanism of earthquakes originating within Eastern Himalaya, northeast India and Indo-Burman Convergence Zone (IBCZ). We compute focal mechanism of 32 moderate-to-large earthquakes (mb >=5.4) by modeling teleseismic P- and SH-waveforms, from GDSN stations, using least-squares inversion algorithm; and 7 small-to-moderate earthquakes (3.5<= mb <5.4) by modeling local P- and S-waveforms, from the NorthEast India Telemetered Network, using non-linear grid search algorithm. We also include source mechanisms from previous studies, either computed by waveform inversion or by first motion polarity from analog data. Depth distribution of modeled earthquakes reveal that the seismogenic layer beneath northeast India is ~45km thick. From source mechanisms we observe that moderate earthquakes in northeast India are spatially clustered in five zones with distinct mechanisms: (a) thrust earthquakes within the Eastern Himalayan wedge, on north dipping low angle faults; (b) thrust earthquakes along the northern edge of Shillong Plateau, on high angle south dipping fault; (c) dextral strike-slip earthquakes along Kopili fault zone, between Shillong Plateau and Mikir Hills, extending southeast beneath Naga Fold belts; (d) dextral strike-slip earthquakes within Bengal Basin, immediately south of Shillong Plateau; and (e) deep focus (>50 km) thrust earthquakes within IBCZ. Combining with GPS geodetic observations, it is evident that the N20E convergence between India and Tibet is accommodated as elastic strain both within eastern Himalaya and regions surrounding the Shillong Plateau. We hypothesize that the strike-slip earthquakes south of the Plateau occur on re-activated continental rifts paralleling the Eocene hinge zone. Distribution of earthquake hypocenters across the IBCZ reveal active subduction of the Indian plate beneath Burma micro-plate.

Analysis and Simulation of 3D Scattering due to Heterogeneous Crustal Structure and Surface Topography on Regional Phases; Magnitude and Discrimination

DTIC Science & Technology

2009-07-07

inversion technique that is based on different weights for relatively high frequency waveform modeling of Pnl and relatively long period surface waves (Tan...et al., 2006). Pnl and surface waves are also allowed to shift in time to take into account of uncertainties in velocity structure. Joint...inversion of Pnl and surface waves provides better constraints on focal depth as well as source mechanisms. The pure strike-slip mechanism of the earthquake

Nuclear test ban treaty verification: Improving test ban monitoring with empirical and model-based signal processing

DOE Office of Scientific and Technical Information (OSTI.GOV)

Harris, David B.; Gibbons, Steven J.; Rodgers, Arthur J.

In this approach, small scale-length medium perturbations not modeled in the tomographic inversion might be described as random fields, characterized by particular distribution functions (e.g., normal with specified spatial covariance). Conceivably, random field parameters (scatterer density or scale length) might themselves be the targets of tomographic inversions of the scattered wave field. As a result, such augmented models may provide processing gain through the use of probabilistic signal sub spaces rather than deterministic waveforms.

Nuclear test ban treaty verification: Improving test ban monitoring with empirical and model-based signal processing

DOE PAGES

Harris, David B.; Gibbons, Steven J.; Rodgers, Arthur J.; ...

2012-05-01

In this approach, small scale-length medium perturbations not modeled in the tomographic inversion might be described as random fields, characterized by particular distribution functions (e.g., normal with specified spatial covariance). Conceivably, random field parameters (scatterer density or scale length) might themselves be the targets of tomographic inversions of the scattered wave field. As a result, such augmented models may provide processing gain through the use of probabilistic signal sub spaces rather than deterministic waveforms.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Dreger, Douglas S.; Ford, Sean R.; Walter, William R.

Research was carried out investigating the feasibility of using a regional distance seismic waveform moment tensor inverse procedure to estimate source parameters of nuclear explosions and to use the source inversion results to develop a source-type discrimination capability. The results of the research indicate that it is possible to robustly determine the seismic moment tensor of nuclear explosions, and when compared to natural seismicity in the context of the a Hudson et al. (1989) source-type diagram they are found to separate from populations of earthquakes and underground cavity collapse seismic sources.

Detection of sinkholes or anomalies using full seismic wave fields : phase II.

DOT National Transportation Integrated Search

2016-08-01

A new 2-D Full Waveform Inversion (FWI) software code was developed to characterize layering and anomalies beneath the ground surface using seismic testing. The software is capable of assessing the shear and compression wave velocities (Vs and Vp) fo...

Nuclear Explosion Monitoring Advances and Challenges

NASA Astrophysics Data System (ADS)

Baker, G. E.

2015-12-01

We address the state-of-the-art in areas important to monitoring, current challenges, specific efforts that illustrate approaches addressing shortcomings in capabilities, and additional approaches that might be helpful. The exponential increase in the number of events that must be screened as magnitude thresholds decrease presents one of the greatest challenges. Ongoing efforts to exploit repeat seismic events using waveform correlation, subspace methods, and empirical matched field processing holds as much "game-changing" promise as anything being done, and further efforts to develop and apply such methods efficiently are critical. Greater accuracy of travel time, signal loss, and full waveform predictions are still needed to better locate and discriminate seismic events. Important developments include methods to model velocities using multiple types of data; to model attenuation with better separation of source, path, and site effects; and to model focusing and defocusing of surface waves. Current efforts to model higher frequency full waveforms are likely to improve source characterization while more effective estimation of attenuation from ambient noise holds promise for filling in gaps. Censoring in attenuation modeling is a critical problem to address. Quantifying uncertainty of discriminants is key to their operational use. Efforts to do so for moment tensor (MT) inversion are particularly important, and fundamental progress on the statistics of MT distributions is the most important advance needed in the near term in this area. Source physics is seeing great progress through theoretical, experimental, and simulation studies. The biggest need is to accurately predict the effects of source conditions on seismic generation. Uniqueness is the challenge here. Progress will depend on studies that probe what distinguishes mechanisms, rather than whether one of many possible mechanisms is consistent with some set of observations.

Seismic Waveform Tomography of the Iranian Region

NASA Astrophysics Data System (ADS)

Maggi, A.; Priestley, K.; Jackson, J.

2001-05-01

Surprisingly little is known about the detailed velocity structure of Iran, despite the region's importance in the tectonics of the Middle East. Previous studies have concentrated mainly on fundamental mode surface wave dispersion measurements along isolated paths (e.g.~Asudeh, 1982; Cong & Mitchell, 1998; Ritzwoller et.~al, 1998), and the propagation characteristics of crust and upper mantle body waves (e.g. Hearn & Ni 1994; Rodgers et.~al 1997). We use the partitioned waveform inversion method of Nolet (1990) on several hundred regional waveforms crossing the Iranian region to produce a 3-D seismic velocity map for the crust and upper mantle of the area. The method consists of using long period seismograms from earthquakes with well determined focal mechanisms and depths to constrain 1-D path-averaged shear wave models along regional paths. The constraints imposed on the 1-D models by the seismograms are then combined with independent constraints from other methods (e.g.~Moho depths from reciever function analysis etc.), to solve for the 3-D seismic velocity structure of the region. A dense coverage of fundamental mode rayleigh waves at a period of 100~s ensures good resolution of lithospheric scale structure. We also use 20~s period fundamental mode rayleigh waves and some Pnl wavetrains to make estimates of crustal thickness variations and average crustal velocities. A few deeper events give us some coverage of higher mode rayleigh waves and mantle S waves, which sample to the base of the upper mantle. Our crustal thickness estimates range from 45~km in the southern Zagros mountains, to 40~km in central Iran and 35~km towards the north of the region. We also find inconsistencies between the 1-D models required to fit the vertical and the tranverse seismograms, indicating the presence of anisotropy.

Full waveform inversion of combined towed streamer and limited OBS seismic data: a theoretical study

NASA Astrophysics Data System (ADS)

Yang, Huachen; Zhang, Jianzhong

2018-06-01

In marine seismic oil exploration, full waveform inversion (FWI) of towed-streamer data is used to reconstruct velocity models. However, the FWI of towed-streamer data easily converges to a local minimum solution due to the lack of low-frequency content. In this paper, we propose a new FWI technique using towed-streamer data, its integrated data sets and limited OBS data. Both integrated towed-streamer seismic data and OBS data have low-frequency components. Therefore, at early iterations in the new FWI technique, the OBS data combined with the integrated towed-streamer data sets reconstruct an appropriate background model. And the towed-streamer seismic data play a major role in later iterations to improve the resolution of the model. The new FWI technique is tested on numerical examples. The results show that when starting models are not accurate enough, the models inverted using the new FWI technique are superior to those inverted using conventional FWI.

Full-waveform Inversion for Localized 3-D S-velocity Structure in D" Beneath the Caribbean using US-Array Data

NASA Astrophysics Data System (ADS)

Borgeaud, A. F. E.; Konishi, K.; Kawai, K.; Geller, R. J.

2015-12-01

The region beneath Central America is known to have significant lateral velocity heterogeneities from the upper mantle down to the lowermost mantle. It is also known for its long history of subducting oceanic plates and fragmented plate remnants that sunk to the lowermost mantle (e.g., Ren et al., 2007). In this study, we use localized full-waveform inversion to invert for the 3-D S-velocity beneath the Caribbean. We use the DSM (Kawai et al., 2006) to compute 1-D synthetic seismograms and the first-order Born approximation to compute the partial derivatives for 3-D structure. We use a larger dataset with better coverage than Kawai et al. (2014), consisting of S and ScS phases from US-Array data for events in South America. The resulting 3-D model can contribute to understanding whether the cause of the velocity anomalies is thermal, chemical, or due to phase transitions.

Lag compensation of optical fibers or thermocouples to achieve waveform fidelity in dynamic gas pyrometry

NASA Technical Reports Server (NTRS)

Warshawsky, I.

1991-01-01

Fidelity of waveform reproduction requires constant amplitude ratio and constant time lag of a temperature sensor's indication, at all frequencies of interest. However, heat-transfer type sensors usually cannot satisfy these requirements. Equations for the actual indication of a thermocouple and an optical-fiber pyrometer are given explicitly, in terms of sensor and flowing-gas properties. A practical, realistic design of each type of sensor behaves like a first-order system with amplitude-ratio attenuation inversely proportional to frequency when the frequency exceeds the corner frequency. Only at much higher frequencies does the amplitude-ratio attenuation for the optical fiber sensor become inversely proportional to the square root of the frequency. Design options for improving the frequency response are discussed. On-line electrical lag compensation, using a linear amplifier and a passive compensation network, can extend the corner frequency of the thermocouple 100-fold or more; a similar passive network can be used for the optical-fiber sensor. Design details for these networks are presented.

Uncertainty Estimation in Elastic Full Waveform Inversion by Utilising the Hessian Matrix

NASA Astrophysics Data System (ADS)

Hagen, V. S.; Arntsen, B.; Raknes, E. B.

2017-12-01

Elastic Full Waveform Inversion (EFWI) is a computationally intensive iterative method for estimating elastic model parameters. A key element of EFWI is the numerical solution of the elastic wave equation which lies as a foundation to quantify the mismatch between synthetic (modelled) and true (real) measured seismic data. The misfit between the modelled and true receiver data is used to update the parameter model to yield a better fit between the modelled and true receiver signal. A common approach to the EFWI model update problem is to use a conjugate gradient search method. In this approach the resolution and cross-coupling for the estimated parameter update can be found by computing the full Hessian matrix. Resolution of the estimated model parameters depend on the chosen parametrisation, acquisition geometry, and temporal frequency range. Although some understanding has been gained, it is still not clear which elastic parameters can be reliably estimated under which conditions. With few exceptions, previous analyses have been based on arguments using radiation pattern analysis. We use the known adjoint-state technique with an expansion to compute the Hessian acting on a model perturbation to conduct our study. The Hessian is used to infer parameter resolution and cross-coupling for different selections of models, acquisition geometries, and data types, including streamer and ocean bottom seismic recordings. Information about the model uncertainty is obtained from the exact Hessian, and is essential when evaluating the quality of estimated parameters due to the strong influence of source-receiver geometry and frequency content. Investigation is done on both a homogeneous model and the Gullfaks model where we illustrate the influence of offset on parameter resolution and cross-coupling as a way of estimating uncertainty.

Deghosting based on the transmission matrix method

NASA Astrophysics Data System (ADS)

Wang, Benfeng; Wu, Ru-Shan; Chen, Xiaohong

2017-12-01

As the developments of seismic exploration and subsequent seismic exploitation advance, marine acquisition systems with towed streamers become an important seismic data acquisition method. But the existing air-water reflective interface can generate surface related multiples, including ghosts, which can affect the accuracy and performance of the following seismic data processing algorithms. Thus, we derive a deghosting method from a new perspective, i.e. using the transmission matrix (T-matrix) method instead of inverse scattering series. The T-matrix-based deghosting algorithm includes all scattering effects and is convergent absolutely. Initially, the effectiveness of the proposed method is demonstrated using synthetic data obtained from a designed layered model, and its noise-resistant property is also illustrated using noisy synthetic data contaminated by random noise. Numerical examples on complicated data from the open SMAART Pluto model and field marine data further demonstrate the validity and flexibility of the proposed method. After deghosting, low frequency components are recovered reasonably and the fake high frequency components are attenuated, and the recovered low frequency components will be useful for the subsequent full waveform inversion. The proposed deghosting method is currently suitable for two-dimensional towed streamer cases with accurate constant depth information and its extension into variable-depth streamers in three-dimensional cases will be studied in the future.

Broadband optical frequency comb generator based on driving N-cascaded modulators by Gaussian-shaped waveform

NASA Astrophysics Data System (ADS)

Hmood, Jassim K.; Harun, Sulaiman W.

2018-05-01

A new approach for realizing a wideband optical frequency comb (OFC) generator based on driving cascaded modulators by a Gaussian-shaped waveform, is proposed and numerically demonstrated. The setup includes N-cascaded MZMs, a single Gaussian-shaped waveform generator, and N-1 electrical time delayer. The first MZM is driven directly by a Gaussian-shaped waveform, while delayed replicas of the Gaussian-shaped waveform drive the other MZMs. An analytical model that describes the proposed OFC generator is provided to study the effect of number and chirp factor of cascaded MZM as well as pulse width on output spectrum. Optical frequency combs at frequency spacing of 1 GHz are generated by applying Gaussian-shaped waveform at pulse widths ranging from 200 to 400 ps. Our results reveal that, the number of comb lines is inversely proportional to the pulse width and directly proportional to both number and chirp factor of cascaded MZMs. At pulse width of 200 ps and chirp factor of 4, 67 frequency lines can be measured at output spectrum of two-cascaded MZMs setup. Whereas, increasing the number of cascaded stages to 3, 4, and 5, the optical spectra counts 89, 109 and 123 frequency lines; respectively. When the delay time is optimized, 61 comb lines can be achieved with power fluctuations of less than 1 dB for five-cascaded MZMs setup.

Mitigating nonlinearity in full waveform inversion using scaled-Sobolev pre-conditioning

NASA Astrophysics Data System (ADS)

Zuberi, M. AH; Pratt, R. G.

2018-04-01

The Born approximation successfully linearizes seismic full waveform inversion if the background velocity is sufficiently accurate. When the background velocity is not known it can be estimated by using model scale separation methods. A frequently used technique is to separate the spatial scales of the model according to the scattering angles present in the data, by using either first- or second-order terms in the Born series. For example, the well-known `banana-donut' and the `rabbit ear' shaped kernels are, respectively, the first- and second-order Born terms in which at least one of the scattering events is associated with a large angle. Whichever term of the Born series is used, all such methods suffer from errors in the starting velocity model because all terms in the Born series assume that the background Green's function is known. An alternative approach to Born-based scale separation is to work in the model domain, for example, by Gaussian smoothing of the update vectors, or some other approach for separation by model wavenumbers. However such model domain methods are usually based on a strict separation in which only the low-wavenumber updates are retained. This implies that the scattered information in the data is not taken into account. This can lead to the inversion being trapped in a false (local) minimum when sharp features are updated incorrectly. In this study we propose a scaled-Sobolev pre-conditioning (SSP) of the updates to achieve a constrained scale separation in the model domain. The SSP is obtained by introducing a scaled Sobolev inner product (SSIP) into the measure of the gradient of the objective function with respect to the model parameters. This modified measure seeks reductions in the L2 norm of the spatial derivatives of the gradient without changing the objective function. The SSP does not rely on the Born prediction of scale based on scattering angles, and requires negligible extra computational cost per iteration. Synthetic examples from the Marmousi model show that the constrained scale separation using SSP is able to keep the background updates in the zone of attraction of the global minimum, in spite of using a poor starting model in which conventional methods fail.

Adjoint tomography of Europe

NASA Astrophysics Data System (ADS)

Zhu, H.; Bozdag, E.; Peter, D. B.; Tromp, J.

2010-12-01

We use spectral-element and adjoint methods to image crustal and upper mantle heterogeneity in Europe. The study area involves the convergent boundaries of the Eurasian, African and Arabian plates and the divergent boundary between the Eurasian and North American plates, making the tectonic structure of this region complex. Our goal is to iteratively fit observed seismograms and improve crustal and upper mantle images by taking advantage of 3D forward and inverse modeling techniques. We use data from 200 earthquakes with magnitudes between 5 and 6 recorded by 262 stations provided by ORFEUS. Crustal model Crust2.0 combined with mantle model S362ANI comprise the initial 3D model. Before the iterative adjoint inversion, we determine earthquake source parameters in the initial 3D model by using 3D Green functions and their Fréchet derivatives with respect to the source parameters (i.e., centroid moment tensor and location). The updated catalog is used in the subsequent structural inversion. Since we concentrate on upper mantle structures which involve anisotropy, transversely isotropic (frequency-dependent) traveltime sensitivity kernels are used in the iterative inversion. Taking advantage of the adjoint method, we use as many measurements as can obtain based on comparisons between observed and synthetic seismograms. FLEXWIN (Maggi et al., 2009) is used to automatically select measurement windows which are analyzed based on a multitaper technique. The bandpass ranges from 15 second to 150 second. Long-period surface waves and short-period body waves are combined in source relocations and structural inversions. A statistical assessments of traveltime anomalies and logarithmic waveform differences is used to characterize the inverted sources and structure.

Active backstop faults in the Mentawai region of Sumatra, Indonesia, revealed by teleseismic broadband waveform modeling

NASA Astrophysics Data System (ADS)

Wang, Xin; Bradley, Kyle Edward; Wei, Shengji; Wu, Wenbo

2018-02-01

Two earthquake sequences that affected the Mentawai islands offshore of central Sumatra in 2005 (Mw 6.9) and 2009 (Mw 6.7) have been highlighted as evidence for active backthrusting of the Sumatran accretionary wedge. However, the geometry of the activated fault planes is not well resolved due to large uncertainties in the locations of the mainshocks and aftershocks. We refine the locations and focal mechanisms of medium size events (Mw > 4.5) of these two earthquake sequences through broadband waveform modeling. In addition to modeling the depth-phases for accurate centroid depths, we use teleseismic surface wave cross-correlation to precisely relocate the relative horizontal locations of the earthquakes. The refined catalog shows that the 2005 and 2009 "backthrust" sequences in Mentawai region actually occurred on steeply (∼60 degrees) landward-dipping faults (Masilo Fault Zone) that intersect the Sunda megathrust beneath the deepest part of the forearc basin, contradicting previous studies that inferred slip on a shallowly seaward-dipping backthrust. Static slip inversion on the newly-proposed fault fits the coseismic GPS offsets for the 2009 mainshock equally well as previous studies, but with a slip distribution more consistent with the mainshock centroid depth (∼20 km) constrained from teleseismic waveform inversion. Rupture of such steeply dipping reverse faults within the forearc crust is rare along the Sumatra-Java margin. We interpret these earthquakes as 'unsticking' of the Sumatran accretionary wedge along a backstop fault separating imbricated material from the stronger Sunda lithosphere. Alternatively, the reverse faults may have originated as pre-Miocene normal faults of the extended continental crust of the western Sunda margin. Our waveform modeling approach can be used to further refine global earthquake catalogs in order to clarify the geometries of active faults.

Stability and uncertainty of finite-fault slip inversions: Application to the 2004 Parkfield, California, earthquake

USGS Publications Warehouse

Hartzell, S.; Liu, P.; Mendoza, C.; Ji, C.; Larson, K.M.

2007-01-01

The 2004 Parkfield, California, earthquake is used to investigate stability and uncertainty aspects of the finite-fault slip inversion problem with different a priori model assumptions. We utilize records from 54 strong ground motion stations and 13 continuous, 1-Hz sampled, geodetic instruments. Two inversion procedures are compared: a linear least-squares subfault-based methodology and a nonlinear global search algorithm. These two methods encompass a wide range of the different approaches that have been used to solve the finite-fault slip inversion problem. For the Parkfield earthquake and the inversion of velocity or displacement waveforms, near-surface related site response (top 100 m, frequencies above 1 Hz) is shown to not significantly affect the solution. Results are also insensitive to selection of slip rate functions with similar duration and to subfault size if proper stabilizing constraints are used. The linear and nonlinear formulations yield consistent results when the same limitations in model parameters are in place and the same inversion norm is used. However, the solution is sensitive to the choice of inversion norm, the bounds on model parameters, such as rake and rupture velocity, and the size of the model fault plane. The geodetic data set for Parkfield gives a slip distribution different from that of the strong-motion data, which may be due to the spatial limitation of the geodetic stations and the bandlimited nature of the strong-motion data. Cross validation and the bootstrap method are used to set limits on the upper bound for rupture velocity and to derive mean slip models and standard deviations in model parameters. This analysis shows that slip on the northwestern half of the Parkfield rupture plane from the inversion of strong-motion data is model dependent and has a greater uncertainty than slip near the hypocenter.

Restoration of clipped seismic waveforms using projection onto convex sets method

PubMed Central

Zhang, Jinhai; Hao, Jinlai; Zhao, Xu; Wang, Shuqin; Zhao, Lianfeng; Wang, Weimin; Yao, Zhenxing

2016-01-01

The seismic waveforms would be clipped when the amplitude exceeds the upper-limit dynamic range of seismometer. Clipped waveforms are typically assumed not useful and seldom used in waveform-based research. Here, we assume the clipped components of the waveform share the same frequency content with the un-clipped components. We leverage this similarity to convert clipped waveforms to true waveforms by iteratively reconstructing the frequency spectrum using the projection onto convex sets method. Using artificially clipped data we find that statistically the restoration error is ~1% and ~5% when clipped at 70% and 40% peak amplitude, respectively. We verify our method using real data recorded at co-located seismometers that have different gain controls, one set to record large amplitudes on scale and the other set to record low amplitudes on scale. Using our restoration method we recover 87 out of 93 clipped broadband records from the 2013 Mw6.6 Lushan earthquake. Estimating that we recover 20 clipped waveforms for each M5.0+ earthquake, so for the ~1,500 M5.0+ events that occur each year we could restore ~30,000 clipped waveforms each year, which would greatly enhance useable waveform data archives. These restored waveform data would also improve the azimuthal station coverage and spatial footprint. PMID:27966618

Bayesian ISOLA: new tool for automated centroid moment tensor inversion

NASA Astrophysics Data System (ADS)

Vackář, Jiří; Burjánek, Jan; Gallovič, František; Zahradník, Jiří; Clinton, John

2017-04-01

Focal mechanisms are important for understanding seismotectonics of a region, and they serve as a basic input for seismic hazard assessment. Usually, the point source approximation and the moment tensor (MT) are used. We have developed a new, fully automated tool for the centroid moment tensor (CMT) inversion in a Bayesian framework. It includes automated data retrieval, data selection where station components with various instrumental disturbances and high signal-to-noise are rejected, and full-waveform inversion in a space-time grid around a provided hypocenter. The method is innovative in the following aspects: (i) The CMT inversion is fully automated, no user interaction is required, although the details of the process can be visually inspected latter on many figures which are automatically plotted.(ii) The automated process includes detection of disturbances based on MouseTrap code, so disturbed recordings do not affect inversion.(iii) A data covariance matrix calculated from pre-event noise yields an automated weighting of the station recordings according to their noise levels and also serves as an automated frequency filter suppressing noisy frequencies.(iv) Bayesian approach is used, so not only the best solution is obtained, but also the posterior probability density function.(v) A space-time grid search effectively combined with the least-squares inversion of moment tensor components speeds up the inversion and allows to obtain more accurate results compared to stochastic methods. The method has been tested on synthetic and observed data. It has been tested by comparison with manually processed moment tensors of all events greater than M≥3 in the Swiss catalogue over 16 years using data available at the Swiss data center (http://arclink.ethz.ch). The quality of the results of the presented automated process is comparable with careful manual processing of data. The software package programmed in Python has been designed to be as versatile as possible in order to be applicable in various networks ranging from local to regional. The method can be applied either to the everyday network data flow, or to process large previously existing earthquake catalogues and data sets.

Frequency dependent Qα and Qβ in the Umbria-Marche (Italy) region using a quadratic approximation of the coda-normalization method

NASA Astrophysics Data System (ADS)

de Lorenzo, Salvatore; Bianco, Francesca; Del Pezzo, Edoardo

2013-06-01

The coda normalization method is one of the most used methods in the inference of attenuation parameters Qα and Qβ. Since, in this method, the geometrical spreading exponent γ is an unknown model parameter, the most part of studies assumes a fixed γ, generally equal to 1. However γ and Q could be also jointly inferred from the non-linear inversion of coda-normalized logarithms of amplitudes, but the trade-off between γ and Q could give rise to unreasonable values of these parameters. To minimize the trade-off between γ and Q, an inversion method based on a parabolic expression of the coda-normalization equation has been developed. The method has been applied to the waveforms recorded during the 1997 Umbria-Marche seismic crisis. The Akaike criterion has been used to compare results of the parabolic model with those of the linear model, corresponding to γ = 1. A small deviation from the spherical geometrical spreading has been inferred, but this is accompanied by a significant variation of Qα and Qβ values. For almost all the considered stations, Qα smaller than Qβ has been inferred, confirming that seismic attenuation, in the Umbria-Marche region, is controlled by crustal pore fluids.

Lower Crstal Reflectity bands and Magma Emplacement in Norweigian sea, NE Atlantic

NASA Astrophysics Data System (ADS)

Rai, A.; Breivik, A. J.; Mjelde, R.

2013-12-01

In this study we present the OBS data collected along seismic profiles in the norweigian sea. The traveltime modelling of the OBS data provides first-hand information about seismic structure of the subsurface. However, waveform modelling is used to further constrain the fine scale structure, velocity constrast and velocity gradients. By forward modelling and inversion of the seismic waveforms, we show that the multiple bands of reflectivity could be due to multiple episodes of magma emplacements that might have frozen in the form of sills. These mafic intrusions probably intruded into the ductile lower crust during the main rifting phase of Europe and Greenland.

Source mechanism of very-long-period signals accompanying dome growth activity at Merapi volcano, Indonesia

USGS Publications Warehouse

Hidayat, D.; Chouet, B.; Voight, B.; Dawson, P.; Ratdomopurbo, Antonius

2002-01-01

Very-long-period (VLP) pulses with period of 6-7s, displaying similar waveforms, were identified in 1998 from broadband seismographs around the summit crater. These pulses accompanied most of multiphase (MP) earthquakes, a type of long-period event locally defined at Merapi Volcano. Source mechanisms for several VLP pulses were examined by applying moment tensor inversion to the waveform data. Solutions were consistent with a crack striking ???70?? and dipping ???50?? SW, 100m under the active dome, suggest pressurized gas transport involving accumulation and sudden release of 10-60 m3 of gas in the crack over a 6s interval.

Visual motion direction is represented in population-level neural response as measured by magnetoencephalography.

PubMed

Kaneoke, Y; Urakawa, T; Kakigi, R

2009-05-19

We investigated whether direction information is represented in the population-level neural response evoked by the visual motion stimulus, as measured by magnetoencephalography. Coherent motions with varied speed, varied direction, and different coherence level were presented using random dot kinematography. Peak latency of responses to motion onset was inversely related to speed in all directions, as previously reported, but no significant effect of direction on latency changes was identified. Mutual information entropy (IE) calculated using four-direction response data increased significantly (>2.14) after motion onset in 41.3% of response data and maximum IE was distributed at approximately 20 ms after peak response latency. When response waveforms showing significant differences (by multivariate discriminant analysis) in distribution of the three waveform parameters (peak amplitude, peak latency, and 75% waveform width) with stimulus directions were analyzed, 87 waveform stimulus directions (80.6%) were correctly estimated using these parameters. Correct estimation rate was unaffected by stimulus speed, but was affected by coherence level, even though both speed and coherence affected response amplitude similarly. Our results indicate that speed and direction of stimulus motion are represented in the distinct properties of a response waveform, suggesting that the human brain processes speed and direction separately, at least in part.

FDM simulation of earthquakes off western Kyushu, Japan, using a land-ocean unified 3D structure model

NASA Astrophysics Data System (ADS)

Okamoto, Taro; Takenaka, Hiroshi; Nakamura, Takeshi; Hara, Tatsuhiko

2017-07-01

Seismic activity occurred off western Kyushu, Japan, at the northern end of the Okinawa Trough on May 6, 2016 (14:11 JST), 22 days after the onset of the 2016 Kumamoto earthquake sequence. The area is adjacent to the Beppu-Shimabara graben where the 2016 Kumamoto earthquake sequence occurred. In the area off western Kyushu, a M7.1 earthquake also occurred on November 14, 2015 (5:51 JST), and a tsunami with a height of 0.3 m was observed. In order to better understand these seismic activity and tsunamis, it is necessary to study the sources of, and strong motions due to, earthquakes in the area off western Kyushu. For such studies, validation of synthetic waveforms is important because of the presence of the oceanic water layer and thick sediments in the source area. We show the validation results for synthetic waveforms through nonlinear inversion analyses of small earthquakes ( M5). We use a land-ocean unified 3D structure model, 3D HOT finite-difference method ("HOT" stands for Heterogeneity, Ocean layer and Topography) and a multi-graphic processing unit (GPU) acceleration to simulate the wave propagations. We estimate the first-motion augmented moment tensor (FAMT) solution based on both the long-period surface waves and short-period body waves. The FAMT solutions systematically shift landward by about 13 km, on average, from the epicenters determined by the Japan Meteorological Agency. The synthetics provide good reproductions of the observed full waveforms with periods of 10 s or longer. On the other hand, for waveforms with shorter periods (down to 4 s), the later surface waves are not reproduced well, while the first parts of the waveforms (comprising P- and S-waves) are reproduced to some extent. These results indicate that the current 3D structure model around Kyushu is effective for generating full waveforms, including surface waves with periods of about 10 s or longer. Based on these findings, we analyze the 2015 M7.1 event using the cross-correlations between the observed and synthetic waveforms. The result suggests a rupture propagation toward the NNE, with a major radiation about 25 km north of the onset point.[Figure not available: see fulltext.

Extrapolation of sonic boom pressure signatures by the waveform parameter method

NASA Technical Reports Server (NTRS)

Thomas, C. L.

1972-01-01

The waveform parameter method of sonic boom extrapolation is derived and shown to be equivalent to the F-function method. A computer program based on the waveform parameter method is presented and discussed, with a sample case demonstrating program input and output.

A Study of Regional Waveform Calibration in the Eastern Mediterranean Region.

NASA Astrophysics Data System (ADS)

di Luccio, F.; Pino, A.; Thio, H.

2002-12-01

We modeled Pnl phases from several moderate magnitude events in the eastern Mediterranean to test methods and to develop path calibrations for source determination. The study region spanning from the eastern part of the Hellenic arc to the eastern Anatolian fault is mostly interested by moderate earthquakes, that can produce relevant damages. The selected area consists of several tectonic environment, which produces increased level of difficulty in waveform modeling. The results of this study are useful for the analysis of regional seismicity and for seismic hazard as well, in particular because very few broadband seismic stations are available in the selected area. The obtained velocity model gives a 30 km crustal tickness and low upper mantle velocities. The applied inversion procedure to determine the source mechanism has been successful, also in terms of discrimination of depth, for the entire range of selected paths. We conclude that using the true calibration of the seismic structure and high quality broadband data, it is possible to determine the seismic source in terms of mechanism, even with a single station.

Parameters Comparsion of Leads Detection in Arctic Sea Ice Using CRYOSAT-2 Waveform Data

NASA Astrophysics Data System (ADS)

Li, J.; Zhang, S.; Xiao, F.; Zhu, C.; Zhang, Y.; Zhu, T.; Yuan, L.

2018-04-01

Leads are only a small part of the polar sea ice structure, but they play a dominant role on the turbulence exchange between the ocean and the atmosphere, they are also important factors about sea ice thickness inversion. Since the early 2000s, Satellite altimetry has been applied to monitor the Arctic sea ice thickness, Satellite altimetry data can be used to distinguish leads and sea ice. In this paper, four parameters including Pulse peakiness (PP), stack standard deviation (SSD), stack kurtosis (SKU) and stack skewness (SSK) are extracted from CryoSat-2 satellite altimetry waveform data. The four parameters are combined into five combinations (PP, PP&SSD, PP&SSD&SKU, PP&SSD&SSK, PP&SSD&SSK&SKU) with constrain conditions to detect the leads. The results of the five methods are compared with MODIS (moderate-resolution imagining spectroradiometer) images and show that, the combination of PP&SSD is better than the single PP, the rest of combinations are the same as the combination of PP&SSD. It turns out, there is no promotion when we add SSK and SKU, successively or simultaneously.

The effects of core-reflected waves on finite fault inversion with teleseismic body wave data

NASA Astrophysics Data System (ADS)

Qian, Y.; Ni, S.; Wei, S.

2016-12-01

Reliable estimation of rupture processes for a large earthquake is valuable for post-seismic rescue, tsunami alert, seismotectonic studies, as well as earthquake physics. Finite-fault inversion has been widely accepted to reconstruct the spatial-temporal distribution of rupture processes, which can be obtained by individual or jointly inversion of seismic, geodetic and tsunami data sets. Among the above observations, teleseismic (30° 90°) body waves, usually P and SH waves, have been used extensively in such inversions because their propagation are well understood and readily available for large earthquakes with good coverages of slowness and azimuth. However, finite fault inversion methods usually assume turning P and SH waves without inclusion of core-reflected waves when calculating the synthetic waveforms, which may result in systematic error in finite-fault inversions. For the core-reflected SH wave ScS, it is expected to be strong due to total reflection from Core-Mantle-Boundary. Moreover, the time interval between direct S and ScS could be smaller than the duration of large earthquakes for large epicentral distances. In order to improve the accuracy of finite fault inversion with teleseismic body waves, we develop a procedure named multitel3 to compute Greens' functions that contain both turning waves (P, pP, sP, S, sS et al.) and core-reflected phases (PcP and ScS) and apply it to finite fault inversions. This ray-based method can rapidly calculate teleseismic body wave synthetics with flexibility for path calibration of 3D mantle structure. The new Green's function is plugged into finite fault inversion package to replace the original Green's function with only turning P and SH waves. With the 2008 Mw7.9 Wenchuan earthquake as example, a series of numerical tests conducted on synthetic data are used to assess the performance of our approach. We also explore this new procedure's stability when there are discrepancies between the parameters of input model and the priori information of inverse model, such as strike, dip of finite fault and so on. With the quantified code, we apply it to study rupture process of the 2016 Mw7.8 Sumatra earthquake.

W-phase estimation of first-order rupture distribution for megathrust earthquakes

NASA Astrophysics Data System (ADS)

Benavente, Roberto; Cummins, Phil; Dettmer, Jan

2014-05-01

Estimating the rupture pattern for large earthquakes during the first hour after the origin time can be crucial for rapid impact assessment and tsunami warning. However, the estimation of coseismic slip distribution models generally involves complex methodologies that are difficult to implement rapidly. Further, while model parameter uncertainty can be crucial for meaningful estimation, they are often ignored. In this work we develop a finite fault inversion for megathrust earthquakes which rapidly generates good first order estimates and uncertainties of spatial slip distributions. The algorithm uses W-phase waveforms and a linear automated regularization approach to invert for rupture models of some recent megathrust earthquakes. The W phase is a long period (100-1000 s) wave which arrives together with the P wave. Because it is fast, has small amplitude and a long-period character, the W phase is regularly used to estimate point source moment tensors by the NEIC and PTWC, among others, within an hour of earthquake occurrence. We use W-phase waveforms processed in a manner similar to that used for such point-source solutions. The inversion makes use of 3 component W-phase records retrieved from the Global Seismic Network. The inverse problem is formulated by a multiple time window method, resulting in a linear over-parametrized problem. The over-parametrization is addressed by Tikhonov regularization and regularization parameters are chosen according to the discrepancy principle by grid search. Noise on the data is addressed by estimating the data covariance matrix from data residuals. The matrix is obtained by starting with an a priori covariance matrix and then iteratively updating the matrix based on the residual errors of consecutive inversions. Then, a covariance matrix for the parameters is computed using a Bayesian approach. The application of this approach to recent megathrust earthquakes produces models which capture the most significant features of their slip distributions. Also, reliable solutions are generally obtained with data in a 30-minute window following the origin time, suggesting that a real-time system could obtain solutions in less than one hour following the origin time.

Spectral filtering of gradient for l2-norm frequency-domain elastic waveform inversion

NASA Astrophysics Data System (ADS)

Oh, Ju-Won; Min, Dong-Joo

2013-05-01

To enhance the robustness of the l2-norm elastic full-waveform inversion (FWI), we propose a denoise function that is incorporated into single-frequency gradients. Because field data are noisy and modelled data are noise-free, the denoise function is designed based on the ratio of modelled data to field data summed over shots and receivers. We first take the sums of the modelled data and field data over shots, then take the sums of the absolute values of the resultant modelled data and field data over the receivers. Due to the monochromatic property of wavefields at each frequency, signals in both modelled and field data tend to be cancelled out or maintained, whereas certain types of noise, particularly random noise, can be amplified in field data. As a result, the spectral distribution of the denoise function is inversely proportional to the ratio of noise to signal at each frequency, which helps prevent the noise-dominant gradients from contributing to model parameter updates. Numerical examples show that the spectral distribution of the denoise function resembles a frequency filter that is determined by the spectrum of the signal-to-noise (S/N) ratio during the inversion process, with little human intervention. The denoise function is applied to the elastic FWI of synthetic data, with three types of random noise generated by the modified version of the Marmousi-2 model: white, low-frequency and high-frequency random noises. Based on the spectrum of S/N ratios at each frequency, the denoise function mainly suppresses noise-dominant single-frequency gradients, which improves the inversion results at the cost of spatial resolution.

Hydraulic Fracturing Induced Seismicity at Preese Hall, UK: Moment Tensors, Uncertainties and Implications for Microseismic Monitoring Strategies

NASA Astrophysics Data System (ADS)

O'toole, T. B.; Woodhouse, J. H.; Verdon, J.; Kendall, J.

2012-12-01

Hydraulic fracturing operations carried out in April and May 2011 by Cuadrilla Resources Ltd. during the exploration of a shale gas reservoir at Preese Hall, near Blackpool, UK, induced a series of microseismic events. The largest of these, with magnitude ML = 2.3, was felt at the surface and recorded by the British Geological Survey regional seismic network. Subsequently, two local seismic stations were installed, which continued to detect seismicity with ML ≤ 1.5 until the hydraulic fracture treatment was suspended due to the anomalously large magnitudes of the induced earthquakes. Here, we present the results of moment tensor inversions of seismic waveforms recorded by these two near-field stations. We determine the best point source description of an event by minimising the least-squares difference between observed and synthetic waveforms. In contrast to source mechanisms obtained from body wave polarity and amplitude picks, which require a good sampling of the focal sphere and typically assume a pure double-couple mechanism, using the whole waveform allows us to place good constraints on the moment tensor even when only a few seismograms are available, and also enables the investigation of possible non-double-couple components and volume changes associated with a source. We discuss our results in the context of the studies commissioned by Cuadrilla after the suspension of hydraulic fracturing operations at Preese Hall. Using synthetic waveform data, we investigate how different monitoring geometries can be used to reduce uncertainties in source parameters of induced microseisms. While our focus is on the monitoring of hydraulic fracturing operations, the methods developed here are general and could equally be applied to determine moment tensors from surface and borehole observations of seismicity induced by other activities.

Determine Earthquake Rupture Directivity Using Taiwan TSMIP Strong Motion Waveforms

NASA Astrophysics Data System (ADS)

Chang, Kaiwen; Chi, Wu-Cheng; Lai, Ying-Ju; Gung, YuanCheng

2013-04-01

Inverting seismic waveforms for the finite fault source parameters is important for studying the physics of earthquake rupture processes. It is also significant to image seismogenic structures in urban areas. Here we analyze the finite-source process and test for the causative fault plane using the accelerograms recorded by the Taiwan Strong-Motion Instrumentation Program (TSMIP) stations. The point source parameters for the mainshock and aftershocks were first obtained by complete waveform moment tensor inversions. We then use the seismograms generated by the aftershocks as empirical Green's functions (EGFs) to retrieve the apparent source time functions (ASTFs) of near-field stations using projected Landweber deconvolution approach. The method for identifying the fault plane relies on the spatial patterns of the apparent source time function durations which depend on the angle between rupture direction and the take-off angle and azimuth of the ray. These derived duration patterns then are compared with the theoretical patterns, which are functions of the following parameters, including focal depth, epicentral distance, average crustal 1D velocity, fault plane attitude, and rupture direction on the fault plane. As a result, the ASTFs derived from EGFs can be used to infer the ruptured fault plane and the rupture direction. Finally we used part of the catalogs to study important seismogenic structures in the area near Chiayi, Taiwan, where a damaging earthquake has occurred about a century ago. The preliminary results show a strike-slip earthquake on 22 October 1999 (Mw 5.6) has ruptured unilaterally toward SSW on a sub-vertical fault. The procedure developed from this study can be applied to other strong motion waveforms recorded from other earthquakes to better understand their kinematic source parameters.

Factorization and resummation: A new paradigm to improve gravitational wave amplitudes

NASA Astrophysics Data System (ADS)

Nagar, Alessandro; Shah, Abhay

2016-11-01

We introduce a new resummed analytical form of the post-Newtonian (PN), factorized, multipolar amplitude corrections fℓm of the effective-one-body (EOB) gravitational waveform of spinning, nonprecessing, circularized, coalescing black hole binaries (BBHs). This stems from the following two-step paradigm: (i) the factorization of the orbital (spin-independent) terms in fℓm; (ii) the resummation of the residual spin (or orbital) factors. We find that resumming the residual spin factor by taking its inverse resummed (iResum) is an efficient way to obtain amplitudes that are more accurate in the strong-field, fast-velocity regime. The performance of the method is illustrated on the ℓ=2 and m =(1 ,2 ) waveform multipoles, both for a test mass orbiting around a Kerr black hole and for comparable-mass BBHs. In the first case, the iResum fℓm's are much closer to the corresponding "exact" functions (obtained by numerically solving the Teukolsky equation) up to the light ring than the nonresummed ones, especially when the black-hole spin is nearly extremal. The iResum paradigm is also more efficient than including higher post-Newtonian terms (up to 20PN order): the resummed 5PN information yields per se a rather good numerical or analytical agreement at the last stable orbit and a well-controlled behavior up to the light ring. For comparable mass binaries (including the highest PN-order information available, 3.5PN), comparing EOB with numerical relativity (NR) data shows that the EOB/NR fractional disagreement at merger, without NR calibration of the EOB waveform, is generically reduced by iResum, from 40% of the usual approach to just a few percent. This suggests that EOBNR waveform models for coalescing BBHs may be improved by using iResum amplitudes.

SGRAPH (SeismoGRAPHer): Seismic waveform analysis and integrated tools in seismology

NASA Astrophysics Data System (ADS)

Abdelwahed, Mohamed F.

2012-03-01

Although numerous seismological programs are currently available, most of them suffer from the inability to manipulate different data formats and the lack of embedded seismological tools. SeismoGRAPHer, or simply SGRAPH, is a new system for maintaining and analyzing seismic waveform data in a stand-alone, Windows-based application that manipulates a wide range of data formats. SGRAPH was intended to be a tool sufficient for performing basic waveform analysis and solving advanced seismological problems. The graphical user interface (GUI) utilities and the Windows functionalities, such as dialog boxes, menus, and toolbars, simplify the user interaction with the data. SGRAPH supports common data formats, such as SAC, SEED, GSE, ASCII, and Nanometrics Y-format, and provides the ability to solve many seismological problems with built-in inversion tools. Loaded traces are maintained, processed, plotted, and saved as SAC, ASCII, or PS (post script) file formats. SGRAPH includes Generalized Ray Theory (GRT), genetic algorithm (GA), least-square fitting, auto-picking, fast Fourier transforms (FFT), and many additional tools. This program provides rapid estimation of earthquake source parameters, location, attenuation, and focal mechanisms. Advanced waveform modeling techniques are provided for crustal structure and focal mechanism estimation. SGRAPH has been employed in the Egyptian National Seismic Network (ENSN) as a tool assisting with routine work and data analysis. More than 30 users have been using previous versions of SGRAPH in their research for more than 3 years. The main features of this application are ease of use, speed, small disk space requirements, and the absence of third-party developed components. Because of its architectural structure, SGRAPH can be interfaced with newly developed methods or applications in seismology. A complete setup file, including the SGRAPH package with the online user guide, is available.

Analysis of seismicity and stress before and after the Mw 8.1 Pisagua, Chile, 2014 earthquake

NASA Astrophysics Data System (ADS)

Grigoli, F.; Cesca, S.; Dahm, T.; Hainzl, S.

2014-12-01

On April 1st, 2014 at 23:46:50 UTC, a powerful earthquake of magnitude Mw 8.1 occurred offshore the Northern Chile in the region of the North Chilean seismic gap. The epicenter of the earthquake was approximately 50 km offshore the Chilean coast, near the town of Pisagua. Two days after the main event a Mw 7.6 aftershock struck approximately the same area. In order to identify spatio-temporal changes of source parameters and stress before and after the mainshock, we analyzed in detail the local seismicity above magnitude Mw 3.0 within the time period 01/01/2013-30/04/2014 and estimated long term trends in b-values and earthquake productivity. We used data from the IPOC (Integrated Plate boundary Observatory Chile) regional seismic network, consisting of 20 "in land" broadband station deployed and managed by the GFZ-Potsdam. The recorded earthquake catalog shows an intense foreshock activity consisting of more than 1000 M3+ events in the source region. Full waveform techniques are used to derive both locations and focal mechanisms of about 435 seismic events. The location process has been performed by using a waveform stacking method (Grigoli et al 2013, 2014) with a layered velocity model based on CRUST 2.0 (see the attached figure for the location results of one of these events). Moment tensor inversion has been performed by using the KIWI tool software (Cesca et al. 2010), which is based on a two-step inversion approach. The first step consists in the inversion of the amplitude spectra to retrieve the best fitting focal planes, while the second inversion step is carried out in time domain to solve the focal mechanism polarity and to obtain the centroid location and time. Both location and moment tensor inversion resulted in agreement with the geodynamical settings of the region. Mapping the b-value reveals a spatiotemporal anomaly of low b-values characterizing the frequency-magnitude distribution of the foreshocks in the source area of the mainshock. Finally, clustering analysis of the retrieved focal mechanism and a stress tensor inversion has been performed in order to analyze the spatio-temporal evolution of the stress, before and after the mainshock. This work has been funded by the German BMBF "Geothecnologien" project MINE (BMBF03G0737A) and by Hazard and Risk Team (HART) and PBO-Chile of GFZ.

Multi-Gaussian fitting for pulse waveform using Weighted Least Squares and multi-criteria decision making method.

PubMed

Wang, Lu; Xu, Lisheng; Feng, Shuting; Meng, Max Q-H; Wang, Kuanquan

2013-11-01

Analysis of pulse waveform is a low cost, non-invasive method for obtaining vital information related to the conditions of the cardiovascular system. In recent years, different Pulse Decomposition Analysis (PDA) methods have been applied to disclose the pathological mechanisms of the pulse waveform. All these methods decompose single-period pulse waveform into a constant number (such as 3, 4 or 5) of individual waves. Furthermore, those methods do not pay much attention to the estimation error of the key points in the pulse waveform. The estimation of human vascular conditions depends on the key points' positions of pulse wave. In this paper, we propose a Multi-Gaussian (MG) model to fit real pulse waveforms using an adaptive number (4 or 5 in our study) of Gaussian waves. The unknown parameters in the MG model are estimated by the Weighted Least Squares (WLS) method and the optimized weight values corresponding to different sampling points are selected by using the Multi-Criteria Decision Making (MCDM) method. Performance of the MG model and the WLS method has been evaluated by fitting 150 real pulse waveforms of five different types. The resulting Normalized Root Mean Square Error (NRMSE) was less than 2.0% and the estimation accuracy for the key points was satisfactory, demonstrating that our proposed method is effective in compressing, synthesizing and analyzing pulse waveforms. Copyright © 2013 Elsevier Ltd. All rights reserved.

A new method to detect event-related potentials based on Pearson's correlation.

PubMed

Giroldini, William; Pederzoli, Luciano; Bilucaglia, Marco; Melloni, Simone; Tressoldi, Patrizio

2016-12-01

Event-related potentials (ERPs) are widely used in brain-computer interface applications and in neuroscience.  Normal EEG activity is rich in background noise, and therefore, in order to detect ERPs, it is usually necessary to take the average from multiple trials to reduce the effects of this noise.  The noise produced by EEG activity itself is not correlated with the ERP waveform and so, by calculating the average, the noise is decreased by a factor inversely proportional to the square root of N , where N is the number of averaged epochs. This is the easiest strategy currently used to detect ERPs, which is based on calculating the average of all ERP's waveform, these waveforms being time- and phase-locked.  In this paper, a new method called GW6 is proposed, which calculates the ERP using a mathematical method based only on Pearson's correlation. The result is a graph with the same time resolution as the classical ERP and which shows only positive peaks representing the increase-in consonance with the stimuli-in EEG signal correlation over all channels.  This new method is also useful for selectively identifying and highlighting some hidden components of the ERP response that are not phase-locked, and that are usually hidden in the standard and simple method based on the averaging of all the epochs.  These hidden components seem to be caused by variations (between each successive stimulus) of the ERP's inherent phase latency period (jitter), although the same stimulus across all EEG channels produces a reasonably constant phase. For this reason, this new method could be very helpful to investigate these hidden components of the ERP response and to develop applications for scientific and medical purposes. Moreover, this new method is more resistant to EEG artifacts than the standard calculations of the average and could be very useful in research and neurology.  The method we are proposing can be directly used in the form of a process written in the well-known Matlab programming language and can be easily and quickly written in any other software language.

Kinematic inversion of the 2008 Mw7 Iwate-Miyagi (Japan) earthquake by two independent methods: Sensitivity and resolution analysis

NASA Astrophysics Data System (ADS)

Gallovic, Frantisek; Cirella, Antonella; Plicka, Vladimir; Piatanesi, Alessio

2013-04-01

On 14 June 2008, UTC 23:43, the border of Iwate and Miyagi prefectures was hit by an Mw7 reverse-fault type crustal earthquake. The event is known to have the largest ground acceleration observed to date (~4g), which was recorded at station IWTH25. We analyze observed strong motion data with the objective to image the event rupture process and the associated uncertainties. Two different slip inversion approaches are used, the difference between the two methods being only in the parameterization of the source model. To minimize mismodeling of the propagation effects we use crustal model obtained by full waveform inversion of aftershock records in the frequency range between 0.05-0.3 Hz. In the first method, based on linear formulation, the parameters are represented by samples of slip velocity functions along the (finely discretized) fault in a time window spanning the whole rupture duration. Such a source description is very general with no prior constraint on the nucleation point, rupture velocity, shape of the velocity function. Thus the inversion could resolve very general (unexpected) features of the rupture evolution, such as multiple rupturing, rupture-propagation reversals, etc. On the other hand, due to the relatively large number of model parameters, the inversion result is highly non-unique, with possibility of obtaining a biased solution. The second method is a non-linear global inversion technique, where each point on the fault can slip only once, following a prescribed functional form of the source time function. We invert simultaneously for peak slip velocity, slip angle, rise time and rupture time by allowing a given range of variability for each kinematic model parameter. For this reason, unlike to the linear inversion approach, the rupture process needs a smaller number of parameters to be retrieved, and is more constrained with a proper control on the allowed range of parameter values. In order to test the resolution and reliability of the retrieved models, we present a thorough analysis of the performance of the two inversion approaches. In fact, depending on the inversion strategy and the intrinsic 'non-uniqueness' of the inverse problem, the final slip maps and distribution of rupture onset times are generally different, sometimes even incompatible with each other. Great emphasis is devoted to the uncertainty estimate of both techniques. Thus we do not compare only the best fitting models, but their 'compatibility' in terms of the uncertainty limits.

Adaptive phase k-means algorithm for waveform classification

NASA Astrophysics Data System (ADS)

Song, Chengyun; Liu, Zhining; Wang, Yaojun; Xu, Feng; Li, Xingming; Hu, Guangmin

2018-01-01

Waveform classification is a powerful technique for seismic facies analysis that describes the heterogeneity and compartments within a reservoir. Horizon interpretation is a critical step in waveform classification. However, the horizon often produces inconsistent waveform phase, and thus results in an unsatisfied classification. To alleviate this problem, an adaptive phase waveform classification method called the adaptive phase k-means is introduced in this paper. Our method improves the traditional k-means algorithm using an adaptive phase distance for waveform similarity measure. The proposed distance is a measure with variable phases as it moves from sample to sample along the traces. Model traces are also updated with the best phase interference in the iterative process. Therefore, our method is robust to phase variations caused by the interpretation horizon. We tested the effectiveness of our algorithm by applying it to synthetic and real data. The satisfactory results reveal that the proposed method tolerates certain waveform phase variation and is a good tool for seismic facies analysis.

Uncertainties in the 2004 Sumatra–Andaman source through nonlinear stochastic inversion of tsunami waves

PubMed Central

Venugopal, M.; Roy, D.; Rajendran, K.; Guillas, S.; Dias, F.

2017-01-01

Numerical inversions for earthquake source parameters from tsunami wave data usually incorporate subjective elements to stabilize the search. In addition, noisy and possibly insufficient data result in instability and non-uniqueness in most deterministic inversions, which are barely acknowledged. Here, we employ the satellite altimetry data for the 2004 Sumatra–Andaman tsunami event to invert the source parameters. We also include kinematic parameters that improve the description of tsunami generation and propagation, especially near the source. Using a finite fault model that represents the extent of rupture and the geometry of the trench, we perform a new type of nonlinear joint inversion of the slips, rupture velocities and rise times with minimal a priori constraints. Despite persistently good waveform fits, large uncertainties in the joint parameter distribution constitute a remarkable feature of the inversion. These uncertainties suggest that objective inversion strategies should incorporate more sophisticated physical models of seabed deformation in order to significantly improve the performance of early warning systems. PMID:28989311

Uncertainties in the 2004 Sumatra-Andaman source through nonlinear stochastic inversion of tsunami waves.

PubMed

Gopinathan, D; Venugopal, M; Roy, D; Rajendran, K; Guillas, S; Dias, F

2017-09-01

Numerical inversions for earthquake source parameters from tsunami wave data usually incorporate subjective elements to stabilize the search. In addition, noisy and possibly insufficient data result in instability and non-uniqueness in most deterministic inversions, which are barely acknowledged. Here, we employ the satellite altimetry data for the 2004 Sumatra-Andaman tsunami event to invert the source parameters. We also include kinematic parameters that improve the description of tsunami generation and propagation, especially near the source. Using a finite fault model that represents the extent of rupture and the geometry of the trench, we perform a new type of nonlinear joint inversion of the slips, rupture velocities and rise times with minimal a priori constraints. Despite persistently good waveform fits, large uncertainties in the joint parameter distribution constitute a remarkable feature of the inversion. These uncertainties suggest that objective inversion strategies should incorporate more sophisticated physical models of seabed deformation in order to significantly improve the performance of early warning systems.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Eken, T; Mayeda, K; Hofstetter, A

A recently developed coda magnitude methodology was applied to selected broadband stations in Turkey for the purpose of testing the coda method in a large, laterally complex region. As found in other, albeit smaller regions, coda envelope amplitude measurements are significantly less variable than distance-corrected direct wave measurements (i.e., L{sub g} and surface waves) by roughly a factor 3-to-4. Despite strong lateral crustal heterogeneity in Turkey, we found that the region could be adequately modeled assuming a simple 1-D, radially symmetric path correction for 10 narrow frequency bands ranging between 0.02 to 2.0 Hz. For higher frequencies however, 2-D pathmore » corrections will be necessary and will be the subject of a future study. After calibrating the stations ISP, ISKB, and MALT for local and regional distances, single-station moment-magnitude estimates (M{sub w}) derived from the coda spectra were in excellent agreement with those determined from multi-station waveform modeling inversions of long-period data, exhibiting a data standard deviation of 0.17. Though the calibration was validated using large events, the results of the calibration will extend M{sub w} estimates to significantly smaller events which could not otherwise be waveform modeled due to poor signal-to-noise ratio at long periods and sparse station coverage. The successful application of the method is remarkable considering the significant lateral complexity in Turkey and the simple assumptions used in the coda method.« less

Seismic Structure of the Antarctic Upper Mantle and Transition Zone Unearthed by Full Waveform Adjoint Tomography

NASA Astrophysics Data System (ADS)

Lloyd, A. J.; Wiens, D.; Zhu, H.; Tromp, J.; Nyblade, A.; Anandakrishnan, S.; Aster, R. C.; Huerta, A. D.; Winberry, J. P.; Wilson, T. J.; Dalziel, I. W. D.; Hansen, S. E.; Shore, P.

2017-12-01

The upper mantle and transition zone beneath Antarctica and the surrounding ocean are among the poorest seismically imaged regions of the Earth's interior. Over the last 1.5 decades researchers have deployed several large temporary broadband seismic arrays focusing on major tectonic features in the Antarctic. The broader international community has also facilitated further instrumentation of the continent, often operating stations in additional regions. As of 2016, waveforms are available from almost 300 unique station locations. Using these stations along with 26 southern mid-latitude seismic stations we have imaged the seismic structure of the upper mantle and transition zone using full waveform adjoint techniques. The full waveform adjoint inversion assimilates phase observations from 3-component seismograms containing P, S, Rayleigh, and Love waves, including reflections and overtones, from 270 earthquakes (5.5 ≤ Mw ≤ 7.0) that occurred between 2001-2003 and 2007-2016. We present the major results of the full waveform adjoint inversion following 20 iterations, resulting in a continental-scale seismic model (ANT_20) with regional-scale resolution. Within East Antarctica, ANT_20 reveals internal seismic heterogeneity and differences in lithospheric thickness. For example, fast seismic velocities extending to 200-300 km depth are imaged beneath both Wilkes Land and the Gamburtsev Subglacial Mountains, whereas fast velocities only extend to 100-200 km depth beneath the Lambert Graben and Enderby Land. Furthermore, fast velocities are not found beneath portions of Dronning Maud Land, suggesting old cratonic lithosphere may be absent. Beneath West Antarctica slow upper mantle seismic velocities are imaged extending from the Balleny Island southward along the Transantarctic Mountains front, and broaden beneath the southern and northern portion of the mountain range. In addition, slow upper mantle velocities are imaged beneath the West Antarctic coast extending from Marie Byrd Land to the Antarctic Peninsula. This region of slow velocity only extends to 150-200 km depth beneath the Antarctic Peninsula, while elsewhere it extends to deeper upper mantle depths and possibly into the transition zone as well as offshore, suggesting two different geodynamic processes are at play.

Acoustic Green's function extraction in the ocean

NASA Astrophysics Data System (ADS)

Zang, Xiaoqin

The acoustic Green's function (GF) is the key to understanding the acoustic properties of ocean environments. With knowledge of the acoustic GF, the physics of sound propagation, such as dispersion, can be analyzed; underwater communication over thousands of miles can be understood; physical properties of the ocean, including ocean temperature, ocean current speed, as well as seafloor bathymetry, can be investigated. Experimental methods of acoustic GF extraction can be categorized as active methods and passive methods. Active methods are based on employment of man-made sound sources. These active methods require less computational complexity and time, but may cause harm to marine mammals. Passive methods cost much less and do not harm marine mammals, but require more theoretical and computational work. Both methods have advantages and disadvantages that should be carefully tailored to fit the need of each specific environment and application. In this dissertation, we study one passive method, the noise interferometry method, and one active method, the inverse filter processing method, to achieve acoustic GF extraction in the ocean. The passive method of noise interferometry makes use of ambient noise to extract an approximation to the acoustic GF. In an environment with a diffusive distribution of sound sources, sound waves that pass through two hydrophones at two locations carry the information of the acoustic GF between these two locations; by listening to the long-term ambient noise signals and cross-correlating the noise data recorded at two locations, the acoustic GF emerges from the noise cross-correlation function (NCF); a coherent stack of many realizations of NCFs yields a good approximation to the acoustic GF between these two locations, with all the deterministic structures clearly exhibited in the waveform. To test the performance of noise interferometry in different types of ocean environments, two field experiments were performed and ambient noise data were collected in a 100-meter deep coastal ocean environment and a 600-meter deep ocean environment. In the coastal ocean environment, the collected noise data were processed by coherently stacking five days of cross-correlation functions between pairs of hydrophones separated by 5 km, 10 km and 15 km, respectively. NCF waveforms were modeled using the KRAKEN normal mode model, with the difference between the NCFs and the acoustic GFs quantified by a weighting function. Through waveform inversion of NCFs, an optimal geoacoustic model was obtained by minimizing the two-norm misfit between the simulation and the measurement. Using a simulated time-reversal mirror, the extracted GF was back propagated from the receiver location to the virtual source, and a strong focus was found in the vicinity of the source, which provides additional support for the optimality of the aforementioned geoacoustic model. With the extracted GF, dispersion in experimental shallow water environment was visualized in the time-frequency representation. Normal modes of GFs were separated using the time-warping transformation. By separating the modes in the frequency domain of the time-warped signal, we isolated modal arrivals and reconstructed the NCF by summing up the isolated modes, thereby significantly improving the signal-to-noise ratio of NCFs. Finally, these reconstructed NCFs were employed to estimate the depth-averaged current speed in the Florida Straits, based on an effective sound speed approximation. In the mid-deep ocean environment, the noise data were processed using the same noise interferometry method, but the obtained NCFs were not as good as those in the coastal ocean environment. Several highly possible reasons of the difference in the noise interferometry performance were investigated and discussed. The first one is the noise source composition, which is different in the spectrograms of noise records in two environments. The second is strong ocean current variability that can result in coherence loss and undermine the utility of coherent stacking. The third one is the downward refracting sound speed profile, which impedes strong coupling between near surface noise sources and the near-bottom instruments. The active method of inverse filter processing was tested in a long-range deep-ocean environment. The high-power sound source, which was located near the sound channel axis, transmitted a pre-designed signal that was composed of a precursor signal and a communication signal. After traveling 1428.5 km distance in the north Pacific Ocean, the transmitted signal was detected by the receiver and was processed using the inverse filter. The probe signal, which was composed of M sequences and was known at the receiver, was utilized for the GF extraction in the inverse filter; the communication signal was then interpreted with the extracted GF. With a glitch in the length of communication signal, the inverse filter processing method was shown to be effective for long-range low-frequency deep ocean acoustic communication. (Abstract shortened by ProQuest.).

Source-Type Identification Analysis Using Regional Seismic Moment Tensors

NASA Astrophysics Data System (ADS)

Chiang, A.; Dreger, D. S.; Ford, S. R.; Walter, W. R.

2012-12-01

Waveform inversion to determine the seismic moment tensor is a standard approach in determining the source mechanism of natural and manmade seismicity, and may be used to identify, or discriminate different types of seismic sources. The successful applications of the regional moment tensor method at the Nevada Test Site (NTS) and the 2006 and 2009 North Korean nuclear tests (Ford et al., 2009a, 2009b, 2010) show that the method is robust and capable for source-type discrimination at regional distances. The well-separated populations of explosions, earthquakes and collapses on a Hudson et al., (1989) source-type diagram enables source-type discrimination; however the question remains whether or not the separation of events is universal in other regions, where we have limited station coverage and knowledge of Earth structure. Ford et al., (2012) have shown that combining regional waveform data and P-wave first motions removes the CLVD-isotropic tradeoff and uniquely discriminating the 2009 North Korean test as an explosion. Therefore, including additional constraints from regional and teleseismic P-wave first motions enables source-type discrimination at regions with limited station coverage. We present moment tensor analysis of earthquakes and explosions (M6) from Lop Nor and Semipalatinsk test sites for station paths crossing Kazakhstan and Western China. We also present analyses of smaller events from industrial sites. In these sparse coverage situations we combine regional long-period waveforms, and high-frequency P-wave polarity from the same stations, as well as from teleseismic arrays to constrain the source type. Discrimination capability with respect to velocity model and station coverage is examined, and additionally we investigate the velocity model dependence of vanishing free-surface traction effects on seismic moment tensor inversion of shallow sources and recovery of explosive scalar moment. Our synthetic data tests indicate that biases in scalar seismic moment and discrimination for shallow sources are small and can be understood in a systematic manner. We are presently investigating the frequency dependence of vanishing traction of a very shallow (10m depth) M2+ chemical explosion recorded at several kilometer distances, and preliminary results indicate at the typical frequency passband we employ the bias does not affect our ability to retrieve the correct source mechanism but may affect the retrieval of the correct scalar seismic moment. Finally, we assess discrimination capability in a composite P-value statistical framework.

Finite-frequency structural sensitivities of short-period compressional body waves

NASA Astrophysics Data System (ADS)

Fuji, Nobuaki; Chevrot, Sébastien; Zhao, Li; Geller, Robert J.; Kawai, Kenji

2012-07-01

We present an extension of the method recently introduced by Zhao & Chevrot for calculating Fréchet kernels from a precomputed database of strain Green's tensors by normal mode summation. The extension involves two aspects: (1) we compute the strain Green's tensors using the Direct Solution Method, which allows us to go up to frequencies as high as 1 Hz; and (2) we develop a spatial interpolation scheme so that the Green's tensors can be computed with a relatively coarse grid, thus improving the efficiency in the computation of the sensitivity kernels. The only requirement is that the Green's tensors be computed with a fine enough spatial sampling rate to avoid spatial aliasing. The Green's tensors can then be interpolated to any location inside the Earth, avoiding the need to store and retrieve strain Green's tensors for a fine sampling grid. The interpolation scheme not only significantly reduces the CPU time required to calculate the Green's tensor database and the disk space to store it, but also enhances the efficiency in computing the kernels by reducing the number of I/O operations needed to retrieve the Green's tensors. Our new implementation allows us to calculate sensitivity kernels for high-frequency teleseismic body waves with very modest computational resources such as a laptop. We illustrate the potential of our approach for seismic tomography by computing traveltime and amplitude sensitivity kernels for high frequency P, PKP and Pdiff phases. A comparison of our PKP kernels with those computed by asymptotic ray theory clearly shows the limits of the latter. With ray theory, it is not possible to model waves diffracted by internal discontinuities such as the core-mantle boundary, and it is also difficult to compute amplitudes for paths close to the B-caustic of the PKP phase. We also compute waveform partial derivatives for different parts of the seismic wavefield, a key ingredient for high resolution imaging by waveform inversion. Our computations of partial derivatives in the time window where PcP precursors are commonly observed show that the distribution of sensitivity is complex and counter-intuitive, with a large contribution from the mid-mantle region. This clearly emphasizes the need to use accurate and complete partial derivatives in waveform inversion.

Rupture history of the 1997 Cariaco, Venezuela, earthquake from teleseismic P waves

USGS Publications Warehouse

Mendoza, C.

2000-01-01

A two-step finite-fault waveform inversion scheme is applied to the broadband teleseismic P waves recorded for the strike-slip, Cariaco, Venezuela, earthquake of 9 July 1997 to recover the distribution of mainshock slip. The earthquake is first analyzed using a long narrow fault with a maximum rise time of 20 sec. This line-source analysis indicates that slip propagated to the west with a constant rupture velocity and a relatively short rise time. The results are then used to constrain a second inversion of the P waveforms using a 60-km by 20-km two-dimensional fault. The rupture shows a zone of large slip (1.3-m peak) near the hypocenter and a second, broader source extending updip and to the west at depths shallower than 5 km. The second source has a peak slip of 2.1 meters and accounts for most of the moment of 1.1 × 1026 dyne-cm (6.6 Mww) estimated from the P waves. The inferred rupture pattern is consistent with macroseismic effects observed in the epicentral area.

Noise suppression in surface microseismic data

USGS Publications Warehouse

Forghani-Arani, Farnoush; Batzle, Mike; Behura, Jyoti; Willis, Mark; Haines, Seth S.; Davidson, Michael

2012-01-01

We introduce a passive noise suppression technique, based on the τ − p transform. In the τ − p domain, one can separate microseismic events from surface noise based on distinct characteristics that are not visible in the time-offset domain. By applying the inverse τ − p transform to the separated microseismic event, we suppress the surface noise in the data. Our technique significantly improves the signal-to-noise ratios of the microseismic events and is superior to existing techniques for passive noise suppression in the sense that it preserves the waveform. We introduce a passive noise suppression technique, based on the τ − p transform. In the τ − p domain, one can separate microseismic events from surface noise based on distinct characteristics that are not visible in the time-offset domain. By applying the inverse τ − p transform to the separated microseismic event, we suppress the surface noise in the data. Our technique significantly improves the signal-to-noise ratios of the microseismic events and is superior to existing techniques for passive noise suppression in the sense that it preserves the waveform.

The tsunami source area of the 2003 Tokachi-oki earthquake estimated from tsunami travel times and its relationship to the 1952 Tokachi-oki earthquake

USGS Publications Warehouse

Hirata, K.; Tanioka, Y.; Satake, K.; Yamaki, S.; Geist, E.L.

2004-01-01

We estimate the tsunami source area of the 2003 Tokachi-oki earthquake (Mw 8.0) from observed tsunami travel times at 17 Japanese tide gauge stations. The estimated tsunami source area (???1.4 ?? 104 km2) coincides with the western-half of the ocean-bottom deformation area (???2.52 ?? 104 km2) of the 1952 Tokachi-oki earthquake (Mw 8.1), previously inferred from tsunami waveform inversion. This suggests that the 2003 event ruptured only the western-half of the 1952 rupture extent. Geographical distribution of the maximum tsunami heights in 2003 differs significantly from that of the 1952 tsunami, supporting this hypothesis. Analysis of first-peak tsunami travel times indicates that a major uplift of the ocean-bottom occurred approximately 30 km to the NNW of the mainshock epicenter, just above a major asperity inferred from seismic waveform inversion. Copyright ?? The Society of Geomagnetism and Earth, Planetary and Space Sciences (SGEPSS); The Seismological Society of Japan; The Volcanological Society of Japan; The Geodetic Society of Japan; The Japanese Society for Planetary Sciences.

Point-source inversion techniques

NASA Astrophysics Data System (ADS)

Langston, Charles A.; Barker, Jeffrey S.; Pavlin, Gregory B.

1982-11-01

A variety of approaches for obtaining source parameters from waveform data using moment-tensor or dislocation point source models have been investigated and applied to long-period body and surface waves from several earthquakes. Generalized inversion techniques have been applied to data for long-period teleseismic body waves to obtain the orientation, time function and depth of the 1978 Thessaloniki, Greece, event, of the 1971 San Fernando event, and of several events associated with the 1963 induced seismicity sequence at Kariba, Africa. The generalized inversion technique and a systematic grid testing technique have also been used to place meaningful constraints on mechanisms determined from very sparse data sets; a single station with high-quality three-component waveform data is often sufficient to discriminate faulting type (e.g., strike-slip, etc.). Sparse data sets for several recent California earthquakes, for a small regional event associated with the Koyna, India, reservoir, and for several events at the Kariba reservoir have been investigated in this way. Although linearized inversion techniques using the moment-tensor model are often robust, even for sparse data sets, there are instances where the simplifying assumption of a single point source is inadequate to model the data successfully. Numerical experiments utilizing synthetic data and actual data for the 1971 San Fernando earthquake graphically demonstrate that severe problems may be encountered if source finiteness effects are ignored. These techniques are generally applicable to on-line processing of high-quality digital data, but source complexity and inadequacy of the assumed Green's functions are major problems which are yet to be fully addressed.

Torsional Ultrasound Sensor Optimization for Soft Tissue Characterization

PubMed Central

Melchor, Juan; Muñoz, Rafael; Rus, Guillermo

2017-01-01

Torsion mechanical waves have the capability to characterize shear stiffness moduli of soft tissue. Under this hypothesis, a computational methodology is proposed to design and optimize a piezoelectrics-based transmitter and receiver to generate and measure the response of torsional ultrasonic waves. The procedure employed is divided into two steps: (i) a finite element method (FEM) is developed to obtain a transmitted and received waveform as well as a resonance frequency of a previous geometry validated with a semi-analytical simplified model and (ii) a probabilistic optimality criteria of the design based on inverse problem from the estimation of robust probability of detection (RPOD) to maximize the detection of the pathology defined in terms of changes of shear stiffness. This study collects different options of design in two separated models, in transmission and contact, respectively. The main contribution of this work describes a framework to establish such as forward, inverse and optimization procedures to choose a set of appropriate parameters of a transducer. This methodological framework may be generalizable for other different applications. PMID:28617353

Spectral-element global waveform tomography: A second-generation upper-mantle model

NASA Astrophysics Data System (ADS)

French, S. W.; Lekic, V.; Romanowicz, B. A.

2012-12-01

The SEMum model of Lekic and Romanowicz (2011a) was the first global upper-mantle VS model obtained using whole-waveform inversion with spectral element (SEM: Komatitsch and Vilotte, 1998) forward modeling of time domain three component waveforms. SEMum exhibits stronger amplitudes of heterogeneity in the upper 200km of the mantle compared to previous global models - particularly with respect to low-velocity anomalies. To make SEM-based waveform inversion tractable at global scales, SEMum was developed using: (1) a version of SEM coupled to 1D mode computation in the earth's core (C-SEM, Capdeville et al., 2003); (2) asymptotic normal-mode sensitivity kernels, incorporating multiple forward scattering and finite-frequency effects in the great-circle plane (NACT: Li and Romanowicz, 1995); and (3) a smooth anisotropic crustal layer of uniform 60km thickness, designed to match global surface-wave dispersion while reducing the cost of time integration in the SEM. The use of asymptotic kernels reduced the number of SEM computations considerably (≥ 3x) relative to purely numerical approaches (e.g. Tarantola, 1984), while remaining sufficiently accurate at the periods of interest (down to 60s). However, while the choice of a 60km crustal-layer thickness is justifiable in the continents, it can complicate interpretation of shallow oceanic upper-mantle structure. We here present an update to the SEMum model, designed primarily to address these concerns. The resulting model, SEMum2, was derived using a crustal layer that again fits global surface-wave dispersion, but with a more geologically consistent laterally varying thickness: approximately honoring Crust2.0 (Bassin, et al., 2000) Moho depth in the continents, while saturating at 30km in the oceans. We demonstrate that this approach does not bias our upper mantle model, which is constrained not only by fundamental mode surface waves, but also by overtone waveforms. We have also improved our data-selection and assimilation scheme, more readily allowing for additional and higher-quality data to be incorporated into our inversion as the model improves. Further, we have been able to refine the parameterization of the isotropic component of our model, previously limited by our ability to solve the large dense linear system that governs model updates (Tarantola and Valette, 1982). The construction of SEMum2 involved 3 additional inversion iterations away from SEMum. Overall, the combined effect of these improvements confirms and validates the general structure of the original SEMum. Model amplitudes remain an impressive feature in SEMum2, wherein peak-to-peak variation in VS can exceed 15% in close lateral juxtaposition. Further, many intriguing structures present in SEMum are now imaged with improved resolution in the updated model. In particular, the geographic extents of the anomalous oceanic cluster identified by Lekic and Romanowicz (2011b) are consistent with our findings and now allow us to further identify alternating bands of lower and higher velocities in the 200-300km depth range beneath the Pacific basin, with a characteristic spacing of ˜2000km normal to absolute plate motion. Possible dynamic interpretation of these and other features in the ocean basins is explored in a companion presentation (Romanowicz et al., this meeting).

Modeling of an electrohydraulic lithotripter with the KZK equation.

PubMed

Averkiou, M A; Cleveland, R O

1999-07-01

The acoustic pressure field of an electrohydraulic extracorporeal shock wave lithotripter is modeled with a nonlinear parabolic wave equation (the KZK equation). The model accounts for diffraction, nonlinearity, and thermoviscous absorption. A numerical algorithm for solving the KZK equation in the time domain is used to model sound propagation from the mouth of the ellipsoidal reflector of the lithotripter. Propagation within the reflector is modeled with geometrical acoustics. It is shown that nonlinear distortion within the ellipsoidal reflector can play an important role for certain parameters. Calculated waveforms are compared with waveforms measured in a clinical lithotripter and good agreement is found. It is shown that the spatial location of the maximum negative pressure occurs pre-focally which suggests that the strongest cavitation activity will also be in front of the focus. Propagation of shock waves from a lithotripter with a pressure release reflector is considered and because of nonlinear propagation the focal waveform is not the inverse of the rigid reflector. Results from propagation through tissue are presented; waveforms are similar to those predicted in water except that the higher absorption in the tissue decreases the peak amplitude and lengthens the rise time of the shock.

A simple accurate chest-compression depth gauge using magnetic coils during cardiopulmonary resuscitation

NASA Astrophysics Data System (ADS)

Kandori, Akihiko; Sano, Yuko; Zhang, Yuhua; Tsuji, Toshio

2015-12-01

This paper describes a new method for calculating chest compression depth and a simple chest-compression gauge for validating the accuracy of the method. The chest-compression gauge has two plates incorporating two magnetic coils, a spring, and an accelerometer. The coils are located at both ends of the spring, and the accelerometer is set on the bottom plate. Waveforms obtained using the magnetic coils (hereafter, "magnetic waveforms"), which are proportional to compression-force waveforms and the acceleration waveforms were measured at the same time. The weight factor expressing the relationship between the second derivatives of the magnetic waveforms and the measured acceleration waveforms was calculated. An estimated-compression-displacement (depth) waveform was obtained by multiplying the weight factor and the magnetic waveforms. Displacements of two large springs (with similar spring constants) within a thorax and displacements of a cardiopulmonary resuscitation training manikin were measured using the gauge to validate the accuracy of the calculated waveform. A laser-displacement detection system was used to compare the real displacement waveform and the estimated waveform. Intraclass correlation coefficients (ICCs) between the real displacement using the laser system and the estimated displacement waveforms were calculated. The estimated displacement error of the compression depth was within 2 mm (<1 standard deviation). All ICCs (two springs and a manikin) were above 0.85 (0.99 in the case of one of the springs). The developed simple chest-compression gauge, based on a new calculation method, provides an accurate compression depth (estimation error < 2 mm).

Waveform fitting and geometry analysis for full-waveform lidar feature extraction

NASA Astrophysics Data System (ADS)

Tsai, Fuan; Lai, Jhe-Syuan; Cheng, Yi-Hsiu

2016-10-01

This paper presents a systematic approach that integrates spline curve fitting and geometry analysis to extract full-waveform LiDAR features for land-cover classification. The cubic smoothing spline algorithm is used to fit the waveform curve of the received LiDAR signals. After that, the local peak locations of the waveform curve are detected using a second derivative method. According to the detected local peak locations, commonly used full-waveform features such as full width at half maximum (FWHM) and amplitude can then be obtained. In addition, the number of peaks, time difference between the first and last peaks, and the average amplitude are also considered as features of LiDAR waveforms with multiple returns. Based on the waveform geometry, dynamic time-warping (DTW) is applied to measure the waveform similarity. The sum of the absolute amplitude differences that remain after time-warping can be used as a similarity feature in a classification procedure. An airborne full-waveform LiDAR data set was used to test the performance of the developed feature extraction method for land-cover classification. Experimental results indicate that the developed spline curve- fitting algorithm and geometry analysis can extract helpful full-waveform LiDAR features to produce better land-cover classification than conventional LiDAR data and feature extraction methods. In particular, the multiple-return features and the dynamic time-warping index can improve the classification results significantly.

Simplified moment tensor analysis and unified decomposition of acoustic emission source: Application to in situ hydrofracturing test

NASA Astrophysics Data System (ADS)

Ohtsu, Masayasu

1991-04-01

An application of a moment tensor analysis to acoustic emission (AE) is studied to elucidate crack types and orientations of AE sources. In the analysis, simplified treatment is desirable, because hundreds of AE records are obtained from just one experiment and thus sophisticated treatment is realistically cumbersome. Consequently, a moment tensor inversion based on P wave amplitude is employed to determine six independent tensor components. Selecting only P wave portion from the full-space Green's function of homogeneous and isotropic material, a computer code named SiGMA (simplified Green's functions for the moment tensor analysis) is developed for the AE inversion analysis. To classify crack type and to determine crack orientation from moment tensor components, a unified decomposition of eigenvalues into a double-couple (DC) part, a compensated linear vector dipole (CLVD) part, and an isotropic part is proposed. The aim of the decomposition is to determine the proportion of shear contribution (DC) and tensile contribution (CLVD + isotropic) on AE sources and to classify cracks into a crack type of the dominant motion. Crack orientations determined from eigenvectors are presented as crack-opening vectors for tensile cracks and fault motion vectors for shear cracks, instead of stereonets. The SiGMA inversion and the unified decomposition are applied to synthetic data and AE waveforms detected during an in situ hydrofracturing test. To check the accuracy of the procedure, numerical experiments are performed on the synthetic waveforms, including cases with 10% random noise added. Results show reasonable agreement with assumed crack configurations. Although the maximum error is approximately 10% with respect to the ratios, the differences on crack orientations are less than 7°. AE waveforms detected by eight accelerometers deployed during the hydrofracturing test are analyzed. Crack types and orientations determined are in reasonable agreement with a predicted failure plane from borehole TV observation. The results suggest that tensile cracks are generated first at weak seams and then shear cracks follow on the opened joints.

Super-resolution Time-Lapse Seismic Waveform Inversion

NASA Astrophysics Data System (ADS)

Ovcharenko, O.; Kazei, V.; Peter, D. B.; Alkhalifah, T.

2017-12-01

Time-lapse seismic waveform inversion is a technique, which allows tracking changes in the reservoirs over time. Such monitoring is relatively computationally extensive and therefore it is barely feasible to perform it on-the-fly. Most of the expenses are related to numerous FWI iterations at high temporal frequencies, which is inevitable since the low-frequency components can not resolve fine scale features of a velocity model. Inverted velocity changes are also blurred when there is noise in the data, so the problem of low-resolution images is widely known. One of the problems intensively tackled by computer vision research community is the recovering of high-resolution images having their low-resolution versions. Usage of artificial neural networks to reach super-resolution from a single downsampled image is one of the leading solutions for this problem. Each pixel of the upscaled image is affected by all the pixels of its low-resolution version, which enables the workflow to recover features that are likely to occur in the corresponding environment. In the present work, we adopt machine learning image enhancement technique to improve the resolution of time-lapse full-waveform inversion. We first invert the baseline model with conventional FWI. Then we run a few iterations of FWI on a set of the monitoring data to find desired model changes. These changes are blurred and we enhance their resolution by using a deep neural network. The network is trained to map low-resolution model updates predicted by FWI into the real perturbations of the baseline model. For supervised training of the network we generate a set of random perturbations in the baseline model and perform FWI on the noisy data from the perturbed models. We test the approach on a realistic perturbation of Marmousi II model and demonstrate that it outperforms conventional convolution-based deblurring techniques.

A fast and robust method for moment tensor and depth determination of shallow seismic events in CTBT related studies.

NASA Astrophysics Data System (ADS)

Baker, Ben; Stachnik, Joshua; Rozhkov, Mikhail

2017-04-01

International Data Center is required to conduct expert technical analysis and special studies to improve event parameters and assist State Parties in identifying the source of specific event according to the protocol to the Protocol to the Comprehensive Nuclear Test Ban Treaty. Determination of seismic event source mechanism and its depth is closely related to these tasks. It is typically done through a strategic linearized inversion of the waveforms for a complete or subset of source parameters, or similarly defined grid search through precomputed Greens Functions created for particular source models. In this presentation we demonstrate preliminary results obtained with the latter approach from an improved software design. In this development we tried to be compliant with different modes of CTBT monitoring regime and cover wide range of source-receiver distances (regional to teleseismic), resolve shallow source depths, provide full moment tensor solution based on body and surface waves recordings, be fast to satisfy both on-demand studies and automatic processing and properly incorporate observed waveforms and any uncertainties a priori as well as accurately estimate posteriori uncertainties. Posterior distributions of moment tensor parameters show narrow peaks where a significant number of reliable surface wave observations are available. For earthquake examples, fault orientation (strike, dip, and rake) posterior distributions also provide results consistent with published catalogues. Inclusion of observations on horizontal components will provide further constraints. In addition, the calculation of teleseismic P wave Green's Functions are improved through prior analysis to determine an appropriate attenuation parameter for each source-receiver path. Implemented HDF5 based Green's Functions pre-packaging allows much greater flexibility in utilizing different software packages and methods for computation. Further additions will have the rapid use of Instaseis/AXISEM full waveform synthetics added to a pre-computed GF archive. Along with traditional post processing analysis of waveform misfits through several objective functions and variance reduction, we follow a probabilistic approach to assess the robustness of moment tensor solution. In a course of this project full moment tensor and depth estimates are determined for DPRK events and shallow earthquakes using a new implementation of teleseismic P waves waveform fitting. A full grid search over the entire moment tensor space is used to appropriately sample all possible solutions. A recent method by Tape & Tape (2012) to discretize the complete moment tensor space from a geometric perspective is used. Probabilistic uncertainty estimates on the moment tensor parameters provide robustness to solution.

Landquake dynamics inferred from seismic source inversion: Greenland and Sichuan events of 2017

NASA Astrophysics Data System (ADS)

Chao, W. A.

2017-12-01

In June 2017 two catastrophic landquake events occurred in Greenland and Sichuan. The Greenland event leads to tsunami hazard in the small town of Nuugaarsiaq. A landquake in Sichuan hit the town, which resulted in over 100 death. Both two events generated the strong seismic signals recorded by the real-time global seismic network. I adopt an inversion algorithm to derive the landquake force time history (LFH) using the long-period waveforms, and the landslide volume ( 76 million m3) can be rapidly estimated, facilitating the tsunami-wave modeling for early warning purpose. Based on an integrated approach involving tsunami forward simulation and seismic waveform inversion, this study has significant implications to issuing actionable warnings before hazardous tsunami waves strike populated areas. Two single-forces (SFs) mechanism (two block model) yields the best explanation for Sichuan event, which demonstrates that secondary event (seismic inferred volume: 8.2 million m3) may be mobilized by collapse-mass hitting from initial rock avalanches ( 5.8 million m3), likely causing a catastrophic disaster. The later source with a force magnitude of 0.9967×1011 N occurred 70 seconds after first mass-movement occurrence. In contrast, first event has the smaller force magnitude of 0.8116×1011 N. In conclusion, seismically inferred physical parameters will substantially contribute to improving our understanding of landquake source mechanisms and mitigating similar hazards in other parts of the world.

Wavefield simulations of earthquakes in Alaska for tomographic inversion

NASA Astrophysics Data System (ADS)

Silwal, V.; Tape, C.; Casarotti, E.

2017-12-01

We assemble a catalog of moment tensors and a three-dimensional seismic velocity model for mainland Alaska, in preparation for an iterative tomographic inversion using spectral-element and adjoint methods. The catalog contains approximately 200 earthquakes with Mw ≥ 4.0 that generate good long-period (≥6 s) signals for stations at distances up to approximately 500 km. To maximize the fraction of usable stations per earthquake, we divide our model into three subregions for simulations: south-central Alaska, central Alaska, and eastern Alaska. The primary geometrical interfaces in the model are the Moho surface, the basement surface of major sedimentary basins, and the topographic surface. The crustal and upper mantle tomographic model is from Eberhart-Phillips et al. (2006), but modified by removing the uppermost slow layer, then embedding sedimentary basin models for Cook Inlet basin, Susitna basin, and Nenana basin. We compute 3D synthetic seismograms using the spectral-element method. We demonstrate the accuracy of the initial three-dimensional reference model in each subregion by comparing 3D synthetics with observed data for several earthquakes originating in the crust and underlying subducting slab. Full waveform similarity between data and synthetics over the period range 6 s to 30 s provides a basis for an iterative inversion. The target resolution of the crustal structure is 4 km vertically and 20 km laterally. We use surface wave and body wave measurements from local earthquakes to obtain moment tensors that will be used within our tomographic inversion. Local slab events down to 180 km depth, in additional to pervasive crustal seismicity, should enhance resolution.

The OSCAR experiment: using full-waveform inversion in the analysis of young oceanic crust

NASA Astrophysics Data System (ADS)

Silverton, Akela; Morgan, Joanna; Wilson, Dean; Hobbs, Richard

2017-04-01

The OSCAR experiment aims to derive an integrated model to better explain the effects of heat loss and alteration by hydrothermal fluids, associated with the cooling of young oceanic crust at an axial ridge. High-resolution seismic imaging of the sediments and basaltic basement can be used to map fluid flow pathways between the oceanic crust and the surrounding ocean. To obtain these high-resolution images, we undertake full-waveform inversion (FWI), an advanced seismic imaging technique capable of resolving velocity heterogeneities at a wide range of length scales, from background trends to fine-scale geological/crustal detail, in a fully data-driven automated manner. This technology is widely used within the petroleum sector due to its potential to obtain high-resolution P-wave velocity models that lead to improvements in migrated seismic images of the subsurface. Here, we use the P-wave velocity model obtained from travel-time tomography as the starting model in the application of acoustic, time-domain FWI to a multichannel streamer field dataset acquired in the east Pacific along a profile between the Costa Rica spreading centre and the Ocean Drilling Program (ODP) borehole 504B, where the crust is approximately six million years old. FWI iteratively improves the velocity model by minimizing the misfit between the predicted data and the field data. It seeks to find a high-fidelity velocity model that is capable of matching individual seismic waveforms of the original raw field dataset, with an initial focus on matching the low-frequency components of the early arriving energy. Quality assurance methods adopted during the inversion ensure convergence in the direction of the global minimum. We demonstrate that FWI is able to recover fine-scale, high-resolution velocity heterogeneities within the young oceanic crust along the profile. The highly resolved FWI velocity model is useful in the identification of the layer 2A/2B interface and low-velocity layers that may represent zones of enhanced fluid movement. With FWI we are able to better explain the non-linear changes in velocity as the crust evolves with distance from the spreading centre and image the effects of any alteration by hydrothermal fluids. This model provides valuable insight and new constraints on the thermal processes involved, at spreading centres, setting a new benchmark for integrated multi-physics experiments at similar ocean ridge systems. This research is part of a major, interdisciplinary NERC-funded collaboration entitled: Oceanographic and Seismic Characterisation of heat dissipation and alteration by hydrothermal fluids at an Axial Ridge (OSCAR).

Progress report on lithium-related geologic investigations in Bolivia

USGS Publications Warehouse

Davis, J.R.; Howard, K.A.; Rettig, S.L.; Smith, R.L.; Ericksen, G.E.; Risacher, Francois; Alarcon, Hugo; Morales, Ricardo

1982-01-01

The September 1, 1981, Samoa Islands Region earthquake occurred at the extreme northern end of the Tonga arc in a region where the Pacific plate may be disjointed along a hinge fault. In the last 50 years, magnitude 7 or greater earthquakes have occurred in this region on the average of once every six years, but four 7+ events have now occurred within the last six years. The mainshock was preceded about two hours earlier by a foreshock that was used as a calibration event for the Joint Epicenter Determination relocation of the mainshock and nearby seismicity occurring within a period seven months prior to and one week after the mainshock. The foreshock, better-located events of the prior seismicity, and most aftershocks are concentrated in a group near the mainshock epicenter, but several more distant aftershocks suggest that the aftershock zone may have been as large as 125 km in length and trended about S35?E. Identification of depth phases from a full suite of broadband records gives source depths of 25-3km for the mainshock and 29.5?3 km for the foreshock using a JB earth model. Source parameters were determined for the mainshock utilizing WWSSN analog and GDSN digital data. The preferred fault plane solution based on P-wave first motion data is a south by southwesterly steeply dipping normal fault, remarkably similar to the mechanism reported by Johnson and Molnar (1972) for the nearby earthquake of April 20, 196B. A waveform inversion technique described by Sipkin (1982), when applied to long-period P waveforms, gives an 'average' point source solution for a purely deviatoric moment rate tensor at a preferred source depth of 22 km. Very similar results were obtained from long-period GDSN body-wave and mantle-wave data using a centroid-moment tensor inversion technique described in Dziewonski, and others (1981). Both techniques provide solutions very close to a double couple source with a south by southwesterly shallow-dipping normal fault mechanism. To obtain the scalar mantle wave moment, GDSN vertical and transverse records 20,000 see in length were processed as described by Buland and Taggart (1981). Averaging all the data from Rayleigh and Love waves yields an estimate of 3.8 x 10^27 dyne-cm (as compared to about 1.9 x 10^27 from body-wave moment tensor inversions) or a moment magnitude (Mr) of 7.6. For the portion of the waveform analysed (50-5B sec), the body-wave inversion performed by Sipkin gives a source time function of duration approximately 28 sec with two peaks in activity. Simultaneous analysis of short-period records, and broadband ground displacements and velocities, a method described by Harvey and Choy (1982) and Choy and Boatwright (1981) revealed a complex rupture consisting of two subevents, of about the same moment, separated in time by about 25 sec, and with durations of about 25 sec each. The two peaks in activity resolved by the body-wave moment tensor inversion correspond to the first of these subevents.

3D magnetotelluric inversion system with static shift correction and theoretical assessment in oil and gas exploration

NASA Astrophysics Data System (ADS)

Dong, H.; Kun, Z.; Zhang, L.

2015-12-01

This magnetotelluric (MT) system contains static shift correction and 3D inversion. The correction method is based on the data study on 3D forward modeling and field test. The static shift can be detected by the quantitative analysis of apparent parameters (apparent resistivity and impedance phase) of MT in high frequency range, and completed correction with inversion. The method is an automatic processing technology of computer with zero-cost, and avoids the additional field work and indoor processing with good results shown in Figure 1a-e. Figure 1a shows a normal model (I) without any local heterogeneity. Figure 1b shows a static-shifted model (II) with two local heterogeneous bodies (10 and 1000 ohm.m). Figure 1c is the inversion result (A) for the synthetic data generated from model I. Figure 1d is the inversion result (B) for the static-shifted data generated from model II. Figure 1e is the inversion result (C) for the static-shifted data from model II, but with static shift correction. The results show that the correction method is useful. The 3D inversion algorithm is improved base on the NLCG method of Newman & Alumbaugh (2000) and Rodi & Mackie (2001). For the algorithm, we added the frequency based parallel structure, improved the computational efficiency, reduced the memory of computer, added the topographic and marine factors, and added the constraints of geology and geophysics. So the 3D inversion could even work in PAD with high efficiency and accuracy. The application example of theoretical assessment in oil and gas exploration is shown in Figure 1f-i. The synthetic geophysical model consists of five layers (from top to downwards): shale, limestone, gas, oil, groundwater and limestone overlying a basement rock. Figure 1f-g show the 3D model and central profile. Figure 1h shows the centrel section of 3D inversion, the resultsd show a high degree of reduction in difference on the synthetic model. Figure 1i shows the seismic waveform reflects the interfaces of every layer overall, but the relative positions of the interface of the two-way travel time vary, and the interface between limestone and oil at the sides of the section is not reflected. So 3-D MT can make up for the deficiency of the seismic results such as the fake sync-phase axis and multiple waves.

Inversion of Attributes and Full Waveforms of Ground Penetrating Radar Data Using PEST

NASA Astrophysics Data System (ADS)

Jazayeri, S.; Kruse, S.; Esmaeili, S.

2015-12-01

We seek to establish a method, based on freely available software, for inverting GPR signals for the underlying physical properties (electrical permittivity, magnetic permeability, target geometries). Such a procedure should be useful for classroom instruction and for analyzing surface GPR surveys over simple targets. We explore the applicability of the PEST parameter estimation software package for GPR inversion (www.pesthomepage.org). PEST is designed to invert data sets with large numbers of parameters, and offers a variety of inversion methods. Although primarily used in hydrogeology, the code has been applied to a wide variety of physical problems. The PEST code requires forward model input; the forward model of the GPR signal is done with the GPRMax package (www.gprmax.com). The problem of extracting the physical characteristics of a subsurface anomaly from the GPR data is highly nonlinear. For synthetic models of simple targets in homogeneous backgrounds, we find PEST's nonlinear Gauss-Marquardt-Levenberg algorithm is preferred. This method requires an initial model, for which the weighted differences between model-generated data and those of the "true" synthetic model (the objective function) are calculated. In order to do this, the Jacobian matrix and the derivatives of the observation data in respect to the model parameters are computed using a finite differences method. Next, the iterative process of building new models by updating the initial values starts in order to minimize the objective function. Another measure of the goodness of the final acceptable model is the correlation coefficient which is calculated based on the method of Cooley and Naff. An accepted final model satisfies both of these conditions. Models to date show that physical properties of simple isolated targets against homogeneous backgrounds can be obtained from multiple traces from common-offset surface surveys. Ongoing work examines the inversion capabilities with more complex target geometries and heterogeneous soils.

Finite Moment Tensors of Southern California Earthquakes

NASA Astrophysics Data System (ADS)

Jordan, T. H.; Chen, P.; Zhao, L.

2003-12-01

We have developed procedures for inverting broadband waveforms for the finite moment tensors (FMTs) of regional earthquakes. The FMT is defined in terms of second-order polynomial moments of the source space-time function and provides the lowest order representation of a finite fault rupture; it removes the fault-plane ambiguity of the centroid moment tensor (CMT) and yields several additional parameters of seismological interest: the characteristic length L{c}, width W{c}, and duration T{c} of the faulting, as well as the directivity vector {v}{d} of the fault slip. To formulate the inverse problem, we follow and extend the methods of McGuire et al. [2001, 2002], who have successfully recovered the second-order moments of large earthquakes using low-frequency teleseismic data. We express the Fourier spectra of a synthetic point-source waveform in its exponential (Rytov) form and represent the observed waveform relative to the synthetic in terms two frequency-dependent differential times, a phase delay δ τ {p}(ω ) and an amplitude-reduction time δ τ {q}(ω ), which we measure using Gee and Jordan's [1992] isolation-filter technique. We numerically calculate the FMT partial derivatives in terms of second-order spatiotemporal gradients, which allows us to use 3D finite-difference seismograms as our isolation filters. We have applied our methodology to a set of small to medium-sized earthquakes in Southern California. The errors in anelastic structure introduced perturbations larger than the signal level caused by finite source effect. We have therefore employed a joint inversion technique that recovers the CMT parameters of the aftershocks, as well as the CMT and FMT parameters of the mainshock, under the assumption that the source finiteness of the aftershocks can be ignored. The joint system of equations relating the δ τ {p} and δ τ {q} data to the source parameters of the mainshock-aftershock cluster is denuisanced for path anomalies in both observables; this projection operation effectively corrects the mainshock data for path-related amplitude anomalies in a way similar to, but more flexible than, empirical Green function (EGF) techniques.

Effect of phase errors in stepped-frequency radar systems

NASA Astrophysics Data System (ADS)

Vanbrundt, H. E.

1988-04-01

Stepped-frequency waveforms are being considered for inverse synthetic aperture radar (ISAR) imaging from ship and airborne platforms and for detailed radar cross section (RCS) measurements of ships and aircraft. These waveforms make it possible to achieve resolutions of 1.0 foot by using existing radar designs and processing technology. One problem not yet fully resolved in using stepped-frequency waveform for ISAR imaging is the deterioration in signal level caused by random frequency error. Random frequency error of the stepped-frequency source results in reduced peak responses and increased null responses. The resulting reduced signal-to-noise ratio is range dependent. Two of the major concerns addressed in this report are radar range limitations for ISAR and the error in calibration for RCS measurements caused by differences in range between a passive reflector used for an RCS reference and the target to be measured. In addressing these concerns, NOSC developed an analysis to assess the tolerable frequency error in terms of resulting power loss in signal power and signal-to-phase noise.

The 2017 North Korea M6 seismic sequence: moment tensor, source time function, and aftershocks

NASA Astrophysics Data System (ADS)

Ni, S.; Zhan, Z.; Chu, R.; He, X.

2017-12-01

On September 3rd, 2017, an M6 seismic event occurred in North Korea, with location near previous nuclear test sites. The event features strong P waves and short period Rayleigh waves are observed in contrast to weak S waves, suggesting mostly explosion mechanism. We performed joint inversion for moment tensor and depth with both local and teleseismic waveforms, and find that the event is shallow with mostly isotropic yet substantial non-isotropic components. Deconvolution of seismic waveforms of this event with respect to previous nuclear test events shows clues of complexity in source time function. The event is followed by smaller earthquakes, as early as 8.5 minutes and lasted at least to October. The later events occurred in a compact region, and show clear S waves, suggesting double couple focal mechanism. Via analyzing Rayleigh wave spectrum, these smaller events are found to be shallow. Relative locations, difference in waveforms of the events are used to infer their possible links and generation mechanism.

Pre-stack full-waveform inversion of multichannel seismic data to retrieve thermohaline ocean structure. Application to the Gulf of Cadiz (SW Iberia).

NASA Astrophysics Data System (ADS)

Dagnino, Daniel; Jiménez Tejero, Clara-Estela; Meléndez, Adrià; Gras, Clàudia; Sallarès, Valentí; Ranero, César R.

2016-04-01

This work demonstrates the feasibility to retrieve high-resolution models of oceanic physical parameters by means of 2D adjoint-state full-waveform inversion (FWI). The proposed method is applied to pre-stack multi-channel seismic (MCS) data acquired in the Gulf of Cadiz (SW Iberia) in the framework of the EU GO (Geophysical Oceanography) project in 2006. We first design and apply a specific data processing flow that allows reducing data noise without modifying trace amplitudes. This step is shown to be essential to obtain accurate results due to the low signal-to-noise ratio (SNR) of water layer reflections, which are typically three-to-four orders of magnitude weaker than those in solid earth. Second, we propose new techniques to improve the inversion results by reducing the artefacts appearing in the gradient and misfit as a consequence of the low SNR. We use a weight and filter operator to focus in the regions where the gradient is reliable. The source wavelet is then inverted together with the sound speed. We demonstrate the efficiency of the proposed method and inversion strategy retrieving a 2D sound speed model along a 50 km-long MCS profile collected in the Gulf of Cadiz during the GO experiment. In this region, the Mediterranean outflow entrains the Atlantic waters, creating a salinity complex thermohaline structure that can be measured by a difference in acoustic impedance. The inverted sound speed model have a resolution of 75m for the horizontal direction, which is two orders of magnitude better than the models obtained using conventional, probe-based oceanographic techniques. In a second step, temperature and salinity are derived from the sound speed by minimizing the difference between the inverted and the theoretical sound speed estimated using the thermodynamic equation of seawater (TEOS-10 software). To apply the TEOS-10 we first calculate a linear-fitting between temperature and salinity using regional data from the National Oceanic and Atmospheric Administration (NOAA) compilation. Pressure is calculated from latitude and depth. In the final step, salinity is calculated using the Temperature-Salinity relation and the previously estimated temperature. The comparison of the inverted temperature, salinity model with measures from XBT and CTD probes deployed simultaneously to the MCS data acquisition shows that the accuracy of the inverted models is ˜0.15°C for temperature and ˜0.1psu for salinity.

Full-waveform seismic tomography of the Vrancea, Romania, subduction region

NASA Astrophysics Data System (ADS)

Baron, Julie; Morelli, Andrea

2017-12-01

The Vrancea region is one of the few locations of deep seismicity in Europe. Seismic tomography has been able to map lithospheric downwelling, but has not been able yet to clearly discriminate between competing geodynamic interpretations of the geological and geophysical evidence available. We study the seismic structure of the Vrancea subduction zone, using adjoint-based, full-waveform tomography to map the 3D vP and vS structure in detail. We use the database that was built during the CALIXTO (Carpathian Arc Lithosphere X-Tomography) temporary experiment, restricted to the broadband sensors and local intermediate-depth events. We fit waveforms with a cross-correlation misfit criterion in separate time windows around the expected P and S arrivals, and perform 17 iterations of vP and vS model updates (altogether, requiring about 16 million CPU hours) before reaching stable convergence. Among other features, our resulting model shows a nearly vertical, high-velocity body, that overlaps with the distribution of seismicity in its northeastern part. In its southwestern part, a slab appears to dip less steeply to the NW, and is suggestive of ongoing - or recently concluded - subduction geodynamic processes. Joint inversion for vP and vS allow us to address the vP/vS ratio distribution, that marks high vP/vS in the crust beneath the Focsani sedimentary basin - possibly due to high fluid pressure - and a low vP/vS edge along the lower plane of the subducting lithosphere, that in other similar environment has been attributed to dehydration of serpentine in the slab. In spite of the restricted amount of data available, and limitations on the usable frequency pass-band, full-waveform inversion reveals its potential to improve the general quality of imaging with respect to other tomographic techniques - although at a sensible cost in terms of computing resources. Our study also shows that re-analysis of legacy data sets with up-to-date techniques may bring new, useful, information.

Temperature analysis with voltage-current time differential operation of electrochemical sensors

DOE Office of Scientific and Technical Information (OSTI.GOV)

Woo, Leta Yar-Li; Glass, Robert Scott; Fitzpatrick, Joseph Jay

A method for temperature analysis of a gas stream. The method includes identifying a temperature parameter of an affected waveform signal. The method also includes calculating a change in the temperature parameter by comparing the affected waveform signal with an original waveform signal. The method also includes generating a value from the calculated change which corresponds to the temperature of the gas stream.

Time domain reflectometry waveform analysis with second order bounded mean oscillation

USDA-ARS?s Scientific Manuscript database

Tangent-line methods and adaptive waveform interpretation with Gaussian filtering (AWIGF) have been proposed for determining reflection positions of time domain reflectometry (TDR) waveforms. However, the accuracy of those methods is limited for short probe TDR sensors. Second order bounded mean osc...

Pulse transmission transceiver architecture for low power communications

DOEpatents

Dress, Jr., William B.; Smith, Stephen F.

2003-08-05

Systems and methods for pulse-transmission low-power communication modes are disclosed. A method of pulse transmission communications includes: generating a modulated pulse signal waveform; transforming said modulated pulse signal waveform into at least one higher-order derivative waveform; and transmitting said at least one higher-order derivative waveform as an emitted pulse. The systems and methods significantly reduce lower-frequency emissions from pulse transmission spread-spectrum communication modes, which reduces potentially harmful interference to existing radio frequency services and users and also simultaneously permit transmission of multiple data bits by utilizing specific pulse shapes.

Full Waveform Modeling of Transient Electromagnetic Response Based on Temporal Interpolation and Convolution Method

NASA Astrophysics Data System (ADS)

Qi, Youzheng; Huang, Ling; Wu, Xin; Zhu, Wanhua; Fang, Guangyou; Yu, Gang

2017-07-01

Quantitative modeling of the transient electromagnetic (TEM) response requires consideration of the full transmitter waveform, i.e., not only the specific current waveform in a half cycle but also the bipolar repetition. In this paper, we present a novel temporal interpolation and convolution (TIC) method to facilitate the accurate TEM modeling. We first calculate the temporal basis response on a logarithmic scale using the fast digital-filter-based methods. Then, we introduce a function named hamlogsinc in the framework of discrete signal processing theory to reconstruct the basis function and to make the convolution with the positive half of the waveform. Finally, a superposition procedure is used to take account of the effect of previous bipolar waveforms. Comparisons with the established fast Fourier transform method demonstrate that our TIC method can get the same accuracy with a shorter computing time.

Time-domain seismic modeling in viscoelastic media for full waveform inversion on heterogeneous computing platforms with OpenCL

NASA Astrophysics Data System (ADS)

Fabien-Ouellet, Gabriel; Gloaguen, Erwan; Giroux, Bernard

2017-03-01

Full Waveform Inversion (FWI) aims at recovering the elastic parameters of the Earth by matching recordings of the ground motion with the direct solution of the wave equation. Modeling the wave propagation for realistic scenarios is computationally intensive, which limits the applicability of FWI. The current hardware evolution brings increasing parallel computing power that can speed up the computations in FWI. However, to take advantage of the diversity of parallel architectures presently available, new programming approaches are required. In this work, we explore the use of OpenCL to develop a portable code that can take advantage of the many parallel processor architectures now available. We present a program called SeisCL for 2D and 3D viscoelastic FWI in the time domain. The code computes the forward and adjoint wavefields using finite-difference and outputs the gradient of the misfit function given by the adjoint state method. To demonstrate the code portability on different architectures, the performance of SeisCL is tested on three different devices: Intel CPUs, NVidia GPUs and Intel Xeon PHI. Results show that the use of GPUs with OpenCL can speed up the computations by nearly two orders of magnitudes over a single threaded application on the CPU. Although OpenCL allows code portability, we show that some device-specific optimization is still required to get the best performance out of a specific architecture. Using OpenCL in conjunction with MPI allows the domain decomposition of large models on several devices located on different nodes of a cluster. For large enough models, the speedup of the domain decomposition varies quasi-linearly with the number of devices. Finally, we investigate two different approaches to compute the gradient by the adjoint state method and show the significant advantages of using OpenCL for FWI.

Factorization and resummation: A new paradigm to improve gravitational wave amplitudes. II. The higher multipolar modes

NASA Astrophysics Data System (ADS)

Messina, Francesco; Maldarella, Alberto; Nagar, Alessandro

2018-04-01

The factorization and resummation approach of Nagar and Shah [Phys. Rev. D 94, 104017 (2016), 10.1103/PhysRevD.94.104017], designed to improve the strong-field behavior of the post-Newtonian (PN) residual waveform amplitudes fℓm's entering the effective-one-body, circularized, gravitational waveform for spinning coalescing binaries, is improved and generalized here to all multipoles up to ℓ=6 . For a test particle orbiting a Kerr black hole, each multipolar amplitude is truncated at relative 6 PN order, both for the orbital (nonspinning) and spin factors. By taking a certain Padé approximant (typically the P24 one) of the orbital factor in conjunction with the inverse Taylor (iResum) representation of the spin factor, it is possible to push the analytical/numerical agreement of the energy flux at the level of 5% at the last-stable orbit for a quasimaximally spinning black hole with dimensionless spin parameter +0.99 . When the procedure is generalized to comparable-mass binaries, each orbital factor is kept at relative 3+3 PN order; i.e., the globally 3 PN-accurate comparable-mass terms are hybridized with higher-PN test-particle terms up to 6 PN relative order in each mode. The same Padé resummation is used for continuity. By contrast, the spin factor is only kept at the highest comparable-mass PN order currently available. We illustrate that the consistency between different truncations in the spin content of the waveform amplitudes is more marked in the resummed case than when using the standard Taylor-expanded form of Pan et al. [Phys. Rev. D 83, 064003 (2011), 10.1103/PhysRevD.83.064003]. We finally introduce a method to consistently hybridize comparable-mass and test-particle information also in the presence of spin (including the spin of the particle), discussing it explicitly for the ℓ=m =2 spin-orbit and spin-square terms. The improved, factorized and resummed, multipolar waveform amplitudes presented here are expected to set a new standard for effective one body-based gravitational waveform models.

Surface Fitting Filtering of LIDAR Point Cloud with Waveform Information

NASA Astrophysics Data System (ADS)

Xing, S.; Li, P.; Xu, Q.; Wang, D.; Li, P.

2017-09-01

Full-waveform LiDAR is an active technology of photogrammetry and remote sensing. It provides more detailed information about objects along the path of a laser pulse than discrete-return topographic LiDAR. The point cloud and waveform information with high quality can be obtained by waveform decomposition, which could make contributions to accurate filtering. The surface fitting filtering method with waveform information is proposed to present such advantage. Firstly, discrete point cloud and waveform parameters are resolved by global convergent Levenberg Marquardt decomposition. Secondly, the ground seed points are selected, of which the abnormal ones are detected by waveform parameters and robust estimation. Thirdly, the terrain surface is fitted and the height difference threshold is determined in consideration of window size and mean square error. Finally, the points are classified gradually with the rising of window size. The filtering process is finished until window size is larger than threshold. The waveform data in urban, farmland and mountain areas from "WATER (Watershed Allied Telemetry Experimental Research)" are selected for experiments. Results prove that compared with traditional method, the accuracy of point cloud filtering is further improved and the proposed method has highly practical value.

Foundations for a multiscale collaborative Earth model

NASA Astrophysics Data System (ADS)

Afanasiev, Michael; Peter, Daniel; Sager, Korbinian; Simutė, Saulė; Ermert, Laura; Krischer, Lion; Fichtner, Andreas

2016-01-01

We present a computational framework for the assimilation of local to global seismic data into a consistent model describing Earth structure on all seismically accessible scales. This Collaborative Seismic Earth Model (CSEM) is designed to meet the following requirements: (i) Flexible geometric parametrization, capable of capturing topography and bathymetry, as well as all aspects of potentially resolvable structure, including small-scale heterogeneities and deformations of internal discontinuities. (ii) Independence of any particular wave equation solver, in order to enable the combination of inversion techniques suitable for different types of seismic data. (iii) Physical parametrization that allows for full anisotropy and for variations in attenuation and density. While not all of these parameters are always resolvable, the assimilation of data that constrain any parameter subset should be possible. (iv) Ability to accommodate successive refinements through the incorporation of updates on any scale as new data or inversion techniques become available. (v) Enable collaborative Earth model construction. The structure of the initial CSEM is represented on a variable-resolution tetrahedral mesh. It is assembled from a long-wavelength 3-D global model into which several regional-scale tomographies are embedded. We illustrate the CSEM workflow of successive updating with two examples from Japan and the Western Mediterranean, where we constrain smaller scale structure using full-waveform inversion. Furthermore, we demonstrate the ability of the CSEM to act as a vehicle for the combination of different tomographic techniques with a joint full-waveform and traveltime ray tomography of Europe. This combination broadens the exploitable frequency range of the individual techniques, thereby improving resolution. We perform two iterations of a whole-Earth full-waveform inversion using a long-period reference data set from 225 globally recorded earthquakes. At this early stage of the CSEM development, the broad global updates mostly act to remove artefacts from the assembly of the initial CSEM. During the future evolution of the CSEM, the reference data set will be used to account for the influence of small-scale refinements on large-scale global structure. The CSEM as a computational framework is intended to help bridging the gap between local, regional and global tomography, and to contribute to the development of a global multiscale Earth model. While the current construction serves as a first proof of concept, future refinements and additions will require community involvement, which is welcome at this stage already.

Resolution testing and limitations of geodetic and tsunami datasets for finite fault inversions along subduction zones

NASA Astrophysics Data System (ADS)

Williamson, A.; Newman, A. V.

2017-12-01

Finite fault inversions utilizing multiple datasets have become commonplace for large earthquakes pending data availability. The mixture of geodetic datasets such as Global Navigational Satellite Systems (GNSS) and InSAR, seismic waveforms, and when applicable, tsunami waveforms from Deep-Ocean Assessment and Reporting of Tsunami (DART) gauges, provide slightly different observations that when incorporated together lead to a more robust model of fault slip distribution. The merging of different datasets is of particular importance along subduction zones where direct observations of seafloor deformation over the rupture area are extremely limited. Instead, instrumentation measures related ground motion from tens to hundreds of kilometers away. The distance from the event and dataset type can lead to a variable degree of resolution, affecting the ability to accurately model the spatial distribution of slip. This study analyzes the spatial resolution attained individually from geodetic and tsunami datasets as well as in a combined dataset. We constrain the importance of distance between estimated parameters and observed data and how that varies between land-based and open ocean datasets. Analysis focuses on accurately scaled subduction zone synthetic models as well as analysis of the relationship between slip and data in recent large subduction zone earthquakes. This study shows that seafloor deformation sensitive datasets, like open-ocean tsunami waveforms or seafloor geodetic instrumentation, can provide unique offshore resolution for understanding most large and particularly tsunamigenic megathrust earthquake activity. In most environments, we simply lack the capability to resolve static displacements using land-based geodetic observations.

Anisotropy in the Australasian upper mantle from Love and Rayleigh waveform inversion

NASA Astrophysics Data System (ADS)

Debayle, Eric; Kennett, B. L. N.

2000-12-01

Records of both Rayleigh and Love waves have been analyzed to determine the pattern of anisotropy in the Australasian region. The approach is based on a two-stage inversion. Starting from a smooth PREM model with transverse isotropy about a vertical symmetry axis, the first step is an inversion of the waveforms of surface waves to produce path specific one-dimensional (1-D) upper mantle models. Under the assumption that the 1-D models represent averages along the paths, the results from 1584 Love and Rayleigh wave seismograms are combined in a tomographic inversion to provide a representation of three-dimensional structure for wavespeed heterogeneities and anisotropy. Polarization anisotropy with SH faster than SV is retrieved in the upper 200-250 km of the mantle for most of Precambrian Australia. In this depth interval, significant lateral variations in the level of polarization anisotropy are present. Locally, the anisotropy can be large, reaching an extreme value of 9% that is difficult to reconcile with current mineralogical models. However, the discrepancy may be explained in part by the presence of strong lateral heterogeneities along the path, or by effects introduced by the simplifying assumption of transverse isotropy for each path. The consistency between the location of polarization and azimuthal anisotropy in depth suggests that both observations share a common origin. The observation of polarization anisotropy down to at least 200 km supports a two-layered anisotropic model as constrained by the azimuthal anisotropy of SV waves. In the upper layer, 150 km thick, anisotropy would be related to past deformation frozen in the lithosphere while in the lower layer, anisotropy would reflect present day deformation due to plate motion.

Imaging of 2-D multichannel land seismic data using an iterative inversion-migration scheme, Naga Thrust and Fold Belt, Assam, India

NASA Astrophysics Data System (ADS)

Jaiswal, Priyank; Dasgupta, Rahul

2010-05-01

We demonstrate that imaging of 2-D multichannel land seismic data can be effectively accomplished by a combination of reflection traveltime tomography and pre-stack depth migration (PSDM); we refer to the combined process as "the unified imaging". The unified imaging comprises cyclic runs of joint reflection and direct arrival inversion and pre-stack depth migration. From one cycle to another, both the inversion and the migration provide mutual feedbacks that are guided by the geological interpretation. The unified imaging is implemented in two broad stages. The first stage is similar to the conventional imaging except that it involves a significant use of velocity model from the inversion of the direct arrivals for both datuming and stacking velocity analysis. The first stage ends with an initial interval velocity model (from the stacking velocity analysis) and a corresponding depth migrated image. The second stage updates the velocity model and the depth image from the first stage in a cyclic manner; a single cycle comprises a single run of reflection traveltime inversion followed by PSDM. Interfaces used in the inversion are interpretations of the PSDM image in the previous cycle and the velocity model used in PSDM is from the joint inversion in the current cycle. Additionally in every cycle interpreted horizons in the stacked data are inverted as zero-offset reflections for constraining the interfaces; the velocity model is maintained stationary for the zero-offset inversion. A congruency factor, j, which measures the discrepancy between interfaces from the interpretation of the PSDM image and their corresponding counterparts from the inversion of the zero-offset reflections within assigned uncertainties, is computed in every cycle. A value of unity for jindicates that images from both the inversion and the migration are equivalent; at this point the unified imaging is said to have converged and is halted. We apply the unified imaging to 2-D multichannel seismic data from the Naga Thrust and Fold Belt (NTFB), India, were several exploratory wells in the last decade targeting sub-thrust leads in the footwall have failed. This failure is speculatively due to incorrect depth images which are in turn attributed to incorrect velocity models that are developed using conventional methods. The 2-D seismic data in this study is acquired perpendicular to the trend of the NTFB where the outcropping hanging wall has a topographic culmination. The acquisition style is split-spread with 30 m shot and receiver spacing and a nominal fold of 90. The data are recorded with a sample interval of 2 ms. Overall the data have a moderate signal-to-noise ratio and a broad frequency bandwidth of 8-80 Hz. The seismic line contains the failed exploratory well in the central part. The final results from unified imaging (both the depth image and the corresponding velocity model) suggest presence of a triangle zone, which was previously undiscovered. Conventional imaging had falsely portrayed the triangle zone as structural high which was interpreted as an anticline. As a result, the exploratory well, meant to target the anticline, met with pressure changes which were neither expected nor explained. The unified imaging results not only explain the observations in the well but also reveal new leads in the region. The velocity model from unified imaging was also found to be adequate for frequency-domain full-waveform imaging of the hanging wall. Results from waveform inversion are further corroborated by the geological interpretation of the exploratory well.

ADVANCED WAVEFORM SIMULATION FOR SEISMIC MONITORING EVENTS

DOE Office of Scientific and Technical Information (OSTI.GOV)

Helmberger, Donald V.; Tromp, Jeroen; Rodgers, Arthur J.

2008-10-17

This quarter, we have focused on several tasks: (1) Building a high-quality catalog of earthquake source parameters for the Middle East and East Asia. In East Asia, we computed source parameters using the CAP method for a set of events studied by Herrman et al., (MRR, 2006) using a complete waveform technique. Results indicated excellent agreement with the moment magnitudes in the range 3.5 -5.5. Below magnitude 3.5 the scatter increases. For events with more than 2-3 observations at different azimuths, we found good agreement of focal mechanisms. Depths were generally consistent, although differences of up to 10 km weremore » found. These results suggest that CAP modeling provides estimates of source parameters at least as reliable as complete waveform modeling techniques. However, East Asia and the Yellow Sea Korean Paraplatform (YSKP) region studied are relatively laterally homogeneous and may not benefit from the CAP method’s flexibility to shift waveform segments to account for path-dependent model errors. A more challenging region to study is the Middle East where strong variations in sedimentary basin, crustal thickness and crustal and mantle seismic velocities greatly impact regional wave propagation. We applied the CAP method to a set of events in and around Iran and found good agreement between estimated focal mechanisms and those reported by the Global Centroid Moment Tensor (CMT) catalog. We found a possible bias in the moment magnitudes that may be due to the thick low-velocity crust in the Iranian Plateau. (2) Testing Methods on a Lifetime Regional Data Set. In particular, the recent 2/21/08 Nevada Event and Aftershock Sequence occurred in the middle of USArray, producing over a thousand records per event. The tectonic setting is quite similar to Central Iran and thus provides an excellent testbed for CAP+ at ranges out to 10°, including extensive observations of crustal thinning and thickening and various Pnl complexities. Broadband modeling in 1D, 2D, and 3D will be presented. (3) Shallow Crustal Structure and Sparse Network Source Inversions for Southern California. We conducted a detailed test of a recently developed technique, CAPloc, in recovering source parameters including location and depth based on tomographic maps. We tested two-station solutions against 160 well determined events which worked well except for paths crossing deep basins and along mountain ridges.« less

Shear wave velocity structure in North America from large-scale waveform inversions of surface waves

USGS Publications Warehouse

Alsina, D.; Woodward, R.L.; Snieder, R.K.

1996-01-01

A two-step nonlinear and linear inversion is carried out to map the lateral heterogeneity beneath North America using surface wave data. The lateral resolution for most areas of the model is of the order of several hundred kilometers. The most obvious feature in the tomographic images is the rapid transition between low velocities in the technically active region west of the Rocky Mountains and high velocities in the stable central and eastern shield of North America. The model also reveals smaller-scale heterogeneous velocity structures. A high-velocity anomaly is imaged beneath the state of Washington that could be explained as the subducting Juan de Fuca plate beneath the Cascades. A large low-velocity structure extends along the coast from the Mendocino to the Rivera triple junction and to the continental interior across the southwestern United States and northwestern Mexico. Its shape changes notably with depth. This anomaly largely coincides with the part of the margin where no lithosphere is consumed since the subduction has been replaced by a transform fault. Evidence for a discontinuous subduction of the Cocos plate along the Middle American Trench is found. In central Mexico a transition is visible from low velocities across the Trans-Mexican Volcanic Belt (TMVB) to high velocities beneath the Yucatan Peninsula. Two elongated low-velocity anomalies beneath the Yellowstone Plateau and the eastern Snake River Plain volcanic system and beneath central Mexico and the TMVB seem to be associated with magmatism and partial melting. Another low-velocity feature is seen at depths of approximately 200 km beneath Florida and the Atlantic Coastal Plain. The inversion technique used is based on a linear surface wave scattering theory, which gives tomographic images of the relative phase velocity perturbations in four period bands ranging from 40 to 150 s. In order to find a smooth reference model a nonlinear inversion based on ray theory is first performed. After correcting for the crustal thickness the phase velocity perturbations obtained from the subsequent linear waveform inversion for the different period bands are converted to a three-layer model of S velocity perturbations (layer 1, 25-100 km; layer 2, 100-200 km) layer 3, 200-300 km). We have applied this method on 275 high-quality Rayleigh waves recorded by a variety of instruments in North America (IRIS/USGS, IRIS/IDA, TERRAscope, RSTN). Sensitivity tests indicate that the lateral resolution is especially good in the densely sampled western continental United States, Mexico, and the Gulf of Mexico.

Surface wave tomography applied to the North American upper mantle

NASA Astrophysics Data System (ADS)

van der Lee, Suzan; Frederiksen, Andrew

Tomographic techniques that invert seismic surface waves for 3-D Earth structure differ in their definitions of data and the forward problem as well as in the parameterization of the tomographic model. However, all such techniques have in common that the tomographic inverse problem involves solving a large and mixed-determined set of linear equations. Consequently these inverse problems have multiple solutions and inherently undefinable accuracy. Smoother and rougher tomographic models are found with rougher (confined to great circle path) and smoother (finite-width) sensitivity kernels, respectively. A powerful, well-tested method of surface wave tomography (Partitioned Waveform Inversion) is based on inverting the waveforms of wave trains comprising regional S and surface waves from at least hundreds of seismograms for 3-D variations in S wave velocity. We apply this method to nearly 1400 seismograms recorded by digital broadband seismic stations in North America. The new 3-D S-velocity model, NA04, is consistent with previous findings that are based on separate, overlapping data sets. The merging of US and Canadian data sets, adding Canadian recordings of Mexican earthquakes, and combining fundamental-mode with higher-mode waveforms provides superior resolution, in particular in the US-Canada border region and the deep upper mantle. NA04 shows that 1) the Atlantic upper mantle is seismically faster than the Pacific upper mantle, 2) the uppermost mantle beneath Precambrian North America could be one and a half times as rigid as the upper mantle beneath Meso- and Cenozoic North America, with the upper mantle beneath Paleozoic North America being intermediate in seismic rigidity, 3) upper-mantle structure varies laterally within these geologic-age domains, and 4) the distribution of high-velocity anomalies in the deep upper mantle aligns with lower mantle images of the subducted Farallon and Kula plates and indicate that trailing fragments of these subducted oceanic plates still reside in the transition zone. The thickness of the high-velocity layer beneath Precambrian North America is estimated to be 250±70 km thick. On a smaller scale NA04 shows 1) high-velocities associated with subduction of the Pacific plate beneath the Aleutian arc, 2) the absence of expected high velocities in the upper mantle beneath the Wyoming craton, 3) a V-shaped dent below 150 km in the high-velocity cratonic lithosphere beneath New England, 4) the cratonic lithosphere beneath Precambrian North America being confined southwest of Baffin Bay, west of the Appalachians, north of the Ouachitas, east of the Rocky Mountains, and south of the Arctic Ocean, 5) the cratonic lithosphere beneath the Canadian shield having higher S-velocities than that beneath Precambrian basement that is covered with Phanerozoic sediments, 6) the lowest S velocities are concentrated beneath the Gulf of California, northern Mexico, and the Basin and Range Province.

Lossy Wavefield Compression for Full-Waveform Inversion

NASA Astrophysics Data System (ADS)

Boehm, C.; Fichtner, A.; de la Puente, J.; Hanzich, M.

2015-12-01

We present lossy compression techniques, tailored to the inexact computation of sensitivity kernels, that significantly reduce the memory requirements of adjoint-based minimization schemes. Adjoint methods are a powerful tool to solve tomography problems in full-waveform inversion (FWI). Yet they face the challenge of massive memory requirements caused by the opposite directions of forward and adjoint simulations and the necessity to access both wavefields simultaneously during the computation of the sensitivity kernel. Thus, storage, I/O operations, and memory bandwidth become key topics in FWI. In this talk, we present strategies for the temporal and spatial compression of the forward wavefield. This comprises re-interpolation with coarse time steps and an adaptive polynomial degree of the spectral element shape functions. In addition, we predict the projection errors on a hierarchy of grids and re-quantize the residuals with an adaptive floating-point accuracy to improve the approximation. Furthermore, we use the first arrivals of adjoint waves to identify "shadow zones" that do not contribute to the sensitivity kernel at all. Updating and storing the wavefield within these shadow zones is skipped, which reduces memory requirements and computational costs at the same time. Compared to check-pointing, our approach has only a negligible computational overhead, utilizing the fact that a sufficiently accurate sensitivity kernel does not require a fully resolved forward wavefield. Furthermore, we use adaptive compression thresholds during the FWI iterations to ensure convergence. Numerical experiments on the reservoir scale and for the Western Mediterranean prove the high potential of this approach with an effective compression factor of 500-1000. Furthermore, it is computationally cheap and easy to integrate in both, finite-differences and finite-element wave propagation codes.

Delineation of Rupture Propagation of Large Earthquakes Using Source-Scanning Algorithm: A Control Study

NASA Astrophysics Data System (ADS)

Kao, H.; Shan, S.

2004-12-01

Determination of the rupture propagation of large earthquakes is important and of wide interest to the seismological research community. The conventional inversion method determines the distribution of slip at a grid of subfaults whose orientations are predefined. As a result, difference choices of fault geometry and dimensions often result in different solutions. In this study, we try to reconstruct the rupture history of an earthquake using the newly developed Source-Scanning Algorithm (SSA) without imposing any a priori constraints on the fault's orientation and dimension. The SSA identifies the distribution of seismic sources in two steps. First, it calculates the theoretical arrival times from all grid points inside the model space to all seismic stations by assuming an origin time. Then, the absolute amplitudes of the observed waveforms at the predicted arrival times are added to give the "brightness" of each time-space pair, and the brightest spots mark the locations of sources. The propagation of the rupture is depicted by the migration of the brightest spots throughout a prescribed time window. A series of experiments are conducted to test the resolution of the SSA inversion. Contrary to the conventional wisdom that seismometers should be placed as close as possible to the fault trace to give the best resolution in delineating rupture details, we found that the best results are obtained if the seismograms are recorded at a distance about half of the total rupture length away from the fault trace. This is especially true when the rupture duration is longer than ~10 s. A possible explanation is that the geometric spreading effects for waveforms from different segments of the rupture are about the same if the stations are sufficiently away from the fault trace, thus giving a uniform resolution to the entire rupture history.

Comparison of Velocity Models for South America through Seismic Wave Modeling of Ten Andean Earthquakes Recorded by the Brazilian Seismographic Network using the Spectral Element Method

NASA Astrophysics Data System (ADS)

Ciardelli, C.; Assumpcao, M.

2016-12-01

In this work, we carried out simulations of seismic waves propagation for ten large earthquakes occurred in Chile between 2011 and 2016,using the SPECFEM3D Global software (Komatitsch and Tromp, 2000) and the Centroid Moment Tensor solutions from the global catalog (Dziewonski, Chou and Woodhouse, 1981; Ekström, Nettles and Dziewonski, 2012). For each event, the complete wave field was calculated using the spectral element method and recorded at the coordinates of the Brazilian Seismographic Network, thus we can compare the synthetic seismograms with the real data. Initially, we assess the differences between CRUST1.0 and CRUST2.0 models using the transversely isotropic PREM for the internal part of the planet. We will also compare the PREM velocity model plus CRUST1.0 with the Feng's velocity model for South America (Feng, Van der Lee and Assumpção, 2007), calculated using Partitioned Waveform Inversion. For each model, we will evaluate the misfit for all stations of the network. The similarity can be estimated by pure RMS or combining it with cross-correlation. Travel-time residuals can also be used to better constrain velocity anomalies and avoid cycle-skipping. The results will help to assess which model is more appropriated to start a Full-waveform Tomography of the South American continent and the surrounding oceans.

Seismology-based early identification of dam-formation landquake events.

PubMed

Chao, Wei-An; Zhao, Li; Chen, Su-Chin; Wu, Yih-Min; Chen, Chi-Hsuan; Huang, Hsin-Hua

2016-01-12

Flooding resulting from the bursting of dams formed by landquake events such as rock avalanches, landslides and debris flows can lead to serious bank erosion and inundation of populated areas near rivers. Seismic waves can be generated by landquake events which can be described as time-dependent forces (unloading/reloading cycles) acting on the Earth. In this study, we conduct inversions of long-period (LP, period ≥20 s) waveforms for the landquake force histories (LFHs) of ten events, which provide quantitative characterization of the initiation, propagation and termination stages of the slope failures. When the results obtained from LP waveforms are analyzed together with high-frequency (HF, 1-3 Hz) seismic signals, we find a relatively strong late-arriving seismic phase (dubbed Dam-forming phase or D-phase) recorded clearly in the HF waveforms at the closest stations, which potentially marks the time when the collapsed masses sliding into river and perhaps even impacting the topographic barrier on the opposite bank. Consequently, our approach to analyzing the LP and HF waveforms developed in this study has a high potential for identifying five dam-forming landquake events (DFLEs) in near real-time using broadband seismic records, which can provide timely warnings of the impending floods to downstream residents.

Collaborative Research: Ground Truth of African and Eastern Mediterranean Shallow Seismicity Using SAR Interferometry and Gibbs Sampling Inversion

DTIC Science & Technology

2006-10-05

the likely existence of a small foreshock . 2. BACKGROUND 2.1. InSAR The most well-known examples of InSAR used as a geodetic tool involve...the event. We have used the seismic waveforms in the Sultan Dag event to identify a small foreshock preceding the main shock by about 3 seconds

Subfemtosecond directional control of chemical processes in molecules

NASA Astrophysics Data System (ADS)

Alnaser, Ali S.; Litvinyuk, Igor V.

2017-02-01

Laser pulses with a waveform-controlled electric field and broken inversion symmetry establish the opportunity to achieve directional control of molecular processes on a subfemtosecond timescale. Several techniques could be used to break the inversion symmetry of an electric field. The most common ones include combining a fundamental laser frequency with its second harmonic or with higher -frequency pulses (or pulse trains) as well as using few-cycle pulses with known carrier-envelope phase (CEP). In the case of CEP, control over chemical transformations, typically occurring on a timescale of many femtoseconds, is driven by much faster sub-cycle processes of subfemtosecond to few-femtosecond duration. This is possible because electrons are much lighter than nuclei and fast electron motion is coupled to the much slower nuclear motion. The control originates from populating coherent superpositions of different electronic or vibrational states with relative phases that are dependent on the CEP or phase offset between components of a two-color pulse. In this paper, we review the recent progress made in the directional control over chemical processes, driven by intense few-cycle laser pulses a of waveform-tailored electric field, in different molecules.

System and Method for Generating a Frequency Modulated Linear Laser Waveform

NASA Technical Reports Server (NTRS)

Pierrottet, Diego F. (Inventor); Petway, Larry B. (Inventor); Amzajerdian, Farzin (Inventor); Barnes, Bruce W. (Inventor); Lockard, George E. (Inventor); Hines, Glenn D. (Inventor)

2017-01-01

A system for generating a frequency modulated linear laser waveform includes a single frequency laser generator to produce a laser output signal. An electro-optical modulator modulates the frequency of the laser output signal to define a linear triangular waveform. An optical circulator passes the linear triangular waveform to a band-pass optical filter to filter out harmonic frequencies created in the waveform during modulation of the laser output signal, to define a pure filtered modulated waveform having a very narrow bandwidth. The optical circulator receives the pure filtered modulated laser waveform and transmits the modulated laser waveform to a target.

System and Method for Generating a Frequency Modulated Linear Laser Waveform

NASA Technical Reports Server (NTRS)

Pierrottet, Diego F. (Inventor); Petway, Larry B. (Inventor); Amzajerdian, Farzin (Inventor); Barnes, Bruce W. (Inventor); Lockard, George E. (Inventor); Hines, Glenn D. (Inventor)

2014-01-01

A system for generating a frequency modulated linear laser waveform includes a single frequency laser generator to produce a laser output signal. An electro-optical modulator modulates the frequency of the laser output signal to define a linear triangular waveform. An optical circulator passes the linear triangular waveform to a band-pass optical filter to filter out harmonic frequencies created in the waveform during modulation of the laser output signal, to define a pure filtered modulated waveform having a very narrow bandwidth. The optical circulator receives the pure filtered modulated laser waveform and transmits the modulated laser waveform to a target.

Frequency-Dependent Rupture Processes for the 2011 Tohoku Earthquake

NASA Astrophysics Data System (ADS)

Miyake, H.

2012-12-01

The 2011 Tohoku earthquake is characterized by frequency-dependent rupture process [e.g., Ide et al., 2011; Wang and Mori, 2011; Yao et al., 2011]. For understanding rupture dynamics of this earthquake, it is extremely important to investigate wave-based source inversions for various frequency bands. The above frequency-dependent characteristics have been derived from teleseismic analyses. This study challenges to infer frequency-dependent rupture processes from strong motion waveforms of K-NET and KiK-net stations. The observations suggested three or more S-wave phases, and ground velocities at several near-source stations showed different arrivals of their long- and short-period components. We performed complex source spectral inversions with frequency-dependent phase weighting developed by Miyake et al. [2002]. The technique idealizes both the coherent and stochastic summation of waveforms using empirical Green's functions. Due to the limitation of signal-to-noise ratio of the empirical Green's functions, the analyzed frequency bands were set within 0.05-10 Hz. We assumed a fault plane with 480 km in length by 180 km in width with a single time window for rupture following Koketsu et al. [2011] and Asano and Iwata [2012]. The inversion revealed source ruptures expanding from the hypocenter, and generated sharp slip-velocity intensities at the down-dip edge. In addition to test the effects of empirical/hybrid Green's functions and with/without rupture front constraints on the inverted solutions, we will discuss distributions of slip-velocity intensity and a progression of wave generation with increasing frequency.

Seafloor Deformation and Localized Source Mechanisms of the 2011 M9 Tohoku Earthquake and Tsunami.

NASA Astrophysics Data System (ADS)

Masterlark, T.; Grilli, S. T.; Tappin, D. R.; Kirby, J. T.

2012-12-01

The 2011 M9 Tohoku Earthquake (TE) ruptured the interface separating the Pacific and Okhotsk Plates. This rupture was about hundred kilometers in the along-strike direction and 200 kilometers in the down-dip direction. The TE was primarily thrust having substantial slip along the up-dip portion of the rupture, near the Japan Trench. The regional-scale seafloor deformation from the TE triggered a tsunami with run-ups of a few tens of meters that caused extensive damage along the east coast of Tohoku, Japan. We construct finite element models (FEMs) to simulate the deformation caused by a distribution of coseismic slip along the curved rupture surface of the TE. The FEMs include a distribution of material properties that accounts for the subduction zone structure -a weak forearc, volcanic arc, and backarc basin of the overriding Okhotsk Plate overriding the relatively strong subducting slab that is capped by basaltic oceanic crust. The coseismic rupture is simulated as a distribution of elastic dislocations along the interface separating the forearc of the overriding plate and the oceanic crust of the subducting slab. The slip distribution is calibrated to both onshore and offshore geodetic data, using linear least-squares inverse methods with FEM-generated Greens Functions and second order regularization. The regularization is imposed with a conductance matrix, constructed using Galerkin's Method to account for the curvilinear relationships among the dislocating node pairs. The estimated slip distribution is generally characterized as a few tens of meters of slip over the entire rupture, with greater slip magnitudes (>50 meters) concentrated up-dip and near the Japan Trench. The offshore geodetic data provide critical constraints for the location of the polarity reversal of predicted seafloor vertical deformation. Wave models excited by the predicted regional-scale seafloor deformation generally well predict observed tsunami run-ups and the vertical displacement magnitudes of low frequency waves of coastal GPS buoys. However, coastal areas near Sanriku, Japan experienced anomalously high run-ups of 40 meters and local offshore GPS buoys indicate high frequency waveforms that are incompatible with the coseismic seafloor deformation of the TE. These observations require a localized deformation source near the Japan Trench and just to the north of the TE rupture zone, which models solely based on tsunami waveform inversion predict. Others suggest that a submarine mass failure at this location, presumably triggered by the TE, can excite such waveforms. In this study, we investigate an alternative hypothesis that localized splay faulting, also presumably triggered by the TE, can excite the anomalous waveforms. To do so, we will estimate plausible suites of splay fault and slip parameters that can account for the anomalously high magnitude and high frequency tsunami waves sourced from a localized area near the Japan Trench and north of the TE rupture.

Beyond HRV: attractor reconstruction using the entire cardiovascular waveform data for novel feature extraction.

PubMed

Aston, Philip J; Christie, Mark I; Huang, Ying H; Nandi, Manasi

2018-03-01

Advances in monitoring technology allow blood pressure waveforms to be collected at sampling frequencies of 250-1000 Hz for long time periods. However, much of the raw data are under-analysed. Heart rate variability (HRV) methods, in which beat-to-beat interval lengths are extracted and analysed, have been extensively studied. However, this approach discards the majority of the raw data. Our aim is to detect changes in the shape of the waveform in long streams of blood pressure data. Our approach involves extracting key features from large complex data sets by generating a reconstructed attractor in a three-dimensional phase space using delay coordinates from a window of the entire raw waveform data. The naturally occurring baseline variation is removed by projecting the attractor onto a plane from which new quantitative measures are obtained. The time window is moved through the data to give a collection of signals which relate to various aspects of the waveform shape. This approach enables visualisation and quantification of changes in the waveform shape and has been applied to blood pressure data collected from conscious unrestrained mice and to human blood pressure data. The interpretation of the attractor measures is aided by the analysis of simple artificial waveforms. We have developed and analysed a new method for analysing blood pressure data that uses all of the waveform data and hence can detect changes in the waveform shape that HRV methods cannot, which is confirmed with an example, and hence our method goes 'beyond HRV'.

Beyond HRV: attractor reconstruction using the entire cardiovascular waveform data for novel feature extraction

PubMed Central

Aston, Philip J; Christie, Mark I; Huang, Ying H; Nandi, Manasi

2018-01-01

Abstract Advances in monitoring technology allow blood pressure waveforms to be collected at sampling frequencies of 250–1000 Hz for long time periods. However, much of the raw data are under-analysed. Heart rate variability (HRV) methods, in which beat-to-beat interval lengths are extracted and analysed, have been extensively studied. However, this approach discards the majority of the raw data. Objective: Our aim is to detect changes in the shape of the waveform in long streams of blood pressure data. Approach: Our approach involves extracting key features from large complex data sets by generating a reconstructed attractor in a three-dimensional phase space using delay coordinates from a window of the entire raw waveform data. The naturally occurring baseline variation is removed by projecting the attractor onto a plane from which new quantitative measures are obtained. The time window is moved through the data to give a collection of signals which relate to various aspects of the waveform shape. Main results: This approach enables visualisation and quantification of changes in the waveform shape and has been applied to blood pressure data collected from conscious unrestrained mice and to human blood pressure data. The interpretation of the attractor measures is aided by the analysis of simple artificial waveforms. Significance: We have developed and analysed a new method for analysing blood pressure data that uses all of the waveform data and hence can detect changes in the waveform shape that HRV methods cannot, which is confirmed with an example, and hence our method goes ‘beyond HRV’. PMID:29350622

Adjoint-tomography for a Local Surface Structure: Methodology and a Blind Test

NASA Astrophysics Data System (ADS)

Kubina, Filip; Michlik, Filip; Moczo, Peter; Kristek, Jozef; Stripajova, Svetlana

2017-04-01

We have developed a multiscale full-waveform adjoint-tomography method for local surface sedimentary structures with complicated interference wavefields. The local surface sedimentary basins and valleys are often responsible for anomalous earthquake ground motions and corresponding damage in earthquakes. In many cases only relatively small number of records of a few local earthquakes is available for a site of interest. Consequently, prediction of earthquake ground motion at the site has to include numerical modeling for a realistic model of the local structure. Though limited, the information about the local structure encoded in the records is important and irreplaceable. It is therefore reasonable to have a method capable of using the limited information in records for improving a model of the local structure. A local surface structure and its interference wavefield require a specific multiscale approach. In order to verify our inversion method, we performed a blind test. We obtained synthetic seismograms at 8 receivers for 2 local sources, complete description of the sources, positions of the receivers and material parameters of the bedrock. We considered the simplest possible starting model - a homogeneous halfspace made of the bedrock. Using our inversion method we obtained an inverted model. Given the starting model, synthetic seismograms simulated for the inverted model are surprisingly close to the synthetic seismograms simulated for the true structure in the target frequency range up to 4.5 Hz. We quantify the level of agreement between the true and inverted seismograms using the L2 and time-frequency misfits, and, more importantly for earthquake-engineering applications, also using the goodness-of-fit criteria based on the earthquake-engineering characteristics of earthquake ground motion. We also verified the inverted model for other source-receiver configurations not used in the inversion.

Modulation Based on Probability Density Functions

NASA Technical Reports Server (NTRS)

Williams, Glenn L.

2009-01-01

A proposed method of modulating a sinusoidal carrier signal to convey digital information involves the use of histograms representing probability density functions (PDFs) that characterize samples of the signal waveform. The method is based partly on the observation that when a waveform is sampled (whether by analog or digital means) over a time interval at least as long as one half cycle of the waveform, the samples can be sorted by frequency of occurrence, thereby constructing a histogram representing a PDF of the waveform during that time interval.

Multivariate Formation Pressure Prediction with Seismic-derived Petrophysical Properties from Prestack AVO inversion and Poststack Seismic Motion Inversion

NASA Astrophysics Data System (ADS)

Yu, H.; Gu, H.

2017-12-01

A novel multivariate seismic formation pressure prediction methodology is presented, which incorporates high-resolution seismic velocity data from prestack AVO inversion, and petrophysical data (porosity and shale volume) derived from poststack seismic motion inversion. In contrast to traditional seismic formation prediction methods, the proposed methodology is based on a multivariate pressure prediction model and utilizes a trace-by-trace multivariate regression analysis on seismic-derived petrophysical properties to calibrate model parameters in order to make accurate predictions with higher resolution in both vertical and lateral directions. With prestack time migration velocity as initial velocity model, an AVO inversion was first applied to prestack dataset to obtain high-resolution seismic velocity with higher frequency that is to be used as the velocity input for seismic pressure prediction, and the density dataset to calculate accurate Overburden Pressure (OBP). Seismic Motion Inversion (SMI) is an inversion technique based on Markov Chain Monte Carlo simulation. Both structural variability and similarity of seismic waveform are used to incorporate well log data to characterize the variability of the property to be obtained. In this research, porosity and shale volume are first interpreted on well logs, and then combined with poststack seismic data using SMI to build porosity and shale volume datasets for seismic pressure prediction. A multivariate effective stress model is used to convert velocity, porosity and shale volume datasets to effective stress. After a thorough study of the regional stratigraphic and sedimentary characteristics, a regional normally compacted interval model is built, and then the coefficients in the multivariate prediction model are determined in a trace-by-trace multivariate regression analysis on the petrophysical data. The coefficients are used to convert velocity, porosity and shale volume datasets to effective stress and then to calculate formation pressure with OBP. Application of the proposed methodology to a research area in East China Sea has proved that the method can bridge the gap between seismic and well log pressure prediction and give predicted pressure values close to pressure meassurements from well testing.

Gas stream analysis using voltage-current time differential operation of electrochemical sensors

DOE Office of Scientific and Technical Information (OSTI.GOV)

Woo, Leta Yar-Li; Glass, Robert Scott; Fitzpatrick, Joseph Jay

A method for analysis of a gas stream. The method includes identifying an affected region of an affected waveform signal corresponding to at least one characteristic of the gas stream. The method also includes calculating a voltage-current time differential between the affected region of the affected waveform signal and a corresponding region of an original waveform signal. The affected region and the corresponding region of the waveform signals have a sensitivity specific to the at least one characteristic of the gas stream. The method also includes generating a value for the at least one characteristic of the gas stream basedmore » on the calculated voltage-current time differential.« less

Radial artery pulse waveform analysis based on curve fitting using discrete Fourier series.

PubMed

Jiang, Zhixing; Zhang, David; Lu, Guangming

2018-04-19

Radial artery pulse diagnosis has been playing an important role in traditional Chinese medicine (TCM). For its non-invasion and convenience, the pulse diagnosis has great significance in diseases analysis of modern medicine. The practitioners sense the pulse waveforms in patients' wrist to make diagnoses based on their non-objective personal experience. With the researches of pulse acquisition platforms and computerized analysis methods, the objective study on pulse diagnosis can help the TCM to keep up with the development of modern medicine. In this paper, we propose a new method to extract feature from pulse waveform based on discrete Fourier series (DFS). It regards the waveform as one kind of signal that consists of a series of sub-components represented by sine and cosine (SC) signals with different frequencies and amplitudes. After the pulse signals are collected and preprocessed, we fit the average waveform for each sample using discrete Fourier series by least squares. The feature vector is comprised by the coefficients of discrete Fourier series function. Compared with the fitting method using Gaussian mixture function, the fitting errors of proposed method are smaller, which indicate that our method can represent the original signal better. The classification performance of proposed feature is superior to the other features extracted from waveform, liking auto-regression model and Gaussian mixture model. The coefficients of optimized DFS function, who is used to fit the arterial pressure waveforms, can obtain better performance in modeling the waveforms and holds more potential information for distinguishing different psychological states. Copyright © 2018 Elsevier B.V. All rights reserved.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Smallwood, D.O.

It is recognized that some dynamic and noise environments are characterized by time histories which are not Gaussian. An example is high intensity acoustic noise. Another example is some transportation vibration. A better simulation of these environments can be generated if a zero mean non-Gaussian time history can be reproduced with a specified auto (or power) spectral density (ASD or PSD) and a specified probability density function (pdf). After the required time history is synthesized, the waveform can be used for simulation purposes. For example, modem waveform reproduction techniques can be used to reproduce the waveform on electrodynamic or electrohydraulicmore » shakers. Or the waveforms can be used in digital simulations. A method is presented for the generation of realizations of zero mean non-Gaussian random time histories with a specified ASD, and pdf. First a Gaussian time history with the specified auto (or power) spectral density (ASD) is generated. A monotonic nonlinear function relating the Gaussian waveform to the desired realization is then established based on the Cumulative Distribution Function (CDF) of the desired waveform and the known CDF of a Gaussian waveform. The established function is used to transform the Gaussian waveform to a realization of the desired waveform. Since the transformation preserves the zero-crossings and peaks of the original Gaussian waveform, and does not introduce any substantial discontinuities, the ASD is not substantially changed. Several methods are available to generate a realization of a Gaussian distributed waveform with a known ASD. The method of Smallwood and Paez (1993) is an example. However, the generation of random noise with a specified ASD but with a non-Gaussian distribution is less well known.« less

Parallelized Three-Dimensional Resistivity Inversion Using Finite Elements And Adjoint State Methods

NASA Astrophysics Data System (ADS)

Schaa, Ralf; Gross, Lutz; Du Plessis, Jaco

2015-04-01

The resistivity method is one of the oldest geophysical exploration methods, which employs one pair of electrodes to inject current into the ground and one or more pairs of electrodes to measure the electrical potential difference. The potential difference is a non-linear function of the subsurface resistivity distribution described by an elliptic partial differential equation (PDE) of the Poisson type. Inversion of measured potentials solves for the subsurface resistivity represented by PDE coefficients. With increasing advances in multichannel resistivity acquisition systems (systems with more than 60 channels and full waveform recording are now emerging), inversion software require efficient storage and solver algorithms. We developed the finite element solver Escript, which provides a user-friendly programming environment in Python to solve large-scale PDE-based problems (see https://launchpad.net/escript-finley). Using finite elements, highly irregular shaped geology and topography can readily be taken into account. For the 3D resistivity problem, we have implemented the secondary potential approach, where the PDE is decomposed into a primary potential caused by the source current and the secondary potential caused by changes in subsurface resistivity. The primary potential is calculated analytically, and the boundary value problem for the secondary potential is solved using nodal finite elements. This approach removes the singularity caused by the source currents and provides more accurate 3D resistivity models. To solve the inversion problem we apply a 'first optimize then discretize' approach using the quasi-Newton scheme in form of the limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) method (see Gross & Kemp 2013). The evaluation of the cost function requires the solution of the secondary potential PDE for each source current and the solution of the corresponding adjoint-state PDE for the cost function gradients with respect to the subsurface resistivity. The Hessian of the regularization term is used as preconditioner which requires an additional PDE solution in each iteration step. As it turns out, the relevant PDEs are naturally formulated in the finite element framework. Using the domain decomposition method provided in Escript, the inversion scheme has been parallelized for distributed memory computers with multi-core shared memory nodes. We show numerical examples from simple layered models to complex 3D models and compare with the results from other methods. The inversion scheme is furthermore tested on a field data example to characterise localised freshwater discharge in a coastal environment.. References: L. Gross and C. Kemp (2013) Large Scale Joint Inversion of Geophysical Data using the Finite Element Method in escript. ASEG Extended Abstracts 2013, http://dx.doi.org/10.1071/ASEG2013ab306

Oscillation of a Shallow Hydrothermal Fissure Inferred from Long-Period Seismic Events at Taal Volcano, the Philippines

NASA Astrophysics Data System (ADS)

Maeda, Y.; Kumagai, H.; Lacson, R.; Figueroa, M. S.; Yamashina, T.

2012-12-01

We installed a multi-parameter monitoring network including five broadband seismometers at Taal volcano, the Philippines, where a high risk of near-future eruption is expected. The network detected more than 40,000 long-period (LP) seismic events which have a peak frequency of 0.8 Hz and a Q value of 6. Most of the events occurred in a 2-month-long swarm period of ~600 events/day. Our travel time analysis pointed to a shallow source (100-200 m) beneath the northeastern flank of the active volcanic island. To determine the source mechanism of the LP events, we performed waveform inversion. We first fixed the source location to that obtained by the travel time analysis, and performed inversion using waveforms with and without site amplification corrections and assuming four simple source geometries (a vertical crack, a horizontal crack, a vertical pipe, and a sphere). We obtained the minimum AIC value for the vertical crack source geometry using the corrected waveforms. We next performed a grid search for dip, azimuth, and the location of the tensile crack source using the corrected waveforms. We obtained small residuals for crack dips between 30 and 60 degrees at similar locations to that of the travel time analysis. We used the fluid-filled crack model to interpret the observed complex frequencies of the events. The observed waveforms of the events show a small Q value (= 6), which may be explained by bubbly basalt, bubbly water, or gas. However, since the source location is estimated to be shallow (100-200 m) and we have no evidence for an ascent of magma to such a shallow depth in the swarm period, bubbly basalt seems to be unrealistic. It seems difficult to maintain bubbly water in the inclined crack. For bubbly water, a peak frequency variation is expected to occur due to a variation of the bubble content, whereas the observed peak frequencies of the events are almost constant. The constant frequency is more easily realized by gas in a crack. We therefore examine H2O gas (vapor) for simplicity. We calculated far-field waveforms generated by an oscillation of a crack containing vapor, and applied the Sompi method to estimate Q and nondimensional frequency. The estimated Q of a fundamental longitudinal mode oscillation was similar to the observation. We obtained a reasonable crack size (188 m) from a comparison of the observed peak frequency (0.8 Hz) with the calculated nondimensional frequency of the mode. In the swarm period of the LP events, other anomalies such as large volcano deformation and significant increase of gas emission from the main crater were not observed. This feature and the crack model result above suggest an active and localized vapor supply from magma at depth to the LP source. Such a localized supply may be realized by a transportation of vapor through a fissure. If we assume that the estimated crack volume (10^2 m^3) corresponds to vapor supplied to the LP source for each event, the total vapor mass supplied throughout the swarm period is ~10^7 kg. If we assume that this amount of vapor was originated by degassing from the magma and transported to the LP source through the fissure, we can estimate a magma volume of ~10^6 m^3. We thus suggest that the LP events at Taal were triggered by degassing and transportation of vapor from a deep magma to a shallow depth through a fissure.

Waveform Tomography Applied to Long Streamer MCS Data from the Scotian Slope

NASA Astrophysics Data System (ADS)

Delescluse, Matthias; Louden, Keith; Nedimovic, Mladen

2010-05-01

Detailed velocity models of the earth subsurface can be obtained through waveform tomography, a method that relies on using information from the full wavefield. Such models can be of significantly higher resolution than the corresponding models formed by more generic traveltime tomography methods, which are constrained only by the wave arrival times. However, to derive the detailed subsurface velocity, the waveform method is sensitive to modelling low-frequency refracted waves that have long paths through target structures. Thus field examples primarily have focused on the analysis of long-offset wide-angle datasets collected using autonomous receivers, in which refractions arrive at earlier times than reflections and there is a significant separation between the two wave arrivals. MCS datasets with shorter offsets typically lack these important features, which result in methodological problems (e.g. Hicks and Pratt, 2001), even though they benefit from a high density of raypaths and uniformity of receiver and shot properties. Modern marine seismic acquisition using long streamers now offers both the ability to record refracted waves at far offsets arriving ahead of the seafloor reflection, and the ability to do this at great density using uniform sources. In this study, we use 2D MCS data acquired with a 9-km-long streamer by ION GX-Technology over the Nova Scotia Slope in water depths of ~1600 m. We show that the refracted arrivals, although restricted to receivers between offsets of 7.5 and 9 km, provide sufficient information to successfully invert for a high-resolution velocity field. Using a frequency domain acoustic code (Pratt, 1999) over frequencies from 8 Hz to 24 Hz on two crossing profiles (45 and 20 km long), we detail how the limited refracted waves can constrain the velocity field above the depth of the turning waves (~1.5 km below seafloor). Several important features are resolved by the waveform velocity model that are not present in the initial travel-time model. In particular, a high velocity layer due to gas hydrates is imaged along the entire profile even where a characteristic BSR is not visible. The velocity increase in the gas hydrate layer is very small (< 100 m/s). In addition, a strong velocity increase of ~ 300 m/s exists below a deeper, gently dipping reflector along which discontinuous low-velocity zones, probably related to gas, are present. Velocity models are consistent at the crossing point between the two profiles. The depth limitation of the detailed MCS waveform tomography imaging could be extended by even longer streamers (e.g. 15 km) or by joint inversion with OBS data.

TSOS and TSOS-FK hybrid methods for modelling the propagation of seismic waves

NASA Astrophysics Data System (ADS)

Ma, Jian; Yang, Dinghui; Tong, Ping; Ma, Xiao

2018-05-01

We develop a new time-space optimized symplectic (TSOS) method for numerically solving elastic wave equations in heterogeneous isotropic media. We use the phase-preserving symplectic partitioned Runge-Kutta method to evaluate the time derivatives and optimized explicit finite-difference (FD) schemes to discretize the space derivatives. We introduce the averaged medium scheme into the TSOS method to further increase its capability of dealing with heterogeneous media and match the boundary-modified scheme for implementing free-surface boundary conditions and the auxiliary differential equation complex frequency-shifted perfectly matched layer (ADE CFS-PML) non-reflecting boundaries with the TSOS method. A comparison of the TSOS method with analytical solutions and standard FD schemes indicates that the waveform generated by the TSOS method is more similar to the analytic solution and has a smaller error than other FD methods, which illustrates the efficiency and accuracy of the TSOS method. Subsequently, we focus on the calculation of synthetic seismograms for teleseismic P- or S-waves entering and propagating in the local heterogeneous region of interest. To improve the computational efficiency, we successfully combine the TSOS method with the frequency-wavenumber (FK) method and apply the ADE CFS-PML to absorb the scattered waves caused by the regional heterogeneity. The TSOS-FK hybrid method is benchmarked against semi-analytical solutions provided by the FK method for a 1-D layered model. Several numerical experiments, including a vertical cross-section of the Chinese capital area crustal model, illustrate that the TSOS-FK hybrid method works well for modelling waves propagating in complex heterogeneous media and remains stable for long-time computation. These numerical examples also show that the TSOS-FK method can tackle the converted and scattered waves of the teleseismic plane waves caused by local heterogeneity. Thus, the TSOS and TSOS-FK methods proposed in this study present an essential tool for the joint inversion of local, regional, and teleseismic waveform data.

Interstation phase speed and amplitude measurements of surface waves with nonlinear waveform fitting: application to USArray

NASA Astrophysics Data System (ADS)

Hamada, K.; Yoshizawa, K.

2015-09-01

A new method of fully nonlinear waveform fitting to measure interstation phase speeds and amplitude ratios is developed and applied to USArray. The Neighbourhood Algorithm is used as a global optimizer, which efficiently searches for model parameters that fit two observed waveforms on a common great-circle path by modulating the phase and amplitude terms of the fundamental-mode surface waves. We introduce the reliability parameter that represents how well the waveforms at two stations can be fitted in a time-frequency domain, which is used as a data selection criterion. The method is applied to observed waveforms of USArray for seismic events in the period from 2007 to 2010 with moment magnitude greater than 6.0. We collect a large number of phase speed data (about 75 000 for Rayleigh and 20 000 for Love) and amplitude ratio data (about 15 000 for Rayleigh waves) in a period range from 30 to 130 s. The majority of the interstation distances of measured dispersion data is less than 1000 km, which is much shorter than the typical average path-length of the conventional single-station measurements for source-receiver pairs. The phase speed models for Rayleigh and Love waves show good correlations on large scales with the recent tomographic maps derived from different approaches for phase speed mapping; for example, significant slow anomalies in volcanic regions in the western Unites States and fast anomalies in the cratonic region. Local-scale phase speed anomalies corresponding to the major tectonic features in the western United States, such as Snake River Plains, Basin and Range, Colorado Plateau and Rio Grande Rift have also been identified clearly in the phase speed models. The short-path information derived from our interstation measurements helps to increase the achievable horizontal resolution. We have also performed joint inversions for phase speed maps using the measured phase and amplitude ratio data of vertical component Rayleigh waves. These maps exhibit better recovery of phase speed perturbations, particularly where the strong lateral velocity gradient exists in which the effects of elastic focussing can be significant; that is, the Yellowstone hotspot, Snake River Plains, and Rio Grande Rift. The enhanced resolution of the phase speed models derived from the interstation phase and amplitude measurements will be of use for the better seismological constraint on the lithospheric structure, in combination with dense broad-band seismic arrays.

Tsunami Wave Height Estimation from GPS-Derived Ionospheric Data

NASA Astrophysics Data System (ADS)

Rakoto, Virgile; Lognonné, Philippe; Rolland, Lucie; Coïsson, P.

2018-05-01

Large underwater earthquakes (Mw>7) can transmit part of their energy to the surrounding ocean through large seafloor motions, generating tsunamis that propagate over long distances. The forcing effect of tsunami waves on the atmosphere generates internal gravity waves that, when they reach the upper atmosphere, produce ionospheric perturbations. These perturbations are frequently observed in the total electron content (TEC) measured by multifrequency Global Navigation Satellite Systems (GNSS) such as GPS, GLONASS, and, in the future, Galileo. This paper describes the first inversion of the variation in sea level derived from GPS TEC data. We used a least squares inversion through a normal-mode summation modeling. This technique was applied to three tsunamis in far field associated to the 2012 Haida Gwaii, 2006 Kuril Islands, and 2011 Tohoku events and for Tohoku also in close field. With the exception of the Tohoku far-field case, for which the tsunami reconstruction by the TEC inversion is less efficient due to the ionospheric noise background associated to geomagnetic storm, which occurred on the earthquake day, we show that the peak-to-peak amplitude of the sea level variation inverted by this method can be compared to the tsunami wave height measured by a DART buoy with an error of less than 20%. This demonstrates that the inversion of TEC data with a tsunami normal-mode summation approach is able to estimate quite accurately the amplitude and waveform of the first tsunami arrival.

Characterizing Geological Facies using Seismic Waveform Classification in Sarawak Basin

NASA Astrophysics Data System (ADS)

Zahraa, Afiqah; Zailani, Ahmad; Prasad Ghosh, Deva

2017-10-01

Numerous effort have been made to build relationship between geology and geophysics using different techniques throughout the years. The integration of these two most important data in oil and gas industry can be used to reduce uncertainty in exploration and production especially for reservoir productivity enhancement and stratigraphic identification. This paper is focusing on seismic waveform classification to different classes using neural network and to link them according to the geological facies which are established using the knowledge on lithology and log motif of well data. Seismic inversion is used as the input for the neural network to act as the direct lithology indicator reducing dependency on well calibration. The interpretation of seismic facies classification map provides a better understanding towards the lithology distribution, depositional environment and help to identify significant reservoir rock

The Crust and Upper Mantle Structure of the Iranian Plateau from Joint Waveform Tomography Imaging of Body and Surface Waves

NASA Astrophysics Data System (ADS)

Roecker, S. W.; Priestley, K. F.; Tatar, M.

2014-12-01

The Iranian Plateau forms a broad zone of deformation between the colliding Arabian and Eurasian plates. The convergence is accommodated in the Zagros Mountains of SW Iran, the Alborz Mountains of northern Iran, and the Kopeh Dagh Mountains of NE Iran. These deforming belts are separated by relatively aseismic depressions such as the Lut Block. It has been suggested that the Arabia-Eurasia collision is similar to the Indo-Eurasia collision but at a early point of development and therefore, it may provide clues to our understanding of the earlier stages of the continent-continent collision process. We present results of the analysis of seismic data collected along two NE-SW trending transects across the Iranian Plateau. The first profile extends from near Bushere on the Persian Gulf coast to near to the Iran-Turkmenistan border north of Mashad, and consists of seismic recordings along the SW portion of the line in 2000-2001 and recording along the NE portion of the line in 2003 and 2006-2008. The second profile extends from near the Iran-Iraq border near the Dezfel embayment to the south Caspian Sea coast north of Tehran. We apply the combined 2.5D finite element waveform tomography algorithm of Baker and Roecker [2014] to jointly invert teleseismic body and surface waves to determine the elastic wavespeed structures of these areas. The joint inversion of these different types of waves affords similar types of advantages that are common to combined surface wave dispersion/receiver function inversions in compensating for intrinsic weaknesses in horizontal and vertical resolution capabilities. We compare results recovered from a finite difference approach to document the effects of various assumptions related to their application, such as the inclusion of topography, on the models recovered. We also apply several different inverse methods, starting with simple gradient techniques to the more sophisticated pseudo-Hessian or L-BFGS approach, and find that the latter are generally more robust. Modeling of receiver functions and surface wave dispersion prior to the analysis is shown to be an efficacious way to generate starting models for this analysis.

First application of tsunami back-projection and source inversion for the 2012 Haida Gwaii earthquake using tsunami data recorded on a dense array of seafloor pressure gauges

NASA Astrophysics Data System (ADS)

Gusman, A. R.; Satake, K.; Sheehan, A. F.; Mulia, I. E.; Heidarzadeh, M.; Maeda, T.

2015-12-01

Adaption of absolute or differential pressure gauges (APG or DPG) to Ocean Bottom Seismometers has provided the opportunity to study tsunamis. Recently we extracted tsunami waveforms of the 28 October 2012 Haida Gwaii earthquake recoded by the APG and DPG of Cascadia Initiative program (Sheehan et al., 2015, SRL). We applied such dense tsunami observations (48 stations) together with other records from DARTs (9 stations) to characterize the tsunami source. This study is the first study that used such a large number of offshore tsunami records for earthquake source study. Conventionally the curves of tsunami travel times are drawn backward from station locations to estimate the tsunami source region. Here we propose a more advanced technique called tsunami back-projection to estimate the source region. Our image produced by tsunami back-projection has the largest value or tsunami centroid that is very close to the epicenter and above the Queen Charlotte transform fault (QCF), whereas the negative values are mostly located east of Haida Gwaii in the Hecate Strait. By using tsunami back-projection we avoid picking initial tsunami phase which is a necessary step in the conventional method that is rather subjective. The slip distribution of the 2012 Haida Gwaii earthquake estimated by tsunami waveform inversion shows large slip near the trench (4-5 m) and also on a plate interface southeast the epicenter (3-4 m) below QCF. From the slip distribution, the calculated seismic moment is 5.4 × 1020 N m (Mw 7.8). The steep bathymetry offshore Haida Gwaii and the horizontal movement caused by the earthquake possibly affects the sea surface deformation. The potential tsunami energy calculated from the sea-surface deformation of pure faulting is 2.20 × 1013 J, while that from the bathymetry effect is 0.12 × 1013 J or about 5% of the total potential energy. The significant deformation above the steep slope is confirmed by another tsunami inversion that disregards fault parameters.

The Rupture Characteristic of 1999 Izmit Sequence Using IRIS Data

NASA Astrophysics Data System (ADS)

Konca, A. O.; Helmberger, D. V.; Ji, C.; Tan, Y.

2003-12-01

The standard source studies use teleseismic data (30° to 90° ) to analyze earthquakes. Therefore, only a limited portion of the focal sphere is involved in source determinations. Furthermore, the locations and origin times of events remain incompatible with local determinations. Here, we attempt to resolve such issues by using IRIS data at all distances, leading to more accurate and detailed rupture properties and accurate relative locations. The 1999 Izmit earthquake sequence is chosen to test our method. The challenge of using data outside the conventional teleseismic distance range is that the arrival times and waveforms are affected more by the Earth structure. We overcome this difficulty by calibrating the path effects for the mainshock using the simpler aftershocks. Therefore, it is crucial to determine the source parameters of the aftershock. We constructed a Green's function library from a regionalized 1-D model and performed a grid search to establish the depth and fault parameters based on waveform matching for the Pnl waves between the synthetics and data, allowing the synthetics in each station to shift separately to account for the path effect. Our results show that the earthquake depth was around 7 km, rather than 19 km from local observatory (Kandilli) and 15 km from the Harvard's CMT solution. The best focal mechanism has a strike of 263° , a dip of 65° , and a rake of 180° , which is very close to the Harvard's CMT solution. The waveform fits of this aftershock is then used as a criterion to select useful source-station paths. A path with a cross-correlation value above 90% between data and synthetics is defined as a "good path" and can be used for studying the Izmit and Duzce earthquakes. We find that the stations in Central Europe and some of the Greek Islands are "good paths", while the stations in Northeast Africa and Italy cannot be used. The time shifts that give the best cross-correlation values are used to calibrate the picks of the Izmit and Duzce events. We realize that this is a very objective way to pick arrival times. However, our preliminary inversions using teleseismic data for Duzce and Izmit events show that handpicked P and S arrival times of the same station from two very close events are not always well correlated. Obviously, how we pick the arrival time governs the rupture pattern and rupture velocity. Therefore, our methodology brings a more objective approach to pick the travel times. To the end, we will invert for the source history of the Duzce and Izmit earthquakes with the regional data and compare with the inversion result using teleseismic data. Moreover, predictions of the teleseismic data, using the solution from the inversion using regional phases will be presented.

A High-Resolution View of Global Seismicity

NASA Astrophysics Data System (ADS)

Waldhauser, F.; Schaff, D. P.

2014-12-01

We present high-precision earthquake relocation results from our global-scale re-analysis of the combined seismic archives of parametric data for the years 1964 to present from the International Seismological Centre (ISC), the USGS's Earthquake Data Report (EDR), and selected waveform data from IRIS. We employed iterative, multistep relocation procedures that initially correct for large location errors present in standard global earthquake catalogs, followed by a simultaneous inversion of delay times formed from regional and teleseismic arrival times of first and later arriving phases. An efficient multi-scale double-difference (DD) algorithm is used to solve for relative event locations to the precision of a few km or less, while incorporating information on absolute hypocenter locations from catalogs such as EHB and GEM. We run the computations on both a 40-core cluster geared towards HTC problems (data processing) and a 500-core HPC cluster for data inversion. Currently, we are incorporating waveform correlation delay time measurements available for events in selected regions, but are continuously building up a comprehensive, global correlation database for densely distributed events recorded at stations with a long history of high-quality waveforms. The current global DD catalog includes nearly one million earthquakes, equivalent to approximately 70% of the number of events in the ISC/EDR catalogs initially selected for relocation. The relocations sharpen the view of seismicity in most active regions around the world, in particular along subduction zones where event density is high, but also along mid-ocean ridges where existing hypocenters are especially poorly located. The new data offers the opportunity to investigate earthquake processes and fault structures along entire plate boundaries at the ~km scale, and provides a common framework that facilitates analysis and comparisons of findings across different plate boundary systems.

Elastic and anelastic structure of the lowermost mantle beneath the Western Pacific from waveform inversion

NASA Astrophysics Data System (ADS)

Konishi, Kensuke; Fuji, Nobuaki; Deschamps, Frédéric

2017-03-01

We investigate the elastic and anelastic structure of the lowermost mantle at the western edge of the Pacific large low shear velocity province (LLSVP) by inverting a collection of S and ScS waveforms. The transverse component data were obtained from F-net for 31 deep earthquakes beneath Tonga and Fiji, filtered between 12.5 and 200 s. We observe a regional variation of S and ScS arrival times and amplitude ratios, according to which we divide our region of interest into three subregions. For each of these subregions, we then perform 1-D (depth-dependent) waveform inversions simultaneously for radial profiles of shear wave velocity (VS) and seismic quality factor (Q). Models for all three subregions show low VS and low Q structures from 2000 km depth down to the core-mantle boundary. We further find that VS and Q in the central subregion, sampling the Caroline plume, are substantially lower than in the surrounding regions, whatever the depth. In the central subregion, VS-anomalies with respect to PREM (dVS) and Q are about -2.5 per cent and 216 at a depth of 2850 km, and -0.6 per cent and 263 at a depth of 2000 km. By contrast, in the two other regions, dVS and Q are -2.2 per cent and 261 at a depth of 2850 km, and -0.3 per cent and 291 at a depth of 2000 km. At depths greater than ∼2500 km, these differences may indicate lateral variations in temperature of ∼100 K within the Pacific LLSVP. At shallower depths, they may be due to the temperature difference between the Caroline plume and its surroundings, and possibly to a small fraction of iron-rich material entrained by the plume.

The Collaborative Seismic Earth Model Project

NASA Astrophysics Data System (ADS)

Fichtner, A.; van Herwaarden, D. P.; Afanasiev, M.

2017-12-01

We present the first generation of the Collaborative Seismic Earth Model (CSEM). This effort is intended to address grand challenges in tomography that currently inhibit imaging the Earth's interior across the seismically accessible scales: [1] For decades to come, computational resources will remain insufficient for the exploitation of the full observable seismic bandwidth. [2] With the man power of individual research groups, only small fractions of available waveform data can be incorporated into seismic tomographies. [3] The limited incorporation of prior knowledge on 3D structure leads to slow progress and inefficient use of resources. The CSEM is a multi-scale model of global 3D Earth structure that evolves continuously through successive regional refinements. Taking the current state of the CSEM as initial model, these refinements are contributed by external collaborators, and used to advance the CSEM to the next state. This mode of operation allows the CSEM to [1] harness the distributed man and computing power of the community, [2] to make consistent use of prior knowledge, and [3] to combine different tomographic techniques, needed to cover the seismic data bandwidth. Furthermore, the CSEM has the potential to serve as a unified and accessible representation of tomographic Earth models. Generation 1 comprises around 15 regional tomographic refinements, computed with full-waveform inversion. These include continental-scale mantle models of North America, Australasia, Europe and the South Atlantic, as well as detailed regional models of the crust beneath the Iberian Peninsula and western Turkey. A global-scale full-waveform inversion ensures that regional refinements are consistent with whole-Earth structure. This first generation will serve as the basis for further automation and methodological improvements concerning validation and uncertainty quantification.

Laplace-Fourier-domain dispersion analysis of an average derivative optimal scheme for scalar-wave equation

NASA Astrophysics Data System (ADS)

Chen, Jing-Bo

2014-06-01

By using low-frequency components of the damped wavefield, Laplace-Fourier-domain full waveform inversion (FWI) can recover a long-wavelength velocity model from the original undamped seismic data lacking low-frequency information. Laplace-Fourier-domain modelling is an important foundation of Laplace-Fourier-domain FWI. Based on the numerical phase velocity and the numerical attenuation propagation velocity, a method for performing Laplace-Fourier-domain numerical dispersion analysis is developed in this paper. This method is applied to an average-derivative optimal scheme. The results show that within the relative error of 1 per cent, the Laplace-Fourier-domain average-derivative optimal scheme requires seven gridpoints per smallest wavelength and smallest pseudo-wavelength for both equal and unequal directional sampling intervals. In contrast, the classical five-point scheme requires 23 gridpoints per smallest wavelength and smallest pseudo-wavelength to achieve the same accuracy. Numerical experiments demonstrate the theoretical analysis.

Gaussian Decomposition of Laser Altimeter Waveforms

NASA Technical Reports Server (NTRS)

Hofton, Michelle A.; Minster, J. Bernard; Blair, J. Bryan

1999-01-01

We develop a method to decompose a laser altimeter return waveform into its Gaussian components assuming that the position of each Gaussian within the waveform can be used to calculate the mean elevation of a specific reflecting surface within the laser footprint. We estimate the number of Gaussian components from the number of inflection points of a smoothed copy of the laser waveform, and obtain initial estimates of the Gaussian half-widths and positions from the positions of its consecutive inflection points. Initial amplitude estimates are obtained using a non-negative least-squares method. To reduce the likelihood of fitting the background noise within the waveform and to minimize the number of Gaussians needed in the approximation, we rank the "importance" of each Gaussian in the decomposition using its initial half-width and amplitude estimates. The initial parameter estimates of all Gaussians ranked "important" are optimized using the Levenburg-Marquardt method. If the sum of the Gaussians does not approximate the return waveform to a prescribed accuracy, then additional Gaussians are included in the optimization procedure. The Gaussian decomposition method is demonstrated on data collected by the airborne Laser Vegetation Imaging Sensor (LVIS) in October 1997 over the Sequoia National Forest, California.

Variable Grid Traveltime Tomography for Near-surface Seismic Imaging

NASA Astrophysics Data System (ADS)

Cai, A.; Zhang, J.

2017-12-01

We present a new algorithm of traveltime tomography for imaging the subsurface with automated variable grids upon geological structures. The nonlinear traveltime tomography along with Tikhonov regularization using conjugate gradient method is a conventional method for near surface imaging. However, model regularization for any regular and even grids assumes uniform resolution. From geophysical point of view, long-wavelength and large scale structures can be reliably resolved, the details along geological boundaries are difficult to resolve. Therefore, we solve a traveltime tomography problem that automatically identifies large scale structures and aggregates grids within the structures for inversion. As a result, the number of velocity unknowns is reduced significantly, and inversion intends to resolve small-scale structures or the boundaries of large-scale structures. The approach is demonstrated by tests on both synthetic and field data. One synthetic model is a buried basalt model with one horizontal layer. Using the variable grid traveltime tomography, the resulted model is more accurate in top layer velocity, and basalt blocks, and leading to a less number of grids. The field data was collected in an oil field in China. The survey was performed in an area where the subsurface structures were predominantly layered. The data set includes 476 shots with a 10 meter spacing and 1735 receivers with a 10 meter spacing. The first-arrival traveltime of the seismogram is picked for tomography. The reciprocal errors of most shots are between 2ms and 6ms. The normal tomography results in fluctuations in layers and some artifacts in the velocity model. In comparison, the implementation of new method with proper threshold provides blocky model with resolved flat layer and less artifacts. Besides, the number of grids reduces from 205,656 to 4,930 and the inversion produces higher resolution due to less unknowns and relatively fine grids in small structures. The variable grid traveltime tomography provides an alternative imaging solution for blocky structures in the subsurface and builds a good starting model for waveform inversion and statics.

Time-marching multi-grid seismic tomography

NASA Astrophysics Data System (ADS)

Tong, P.; Yang, D.; Liu, Q.

2016-12-01

From the classic ray-based traveltime tomography to the state-of-the-art full waveform inversion, because of the nonlinearity of seismic inverse problems, a good starting model is essential for preventing the convergence of the objective function toward local minima. With a focus on building high-accuracy starting models, we propose the so-called time-marching multi-grid seismic tomography method in this study. The new seismic tomography scheme consists of a temporal time-marching approach and a spatial multi-grid strategy. We first divide the recording period of seismic data into a series of time windows. Sequentially, the subsurface properties in each time window are iteratively updated starting from the final model of the previous time window. There are at least two advantages of the time-marching approach: (1) the information included in the seismic data of previous time windows has been explored to build the starting models of later time windows; (2) seismic data of later time windows could provide extra information to refine the subsurface images. Within each time window, we use a multi-grid method to decompose the scale of the inverse problem. Specifically, the unknowns of the inverse problem are sampled on a coarse mesh to capture the macro-scale structure of the subsurface at the beginning. Because of the low dimensionality, it is much easier to reach the global minimum on a coarse mesh. After that, finer meshes are introduced to recover the micro-scale properties. That is to say, the subsurface model is iteratively updated on multi-grid in every time window. We expect that high-accuracy starting models should be generated for the second and later time windows. We will test this time-marching multi-grid method by using our newly developed eikonal-based traveltime tomography software package tomoQuake. Real application results in the 2016 Kumamoto earthquake (Mw 7.0) region in Japan will be demonstrated.

Seismic tomography of northeastern Tibetan Plateau from body wave arrival times and surface wave dispersion data

NASA Astrophysics Data System (ADS)

Fang, H.; Yao, H.; Zhang, H.

2017-12-01

Reliable crustal and upper mantle structure is important to understand expansion of material from the Tibetan plateau to its northeastern margin. Previous studies have used either ambient noise tomography or body wave traveltime tomography to obtain the crustal velocity models in northeastern Tibetan Plateau. However, clear differences appear in these models obtained using different datasets. Here we propose to jointly invert local and teleseismic body wave arrival times and surface wave dispersion data from ambient noise cross correlation to obtain a unified P and S wavespeed model of the crust and upper mantle in NE Tibetan Plateau. Following Fang et al. (2016), we adopt the direct inversion strategy for surface wave data (Fang et al., 2015), which eliminates the need to construct the phase/group velocity maps and allows the straightforward incorporation of surface wave dispersion data into the body wave inversion framework. For body wave data including both local and teleseismic arrival times, we use the fast marching method (Rawlinson et al., 2004) in order to trace multiple seismic phases simultaneously. The joint inversion method takes advantage of the complementary strengths of different data types, with local body wave data constraining more on the P than S wavespeed in the crust, surface wave data most sensitive to S wavespeed in the crust and upper mantle, teleseismic body wave data resolving the upper mantle structure. A series of synthetic tests will be used to show the robustness and superiority of the joint inversion method. Besides, the inverted model will be validated by waveform simulation and comparison with other studies, like receiver function imaging. The resultant P and S wavespeed models, as well as the derived Vp/Vs model, will be essential to understand the regional tectonics of the northeastern Tibetan Plateau, and to address the related geodynamic questions of the Tibetan Plateau formation and expansion.

A waveform diversity method for optimizing 3-d power depositions generated by ultrasound phased arrays.

PubMed

Zeng, Xiaozheng Jenny; Li, Jian; McGough, Robert J

2010-01-01

A waveform-diversity-based approach for 3-D tumor heating is compared to spot scanning for hyperthermia applications. The waveform diversity method determines the excitation signals applied to the phased array elements and produces a beam pattern that closely matches the desired power distribution. The optimization algorithm solves the covariance matrix of the excitation signals through semidefinite programming subject to a series of quadratic cost functions and constraints on the control points. A numerical example simulates a 1444-element spherical-section phased array that delivers heat to a 3-cm-diameter spherical tumor located 12 cm from the array aperture, and the results show that waveform diversity combined with mode scanning increases the heated volume within the tumor while simultaneously decreasing normal tissue heating. Whereas standard single focus and multiple focus methods are often associated with unwanted intervening tissue heating, the waveform diversity method combined with mode scanning shifts energy away from intervening tissues where hotspots otherwise accumulate to improve temperature localization in deep-seated tumors.

Effect of positive pulse charge waveforms on the energy efficiency of lead-acid traction cells

NASA Technical Reports Server (NTRS)

Smithrick, J. J.

1981-01-01

The effects of four different charge methods on the energy conversion efficiency of 300 ampere hour lead acid traction cells were investigated. Three of the methods were positive pulse charge waveforms; the fourth, a constant current method, was used as a baseline of comparison. The positive pulse charge waveforms were: 120 Hz full wave rectified sinusoidal; 120 Hz silicon controlled rectified; and 1 kHz square wave. The constant current charger was set at the time average pulse current of each pulse waveform, which was 150 amps. The energy efficiency does not include charger losses. The lead acid traction cells were charged to 70 percent of rated ampere hour capacity in each case. The results of charging the cells using the three different pulse charge waveforms indicate there was no significant difference in energy conversion efficiency when compared to constant current charging at the time average pulse current value.

3D frequency-domain finite-difference modeling of acoustic wave propagation

NASA Astrophysics Data System (ADS)

Operto, S.; Virieux, J.

2006-12-01

We present a 3D frequency-domain finite-difference method for acoustic wave propagation modeling. This method is developed as a tool to perform 3D frequency-domain full-waveform inversion of wide-angle seismic data. For wide-angle data, frequency-domain full-waveform inversion can be applied only to few discrete frequencies to develop reliable velocity model. Frequency-domain finite-difference (FD) modeling of wave propagation requires resolution of a huge sparse system of linear equations. If this system can be solved with a direct method, solutions for multiple sources can be computed efficiently once the underlying matrix has been factorized. The drawback of the direct method is the memory requirement resulting from the fill-in of the matrix during factorization. We assess in this study whether representative problems can be addressed in 3D geometry with such approach. We start from the velocity-stress formulation of the 3D acoustic wave equation. The spatial derivatives are discretized with second-order accurate staggered-grid stencil on different coordinate systems such that the axis span over as many directions as possible. Once the discrete equations were developed on each coordinate system, the particle velocity fields are eliminated from the first-order hyperbolic system (following the so-called parsimonious staggered-grid method) leading to second-order elliptic wave equations in pressure. The second-order wave equations discretized on each coordinate system are combined linearly to mitigate the numerical anisotropy. Secondly, grid dispersion is minimized by replacing the mass term at the collocation point by its weighted averaging over all the grid points of the stencil. Use of second-order accurate staggered- grid stencil allows to reduce the bandwidth of the matrix to be factorized. The final stencil incorporates 27 points. Absorbing conditions are PML. The system is solved using the parallel direct solver MUMPS developed for distributed-memory computers. The MUMPS solver is based on a multifrontal method for LU factorization. We used the METIS algorithm to perform re-ordering of the matrix coefficients before factorization. Four grid points per minimum wavelength is used for discretization. We applied our algorithm to the 3D SEG/EAGE synthetic onshore OVERTHRUST model of dimensions 20 x 20 x 4.65 km. The velocities range between 2 and 6 km/s. We performed the simulations using 192 processors with 2 Gbytes of RAM memory per processor. We performed simulations for the 5 Hz, 7 Hz and 10 Hz frequencies in some fractions of the OVERTHRUST model. The grid interval was 100 m, 75 m and 50 m respectively. The grid dimensions were 207x207x53, 275x218x71 and 409x109x102 respectively corresponding to 100, 80 and 25 percents of the model respectively. The time for factorization is 20 mn, 108 mn and 163 mn respectively. The time for resolution was 3.8, 9.3 and 10.3 s per source. The total memory used during factorization is 143, 384 and 449 Gbytes respectively. One can note the huge memory requirement for factorization and the efficiency of the direct method to compute solutions for a large number of sources. This highlights the respective drawback and merit of the frequency-domain approach with respect to the time- domain counterpart. These results show that 3D acoustic frequency-domain wave propagation modeling can be performed at low frequencies using direct solver on large clusters of Pcs. This forward modeling algorithm may be used in the future as a tool to image the first kilometers of the crust by frequency-domain full-waveform inversion. For larger problems, we will use the out-of-core memory during factorization that has been implemented by the authors of MUMPS.