Accounting for model error in Bayesian solutions to hydrogeophysical inverse problems using a local basis approach

NASA Astrophysics Data System (ADS)

Köpke, Corinna; Irving, James; Elsheikh, Ahmed H.

2018-06-01

Bayesian solutions to geophysical and hydrological inverse problems are dependent upon a forward model linking subsurface physical properties to measured data, which is typically assumed to be perfectly known in the inversion procedure. However, to make the stochastic solution of the inverse problem computationally tractable using methods such as Markov-chain-Monte-Carlo (MCMC), fast approximations of the forward model are commonly employed. This gives rise to model error, which has the potential to significantly bias posterior statistics if not properly accounted for. Here, we present a new methodology for dealing with the model error arising from the use of approximate forward solvers in Bayesian solutions to hydrogeophysical inverse problems. Our approach is geared towards the common case where this error cannot be (i) effectively characterized through some parametric statistical distribution; or (ii) estimated by interpolating between a small number of computed model-error realizations. To this end, we focus on identification and removal of the model-error component of the residual during MCMC using a projection-based approach, whereby the orthogonal basis employed for the projection is derived in each iteration from the K-nearest-neighboring entries in a model-error dictionary. The latter is constructed during the inversion and grows at a specified rate as the iterations proceed. We demonstrate the performance of our technique on the inversion of synthetic crosshole ground-penetrating radar travel-time data considering three different subsurface parameterizations of varying complexity. Synthetic data are generated using the eikonal equation, whereas a straight-ray forward model is assumed for their inversion. In each case, our developed approach enables us to remove posterior bias and obtain a more realistic characterization of uncertainty.

Probabilistic numerical methods for PDE-constrained Bayesian inverse problems

NASA Astrophysics Data System (ADS)

Cockayne, Jon; Oates, Chris; Sullivan, Tim; Girolami, Mark

2017-06-01

This paper develops meshless methods for probabilistically describing discretisation error in the numerical solution of partial differential equations. This construction enables the solution of Bayesian inverse problems while accounting for the impact of the discretisation of the forward problem. In particular, this drives statistical inferences to be more conservative in the presence of significant solver error. Theoretical results are presented describing rates of convergence for the posteriors in both the forward and inverse problems. This method is tested on a challenging inverse problem with a nonlinear forward model.

Estimation of biological parameters of marine organisms using linear and nonlinear acoustic scattering model-based inversion methods.

PubMed

Chu, Dezhang; Lawson, Gareth L; Wiebe, Peter H

2016-05-01

The linear inversion commonly used in fisheries and zooplankton acoustics assumes a constant inversion kernel and ignores the uncertainties associated with the shape and behavior of the scattering targets, as well as other relevant animal parameters. Here, errors of the linear inversion due to uncertainty associated with the inversion kernel are quantified. A scattering model-based nonlinear inversion method is presented that takes into account the nonlinearity of the inverse problem and is able to estimate simultaneously animal abundance and the parameters associated with the scattering model inherent to the kernel. It uses sophisticated scattering models to estimate first, the abundance, and second, the relevant shape and behavioral parameters of the target organisms. Numerical simulations demonstrate that the abundance, size, and behavior (tilt angle) parameters of marine animals (fish or zooplankton) can be accurately inferred from the inversion by using multi-frequency acoustic data. The influence of the singularity and uncertainty in the inversion kernel on the inversion results can be mitigated by examining the singular values for linear inverse problems and employing a non-linear inversion involving a scattering model-based kernel.

An inverse finance problem for estimation of the volatility

NASA Astrophysics Data System (ADS)

Neisy, A.; Salmani, K.

2013-01-01

Black-Scholes model, as a base model for pricing in derivatives markets has some deficiencies, such as ignoring market jumps, and considering market volatility as a constant factor. In this article, we introduce a pricing model for European-Options under jump-diffusion underlying asset. Then, using some appropriate numerical methods we try to solve this model with integral term, and terms including derivative. Finally, considering volatility as an unknown parameter, we try to estimate it by using our proposed model. For the purpose of estimating volatility, in this article, we utilize inverse problem, in which inverse problem model is first defined, and then volatility is estimated using minimization function with Tikhonov regularization.

Appraisal of geodynamic inversion results: a data mining approach

NASA Astrophysics Data System (ADS)

Baumann, T. S.

2016-11-01

Bayesian sampling based inversions require many thousands or even millions of forward models, depending on how nonlinear or non-unique the inverse problem is, and how many unknowns are involved. The result of such a probabilistic inversion is not a single `best-fit' model, but rather a probability distribution that is represented by the entire model ensemble. Often, a geophysical inverse problem is non-unique, and the corresponding posterior distribution is multimodal, meaning that the distribution consists of clusters with similar models that represent the observations equally well. In these cases, we would like to visualize the characteristic model properties within each of these clusters of models. However, even for a moderate number of inversion parameters, a manual appraisal for a large number of models is not feasible. This poses the question whether it is possible to extract end-member models that represent each of the best-fit regions including their uncertainties. Here, I show how a machine learning tool can be used to characterize end-member models, including their uncertainties, from a complete model ensemble that represents a posterior probability distribution. The model ensemble used here results from a nonlinear geodynamic inverse problem, where rheological properties of the lithosphere are constrained from multiple geophysical observations. It is demonstrated that by taking vertical cross-sections through the effective viscosity structure of each of the models, the entire model ensemble can be classified into four end-member model categories that have a similar effective viscosity structure. These classification results are helpful to explore the non-uniqueness of the inverse problem and can be used to compute representative data fits for each of the end-member models. Conversely, these insights also reveal how new observational constraints could reduce the non-uniqueness. The method is not limited to geodynamic applications and a generalized MATLAB code is provided to perform the appraisal analysis.

A systematic linear space approach to solving partially described inverse eigenvalue problems

NASA Astrophysics Data System (ADS)

Hu, Sau-Lon James; Li, Haujun

2008-06-01

Most applications of the inverse eigenvalue problem (IEP), which concerns the reconstruction of a matrix from prescribed spectral data, are associated with special classes of structured matrices. Solving the IEP requires one to satisfy both the spectral constraint and the structural constraint. If the spectral constraint consists of only one or few prescribed eigenpairs, this kind of inverse problem has been referred to as the partially described inverse eigenvalue problem (PDIEP). This paper develops an efficient, general and systematic approach to solve the PDIEP. Basically, the approach, applicable to various structured matrices, converts the PDIEP into an ordinary inverse problem that is formulated as a set of simultaneous linear equations. While solving simultaneous linear equations for model parameters, the singular value decomposition method is applied. Because of the conversion to an ordinary inverse problem, other constraints associated with the model parameters can be easily incorporated into the solution procedure. The detailed derivation and numerical examples to implement the newly developed approach to symmetric Toeplitz and quadratic pencil (including mass, damping and stiffness matrices of a linear dynamic system) PDIEPs are presented. Excellent numerical results for both kinds of problem are achieved under the situations that have either unique or infinitely many solutions.

Mixed linear-non-linear inversion of crustal deformation data: Bayesian inference of model, weighting and regularization parameters

NASA Astrophysics Data System (ADS)

Fukuda, Jun'ichi; Johnson, Kaj M.

2010-06-01

We present a unified theoretical framework and solution method for probabilistic, Bayesian inversions of crustal deformation data. The inversions involve multiple data sets with unknown relative weights, model parameters that are related linearly or non-linearly through theoretic models to observations, prior information on model parameters and regularization priors to stabilize underdetermined problems. To efficiently handle non-linear inversions in which some of the model parameters are linearly related to the observations, this method combines both analytical least-squares solutions and a Monte Carlo sampling technique. In this method, model parameters that are linearly and non-linearly related to observations, relative weights of multiple data sets and relative weights of prior information and regularization priors are determined in a unified Bayesian framework. In this paper, we define the mixed linear-non-linear inverse problem, outline the theoretical basis for the method, provide a step-by-step algorithm for the inversion, validate the inversion method using synthetic data and apply the method to two real data sets. We apply the method to inversions of multiple geodetic data sets with unknown relative data weights for interseismic fault slip and locking depth. We also apply the method to the problem of estimating the spatial distribution of coseismic slip on faults with unknown fault geometry, relative data weights and smoothing regularization weight.

Frechet derivatives for shallow water ocean acoustic inverse problems

NASA Astrophysics Data System (ADS)

Odom, Robert I.

2003-04-01

For any inverse problem, finding a model fitting the data is only half the problem. Most inverse problems of interest in ocean acoustics yield nonunique model solutions, and involve inevitable trade-offs between model and data resolution and variance. Problems of uniqueness and resolution and variance trade-offs can be addressed by examining the Frechet derivatives of the model-data functional with respect to the model variables. Tarantola [Inverse Problem Theory (Elsevier, Amsterdam, 1987), p. 613] published analytical formulas for the basic derivatives, e.g., derivatives of pressure with respect to elastic moduli and density. Other derivatives of interest, such as the derivative of transmission loss with respect to attenuation, can be easily constructed using the chain rule. For a range independent medium the analytical formulas involve only the Green's function and the vertical derivative of the Green's function for the medium. A crucial advantage of the analytical formulas for the Frechet derivatives over numerical differencing is that they can be computed with a single pass of any program which supplies the Green's function. Various derivatives of interest in shallow water ocean acoustics are presented and illustrated by an application to the sensitivity of measured pressure to shallow water sediment properties. [Work supported by ONR.

The neural network approximation method for solving multidimensional nonlinear inverse problems of geophysics

NASA Astrophysics Data System (ADS)

Shimelevich, M. I.; Obornev, E. A.; Obornev, I. E.; Rodionov, E. A.

2017-07-01

The iterative approximation neural network method for solving conditionally well-posed nonlinear inverse problems of geophysics is presented. The method is based on the neural network approximation of the inverse operator. The inverse problem is solved in the class of grid (block) models of the medium on a regularized parameterization grid. The construction principle of this grid relies on using the calculated values of the continuity modulus of the inverse operator and its modifications determining the degree of ambiguity of the solutions. The method provides approximate solutions of inverse problems with the maximal degree of detail given the specified degree of ambiguity with the total number of the sought parameters n × 103 of the medium. The a priori and a posteriori estimates of the degree of ambiguity of the approximated solutions are calculated. The work of the method is illustrated by the example of the three-dimensional (3D) inversion of the synthesized 2D areal geoelectrical (audio magnetotelluric sounding, AMTS) data corresponding to the schematic model of a kimberlite pipe.

A geostatistical approach to the change-of-support problem and variable-support data fusion in spatial analysis

NASA Astrophysics Data System (ADS)

Wang, Jun; Wang, Yang; Zeng, Hui

2016-01-01

A key issue to address in synthesizing spatial data with variable-support in spatial analysis and modeling is the change-of-support problem. We present an approach for solving the change-of-support and variable-support data fusion problems. This approach is based on geostatistical inverse modeling that explicitly accounts for differences in spatial support. The inverse model is applied here to produce both the best predictions of a target support and prediction uncertainties, based on one or more measurements, while honoring measurements. Spatial data covering large geographic areas often exhibit spatial nonstationarity and can lead to computational challenge due to the large data size. We developed a local-window geostatistical inverse modeling approach to accommodate these issues of spatial nonstationarity and alleviate computational burden. We conducted experiments using synthetic and real-world raster data. Synthetic data were generated and aggregated to multiple supports and downscaled back to the original support to analyze the accuracy of spatial predictions and the correctness of prediction uncertainties. Similar experiments were conducted for real-world raster data. Real-world data with variable-support were statistically fused to produce single-support predictions and associated uncertainties. The modeling results demonstrate that geostatistical inverse modeling can produce accurate predictions and associated prediction uncertainties. It is shown that the local-window geostatistical inverse modeling approach suggested offers a practical way to solve the well-known change-of-support problem and variable-support data fusion problem in spatial analysis and modeling.

Joint Geophysical Inversion With Multi-Objective Global Optimization Methods

NASA Astrophysics Data System (ADS)

Lelievre, P. G.; Bijani, R.; Farquharson, C. G.

2015-12-01

Pareto multi-objective global optimization (PMOGO) methods generate a suite of solutions that minimize multiple objectives (e.g. data misfits and regularization terms) in a Pareto-optimal sense. Providing a suite of models, as opposed to a single model that minimizes a weighted sum of objectives, allows a more complete assessment of the possibilities and avoids the often difficult choice of how to weight each objective. We are applying PMOGO methods to three classes of inverse problems. The first class are standard mesh-based problems where the physical property values in each cell are treated as continuous variables. The second class of problems are also mesh-based but cells can only take discrete physical property values corresponding to known or assumed rock units. In the third class we consider a fundamentally different type of inversion in which a model comprises wireframe surfaces representing contacts between rock units; the physical properties of each rock unit remain fixed while the inversion controls the position of the contact surfaces via control nodes. This third class of problem is essentially a geometry inversion, which can be used to recover the unknown geometry of a target body or to investigate the viability of a proposed Earth model. Joint inversion is greatly simplified for the latter two problem classes because no additional mathematical coupling measure is required in the objective function. PMOGO methods can solve numerically complicated problems that could not be solved with standard descent-based local minimization methods. This includes the latter two classes of problems mentioned above. There are significant increases in the computational requirements when PMOGO methods are used but these can be ameliorated using parallelization and problem dimension reduction strategies.

Inverse problems in the modeling of vibrations of flexible beams

NASA Technical Reports Server (NTRS)

Banks, H. T.; Powers, R. K.; Rosen, I. G.

1987-01-01

The formulation and solution of inverse problems for the estimation of parameters which describe damping and other dynamic properties in distributed models for the vibration of flexible structures is considered. Motivated by a slewing beam experiment, the identification of a nonlinear velocity dependent term which models air drag damping in the Euler-Bernoulli equation is investigated. Galerkin techniques are used to generate finite dimensional approximations. Convergence estimates and numerical results are given. The modeling of, and related inverse problems for the dynamics of a high pressure hose line feeding a gas thruster actuator at the tip of a cantilevered beam are then considered. Approximation and convergence are discussed and numerical results involving experimental data are presented.

Modular Approaches to Earth Science Scientific Computing: 3D Electromagnetic Induction Modeling as an Example

NASA Astrophysics Data System (ADS)

Tandon, K.; Egbert, G.; Siripunvaraporn, W.

2003-12-01

We are developing a modular system for three-dimensional inversion of electromagnetic (EM) induction data, using an object oriented programming approach. This approach allows us to modify the individual components of the inversion scheme proposed, and also reuse the components for variety of problems in earth science computing howsoever diverse they might be. In particular, the modularity allows us to (a) change modeling codes independently of inversion algorithm details; (b) experiment with new inversion algorithms; and (c) modify the way prior information is imposed in the inversion to test competing hypothesis and techniques required to solve an earth science problem. Our initial code development is for EM induction equations on a staggered grid, using iterative solution techniques in 3D. An example illustrated here is an experiment with the sensitivity of 3D magnetotelluric inversion to uncertainties in the boundary conditions required for regional induction problems. These boundary conditions should reflect the large-scale geoelectric structure of the study area, which is usually poorly constrained. In general for inversion of MT data, one fixes boundary conditions at the edge of the model domain, and adjusts the earth?s conductivity structure within the modeling domain. Allowing for errors in specification of the open boundary values is simple in principle, but no existing inversion codes that we are aware of have this feature. Adding a feature such as this is straightforward within the context of the modular approach. More generally, a modular approach provides an efficient methodology for setting up earth science computing problems to test various ideas. As a concrete illustration relevant to EM induction problems, we investigate the sensitivity of MT data near San Andreas Fault at Parkfield (California) to uncertainties in the regional geoelectric structure.

Rapid processing of data based on high-performance algorithms for solving inverse problems and 3D-simulation of the tsunami and earthquakes

NASA Astrophysics Data System (ADS)

Marinin, I. V.; Kabanikhin, S. I.; Krivorotko, O. I.; Karas, A.; Khidasheli, D. G.

2012-04-01

We consider new techniques and methods for earthquake and tsunami related problems, particularly - inverse problems for the determination of tsunami source parameters, numerical simulation of long wave propagation in soil and water and tsunami risk estimations. In addition, we will touch upon the issue of database management and destruction scenario visualization. New approaches and strategies, as well as mathematical tools and software are to be shown. The long joint investigations by researchers of the Institute of Mathematical Geophysics and Computational Mathematics SB RAS and specialists from WAPMERR and Informap have produced special theoretical approaches, numerical methods, and software tsunami and earthquake modeling (modeling of propagation and run-up of tsunami waves on coastal areas), visualization, risk estimation of tsunami, and earthquakes. Algorithms are developed for the operational definition of the origin and forms of the tsunami source. The system TSS numerically simulates the source of tsunami and/or earthquakes and includes the possibility to solve the direct and the inverse problem. It becomes possible to involve advanced mathematical results to improve models and to increase the resolution of inverse problems. Via TSS one can construct maps of risks, the online scenario of disasters, estimation of potential damage to buildings and roads. One of the main tools for the numerical modeling is the finite volume method (FVM), which allows us to achieve stability with respect to possible input errors, as well as to achieve optimum computing speed. Our approach to the inverse problem of tsunami and earthquake determination is based on recent theoretical results concerning the Dirichlet problem for the wave equation. This problem is intrinsically ill-posed. We use the optimization approach to solve this problem and SVD-analysis to estimate the degree of ill-posedness and to find the quasi-solution. The software system we developed is intended to create technology «no frost», realizing a steady stream of direct and inverse problems: solving the direct problem, the visualization and comparison with observed data, to solve the inverse problem (correction of the model parameters). The main objective of further work is the creation of a workstation operating emergency tool that could be used by an emergency duty person in real time.

Distorted Born iterative T-matrix method for inversion of CSEM data in anisotropic media

NASA Astrophysics Data System (ADS)

Jakobsen, Morten; Tveit, Svenn

2018-05-01

We present a direct iterative solutions to the nonlinear controlled-source electromagnetic (CSEM) inversion problem in the frequency domain, which is based on a volume integral equation formulation of the forward modelling problem in anisotropic conductive media. Our vectorial nonlinear inverse scattering approach effectively replaces an ill-posed nonlinear inverse problem with a series of linear ill-posed inverse problems, for which there already exist efficient (regularized) solution methods. The solution update the dyadic Green's function's from the source to the scattering-volume and from the scattering-volume to the receivers, after each iteration. The T-matrix approach of multiple scattering theory is used for efficient updating of all dyadic Green's functions after each linearized inversion step. This means that we have developed a T-matrix variant of the Distorted Born Iterative (DBI) method, which is often used in the acoustic and electromagnetic (medical) imaging communities as an alternative to contrast-source inversion. The main advantage of using the T-matrix approach in this context, is that it eliminates the need to perform a full forward simulation at each iteration of the DBI method, which is known to be consistent with the Gauss-Newton method. The T-matrix allows for a natural domain decomposition, since in the sense that a large model can be decomposed into an arbitrary number of domains that can be treated independently and in parallel. The T-matrix we use for efficient model updating is also independent of the source-receiver configuration, which could be an advantage when performing fast-repeat modelling and time-lapse inversion. The T-matrix is also compatible with the use of modern renormalization methods that can potentially help us to reduce the sensitivity of the CSEM inversion results on the starting model. To illustrate the performance and potential of our T-matrix variant of the DBI method for CSEM inversion, we performed a numerical experiments based on synthetic CSEM data associated with 2D VTI and 3D orthorombic model inversions. The results of our numerical experiment suggest that the DBIT method for inversion of CSEM data in anisotropic media is both accurate and efficient.

FOREWORD: 5th International Workshop on New Computational Methods for Inverse Problems

NASA Astrophysics Data System (ADS)

Vourc'h, Eric; Rodet, Thomas

2015-11-01

This volume of Journal of Physics: Conference Series is dedicated to the scientific research presented during the 5th International Workshop on New Computational Methods for Inverse Problems, NCMIP 2015 (http://complement.farman.ens-cachan.fr/NCMIP_2015.html). This workshop took place at Ecole Normale Supérieure de Cachan, on May 29, 2015. The prior editions of NCMIP also took place in Cachan, France, firstly within the scope of ValueTools Conference, in May 2011, and secondly at the initiative of Institut Farman, in May 2012, May 2013 and May 2014. The New Computational Methods for Inverse Problems (NCMIP) workshop focused on recent advances in the resolution of inverse problems. Indeed, inverse problems appear in numerous scientific areas such as geophysics, biological and medical imaging, material and structure characterization, electrical, mechanical and civil engineering, and finances. The resolution of inverse problems consists of estimating the parameters of the observed system or structure from data collected by an instrumental sensing or imaging device. Its success firstly requires the collection of relevant observation data. It also requires accurate models describing the physical interactions between the instrumental device and the observed system, as well as the intrinsic properties of the solution itself. Finally, it requires the design of robust, accurate and efficient inversion algorithms. Advanced sensor arrays and imaging devices provide high rate and high volume data; in this context, the efficient resolution of the inverse problem requires the joint development of new models and inversion methods, taking computational and implementation aspects into account. During this one-day workshop, researchers had the opportunity to bring to light and share new techniques and results in the field of inverse problems. The topics of the workshop were: algorithms and computational aspects of inversion, Bayesian estimation, Kernel methods, learning methods, convex optimization, free discontinuity problems, metamodels, proper orthogonal decomposition, reduced models for the inversion, non-linear inverse scattering, image reconstruction and restoration, and applications (bio-medical imaging, non-destructive evaluation...). NCMIP 2015 was a one-day workshop held in May 2015 which attracted around 70 attendees. Each of the submitted papers has been reviewed by two reviewers. There have been 15 accepted papers. In addition, three international speakers were invited to present a longer talk. The workshop was supported by Institut Farman (ENS Cachan, CNRS) and endorsed by the following French research networks: GDR ISIS, GDR MIA, GDR MOA and GDR Ondes. The program committee acknowledges the following research laboratories: CMLA, LMT, LURPA and SATIE.

Direct and inverse problems of studying the properties of multilayer nanostructures based on a two-dimensional model of X-ray reflection and scattering

NASA Astrophysics Data System (ADS)

Khachaturov, R. V.

2014-06-01

A mathematical model of X-ray reflection and scattering by multilayered nanostructures in the quasi-optical approximation is proposed. X-ray propagation and the electric field distribution inside the multilayered structure are considered with allowance for refraction, which is taken into account via the second derivative with respect to the depth of the structure. This model is used to demonstrate the possibility of solving inverse problems in order to determine the characteristics of irregularities not only over the depth (as in the one-dimensional problem) but also over the length of the structure. An approximate combinatorial method for system decomposition and composition is proposed for solving the inverse problems.

Computational inverse methods of heat source in fatigue damage problems

NASA Astrophysics Data System (ADS)

Chen, Aizhou; Li, Yuan; Yan, Bo

2018-04-01

Fatigue dissipation energy is the research focus in field of fatigue damage at present. It is a new idea to solve the problem of calculating fatigue dissipation energy by introducing inverse method of heat source into parameter identification of fatigue dissipation energy model. This paper introduces the research advances on computational inverse method of heat source and regularization technique to solve inverse problem, as well as the existing heat source solution method in fatigue process, prospects inverse method of heat source applying in fatigue damage field, lays the foundation for further improving the effectiveness of fatigue dissipation energy rapid prediction.

A Computationally Efficient Parallel Levenberg-Marquardt Algorithm for Large-Scale Big-Data Inversion

NASA Astrophysics Data System (ADS)

Lin, Y.; O'Malley, D.; Vesselinov, V. V.

2015-12-01

Inverse modeling seeks model parameters given a set of observed state variables. However, for many practical problems due to the facts that the observed data sets are often large and model parameters are often numerous, conventional methods for solving the inverse modeling can be computationally expensive. We have developed a new, computationally-efficient Levenberg-Marquardt method for solving large-scale inverse modeling. Levenberg-Marquardt methods require the solution of a dense linear system of equations which can be prohibitively expensive to compute for large-scale inverse problems. Our novel method projects the original large-scale linear problem down to a Krylov subspace, such that the dimensionality of the measurements can be significantly reduced. Furthermore, instead of solving the linear system for every Levenberg-Marquardt damping parameter, we store the Krylov subspace computed when solving the first damping parameter and recycle it for all the following damping parameters. The efficiency of our new inverse modeling algorithm is significantly improved by using these computational techniques. We apply this new inverse modeling method to invert for a random transitivity field. Our algorithm is fast enough to solve for the distributed model parameters (transitivity) at each computational node in the model domain. The inversion is also aided by the use regularization techniques. The algorithm is coded in Julia and implemented in the MADS computational framework (http://mads.lanl.gov). Julia is an advanced high-level scientific programing language that allows for efficient memory management and utilization of high-performance computational resources. By comparing with a Levenberg-Marquardt method using standard linear inversion techniques, our Levenberg-Marquardt method yields speed-up ratio of 15 in a multi-core computational environment and a speed-up ratio of 45 in a single-core computational environment. Therefore, our new inverse modeling method is a powerful tool for large-scale applications.

Fractional Gaussian model in global optimization

NASA Astrophysics Data System (ADS)

Dimri, V. P.; Srivastava, R. P.

2009-12-01

Earth system is inherently non-linear and it can be characterized well if we incorporate no-linearity in the formulation and solution of the problem. General tool often used for characterization of the earth system is inversion. Traditionally inverse problems are solved using least-square based inversion by linearizing the formulation. The initial model in such inversion schemes is often assumed to follow posterior Gaussian probability distribution. It is now well established that most of the physical properties of the earth follow power law (fractal distribution). Thus, the selection of initial model based on power law probability distribution will provide more realistic solution. We present a new method which can draw samples of posterior probability density function very efficiently using fractal based statistics. The application of the method has been demonstrated to invert band limited seismic data with well control. We used fractal based probability density function which uses mean, variance and Hurst coefficient of the model space to draw initial model. Further this initial model is used in global optimization inversion scheme. Inversion results using initial models generated by our method gives high resolution estimates of the model parameters than the hitherto used gradient based liner inversion method.

Two-dimensional probabilistic inversion of plane-wave electromagnetic data: methodology, model constraints and joint inversion with electrical resistivity data

NASA Astrophysics Data System (ADS)

Rosas-Carbajal, Marina; Linde, Niklas; Kalscheuer, Thomas; Vrugt, Jasper A.

2014-03-01

Probabilistic inversion methods based on Markov chain Monte Carlo (MCMC) simulation are well suited to quantify parameter and model uncertainty of nonlinear inverse problems. Yet, application of such methods to CPU-intensive forward models can be a daunting task, particularly if the parameter space is high dimensional. Here, we present a 2-D pixel-based MCMC inversion of plane-wave electromagnetic (EM) data. Using synthetic data, we investigate how model parameter uncertainty depends on model structure constraints using different norms of the likelihood function and the model constraints, and study the added benefits of joint inversion of EM and electrical resistivity tomography (ERT) data. Our results demonstrate that model structure constraints are necessary to stabilize the MCMC inversion results of a highly discretized model. These constraints decrease model parameter uncertainty and facilitate model interpretation. A drawback is that these constraints may lead to posterior distributions that do not fully include the true underlying model, because some of its features exhibit a low sensitivity to the EM data, and hence are difficult to resolve. This problem can be partly mitigated if the plane-wave EM data is augmented with ERT observations. The hierarchical Bayesian inverse formulation introduced and used herein is able to successfully recover the probabilistic properties of the measurement data errors and a model regularization weight. Application of the proposed inversion methodology to field data from an aquifer demonstrates that the posterior mean model realization is very similar to that derived from a deterministic inversion with similar model constraints.

Geostatistical regularization operators for geophysical inverse problems on irregular meshes

NASA Astrophysics Data System (ADS)

Jordi, C.; Doetsch, J.; Günther, T.; Schmelzbach, C.; Robertsson, J. OA

2018-05-01

Irregular meshes allow to include complicated subsurface structures into geophysical modelling and inverse problems. The non-uniqueness of these inverse problems requires appropriate regularization that can incorporate a priori information. However, defining regularization operators for irregular discretizations is not trivial. Different schemes for calculating smoothness operators on irregular meshes have been proposed. In contrast to classical regularization constraints that are only defined using the nearest neighbours of a cell, geostatistical operators include a larger neighbourhood around a particular cell. A correlation model defines the extent of the neighbourhood and allows to incorporate information about geological structures. We propose an approach to calculate geostatistical operators for inverse problems on irregular meshes by eigendecomposition of a covariance matrix that contains the a priori geological information. Using our approach, the calculation of the operator matrix becomes tractable for 3-D inverse problems on irregular meshes. We tested the performance of the geostatistical regularization operators and compared them against the results of anisotropic smoothing in inversions of 2-D surface synthetic electrical resistivity tomography (ERT) data as well as in the inversion of a realistic 3-D cross-well synthetic ERT scenario. The inversions of 2-D ERT and seismic traveltime field data with geostatistical regularization provide results that are in good accordance with the expected geology and thus facilitate their interpretation. In particular, for layered structures the geostatistical regularization provides geologically more plausible results compared to the anisotropic smoothness constraints.

Identification of groundwater flow parameters using reciprocal data from hydraulic interference tests

NASA Astrophysics Data System (ADS)

Marinoni, Marianna; Delay, Frederick; Ackerer, Philippe; Riva, Monica; Guadagnini, Alberto

2016-08-01

We investigate the effect of considering reciprocal drawdown curves for the characterization of hydraulic properties of aquifer systems through inverse modeling based on interference well testing. Reciprocity implies that drawdown observed in a well B when pumping takes place from well A should strictly coincide with the drawdown observed in A when pumping in B with the same flow rate as in A. In this context, a critical point related to applications of hydraulic tomography is the assessment of the number of available independent drawdown data and their impact on the solution of the inverse problem. The issue arises when inverse modeling relies upon mathematical formulations of the classical single-continuum approach to flow in porous media grounded on Darcy's law. In these cases, introducing reciprocal drawdown curves in the database of an inverse problem is equivalent to duplicate some information, to a certain extent. We present a theoretical analysis of the way a Least-Square objective function and a Levenberg-Marquardt minimization algorithm are affected by the introduction of reciprocal information in the inverse problem. We also investigate the way these reciprocal data, eventually corrupted by measurement errors, influence model parameter identification in terms of: (a) the convergence of the inverse model, (b) the optimal values of parameter estimates, and (c) the associated estimation uncertainty. Our theoretical findings are exemplified through a suite of computational examples focused on block-heterogeneous systems with increased complexity level. We find that the introduction of noisy reciprocal information in the objective function of the inverse problem has a very limited influence on the optimal parameter estimates. Convergence of the inverse problem improves when adding diverse (nonreciprocal) drawdown series, but does not improve when reciprocal information is added to condition the flow model. The uncertainty on optimal parameter estimates is influenced by the strength of measurement errors and it is not significantly diminished or increased by adding noisy reciprocal information.

Identifing Atmospheric Pollutant Sources Using Artificial Neural Networks

NASA Astrophysics Data System (ADS)

Paes, F. F.; Campos, H. F.; Luz, E. P.; Carvalho, A. R.

2008-05-01

The estimation of the area source pollutant strength is a relevant issue for atmospheric environment. This characterizes an inverse problem in the atmospheric pollution dispersion. In the inverse analysis, an area source domain is considered, where the strength of such area source term is assumed unknown. The inverse problem is solved by using a supervised artificial neural network: multi-layer perceptron. The conection weights of the neural network are computed from delta rule - learning process. The neural network inversion is compared with results from standard inverse analysis (regularized inverse solution). In the regularization method, the inverse problem is formulated as a non-linear optimization approach, whose the objective function is given by the square difference between the measured pollutant concentration and the mathematical models, associated with a regularization operator. In our numerical experiments, the forward problem is addressed by a source-receptor scheme, where a regressive Lagrangian model is applied to compute the transition matrix. The second order maximum entropy regularization is used, and the regularization parameter is calculated by the L-curve technique. The objective function is minimized employing a deterministic scheme (a quasi-Newton algorithm) [1] and a stochastic technique (PSO: particle swarm optimization) [2]. The inverse problem methodology is tested with synthetic observational data, from six measurement points in the physical domain. The best inverse solutions were obtained with neural networks. References: [1] D. R. Roberti, D. Anfossi, H. F. Campos Velho, G. A. Degrazia (2005): Estimating Emission Rate and Pollutant Source Location, Ciencia e Natura, p. 131-134. [2] E.F.P. da Luz, H.F. de Campos Velho, J.C. Becceneri, D.R. Roberti (2007): Estimating Atmospheric Area Source Strength Through Particle Swarm Optimization. Inverse Problems, Desing and Optimization Symposium IPDO-2007, April 16-18, Miami (FL), USA, vol 1, p. 354-359.

The inverse electroencephalography pipeline

NASA Astrophysics Data System (ADS)

Weinstein, David Michael

The inverse electroencephalography (EEG) problem is defined as determining which regions of the brain are active based on remote measurements recorded with scalp EEG electrodes. An accurate solution to this problem would benefit both fundamental neuroscience research and clinical neuroscience applications. However, constructing accurate patient-specific inverse EEG solutions requires complex modeling, simulation, and visualization algorithms, and to date only a few systems have been developed that provide such capabilities. In this dissertation, a computational system for generating and investigating patient-specific inverse EEG solutions is introduced, and the requirements for each stage of this Inverse EEG Pipeline are defined and discussed. While the requirements of many of the stages are satisfied with existing algorithms, others have motivated research into novel modeling and simulation methods. The principal technical results of this work include novel surface-based volume modeling techniques, an efficient construction for the EEG lead field, and the Open Source release of the Inverse EEG Pipeline software for use by the bioelectric field research community. In this work, the Inverse EEG Pipeline is applied to three research problems in neurology: comparing focal and distributed source imaging algorithms; separating measurements into independent activation components for multifocal epilepsy; and localizing the cortical activity that produces the P300 effect in schizophrenia.

A practical method to assess model sensitivity and parameter uncertainty in C cycle models

NASA Astrophysics Data System (ADS)

Delahaies, Sylvain; Roulstone, Ian; Nichols, Nancy

2015-04-01

The carbon cycle combines multiple spatial and temporal scales, from minutes to hours for the chemical processes occurring in plant cells to several hundred of years for the exchange between the atmosphere and the deep ocean and finally to millennia for the formation of fossil fuels. Together with our knowledge of the transformation processes involved in the carbon cycle, many Earth Observation systems are now available to help improving models and predictions using inverse modelling techniques. A generic inverse problem consists in finding a n-dimensional state vector x such that h(x) = y, for a given N-dimensional observation vector y, including random noise, and a given model h. The problem is well posed if the three following conditions hold: 1) there exists a solution, 2) the solution is unique and 3) the solution depends continuously on the input data. If at least one of these conditions is violated the problem is said ill-posed. The inverse problem is often ill-posed, a regularization method is required to replace the original problem with a well posed problem and then a solution strategy amounts to 1) constructing a solution x, 2) assessing the validity of the solution, 3) characterizing its uncertainty. The data assimilation linked ecosystem carbon (DALEC) model is a simple box model simulating the carbon budget allocation for terrestrial ecosystems. Intercomparison experiments have demonstrated the relative merit of various inverse modelling strategies (MCMC, ENKF) to estimate model parameters and initial carbon stocks for DALEC using eddy covariance measurements of net ecosystem exchange of CO2 and leaf area index observations. Most results agreed on the fact that parameters and initial stocks directly related to fast processes were best estimated with narrow confidence intervals, whereas those related to slow processes were poorly estimated with very large uncertainties. While other studies have tried to overcome this difficulty by adding complementary data streams or by considering longer observation windows no systematic analysis has been carried out so far to explain the large differences among results. We consider adjoint based methods to investigate inverse problems using DALEC and various data streams. Using resolution matrices we study the nature of the inverse problems (solution existence, uniqueness and stability) and show how standard regularization techniques affect resolution and stability properties. Instead of using standard prior information as a penalty term in the cost function to regularize the problems we constraint the parameter space using ecological balance conditions and inequality constraints. The efficiency and rapidity of this approach allows us to compute ensembles of solutions to the inverse problems from which we can establish the robustness of the variational method and obtain non Gaussian posterior distributions for the model parameters and initial carbon stocks.

Restart Operator Meta-heuristics for a Problem-Oriented Evolutionary Strategies Algorithm in Inverse Mathematical MISO Modelling Problem Solving

NASA Astrophysics Data System (ADS)

Ryzhikov, I. S.; Semenkin, E. S.

2017-02-01

This study is focused on solving an inverse mathematical modelling problem for dynamical systems based on observation data and control inputs. The mathematical model is being searched in the form of a linear differential equation, which determines the system with multiple inputs and a single output, and a vector of the initial point coordinates. The described problem is complex and multimodal and for this reason the proposed evolutionary-based optimization technique, which is oriented on a dynamical system identification problem, was applied. To improve its performance an algorithm restart operator was implemented.

The inverse problem of refraction travel times, part I: Types of Geophysical Nonuniqueness through Minimization

USGS Publications Warehouse

Ivanov, J.; Miller, R.D.; Xia, J.; Steeples, D.; Park, C.B.

2005-01-01

In a set of two papers we study the inverse problem of refraction travel times. The purpose of this work is to use the study as a basis for development of more sophisticated methods for finding more reliable solutions to the inverse problem of refraction travel times, which is known to be nonunique. The first paper, "Types of Geophysical Nonuniqueness through Minimization," emphasizes the existence of different forms of nonuniqueness in the realm of inverse geophysical problems. Each type of nonuniqueness requires a different type and amount of a priori information to acquire a reliable solution. Based on such coupling, a nonuniqueness classification is designed. Therefore, since most inverse geophysical problems are nonunique, each inverse problem must be studied to define what type of nonuniqueness it belongs to and thus determine what type of a priori information is necessary to find a realistic solution. The second paper, "Quantifying Refraction Nonuniqueness Using a Three-layer Model," serves as an example of such an approach. However, its main purpose is to provide a better understanding of the inverse refraction problem by studying the type of nonuniqueness it possesses. An approach for obtaining a realistic solution to the inverse refraction problem is planned to be offered in a third paper that is in preparation. The main goal of this paper is to redefine the existing generalized notion of nonuniqueness and a priori information by offering a classified, discriminate structure. Nonuniqueness is often encountered when trying to solve inverse problems. However, possible nonuniqueness diversity is typically neglected and nonuniqueness is regarded as a whole, as an unpleasant "black box" and is approached in the same manner by applying smoothing constraints, damping constraints with respect to the solution increment and, rarely, damping constraints with respect to some sparse reference information about the true parameters. In practice, when solving geophysical problems different types of nonuniqueness exist, and thus there are different ways to solve the problems. Nonuniqueness is usually regarded as due to data error, assuming the true geology is acceptably approximated by simple mathematical models. Compounding the nonlinear problems, geophysical applications routinely exhibit exact-data nonuniqueness even for models with very few parameters adding to the nonuniqueness due to data error. While nonuniqueness variations have been defined earlier, they have not been linked to specific use of a priori information necessary to resolve each case. Four types of nonuniqueness, typical for minimization problems are defined with the corresponding methods for inclusion of a priori information to find a realistic solution without resorting to a non-discriminative approach. The above-developed stand-alone classification is expected to be helpful when solving any geophysical inverse problems. ?? Birkha??user Verlag, Basel, 2005.

Fast, Nonlinear, Fully Probabilistic Inversion of Large Geophysical Problems

NASA Astrophysics Data System (ADS)

Curtis, A.; Shahraeeni, M.; Trampert, J.; Meier, U.; Cho, G.

2010-12-01

Almost all Geophysical inverse problems are in reality nonlinear. Fully nonlinear inversion including non-approximated physics, and solving for probability distribution functions (pdf’s) that describe the solution uncertainty, generally requires sampling-based Monte-Carlo style methods that are computationally intractable in most large problems. In order to solve such problems, physical relationships are usually linearized leading to efficiently-solved, (possibly iterated) linear inverse problems. However, it is well known that linearization can lead to erroneous solutions, and in particular to overly optimistic uncertainty estimates. What is needed across many Geophysical disciplines is a method to invert large inverse problems (or potentially tens of thousands of small inverse problems) fully probabilistically and without linearization. This talk shows how very large nonlinear inverse problems can be solved fully probabilistically and incorporating any available prior information using mixture density networks (driven by neural network banks), provided the problem can be decomposed into many small inverse problems. In this talk I will explain the methodology, compare multi-dimensional pdf inversion results to full Monte Carlo solutions, and illustrate the method with two applications: first, inverting surface wave group and phase velocities for a fully-probabilistic global tomography model of the Earth’s crust and mantle, and second inverting industrial 3D seismic data for petrophysical properties throughout and around a subsurface hydrocarbon reservoir. The latter problem is typically decomposed into 104 to 105 individual inverse problems, each solved fully probabilistically and without linearization. The results in both cases are sufficiently close to the Monte Carlo solution to exhibit realistic uncertainty, multimodality and bias. This provides far greater confidence in the results, and in decisions made on their basis.

Parana Basin Structure from Multi-Objective Inversion of Surface Wave and Receiver Function by Competent Genetic Algorithm

NASA Astrophysics Data System (ADS)

An, M.; Assumpcao, M.

2003-12-01

The joint inversion of receiver function and surface wave is an effective way to diminish the influences of the strong tradeoff among parameters and the different sensitivity to the model parameters in their respective inversions, but the inversion problem becomes more complex. Multi-objective problems can be much more complicated than single-objective inversion in the model selection and optimization. If objectives are involved and conflicting, models can be ordered only partially. In this case, Pareto-optimal preference should be used to select solutions. On the other hand, the inversion to get only a few optimal solutions can not deal properly with the strong tradeoff between parameters, the uncertainties in the observation, the geophysical complexities and even the incompetency of the inversion technique. The effective way is to retrieve the geophysical information statistically from many acceptable solutions, which requires more competent global algorithms. Competent genetic algorithms recently proposed are far superior to the conventional genetic algorithm and can solve hard problems quickly, reliably and accurately. In this work we used one of competent genetic algorithms, Bayesian Optimization Algorithm as the main inverse procedure. This algorithm uses Bayesian networks to draw out inherited information and can use Pareto-optimal preference in the inversion. With this algorithm, the lithospheric structure of Paran"› basin is inverted to fit both the observations of inter-station surface wave dispersion and receiver function.

FOREWORD: 4th International Workshop on New Computational Methods for Inverse Problems (NCMIP2014)

NASA Astrophysics Data System (ADS)

2014-10-01

This volume of Journal of Physics: Conference Series is dedicated to the scientific contributions presented during the 4th International Workshop on New Computational Methods for Inverse Problems, NCMIP 2014 (http://www.farman.ens-cachan.fr/NCMIP_2014.html). This workshop took place at Ecole Normale Supérieure de Cachan, on May 23, 2014. The prior editions of NCMIP also took place in Cachan, France, firstly within the scope of ValueTools Conference, in May 2011 (http://www.ncmip.org/2011/), and secondly at the initiative of Institut Farman, in May 2012 and May 2013, (http://www.farman.ens-cachan.fr/NCMIP_2012.html), (http://www.farman.ens-cachan.fr/NCMIP_2013.html). The New Computational Methods for Inverse Problems (NCMIP) Workshop focused on recent advances in the resolution of inverse problems. Indeed, inverse problems appear in numerous scientific areas such as geophysics, biological and medical imaging, material and structure characterization, electrical, mechanical and civil engineering, and finances. The resolution of inverse problems consists of estimating the parameters of the observed system or structure from data collected by an instrumental sensing or imaging device. Its success firstly requires the collection of relevant observation data. It also requires accurate models describing the physical interactions between the instrumental device and the observed system, as well as the intrinsic properties of the solution itself. Finally, it requires the design of robust, accurate and efficient inversion algorithms. Advanced sensor arrays and imaging devices provide high rate and high volume data; in this context, the efficient resolution of the inverse problem requires the joint development of new models and inversion methods, taking computational and implementation aspects into account. During this one-day workshop, researchers had the opportunity to bring to light and share new techniques and results in the field of inverse problems. The topics of the workshop were: algorithms and computational aspects of inversion, Bayesian estimation, Kernel methods, learning methods, convex optimization, free discontinuity problems, metamodels, proper orthogonal decomposition, reduced models for the inversion, non-linear inverse scattering, image reconstruction and restoration, and applications (bio-medical imaging, non-destructive evaluation...). NCMIP 2014 was a one-day workshop held in May 2014 which attracted around sixty attendees. Each of the submitted papers has been reviewed by two reviewers. There have been nine accepted papers. In addition, three international speakers were invited to present a longer talk. The workshop was supported by Institut Farman (ENS Cachan, CNRS) and endorsed by the following French research networks (GDR ISIS, GDR MIA, GDR MOA, GDR Ondes). The program committee acknowledges the following research laboratories: CMLA, LMT, LURPA, SATIE. Eric Vourc'h and Thomas Rodet

ANNIT - An Efficient Inversion Algorithm based on Prediction Principles

NASA Astrophysics Data System (ADS)

Růžek, B.; Kolář, P.

2009-04-01

Solution of inverse problems represents meaningful job in geophysics. The amount of data is continuously increasing, methods of modeling are being improved and the computer facilities are also advancing great technical progress. Therefore the development of new and efficient algorithms and computer codes for both forward and inverse modeling is still up to date. ANNIT is contributing to this stream since it is a tool for efficient solution of a set of non-linear equations. Typical geophysical problems are based on parametric approach. The system is characterized by a vector of parameters p, the response of the system is characterized by a vector of data d. The forward problem is usually represented by unique mapping F(p)=d. The inverse problem is much more complex and the inverse mapping p=G(d) is available in an analytical or closed form only exceptionally and generally it may not exist at all. Technically, both forward and inverse mapping F and G are sets of non-linear equations. ANNIT solves such situation as follows: (i) joint subspaces {pD, pM} of original data and model spaces D, M, resp. are searched for, within which the forward mapping F is sufficiently smooth that the inverse mapping G does exist, (ii) numerical approximation of G in subspaces {pD, pM} is found, (iii) candidate solution is predicted by using this numerical approximation. ANNIT is working in an iterative way in cycles. The subspaces {pD, pM} are searched for by generating suitable populations of individuals (models) covering data and model spaces. The approximation of the inverse mapping is made by using three methods: (a) linear regression, (b) Radial Basis Function Network technique, (c) linear prediction (also known as "Kriging"). The ANNIT algorithm has built in also an archive of already evaluated models. Archive models are re-used in a suitable way and thus the number of forward evaluations is minimized. ANNIT is now implemented both in MATLAB and SCILAB. Numerical tests show good performance of the algorithm. Both versions and documentation are available on Internet and anybody can download them. The goal of this presentation is to offer the algorithm and computer codes for anybody interested in the solution to inverse problems.

Assimilating data into open ocean tidal models

NASA Astrophysics Data System (ADS)

Kivman, Gennady A.

The problem of deriving tidal fields from observations by reason of incompleteness and imperfectness of every data set practically available has an infinitely large number of allowable solutions fitting the data within measurement errors and hence can be treated as ill-posed. Therefore, interpolating the data always relies on some a priori assumptions concerning the tides, which provide a rule of sampling or, in other words, a regularization of the ill-posed problem. Data assimilation procedures used in large scale tide modeling are viewed in a common mathematical framework as such regularizations. It is shown that they all (basis functions expansion, parameter estimation, nudging, objective analysis, general inversion, and extended general inversion), including those (objective analysis and general inversion) originally formulated in stochastic terms, may be considered as utilizations of one of the three general methods suggested by the theory of ill-posed problems. The problem of grid refinement critical for inverse methods and nudging is discussed.

EIT image reconstruction based on a hybrid FE-EFG forward method and the complete-electrode model.

PubMed

Hadinia, M; Jafari, R; Soleimani, M

2016-06-01

This paper presents the application of the hybrid finite element-element free Galerkin (FE-EFG) method for the forward and inverse problems of electrical impedance tomography (EIT). The proposed method is based on the complete electrode model. Finite element (FE) and element-free Galerkin (EFG) methods are accurate numerical techniques. However, the FE technique has meshing task problems and the EFG method is computationally expensive. In this paper, the hybrid FE-EFG method is applied to take both advantages of FE and EFG methods, the complete electrode model of the forward problem is solved, and an iterative regularized Gauss-Newton method is adopted to solve the inverse problem. The proposed method is applied to compute Jacobian in the inverse problem. Utilizing 2D circular homogenous models, the numerical results are validated with analytical and experimental results and the performance of the hybrid FE-EFG method compared with the FE method is illustrated. Results of image reconstruction are presented for a human chest experimental phantom.

Large-scale inverse model analyses employing fast randomized data reduction

NASA Astrophysics Data System (ADS)

Lin, Youzuo; Le, Ellen B.; O'Malley, Daniel; Vesselinov, Velimir V.; Bui-Thanh, Tan

2017-08-01

When the number of observations is large, it is computationally challenging to apply classical inverse modeling techniques. We have developed a new computationally efficient technique for solving inverse problems with a large number of observations (e.g., on the order of 107 or greater). Our method, which we call the randomized geostatistical approach (RGA), is built upon the principal component geostatistical approach (PCGA). We employ a data reduction technique combined with the PCGA to improve the computational efficiency and reduce the memory usage. Specifically, we employ a randomized numerical linear algebra technique based on a so-called "sketching" matrix to effectively reduce the dimension of the observations without losing the information content needed for the inverse analysis. In this way, the computational and memory costs for RGA scale with the information content rather than the size of the calibration data. Our algorithm is coded in Julia and implemented in the MADS open-source high-performance computational framework (http://mads.lanl.gov). We apply our new inverse modeling method to invert for a synthetic transmissivity field. Compared to a standard geostatistical approach (GA), our method is more efficient when the number of observations is large. Most importantly, our method is capable of solving larger inverse problems than the standard GA and PCGA approaches. Therefore, our new model inversion method is a powerful tool for solving large-scale inverse problems. The method can be applied in any field and is not limited to hydrogeological applications such as the characterization of aquifer heterogeneity.

An Inverse Problem for a Class of Conditional Probability Measure-Dependent Evolution Equations

PubMed Central

Mirzaev, Inom; Byrne, Erin C.; Bortz, David M.

2016-01-01

We investigate the inverse problem of identifying a conditional probability measure in measure-dependent evolution equations arising in size-structured population modeling. We formulate the inverse problem as a least squares problem for the probability measure estimation. Using the Prohorov metric framework, we prove existence and consistency of the least squares estimates and outline a discretization scheme for approximating a conditional probability measure. For this scheme, we prove general method stability. The work is motivated by Partial Differential Equation (PDE) models of flocculation for which the shape of the post-fragmentation conditional probability measure greatly impacts the solution dynamics. To illustrate our methodology, we apply the theory to a particular PDE model that arises in the study of population dynamics for flocculating bacterial aggregates in suspension, and provide numerical evidence for the utility of the approach. PMID:28316360

Backus-Gilbert inversion of travel time data

NASA Technical Reports Server (NTRS)

Johnson, L. E.

1972-01-01

Application of the Backus-Gilbert theory for geophysical inverse problems to the seismic body wave travel-time problem is described. In particular, it is shown how to generate earth models that fit travel-time data to within one standard error and having generated such models how to describe their degree of uniqueness. An example is given to illustrate the process.

Determination of thermophysical characteristics of solid materials by electrical modelling of the solutions to the inverse problems in nonsteady heat conduction

NASA Technical Reports Server (NTRS)

Kozdoba, L. A.; Krivoshei, F. A.

1985-01-01

The solution of the inverse problem of nonsteady heat conduction is discussed, based on finding the coefficient of the heat conduction and the coefficient of specific volumetric heat capacity. These findings are included in the equation used for the electrical model of this phenomenon.

Damage identification using inverse methods.

PubMed

Friswell, Michael I

2007-02-15

This paper gives an overview of the use of inverse methods in damage detection and location, using measured vibration data. Inverse problems require the use of a model and the identification of uncertain parameters of this model. Damage is often local in nature and although the effect of the loss of stiffness may require only a small number of parameters, the lack of knowledge of the location means that a large number of candidate parameters must be included. This paper discusses a number of problems that exist with this approach to health monitoring, including modelling error, environmental effects, damage localization and regularization.

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.

Black hole algorithm for determining model parameter in self-potential data

NASA Astrophysics Data System (ADS)

Sungkono; Warnana, Dwa Desa

2018-01-01

Analysis of self-potential (SP) data is increasingly popular in geophysical method due to its relevance in many cases. However, the inversion of SP data is often highly nonlinear. Consequently, local search algorithms commonly based on gradient approaches have often failed to find the global optimum solution in nonlinear problems. Black hole algorithm (BHA) was proposed as a solution to such problems. As the name suggests, the algorithm was constructed based on the black hole phenomena. This paper investigates the application of BHA to solve inversions of field and synthetic self-potential (SP) data. The inversion results show that BHA accurately determines model parameters and model uncertainty. This indicates that BHA is highly potential as an innovative approach for SP data inversion.

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.

The incomplete inverse and its applications to the linear least squares problem

NASA Technical Reports Server (NTRS)

Morduch, G. E.

1977-01-01

A modified matrix product is explained, and it is shown that this product defiles a group whose inverse is called the incomplete inverse. It was proven that the incomplete inverse of an augmented normal matrix includes all the quantities associated with the least squares solution. An answer is provided to the problem that occurs when the data residuals are too large and when insufficient data to justify augmenting the model are available.

The Lanchester square-law model extended to a (2,2) conflict

NASA Astrophysics Data System (ADS)

Colegrave, R. K.; Hyde, J. M.

1993-01-01

A natural extension of the Lanchester (1,1) square-law model is the (M,N) linear model in which M forces oppose N forces with constant attrition rates. The (2,2) model is treated from both direct and inverse viewpoints. The inverse problem means that the model is to be fitted to a minimum number of observed force levels, i.e. the attrition rates are to be found from the initial force levels together with the levels observed at two subsequent times. An approach based on Hamiltonian dynamics has enabled the authors to derive a procedure for solving the inverse problem, which is readily computerized. Conflicts in which participants unexpectedly rally or weaken must be excluded.

Inverse kinematics problem in robotics using neural networks

NASA Technical Reports Server (NTRS)

Choi, Benjamin B.; Lawrence, Charles

1992-01-01

In this paper, Multilayer Feedforward Networks are applied to the robot inverse kinematic problem. The networks are trained with endeffector position and joint angles. After training, performance is measured by having the network generate joint angles for arbitrary endeffector trajectories. A 3-degree-of-freedom (DOF) spatial manipulator is used for the study. It is found that neural networks provide a simple and effective way to both model the manipulator inverse kinematics and circumvent the problems associated with algorithmic solution methods.

Model based inversion of ultrasound data in composites

NASA Astrophysics Data System (ADS)

Roberts, R. A.

2018-04-01

Work is reported on model-based defect characterization in CFRP composites. The work utilizes computational models of ultrasound interaction with defects in composites, to determine 1) the measured signal dependence on material and defect properties (forward problem), and 2) an assessment of defect properties from analysis of measured ultrasound signals (inverse problem). Work is reported on model implementation for inspection of CFRP laminates containing multi-ply impact-induced delamination, in laminates displaying irregular surface geometry (roughness), as well as internal elastic heterogeneity (varying fiber density, porosity). Inversion of ultrasound data is demonstrated showing the quantitative extraction of delamination geometry and surface transmissivity. Additionally, data inversion is demonstrated for determination of surface roughness and internal heterogeneity, and the influence of these features on delamination characterization is examined. Estimation of porosity volume fraction is demonstrated when internal heterogeneity is attributed to porosity.

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.

Developing a particle tracking surrogate model to improve inversion of ground water - Surface water models

NASA Astrophysics Data System (ADS)

Cousquer, Yohann; Pryet, Alexandre; Atteia, Olivier; Ferré, Ty P. A.; Delbart, Célestine; Valois, Rémi; Dupuy, Alain

2018-03-01

The inverse problem of groundwater models is often ill-posed and model parameters are likely to be poorly constrained. Identifiability is improved if diverse data types are used for parameter estimation. However, some models, including detailed solute transport models, are further limited by prohibitive computation times. This often precludes the use of concentration data for parameter estimation, even if those data are available. In the case of surface water-groundwater (SW-GW) models, concentration data can provide SW-GW mixing ratios, which efficiently constrain the estimate of exchange flow, but are rarely used. We propose to reduce computational limits by simulating SW-GW exchange at a sink (well or drain) based on particle tracking under steady state flow conditions. Particle tracking is used to simulate advective transport. A comparison between the particle tracking surrogate model and an advective-dispersive model shows that dispersion can often be neglected when the mixing ratio is computed for a sink, allowing for use of the particle tracking surrogate model. The surrogate model was implemented to solve the inverse problem for a real SW-GW transport problem with heads and concentrations combined in a weighted hybrid objective function. The resulting inversion showed markedly reduced uncertainty in the transmissivity field compared to calibration on head data alone.

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

Scalable posterior approximations for large-scale Bayesian inverse problems via likelihood-informed parameter and state reduction

NASA Astrophysics Data System (ADS)

Cui, Tiangang; Marzouk, Youssef; Willcox, Karen

2016-06-01

Two major bottlenecks to the solution of large-scale Bayesian inverse problems are the scaling of posterior sampling algorithms to high-dimensional parameter spaces and the computational cost of forward model evaluations. Yet incomplete or noisy data, the state variation and parameter dependence of the forward model, and correlations in the prior collectively provide useful structure that can be exploited for dimension reduction in this setting-both in the parameter space of the inverse problem and in the state space of the forward model. To this end, we show how to jointly construct low-dimensional subspaces of the parameter space and the state space in order to accelerate the Bayesian solution of the inverse problem. As a byproduct of state dimension reduction, we also show how to identify low-dimensional subspaces of the data in problems with high-dimensional observations. These subspaces enable approximation of the posterior as a product of two factors: (i) a projection of the posterior onto a low-dimensional parameter subspace, wherein the original likelihood is replaced by an approximation involving a reduced model; and (ii) the marginal prior distribution on the high-dimensional complement of the parameter subspace. We present and compare several strategies for constructing these subspaces using only a limited number of forward and adjoint model simulations. The resulting posterior approximations can rapidly be characterized using standard sampling techniques, e.g., Markov chain Monte Carlo. Two numerical examples demonstrate the accuracy and efficiency of our approach: inversion of an integral equation in atmospheric remote sensing, where the data dimension is very high; and the inference of a heterogeneous transmissivity field in a groundwater system, which involves a partial differential equation forward model with high dimensional state and parameters.

Inverse modeling for seawater intrusion in coastal aquifers: Insights about parameter sensitivities, variances, correlations and estimation procedures derived from the Henry problem

USGS Publications Warehouse

Sanz, E.; Voss, C.I.

2006-01-01

Inverse modeling studies employing data collected from the classic Henry seawater intrusion problem give insight into several important aspects of inverse modeling of seawater intrusion problems and effective measurement strategies for estimation of parameters for seawater intrusion. Despite the simplicity of the Henry problem, it embodies the behavior of a typical seawater intrusion situation in a single aquifer. Data collected from the numerical problem solution are employed without added noise in order to focus on the aspects of inverse modeling strategies dictated by the physics of variable-density flow and solute transport during seawater intrusion. Covariances of model parameters that can be estimated are strongly dependent on the physics. The insights gained from this type of analysis may be directly applied to field problems in the presence of data errors, using standard inverse modeling approaches to deal with uncertainty in data. Covariance analysis of the Henry problem indicates that in order to generally reduce variance of parameter estimates, the ideal places to measure pressure are as far away from the coast as possible, at any depth, and the ideal places to measure concentration are near the bottom of the aquifer between the center of the transition zone and its inland fringe. These observations are located in and near high-sensitivity regions of system parameters, which may be identified in a sensitivity analysis with respect to several parameters. However, both the form of error distribution in the observations and the observation weights impact the spatial sensitivity distributions, and different choices for error distributions or weights can result in significantly different regions of high sensitivity. Thus, in order to design effective sampling networks, the error form and weights must be carefully considered. For the Henry problem, permeability and freshwater inflow can be estimated with low estimation variance from only pressure or only concentration observations. Permeability, freshwater inflow, solute molecular diffusivity, and porosity can be estimated with roughly equivalent confidence using observations of only the logarithm of concentration. Furthermore, covariance analysis allows a logical reduction of the number of estimated parameters for ill-posed inverse seawater intrusion problems. Ill-posed problems may exhibit poor estimation convergence, have a non-unique solution, have multiple minima, or require excessive computational effort, and the condition often occurs when estimating too many or co-dependent parameters. For the Henry problem, such analysis allows selection of the two parameters that control system physics from among all possible system parameters. ?? 2005 Elsevier Ltd. All rights reserved.

Inverse problem of HIV cell dynamics using Genetic Algorithms

NASA Astrophysics Data System (ADS)

González, J. A.; Guzmán, F. S.

2017-01-01

In order to describe the cell dynamics of T-cells in a patient infected with HIV, we use a flavour of Perelson's model. This is a non-linear system of Ordinary Differential Equations that describes the evolution of healthy, latently infected, infected T-cell concentrations and the free viral cells. Different parameters in the equations give different dynamics. Considering the concentration of these types of cells is known for a particular patient, the inverse problem consists in estimating the parameters in the model. We solve this inverse problem using a Genetic Algorithm (GA) that minimizes the error between the solutions of the model and the data from the patient. These errors depend on the parameters of the GA, like mutation rate and population, although a detailed analysis of this dependence will be described elsewhere.

A model reduction approach to numerical inversion for a parabolic partial differential equation

NASA Astrophysics Data System (ADS)

Borcea, Liliana; Druskin, Vladimir; Mamonov, Alexander V.; Zaslavsky, Mikhail

2014-12-01

We propose a novel numerical inversion algorithm for the coefficients of parabolic partial differential equations, based on model reduction. The study is motivated by the application of controlled source electromagnetic exploration, where the unknown is the subsurface electrical resistivity and the data are time resolved surface measurements of the magnetic field. The algorithm presented in this paper considers inversion in one and two dimensions. The reduced model is obtained with rational interpolation in the frequency (Laplace) domain and a rational Krylov subspace projection method. It amounts to a nonlinear mapping from the function space of the unknown resistivity to the small dimensional space of the parameters of the reduced model. We use this mapping as a nonlinear preconditioner for the Gauss-Newton iterative solution of the inverse problem. The advantage of the inversion algorithm is twofold. First, the nonlinear preconditioner resolves most of the nonlinearity of the problem. Thus the iterations are less likely to get stuck in local minima and the convergence is fast. Second, the inversion is computationally efficient because it avoids repeated accurate simulations of the time-domain response. We study the stability of the inversion algorithm for various rational Krylov subspaces, and assess its performance with numerical experiments.

Variational approach to direct and inverse problems of atmospheric pollution studies

NASA Astrophysics Data System (ADS)

Penenko, Vladimir; Tsvetova, Elena; Penenko, Alexey

2016-04-01

We present the development of a variational approach for solving interrelated problems of atmospheric hydrodynamics and chemistry concerning air pollution transport and transformations. The proposed approach allows us to carry out complex studies of different-scale physical and chemical processes using the methods of direct and inverse modeling [1-3]. We formulate the problems of risk/vulnerability and uncertainty assessment, sensitivity studies, variational data assimilation procedures [4], etc. A computational technology of constructing consistent mathematical models and methods of their numerical implementation is based on the variational principle in the weak constraint formulation specifically designed to account for uncertainties in models and observations. Algorithms for direct and inverse modeling are designed with the use of global and local adjoint problems. Implementing the idea of adjoint integrating factors provides unconditionally monotone and stable discrete-analytic approximations for convection-diffusion-reaction problems [5,6]. The general framework is applied to the direct and inverse problems for the models of transport and transformation of pollutants in Siberian and Arctic regions. The work has been partially supported by the RFBR grant 14-01-00125 and RAS Presidium Program I.33P. References: 1. V. Penenko, A.Baklanov, E. Tsvetova and A. Mahura . Direct and inverse problems in a variational concept of environmental modeling //Pure and Applied Geoph.(2012) v.169: 447-465. 2. V. V. Penenko, E. A. Tsvetova, and A. V. Penenko Development of variational approach for direct and inverse problems of atmospheric hydrodynamics and chemistry, Izvestiya, Atmospheric and Oceanic Physics, 2015, Vol. 51, No. 3, p. 311-319, DOI: 10.1134/S0001433815030093. 3. V.V. Penenko, E.A. Tsvetova, A.V. Penenko. Methods based on the joint use of models and observational data in the framework of variational approach to forecasting weather and atmospheric composition quality// Russian meteorology and hydrology, V. 40, Issue: 6, Pages: 365-373, DOI: 10.3103/S1068373915060023. 4. A.V. Penenko and V.V. Penenko. Direct data assimilation method for convection-diffusion models based on splitting scheme. Computational technologies, 19(4):69-83, 2014. 5. V.V. Penenko, E.A. Tsvetova, A.V. Penenko Variational approach and Euler's integrating factors for environmental studies// Computers and Mathematics with Applications, 2014, V.67, Issue 12, Pages 2240-2256, DOI:10.1016/j.camwa.2014.04.004 6. V.V. Penenko, E.A. Tsvetova. Variational methods of constructing monotone approximations for atmospheric chemistry models // Numerical analysis and applications, 2013, V. 6, Issue 3, pp 210-220, DOI 10.1134/S199542391303004X

Gravity inversion of a fault by Particle swarm optimization (PSO).

PubMed

Toushmalani, Reza

2013-01-01

Particle swarm optimization is a heuristic global optimization method and also an optimization algorithm, which is based on swarm intelligence. It comes from the research on the bird and fish flock movement behavior. In this paper we introduce and use this method in gravity inverse problem. We discuss the solution for the inverse problem of determining the shape of a fault whose gravity anomaly is known. Application of the proposed algorithm to this problem has proven its capability to deal with difficult optimization problems. The technique proved to work efficiently when tested to a number of models.

Characterizing the impact of model error in hydrologic time series recovery inverse problems

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

Hansen, Scott K.; He, Jiachuan; Vesselinov, Velimir V.

Hydrologic models are commonly over-smoothed relative to reality, owing to computational limitations and to the difficulty of obtaining accurate high-resolution information. When used in an inversion context, such models may introduce systematic biases which cannot be encapsulated by an unbiased “observation noise” term of the type assumed by standard regularization theory and typical Bayesian formulations. Despite its importance, model error is difficult to encapsulate systematically and is often neglected. In this paper, model error is considered for an important class of inverse problems that includes interpretation of hydraulic transients and contaminant source history inference: reconstruction of a time series thatmore » has been convolved against a transfer function (i.e., impulse response) that is only approximately known. Using established harmonic theory along with two results established here regarding triangular Toeplitz matrices, upper and lower error bounds are derived for the effect of systematic model error on time series recovery for both well-determined and over-determined inverse problems. It is seen that use of additional measurement locations does not improve expected performance in the face of model error. A Monte Carlo study of a realistic hydraulic reconstruction problem is presented, and the lower error bound is seen informative about expected behavior. Finally, a possible diagnostic criterion for blind transfer function characterization is also uncovered.« less

Characterizing the impact of model error in hydrologic time series recovery inverse problems

DOE PAGES

Hansen, Scott K.; He, Jiachuan; Vesselinov, Velimir V.

2017-10-28

Hydrologic models are commonly over-smoothed relative to reality, owing to computational limitations and to the difficulty of obtaining accurate high-resolution information. When used in an inversion context, such models may introduce systematic biases which cannot be encapsulated by an unbiased “observation noise” term of the type assumed by standard regularization theory and typical Bayesian formulations. Despite its importance, model error is difficult to encapsulate systematically and is often neglected. In this paper, model error is considered for an important class of inverse problems that includes interpretation of hydraulic transients and contaminant source history inference: reconstruction of a time series thatmore » has been convolved against a transfer function (i.e., impulse response) that is only approximately known. Using established harmonic theory along with two results established here regarding triangular Toeplitz matrices, upper and lower error bounds are derived for the effect of systematic model error on time series recovery for both well-determined and over-determined inverse problems. It is seen that use of additional measurement locations does not improve expected performance in the face of model error. A Monte Carlo study of a realistic hydraulic reconstruction problem is presented, and the lower error bound is seen informative about expected behavior. Finally, a possible diagnostic criterion for blind transfer function characterization is also uncovered.« less

Towards adjoint-based inversion for rheological parameters in nonlinear viscous mantle flow

NASA Astrophysics Data System (ADS)

Worthen, Jennifer; Stadler, Georg; Petra, Noemi; Gurnis, Michael; Ghattas, Omar

2014-09-01

We address the problem of inferring mantle rheological parameter fields from surface velocity observations and instantaneous nonlinear mantle flow models. We formulate this inverse problem as an infinite-dimensional nonlinear least squares optimization problem governed by nonlinear Stokes equations. We provide expressions for the gradient of the cost functional of this optimization problem with respect to two spatially-varying rheological parameter fields: the viscosity prefactor and the exponent of the second invariant of the strain rate tensor. Adjoint (linearized) Stokes equations, which are characterized by a 4th order anisotropic viscosity tensor, facilitates efficient computation of the gradient. A quasi-Newton method for the solution of this optimization problem is presented, which requires the repeated solution of both nonlinear forward Stokes and linearized adjoint Stokes equations. For the solution of the nonlinear Stokes equations, we find that Newton’s method is significantly more efficient than a Picard fixed point method. Spectral analysis of the inverse operator given by the Hessian of the optimization problem reveals that the numerical eigenvalues collapse rapidly to zero, suggesting a high degree of ill-posedness of the inverse problem. To overcome this ill-posedness, we employ Tikhonov regularization (favoring smooth parameter fields) or total variation (TV) regularization (favoring piecewise-smooth parameter fields). Solution of two- and three-dimensional finite element-based model inverse problems show that a constant parameter in the constitutive law can be recovered well from surface velocity observations. Inverting for a spatially-varying parameter field leads to its reasonable recovery, in particular close to the surface. When inferring two spatially varying parameter fields, only an effective viscosity field and the total viscous dissipation are recoverable. Finally, a model of a subducting plate shows that a localized weak zone at the plate boundary can be partially recovered, especially with TV regularization.

Convex blind image deconvolution with inverse filtering

NASA Astrophysics Data System (ADS)

Lv, Xiao-Guang; Li, Fang; Zeng, Tieyong

2018-03-01

Blind image deconvolution is the process of estimating both the original image and the blur kernel from the degraded image with only partial or no information about degradation and the imaging system. It is a bilinear ill-posed inverse problem corresponding to the direct problem of convolution. Regularization methods are used to handle the ill-posedness of blind deconvolution and get meaningful solutions. In this paper, we investigate a convex regularized inverse filtering method for blind deconvolution of images. We assume that the support region of the blur object is known, as has been done in a few existing works. By studying the inverse filters of signal and image restoration problems, we observe the oscillation structure of the inverse filters. Inspired by the oscillation structure of the inverse filters, we propose to use the star norm to regularize the inverse filter. Meanwhile, we use the total variation to regularize the resulting image obtained by convolving the inverse filter with the degraded image. The proposed minimization model is shown to be convex. We employ the first-order primal-dual method for the solution of the proposed minimization model. Numerical examples for blind image restoration are given to show that the proposed method outperforms some existing methods in terms of peak signal-to-noise ratio (PSNR), structural similarity (SSIM), visual quality and time consumption.

Atmospheric inverse modeling via sparse reconstruction

NASA Astrophysics Data System (ADS)

Hase, Nils; Miller, Scot M.; Maaß, Peter; Notholt, Justus; Palm, Mathias; Warneke, Thorsten

2017-10-01

Many applications in atmospheric science involve ill-posed inverse problems. A crucial component of many inverse problems is the proper formulation of a priori knowledge about the unknown parameters. In most cases, this knowledge is expressed as a Gaussian prior. This formulation often performs well at capturing smoothed, large-scale processes but is often ill equipped to capture localized structures like large point sources or localized hot spots. Over the last decade, scientists from a diverse array of applied mathematics and engineering fields have developed sparse reconstruction techniques to identify localized structures. In this study, we present a new regularization approach for ill-posed inverse problems in atmospheric science. It is based on Tikhonov regularization with sparsity constraint and allows bounds on the parameters. We enforce sparsity using a dictionary representation system. We analyze its performance in an atmospheric inverse modeling scenario by estimating anthropogenic US methane (CH4) emissions from simulated atmospheric measurements. Different measures indicate that our sparse reconstruction approach is better able to capture large point sources or localized hot spots than other methods commonly used in atmospheric inversions. It captures the overall signal equally well but adds details on the grid scale. This feature can be of value for any inverse problem with point or spatially discrete sources. We show an example for source estimation of synthetic methane emissions from the Barnett shale formation.

Fast Component Pursuit for Large-Scale Inverse Covariance Estimation.

PubMed

Han, Lei; Zhang, Yu; Zhang, Tong

2016-08-01

The maximum likelihood estimation (MLE) for the Gaussian graphical model, which is also known as the inverse covariance estimation problem, has gained increasing interest recently. Most existing works assume that inverse covariance estimators contain sparse structure and then construct models with the ℓ 1 regularization. In this paper, different from existing works, we study the inverse covariance estimation problem from another perspective by efficiently modeling the low-rank structure in the inverse covariance, which is assumed to be a combination of a low-rank part and a diagonal matrix. One motivation for this assumption is that the low-rank structure is common in many applications including the climate and financial analysis, and another one is that such assumption can reduce the computational complexity when computing its inverse. Specifically, we propose an efficient COmponent Pursuit (COP) method to obtain the low-rank part, where each component can be sparse. For optimization, the COP method greedily learns a rank-one component in each iteration by maximizing the log-likelihood. Moreover, the COP algorithm enjoys several appealing properties including the existence of an efficient solution in each iteration and the theoretical guarantee on the convergence of this greedy approach. Experiments on large-scale synthetic and real-world datasets including thousands of millions variables show that the COP method is faster than the state-of-the-art techniques for the inverse covariance estimation problem when achieving comparable log-likelihood on test data.

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.

A MATLAB based 3D modeling and inversion code for MT data

NASA Astrophysics Data System (ADS)

Singh, Arun; Dehiya, Rahul; Gupta, Pravin K.; Israil, M.

2017-07-01

The development of a MATLAB based computer code, AP3DMT, for modeling and inversion of 3D Magnetotelluric (MT) data is presented. The code comprises two independent components: grid generator code and modeling/inversion code. The grid generator code performs model discretization and acts as an interface by generating various I/O files. The inversion code performs core computations in modular form - forward modeling, data functionals, sensitivity computations and regularization. These modules can be readily extended to other similar inverse problems like Controlled-Source EM (CSEM). The modular structure of the code provides a framework useful for implementation of new applications and inversion algorithms. The use of MATLAB and its libraries makes it more compact and user friendly. The code has been validated on several published models. To demonstrate its versatility and capabilities the results of inversion for two complex models are presented.

Implement Method for Automated Testing of Markov Chain Convergence into INVERSE for ORNL12-RS-108J: Advanced Multi-Dimensional Forward and Inverse Modeling

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

Bledsoe, Keith C.

2015-04-01

The DiffeRential Evolution Adaptive Metropolis (DREAM) method is a powerful optimization/uncertainty quantification tool used to solve inverse transport problems in Los Alamos National Laboratory’s INVERSE code system. The DREAM method has been shown to be adept at accurate uncertainty quantification, but it can be very computationally demanding. Previously, the DREAM method in INVERSE performed a user-defined number of particle transport calculations. This placed a burden on the user to guess the number of calculations that would be required to accurately solve any given problem. This report discusses a new approach that has been implemented into INVERSE, the Gelman-Rubin convergence metric.more » This metric automatically detects when an appropriate number of transport calculations have been completed and the uncertainty in the inverse problem has been accurately calculated. In a test problem with a spherical geometry, this method was found to decrease the number of transport calculations (and thus time required) to solve a problem by an average of over 90%. In a cylindrical test geometry, a 75% decrease was obtained.« less

Steady induction effects in geomagnetism. Part 1B: Geomagnetic estimation of steady surficial core motions: A non-linear inverse problem

NASA Technical Reports Server (NTRS)

Voorhies, Coerte V.

1993-01-01

The problem of estimating a steady fluid velocity field near the top of Earth's core which induces the secular variation (SV) indicated by models of the observed geomagnetic field is examined in the source-free mantle/frozen-flux core (SFI/VFFC) approximation. This inverse problem is non-linear because solutions of the forward problem are deterministically chaotic. The SFM/FFC approximation is inexact, and neither the models nor the observations they represent are either complete or perfect. A method is developed for solving the non-linear inverse motional induction problem posed by the hypothesis of (piecewise, statistically) steady core surface flow and the supposition of a complete initial geomagnetic condition. The method features iterative solution of the weighted, linearized least-squares problem and admits optional biases favoring surficially geostrophic flow and/or spatially simple flow. Two types of weights are advanced radial field weights for fitting the evolution of the broad-scale portion of the radial field component near Earth's surface implied by the models, and generalized weights for fitting the evolution of the broad-scale portion of the scalar potential specified by the models.

Mantle Circulation Models with variational data assimilation: Inferring past mantle flow and structure from plate motion histories and seismic tomography

NASA Astrophysics Data System (ADS)

Bunge, H.; Hagelberg, C.; Travis, B.

2002-12-01

EarthScope will deliver data on structure and dynamics of continental North America and the underlying mantle on an unprecedented scale. Indeed, the scope of EarthScope makes its mission comparable to the large remote sensing efforts that are transforming the oceanographic and atmospheric sciences today. Arguably the main impact of new solid Earth observing systems is to transform our use of geodynamic models increasingly from conditions that are data poor to an environment that is data rich. Oceanographers and meteorologists already have made substantial progress in adapting to this environment, by developing new approaches of interpreting oceanographic and atmospheric data objectively through data assimilation methods in their models. However, a similarly rigorous theoretical framework for merging EarthScope derived solid Earth data with geodynamic models has yet to be devised. Here we explore the feasibility of data assimilation in mantle convection studies in an attempt to fit global geodynamic model calculations explicitly to tomographic and tectonic constraints. This is an inverse problem not quite unlike the inverse problem of finding optimal seismic velocity structures faced by seismologists. We derive the generalized inverse of mantle convection from a variational approach and present the adjoint equations of mantle flow. The substantial computational burden associated with solutions to the generalized inverse problem of mantle convection is made feasible using a highly efficient finite element approach based on the 3-D spherical fully parallelized mantle dynamics code TERRA, implemented on a cost-effective topical PC-cluster (geowulf) dedicated specifically to large-scale geophysical simulations. This dedicated geophysical modeling computer allows us to investigate global inverse convection problems having a spatial discretization of less than 50 km throughout the mantle. We present a synthetic high-resolution modeling experiment to demonstrate that mid-Cretaceous mantle structure can be inferred accurately from our inverse approach assuming present-day mantle structure is well-known, even if an initial first guess assumption about the mid-Cretaceous mantle involved only a simple 1-D radial temperature profile. We suggest that geodynamic inverse modeling should make it possible to infer a number of flow parameters from observational constraints of the mantle.

Model Order Reduction for the fast solution of 3D Stokes problems and its application in geophysical inversion

NASA Astrophysics Data System (ADS)

Ortega Gelabert, Olga; Zlotnik, Sergio; Afonso, Juan Carlos; Díez, Pedro

2017-04-01

The determination of the present-day physical state of the thermal and compositional structure of the Earth's lithosphere and sub-lithospheric mantle is one of the main goals in modern lithospheric research. All this data is essential to build Earth's evolution models and to reproduce many geophysical observables (e.g. elevation, gravity anomalies, travel time data, heat flow, etc) together with understanding the relationship between them. Determining the lithospheric state involves the solution of high-resolution inverse problems and, consequently, the solution of many direct models is required. The main objective of this work is to contribute to the existing inversion techniques in terms of improving the estimation of the elevation (topography) by including a dynamic component arising from sub-lithospheric mantle flow. In order to do so, we implement an efficient Reduced Order Method (ROM) built upon classic Finite Elements. ROM allows to reduce significantly the computational cost of solving a family of problems, for example all the direct models that are required in the solution of the inverse problem. The strategy of the method consists in creating a (reduced) basis of solutions, so that when a new problem has to be solved, its solution is sought within the basis instead of attempting to solve the problem itself. In order to check the Reduced Basis approach, we implemented the method in a 3D domain reproducing a portion of Earth that covers up to 400 km depth. Within the domain the Stokes equation is solved with realistic viscosities and densities. The different realizations (the family of problems) is created by varying viscosities and densities in a similar way as it would happen in an inversion problem. The Reduced Basis method is shown to be an extremely efficiently solver for the Stokes equation in this context.

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.

Exact solution for an optimal impermeable parachute problem

NASA Astrophysics Data System (ADS)

Lupu, Mircea; Scheiber, Ernest

2002-10-01

In the paper there are solved direct and inverse boundary problems and analytical solutions are obtained for optimization problems in the case of some nonlinear integral operators. It is modeled the plane potential flow of an inviscid, incompressible and nonlimited fluid jet, witch encounters a symmetrical, curvilinear obstacle--the deflector of maximal drag. There are derived integral singular equations, for direct and inverse problems and the movement in the auxiliary canonical half-plane is obtained. Next, the optimization problem is solved in an analytical manner. The design of the optimal airfoil is performed and finally, numerical computations concerning the drag coefficient and other geometrical and aerodynamical parameters are carried out. This model corresponds to the Helmholtz impermeable parachute problem.

Time-domain full waveform inversion using instantaneous phase information with damping

NASA Astrophysics Data System (ADS)

Luo, Jingrui; Wu, Ru-Shan; Gao, Fuchun

2018-06-01

In time domain, the instantaneous phase can be obtained from the complex seismic trace using Hilbert transform. The instantaneous phase information has great potential in overcoming the local minima problem and improving the result of full waveform inversion. However, the phase wrapping problem, which comes from numerical calculation, prevents its application. In order to avoid the phase wrapping problem, we choose to use the exponential phase combined with the damping method, which gives instantaneous phase-based multi-stage inversion. We construct the objective functions based on the exponential instantaneous phase, and also derive the corresponding gradient operators. Conventional full waveform inversion and the instantaneous phase-based inversion are compared with numerical examples, which indicates that in the case without low frequency information in seismic data, our method is an effective and efficient approach for initial model construction for full waveform inversion.

Effect of conductor geometry on source localization: Implications for epilepsy studies

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

Schlitt, H.; Heller, L.; Best, E.

1994-07-01

We shall discuss the effects of conductor geometry on source localization for applications in epilepsy studies. The most popular conductor model for clinical MEG studies is a homogeneous sphere. However, several studies have indicated that a sphere is a poor model for the head when the sources are deep, as is the case for epileptic foci in the mesial temporal lobe. We believe that replacing the spherical model with a more realistic one in the inverse fitting procedure will improve the accuracy of localizing epileptic sources. In order to include a realistic head model in the inverse problem, we mustmore » first solve the forward problem for the realistic conductor geometry. We create a conductor geometry model from MR images, and then solve the forward problem via a boundary integral equation for the electric potential due to a specified primary source. One the electric potential is known, the magnetic field can be calculated directly. The most time-intensive part of the problem is generating the conductor model; fortunately, this needs to be done only once for each patient. It takes little time to change the primary current and calculate a new magnetic field for use in the inverse fitting procedure. We present the results of a series of computer simulations in which we investigate the localization accuracy due to replacing the spherical model with the realistic head model in the inverse fitting procedure. The data to be fit consist of a computer generated magnetic field due to a known current dipole in a realistic head model, with added noise. We compare the localization errors when this field is fit using a spherical model to the fit using a realistic head model. Using a spherical model is comparable to what is usually done when localizing epileptic sources in humans, where the conductor model used in the inverse fitting procedure does not correspond to the actual head.« less

Genetic algorithms and their use in Geophysical Problems

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

Parker, Paul B.

1999-04-01

Genetic algorithms (GAs), global optimization methods that mimic Darwinian evolution are well suited to the nonlinear inverse problems of geophysics. A standard genetic algorithm selects the best or ''fittest'' models from a ''population'' and then applies operators such as crossover and mutation in order to combine the most successful characteristics of each model and produce fitter models. More sophisticated operators have been developed, but the standard GA usually provides a robust and efficient search. Although the choice of parameter settings such as crossover and mutation rate may depend largely on the type of problem being solved, numerous results show thatmore » certain parameter settings produce optimal performance for a wide range of problems and difficulties. In particular, a low (about half of the inverse of the population size) mutation rate is crucial for optimal results, but the choice of crossover method and rate do not seem to affect performance appreciably. Optimal efficiency is usually achieved with smaller (< 50) populations. Lastly, tournament selection appears to be the best choice of selection methods due to its simplicity and its autoscaling properties. However, if a proportional selection method is used such as roulette wheel selection, fitness scaling is a necessity, and a high scaling factor (> 2.0) should be used for the best performance. Three case studies are presented in which genetic algorithms are used to invert for crustal parameters. The first is an inversion for basement depth at Yucca mountain using gravity data, the second an inversion for velocity structure in the crust of the south island of New Zealand using receiver functions derived from teleseismic events, and the third is a similar receiver function inversion for crustal velocities beneath the Mendocino Triple Junction region of Northern California. The inversions demonstrate that genetic algorithms are effective in solving problems with reasonably large numbers of free parameters and with computationally expensive objective function calculations. More sophisticated techniques are presented for special problems. Niching and island model algorithms are introduced as methods to find multiple, distinct solutions to the nonunique problems that are typically seen in geophysics. Finally, hybrid algorithms are investigated as a way to improve the efficiency of the standard genetic algorithm.« less

Genetic algorithms and their use in geophysical problems

NASA Astrophysics Data System (ADS)

Parker, Paul Bradley

Genetic algorithms (GAs), global optimization methods that mimic Darwinian evolution are well suited to the nonlinear inverse problems of geophysics. A standard genetic algorithm selects the best or "fittest" models from a "population" and then applies operators such as crossover and mutation in order to combine the most successful characteristics of each model and produce fitter models. More sophisticated operators have been developed, but the standard GA usually provides a robust and efficient search. Although the choice of parameter settings such as crossover and mutation rate may depend largely on the type of problem being solved, numerous results show that certain parameter settings produce optimal performance for a wide range of problems and difficulties. In particular, a low (about half of the inverse of the population size) mutation rate is crucial for optimal results, but the choice of crossover method and rate do not seem to affect performance appreciably. Also, optimal efficiency is usually achieved with smaller (<50) populations. Lastly, tournament selection appears to be the best choice of selection methods due to its simplicity and its autoscaling properties. However, if a proportional selection method is used such as roulette wheel selection, fitness scaling is a necessity, and a high scaling factor (>2.0) should be used for the best performance. Three case studies are presented in which genetic algorithms are used to invert for crustal parameters. The first is an inversion for basement depth at Yucca mountain using gravity data, the second an inversion for velocity structure in the crust of the south island of New Zealand using receiver functions derived from teleseismic events, and the third is a similar receiver function inversion for crustal velocities beneath the Mendocino Triple Junction region of Northern California. The inversions demonstrate that genetic algorithms are effective in solving problems with reasonably large numbers of free parameters and with computationally expensive objective function calculations. More sophisticated techniques are presented for special problems. Niching and island model algorithms are introduced as methods to find multiple, distinct solutions to the nonunique problems that are typically seen in geophysics. Finally, hybrid algorithms are investigated as a way to improve the efficiency of the standard genetic algorithm.

Inversion of geothermal heat flux in a thermomechanically coupled nonlinear Stokes ice sheet model

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

Zhu, Hongyu; Petra, Noemi; Stadler, Georg

We address the inverse problem of inferring the basal geothermal heat flux from surface velocity observations using a steady-state thermomechanically coupled nonlinear Stokes ice flow model. This is a challenging inverse problem since the map from basal heat flux to surface velocity observables is indirect: the heat flux is a boundary condition for the thermal advection–diffusion equation, which couples to the nonlinear Stokes ice flow equations; together they determine the surface ice flow velocity. This multiphysics inverse problem is formulated as a nonlinear least-squares optimization problem with a cost functional that includes the data misfit between surface velocity observations andmore » model predictions. A Tikhonov regularization term is added to render the problem well posed. We derive adjoint-based gradient and Hessian expressions for the resulting partial differential equation (PDE)-constrained optimization problem and propose an inexact Newton method for its solution. As a consequence of the Petrov–Galerkin discretization of the energy equation, we show that discretization and differentiation do not commute; that is, the order in which we discretize the cost functional and differentiate it affects the correctness of the gradient. Using two- and three-dimensional model problems, we study the prospects for and limitations of the inference of the geothermal heat flux field from surface velocity observations. The results show that the reconstruction improves as the noise level in the observations decreases and that short-wavelength variations in the geothermal heat flux are difficult to recover. We analyze the ill-posedness of the inverse problem as a function of the number of observations by examining the spectrum of the Hessian of the cost functional. Motivated by the popularity of operator-split or staggered solvers for forward multiphysics problems – i.e., those that drop two-way coupling terms to yield a one-way coupled forward Jacobian – we study the effect on the inversion of a one-way coupling of the adjoint energy and Stokes equations. Here, we show that taking such a one-way coupled approach for the adjoint equations can lead to an incorrect gradient and premature termination of optimization iterations. This is due to loss of a descent direction stemming from inconsistency of the gradient with the contours of the cost functional. Nevertheless, one may still obtain a reasonable approximate inverse solution particularly if important features of the reconstructed solution emerge early in optimization iterations, before the premature termination.« less

Inversion of geothermal heat flux in a thermomechanically coupled nonlinear Stokes ice sheet model

DOE PAGES

Zhu, Hongyu; Petra, Noemi; Stadler, Georg; ...

2016-07-13

We address the inverse problem of inferring the basal geothermal heat flux from surface velocity observations using a steady-state thermomechanically coupled nonlinear Stokes ice flow model. This is a challenging inverse problem since the map from basal heat flux to surface velocity observables is indirect: the heat flux is a boundary condition for the thermal advection–diffusion equation, which couples to the nonlinear Stokes ice flow equations; together they determine the surface ice flow velocity. This multiphysics inverse problem is formulated as a nonlinear least-squares optimization problem with a cost functional that includes the data misfit between surface velocity observations andmore » model predictions. A Tikhonov regularization term is added to render the problem well posed. We derive adjoint-based gradient and Hessian expressions for the resulting partial differential equation (PDE)-constrained optimization problem and propose an inexact Newton method for its solution. As a consequence of the Petrov–Galerkin discretization of the energy equation, we show that discretization and differentiation do not commute; that is, the order in which we discretize the cost functional and differentiate it affects the correctness of the gradient. Using two- and three-dimensional model problems, we study the prospects for and limitations of the inference of the geothermal heat flux field from surface velocity observations. The results show that the reconstruction improves as the noise level in the observations decreases and that short-wavelength variations in the geothermal heat flux are difficult to recover. We analyze the ill-posedness of the inverse problem as a function of the number of observations by examining the spectrum of the Hessian of the cost functional. Motivated by the popularity of operator-split or staggered solvers for forward multiphysics problems – i.e., those that drop two-way coupling terms to yield a one-way coupled forward Jacobian – we study the effect on the inversion of a one-way coupling of the adjoint energy and Stokes equations. Here, we show that taking such a one-way coupled approach for the adjoint equations can lead to an incorrect gradient and premature termination of optimization iterations. This is due to loss of a descent direction stemming from inconsistency of the gradient with the contours of the cost functional. Nevertheless, one may still obtain a reasonable approximate inverse solution particularly if important features of the reconstructed solution emerge early in optimization iterations, before the premature termination.« less

Inversion of geothermal heat flux in a thermomechanically coupled nonlinear Stokes ice sheet model

NASA Astrophysics Data System (ADS)

Zhu, Hongyu; Petra, Noemi; Stadler, Georg; Isaac, Tobin; Hughes, Thomas J. R.; Ghattas, Omar

2016-07-01

We address the inverse problem of inferring the basal geothermal heat flux from surface velocity observations using a steady-state thermomechanically coupled nonlinear Stokes ice flow model. This is a challenging inverse problem since the map from basal heat flux to surface velocity observables is indirect: the heat flux is a boundary condition for the thermal advection-diffusion equation, which couples to the nonlinear Stokes ice flow equations; together they determine the surface ice flow velocity. This multiphysics inverse problem is formulated as a nonlinear least-squares optimization problem with a cost functional that includes the data misfit between surface velocity observations and model predictions. A Tikhonov regularization term is added to render the problem well posed. We derive adjoint-based gradient and Hessian expressions for the resulting partial differential equation (PDE)-constrained optimization problem and propose an inexact Newton method for its solution. As a consequence of the Petrov-Galerkin discretization of the energy equation, we show that discretization and differentiation do not commute; that is, the order in which we discretize the cost functional and differentiate it affects the correctness of the gradient. Using two- and three-dimensional model problems, we study the prospects for and limitations of the inference of the geothermal heat flux field from surface velocity observations. The results show that the reconstruction improves as the noise level in the observations decreases and that short-wavelength variations in the geothermal heat flux are difficult to recover. We analyze the ill-posedness of the inverse problem as a function of the number of observations by examining the spectrum of the Hessian of the cost functional. Motivated by the popularity of operator-split or staggered solvers for forward multiphysics problems - i.e., those that drop two-way coupling terms to yield a one-way coupled forward Jacobian - we study the effect on the inversion of a one-way coupling of the adjoint energy and Stokes equations. We show that taking such a one-way coupled approach for the adjoint equations can lead to an incorrect gradient and premature termination of optimization iterations. This is due to loss of a descent direction stemming from inconsistency of the gradient with the contours of the cost functional. Nevertheless, one may still obtain a reasonable approximate inverse solution particularly if important features of the reconstructed solution emerge early in optimization iterations, before the premature termination.

Trimming and procrastination as inversion techniques

NASA Astrophysics Data System (ADS)

Backus, George E.

1996-12-01

By examining the processes of truncating and approximating the model space (trimming it), and by committing to neither the objectivist nor the subjectivist interpretation of probability (procrastinating), we construct a formal scheme for solving linear and non-linear geophysical inverse problems. The necessary prior information about the correct model xE can be either a collection of inequalities or a probability measure describing where xE was likely to be in the model space X before the data vector y0 was measured. The results of the inversion are (1) a vector z0 that estimates some numerical properties zE of xE; (2) an estimate of the error δz = z0 - zE. As y0 is finite dimensional, so is z0, and hence in principle inversion cannot describe all of xE. The error δz is studied under successively more specialized assumptions about the inverse problem, culminating in a complete analysis of the linear inverse problem with a prior quadratic bound on xE. Our formalism appears to encompass and provide error estimates for many of the inversion schemes current in geomagnetism, and would be equally applicable in geodesy and seismology if adequate prior information were available there. As an idealized example we study the magnetic field at the core-mantle boundary, using satellite measurements of field elements at sites assumed to be almost uniformly distributed on a single spherical surface. Magnetospheric currents are neglected and the crustal field is idealized as a random process with rotationally invariant statistics. We find that an appropriate data compression diagonalizes the variance matrix of the crustal signal and permits an analytic trimming of the idealized problem.

Model based approach to UXO imaging using the time domain electromagnetic method

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

Lavely, E.M.

1999-04-01

Time domain electromagnetic (TDEM) sensors have emerged as a field-worthy technology for UXO detection in a variety of geological and environmental settings. This success has been achieved with commercial equipment that was not optimized for UXO detection and discrimination. The TDEM response displays a rich spatial and temporal behavior which is not currently utilized. Therefore, in this paper the author describes a research program for enhancing the effectiveness of the TDEM method for UXO detection and imaging. Fundamental research is required in at least three major areas: (a) model based imaging capability i.e. the forward and inverse problem, (b) detectormore » modeling and instrument design, and (c) target recognition and discrimination algorithms. These research problems are coupled and demand a unified treatment. For example: (1) the inverse solution depends on solution of the forward problem and knowledge of the instrument response; (2) instrument design with improved diagnostic power requires forward and inverse modeling capability; and (3) improved target recognition algorithms (such as neural nets) must be trained with data collected from the new instrument and with synthetic data computed using the forward model. Further, the design of the appropriate input and output layers of the net will be informed by the results of the forward and inverse modeling. A more fully developed model of the TDEM response would enable the joint inversion of data collected from multiple sensors (e.g., TDEM sensors and magnetometers). Finally, the author suggests that a complementary approach to joint inversions is the statistical recombination of data using principal component analysis. The decomposition into principal components is useful since the first principal component contains those features that are most strongly correlated from image to image.« less

Iterative algorithms for a non-linear inverse problem in atmospheric lidar

NASA Astrophysics Data System (ADS)

Denevi, Giulia; Garbarino, Sara; Sorrentino, Alberto

2017-08-01

We consider the inverse problem of retrieving aerosol extinction coefficients from Raman lidar measurements. In this problem the unknown and the data are related through the exponential of a linear operator, the unknown is non-negative and the data follow the Poisson distribution. Standard methods work on the log-transformed data and solve the resulting linear inverse problem, but neglect to take into account the noise statistics. In this study we show that proper modelling of the noise distribution can improve substantially the quality of the reconstructed extinction profiles. To achieve this goal, we consider the non-linear inverse problem with non-negativity constraint, and propose two iterative algorithms derived using the Karush-Kuhn-Tucker conditions. We validate the algorithms with synthetic and experimental data. As expected, the proposed algorithms out-perform standard methods in terms of sensitivity to noise and reliability of the estimated profile.

Retrieval of LAI and leaf chlorophyll content from remote sensing data by agronomy mechanism knowledge to solve the ill-posed inverse problem

NASA Astrophysics Data System (ADS)

Li, Zhenhai; Nie, Chenwei; Yang, Guijun; Xu, Xingang; Jin, Xiuliang; Gu, Xiaohe

2014-10-01

Leaf area index (LAI) and LCC, as the two most important crop growth variables, are major considerations in management decisions, agricultural planning and policy making. Estimation of canopy biophysical variables from remote sensing data was investigated using a radiative transfer model. However, the ill-posed problem is unavoidable for the unique solution of the inverse problem and the uncertainty of measurements and model assumptions. This study focused on the use of agronomy mechanism knowledge to restrict and remove the ill-posed inversion results. For this purpose, the inversion results obtained using the PROSAIL model alone (NAMK) and linked with agronomic mechanism knowledge (AMK) were compared. The results showed that AMK did not significantly improve the accuracy of LAI inversion. LAI was estimated with high accuracy, and there was no significant improvement after considering AMK. The validation results of the determination coefficient (R2) and the corresponding root mean square error (RMSE) between measured LAI and estimated LAI were 0.635 and 1.022 for NAMK, and 0.637 and 0.999 for AMK, respectively. LCC estimation was significantly improved with agronomy mechanism knowledge; the R2 and RMSE values were 0.377 and 14.495 μg cm-2 for NAMK, and 0.503 and 10.661 μg cm-2 for AMK, respectively. Results of the comparison demonstrated the need for agronomy mechanism knowledge in radiative transfer model inversion.

Final Report: Large-Scale Optimization for Bayesian Inference in Complex Systems

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

Ghattas, Omar

2013-10-15

The SAGUARO (Scalable Algorithms for Groundwater Uncertainty Analysis and Robust Optimiza- tion) Project focuses on the development of scalable numerical algorithms for large-scale Bayesian inversion in complex systems that capitalize on advances in large-scale simulation-based optimiza- tion and inversion methods. Our research is directed in three complementary areas: efficient approximations of the Hessian operator, reductions in complexity of forward simulations via stochastic spectral approximations and model reduction, and employing large-scale optimization concepts to accelerate sampling. Our efforts are integrated in the context of a challenging testbed problem that considers subsurface reacting flow and transport. The MIT component of the SAGUAROmore » Project addresses the intractability of conventional sampling methods for large-scale statistical inverse problems by devising reduced-order models that are faithful to the full-order model over a wide range of parameter values; sampling then employs the reduced model rather than the full model, resulting in very large computational savings. Results indicate little effect on the computed posterior distribution. On the other hand, in the Texas-Georgia Tech component of the project, we retain the full-order model, but exploit inverse problem structure (adjoint-based gradients and partial Hessian information of the parameter-to- observation map) to implicitly extract lower dimensional information on the posterior distribution; this greatly speeds up sampling methods, so that fewer sampling points are needed. We can think of these two approaches as "reduce then sample" and "sample then reduce." In fact, these two approaches are complementary, and can be used in conjunction with each other. Moreover, they both exploit deterministic inverse problem structure, in the form of adjoint-based gradient and Hessian information of the underlying parameter-to-observation map, to achieve their speedups.« less

Lithological and Surface Geometry Joint Inversions Using Multi-Objective Global Optimization Methods

NASA Astrophysics Data System (ADS)

Lelièvre, Peter; Bijani, Rodrigo; Farquharson, Colin

2016-04-01

Geologists' interpretations about the Earth typically involve distinct rock units with contacts (interfaces) between them. In contrast, standard minimum-structure geophysical inversions are performed on meshes of space-filling cells (typically prisms or tetrahedra) and recover smoothly varying physical property distributions that are inconsistent with typical geological interpretations. There are several approaches through which mesh-based minimum-structure geophysical inversion can help recover models with some of the desired characteristics. However, a more effective strategy may be to consider two fundamentally different types of inversions: lithological and surface geometry inversions. A major advantage of these two inversion approaches is that joint inversion of multiple types of geophysical data is greatly simplified. In a lithological inversion, the subsurface is discretized into a mesh and each cell contains a particular rock type. A lithological model must be translated to a physical property model before geophysical data simulation. Each lithology may map to discrete property values or there may be some a priori probability density function associated with the mapping. Through this mapping, lithological inverse problems limit the parameter domain and consequently reduce the non-uniqueness from that presented by standard mesh-based inversions that allow physical property values on continuous ranges. Furthermore, joint inversion is greatly simplified because no additional mathematical coupling measure is required in the objective function to link multiple physical property models. In a surface geometry inversion, the model comprises wireframe surfaces representing contacts between rock units. This parameterization is then fully consistent with Earth models built by geologists, which in 3D typically comprise wireframe contact surfaces of tessellated triangles. As for the lithological case, the physical properties of the units lying between the contact surfaces are set to a priori values. The inversion is tasked with calculating the geometry of the contact surfaces instead of some piecewise distribution of properties in a mesh. Again, no coupling measure is required and joint inversion is simplified. Both of these inverse problems involve high nonlinearity and discontinuous or non-obtainable derivatives. They can also involve the existence of multiple minima. Hence, one can not apply the standard descent-based local minimization methods used to solve typical minimum-structure inversions. Instead, we are applying Pareto multi-objective global optimization (PMOGO) methods, which generate a suite of solutions that minimize multiple objectives (e.g. data misfits and regularization terms) in a Pareto-optimal sense. Providing a suite of models, as opposed to a single model that minimizes a weighted sum of objectives, allows a more complete assessment of the possibilities and avoids the often difficult choice of how to weight each objective. While there are definite advantages to PMOGO joint inversion approaches, the methods come with significantly increased computational requirements. We are researching various strategies to ameliorate these computational issues including parallelization and problem dimension reduction.

Estimating uncertainties in complex joint inverse problems

NASA Astrophysics Data System (ADS)

Afonso, Juan Carlos

2016-04-01

Sources of uncertainty affecting geophysical inversions can be classified either as reflective (i.e. the practitioner is aware of her/his ignorance) or non-reflective (i.e. the practitioner does not know that she/he does not know!). Although we should be always conscious of the latter, the former are the ones that, in principle, can be estimated either empirically (by making measurements or collecting data) or subjectively (based on the experience of the researchers). For complex parameter estimation problems in geophysics, subjective estimation of uncertainty is the most common type. In this context, probabilistic (aka Bayesian) methods are commonly claimed to offer a natural and realistic platform from which to estimate model uncertainties. This is because in the Bayesian approach, errors (whatever their nature) can be naturally included as part of the global statistical model, the solution of which represents the actual solution to the inverse problem. However, although we agree that probabilistic inversion methods are the most powerful tool for uncertainty estimation, the common claim that they produce "realistic" or "representative" uncertainties is not always justified. Typically, ALL UNCERTAINTY ESTIMATES ARE MODEL DEPENDENT, and therefore, besides a thorough characterization of experimental uncertainties, particular care must be paid to the uncertainty arising from model errors and input uncertainties. We recall here two quotes by G. Box and M. Gunzburger, respectively, of special significance for inversion practitioners and for this session: "…all models are wrong, but some are useful" and "computational results are believed by no one, except the person who wrote the code". In this presentation I will discuss and present examples of some problems associated with the estimation and quantification of uncertainties in complex multi-observable probabilistic inversions, and how to address them. Although the emphasis will be on sources of uncertainty related to the forward and statistical models, I will also address other uncertainties associated with data and uncertainty propagation.

Visco-acoustic wave-equation traveltime inversion and its sensitivity to attenuation errors

NASA Astrophysics Data System (ADS)

Yu, Han; Chen, Yuqing; Hanafy, Sherif M.; Huang, Jiangping

2018-04-01

A visco-acoustic wave-equation traveltime inversion method is presented that inverts for the shallow subsurface velocity distribution. Similar to the classical wave equation traveltime inversion, this method finds the velocity model that minimizes the squared sum of the traveltime residuals. Even though, wave-equation traveltime inversion can partly avoid the cycle skipping problem, a good initial velocity model is required for the inversion to converge to a reasonable tomogram with different attenuation profiles. When Q model is far away from the real model, the final tomogram is very sensitive to the starting velocity model. Nevertheless, a minor or moderate perturbation of the Q model from the true one does not strongly affect the inversion if the low wavenumber information of the initial velocity model is mostly correct. These claims are validated with numerical tests on both the synthetic and field data sets.

Model-based elastography: a survey of approaches to the inverse elasticity problem

PubMed Central

Doyley, M M

2012-01-01

Elastography is emerging as an imaging modality that can distinguish normal versus diseased tissues via their biomechanical properties. This article reviews current approaches to elastography in three areas — quasi-static, harmonic, and transient — and describes inversion schemes for each elastographic imaging approach. Approaches include: first-order approximation methods; direct and iterative inversion schemes for linear elastic; isotropic materials; and advanced reconstruction methods for recovering parameters that characterize complex mechanical behavior. The paper’s objective is to document efforts to develop elastography within the framework of solving an inverse problem, so that elastography may provide reliable estimates of shear modulus and other mechanical parameters. We discuss issues that must be addressed if model-based elastography is to become the prevailing approach to quasi-static, harmonic, and transient elastography: (1) developing practical techniques to transform the ill-posed problem with a well-posed one; (2) devising better forward models to capture the transient behavior of soft tissue; and (3) developing better test procedures to evaluate the performance of modulus elastograms. PMID:22222839

Resolution enhancement of robust Bayesian pre-stack inversion in the frequency domain

NASA Astrophysics Data System (ADS)

Yin, Xingyao; Li, Kun; Zong, Zhaoyun

2016-10-01

AVO/AVA (amplitude variation with an offset or angle) inversion is one of the most practical and useful approaches to estimating model parameters. So far, publications on AVO inversion in the Fourier domain have been quite limited in view of its poor stability and sensitivity to noise compared with time-domain inversion. For the resolution and stability of AVO inversion in the Fourier domain, a novel robust Bayesian pre-stack AVO inversion based on the mixed domain formulation of stationary convolution is proposed which could solve the instability and achieve superior resolution. The Fourier operator will be integrated into the objective equation and it avoids the Fourier inverse transform in our inversion process. Furthermore, the background constraints of model parameters are taken into consideration to improve the stability and reliability of inversion which could compensate for the low-frequency components of seismic signals. Besides, the different frequency components of seismic signals can realize decoupling automatically. This will help us to solve the inverse problem by means of multi-component successive iterations and the convergence precision of the inverse problem could be improved. So, superior resolution compared with the conventional time-domain pre-stack inversion could be achieved easily. Synthetic tests illustrate that the proposed method could achieve high-resolution results with a high degree of agreement with the theoretical model and verify the quality of anti-noise. Finally, applications on a field data case demonstrate that the proposed method could obtain stable inversion results of elastic parameters from pre-stack seismic data in conformity with the real logging data.

An inverse problem for a semilinear parabolic equation arising from cardiac electrophysiology

NASA Astrophysics Data System (ADS)

Beretta, Elena; Cavaterra, Cecilia; Cerutti, M. Cristina; Manzoni, Andrea; Ratti, Luca

2017-10-01

In this paper we develop theoretical analysis and numerical reconstruction techniques for the solution of an inverse boundary value problem dealing with the nonlinear, time-dependent monodomain equation, which models the evolution of the electric potential in the myocardial tissue. The goal is the detection of an inhomogeneity \

A posteriori error estimates in voice source recovery

NASA Astrophysics Data System (ADS)

Leonov, A. S.; Sorokin, V. N.

2017-12-01

The inverse problem of voice source pulse recovery from a segment of a speech signal is under consideration. A special mathematical model is used for the solution that relates these quantities. A variational method of solving inverse problem of voice source recovery for a new parametric class of sources, that is for piecewise-linear sources (PWL-sources), is proposed. Also, a technique for a posteriori numerical error estimation for obtained solutions is presented. A computer study of the adequacy of adopted speech production model with PWL-sources is performed in solving the inverse problems for various types of voice signals, as well as corresponding study of a posteriori error estimates. Numerical experiments for speech signals show satisfactory properties of proposed a posteriori error estimates, which represent the upper bounds of possible errors in solving the inverse problem. The estimate of the most probable error in determining the source-pulse shapes is about 7-8% for the investigated speech material. It is noted that a posteriori error estimates can be used as a criterion of the quality for obtained voice source pulses in application to speaker recognition.

An optimization method for the problems of thermal cloaking of material bodies

NASA Astrophysics Data System (ADS)

Alekseev, G. V.; Levin, V. A.

2016-11-01

Inverse heat-transfer problems related to constructing special thermal devices such as cloaking shells, thermal-illusion or thermal-camouflage devices, and heat-flux concentrators are studied. The heatdiffusion equation with a variable heat-conductivity coefficient is used as the initial heat-transfer model. An optimization method is used to reduce the above inverse problems to the respective control problem. The solvability of the above control problem is proved, an optimality system that describes necessary extremum conditions is derived, and a numerical algorithm for solving the control problem is proposed.

Applications of Bayesian spectrum representation in acoustics

NASA Astrophysics Data System (ADS)

Botts, Jonathan M.

This dissertation utilizes a Bayesian inference framework to enhance the solution of inverse problems where the forward model maps to acoustic spectra. A Bayesian solution to filter design inverts a acoustic spectra to pole-zero locations of a discrete-time filter model. Spatial sound field analysis with a spherical microphone array is a data analysis problem that requires inversion of spatio-temporal spectra to directions of arrival. As with many inverse problems, a probabilistic analysis results in richer solutions than can be achieved with ad-hoc methods. In the filter design problem, the Bayesian inversion results in globally optimal coefficient estimates as well as an estimate the most concise filter capable of representing the given spectrum, within a single framework. This approach is demonstrated on synthetic spectra, head-related transfer function spectra, and measured acoustic reflection spectra. The Bayesian model-based analysis of spatial room impulse responses is presented as an analogous problem with equally rich solution. The model selection mechanism provides an estimate of the number of arrivals, which is necessary to properly infer the directions of simultaneous arrivals. Although, spectrum inversion problems are fairly ubiquitous, the scope of this dissertation has been limited to these two and derivative problems. The Bayesian approach to filter design is demonstrated on an artificial spectrum to illustrate the model comparison mechanism and then on measured head-related transfer functions to show the potential range of application. Coupled with sampling methods, the Bayesian approach is shown to outperform least-squares filter design methods commonly used in commercial software, confirming the need for a global search of the parameter space. The resulting designs are shown to be comparable to those that result from global optimization methods, but the Bayesian approach has the added advantage of a filter length estimate within the same unified framework. The application to reflection data is useful for representing frequency-dependent impedance boundaries in finite difference acoustic simulations. Furthermore, since the filter transfer function is a parametric model, it can be modified to incorporate arbitrary frequency weighting and account for the band-limited nature of measured reflection spectra. Finally, the model is modified to compensate for dispersive error in the finite difference simulation, from the filter design process. Stemming from the filter boundary problem, the implementation of pressure sources in finite difference simulation is addressed in order to assure that schemes properly converge. A class of parameterized source functions is proposed and shown to offer straightforward control of residual error in the simulation. Guided by the notion that the solution to be approximated affects the approximation error, sources are designed which reduce residual dispersive error to the size of round-off errors. The early part of a room impulse response can be characterized by a series of isolated plane waves. Measured with an array of microphones, plane waves map to a directional response of the array or spatial intensity map. Probabilistic inversion of this response results in estimates of the number and directions of image source arrivals. The model-based inversion is shown to avoid ambiguities associated with peak-finding or inspection of the spatial intensity map. For this problem, determining the number of arrivals in a given frame is critical for properly inferring the state of the sound field. This analysis is effectively compression of the spatial room response, which is useful for analysis or encoding of the spatial sound field. Parametric, model-based formulations of these problems enhance the solution in all cases, and a Bayesian interpretation provides a principled approach to model comparison and parameter estimation. v

Probability density of spatially distributed soil moisture inferred from crosshole georadar traveltime measurements

NASA Astrophysics Data System (ADS)

Linde, N.; Vrugt, J. A.

2009-04-01

Geophysical models are increasingly used in hydrological simulations and inversions, where they are typically treated as an artificial data source with known uncorrelated "data errors". The model appraisal problem in classical deterministic linear and non-linear inversion approaches based on linearization is often addressed by calculating model resolution and model covariance matrices. These measures offer only a limited potential to assign a more appropriate "data covariance matrix" for future hydrological applications, simply because the regularization operators used to construct a stable inverse solution bear a strong imprint on such estimates and because the non-linearity of the geophysical inverse problem is not explored. We present a parallelized Markov Chain Monte Carlo (MCMC) scheme to efficiently derive the posterior spatially distributed radar slowness and water content between boreholes given first-arrival traveltimes. This method is called DiffeRential Evolution Adaptive Metropolis (DREAM_ZS) with snooker updater and sampling from past states. Our inverse scheme does not impose any smoothness on the final solution, and uses uniform prior ranges of the parameters. The posterior distribution of radar slowness is converted into spatially distributed soil moisture values using a petrophysical relationship. To benchmark the performance of DREAM_ZS, we first apply our inverse method to a synthetic two-dimensional infiltration experiment using 9421 traveltimes contaminated with Gaussian errors and 80 different model parameters, corresponding to a model discretization of 0.3 m × 0.3 m. After this, the method is applied to field data acquired in the vadose zone during snowmelt. This work demonstrates that fully non-linear stochastic inversion can be applied with few limiting assumptions to a range of common two-dimensional tomographic geophysical problems. The main advantage of DREAM_ZS is that it provides a full view of the posterior distribution of spatially distributed soil moisture, which is key to appropriately treat geophysical parameter uncertainty and infer hydrologic models.

Inverse random source scattering for the Helmholtz equation in inhomogeneous media

NASA Astrophysics Data System (ADS)

Li, Ming; Chen, Chuchu; Li, Peijun

2018-01-01

This paper is concerned with an inverse random source scattering problem in an inhomogeneous background medium. The wave propagation is modeled by the stochastic Helmholtz equation with the source driven by additive white noise. The goal is to reconstruct the statistical properties of the random source such as the mean and variance from the boundary measurement of the radiated random wave field at multiple frequencies. Both the direct and inverse problems are considered. We show that the direct problem has a unique mild solution by a constructive proof. For the inverse problem, we derive Fredholm integral equations, which connect the boundary measurement of the radiated wave field with the unknown source function. A regularized block Kaczmarz method is developed to solve the ill-posed integral equations. Numerical experiments are included to demonstrate the effectiveness of the proposed method.

Viscoelastic material inversion using Sierra-SD and ROL

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

Walsh, Timothy; Aquino, Wilkins; Ridzal, Denis

2014-11-01

In this report we derive frequency-domain methods for inverse characterization of the constitutive parameters of viscoelastic materials. The inverse problem is cast in a PDE-constrained optimization framework with efficient computation of gradients and Hessian vector products through matrix free operations. The abstract optimization operators for first and second derivatives are derived from first principles. Various methods from the Rapid Optimization Library (ROL) are tested on the viscoelastic inversion problem. The methods described herein are applied to compute the viscoelastic bulk and shear moduli of a foam block model, which was recently used in experimental testing for viscoelastic property characterization.

Adaptive multi-step Full Waveform Inversion based on Waveform Mode Decomposition

NASA Astrophysics Data System (ADS)

Hu, Yong; Han, Liguo; Xu, Zhuo; Zhang, Fengjiao; Zeng, Jingwen

2017-04-01

Full Waveform Inversion (FWI) can be used to build high resolution velocity models, but there are still many challenges in seismic field data processing. The most difficult problem is about how to recover long-wavelength components of subsurface velocity models when seismic data is lacking of low frequency information and without long-offsets. To solve this problem, we propose to use Waveform Mode Decomposition (WMD) method to reconstruct low frequency information for FWI to obtain a smooth model, so that the initial model dependence of FWI can be reduced. In this paper, we use adjoint-state method to calculate the gradient for Waveform Mode Decomposition Full Waveform Inversion (WMDFWI). Through the illustrative numerical examples, we proved that the low frequency which is reconstructed by WMD method is very reliable. WMDFWI in combination with the adaptive multi-step inversion strategy can obtain more faithful and accurate final inversion results. Numerical examples show that even if the initial velocity model is far from the true model and lacking of low frequency information, we still can obtain good inversion results with WMD method. From numerical examples of anti-noise test, we see that the adaptive multi-step inversion strategy for WMDFWI has strong ability to resist Gaussian noise. WMD method is promising to be able to implement for the land seismic FWI, because it can reconstruct the low frequency information, lower the dominant frequency in the adjoint source, and has a strong ability to resist noise.

Atmospheric Tracer Inverse Modeling Using Markov Chain Monte Carlo (MCMC)

NASA Astrophysics Data System (ADS)

Kasibhatla, P.

2004-12-01

In recent years, there has been an increasing emphasis on the use of Bayesian statistical estimation techniques to characterize the temporal and spatial variability of atmospheric trace gas sources and sinks. The applications have been varied in terms of the particular species of interest, as well as in terms of the spatial and temporal resolution of the estimated fluxes. However, one common characteristic has been the use of relatively simple statistical models for describing the measurement and chemical transport model error statistics and prior source statistics. For example, multivariate normal probability distribution functions (pdfs) are commonly used to model these quantities and inverse source estimates are derived for fixed values of pdf paramaters. While the advantage of this approach is that closed form analytical solutions for the a posteriori pdfs of interest are available, it is worth exploring Bayesian analysis approaches which allow for a more general treatment of error and prior source statistics. Here, we present an application of the Markov Chain Monte Carlo (MCMC) methodology to an atmospheric tracer inversion problem to demonstrate how more gereral statistical models for errors can be incorporated into the analysis in a relatively straightforward manner. The MCMC approach to Bayesian analysis, which has found wide application in a variety of fields, is a statistical simulation approach that involves computing moments of interest of the a posteriori pdf by efficiently sampling this pdf. The specific inverse problem that we focus on is the annual mean CO2 source/sink estimation problem considered by the TransCom3 project. TransCom3 was a collaborative effort involving various modeling groups and followed a common modeling and analysis protocoal. As such, this problem provides a convenient case study to demonstrate the applicability of the MCMC methodology to atmospheric tracer source/sink estimation problems.

The inverse problems of wing panel manufacture processes

NASA Astrophysics Data System (ADS)

Oleinikov, A. I.; Bormotin, K. S.

2013-12-01

It is shown that inverse problems of steady-state creep bending of plates in both the geometrically linear and nonlinear formulations can be represented in a variational formulation. Steady-state values of the obtained functionals corresponding to the solutions of the problems of inelastic deformation and springback are determined by applying a finite element procedure to the functionals. Optimal laws of creep deformation are formulated using the criterion of minimizing damage in the functionals of the inverse problems. The formulated problems are reduced to the problems solved by the finite element method using MSC.Marc software. Currently, forming of light metals poses tremendous challenges due to their low ductility at room temperature and their unusual deformation characteristics at hot-cold work: strong asymmetry between tensile and compressive behavior, and a very pronounced anisotropy. We used the constitutive models of steady-state creep of initially transverse isotropy structural materials the kind of the stress state has influence. The paper gives basics of the developed computer-aided system of design, modeling, and electronic simulation targeting the processes of manufacture of wing integral panels. The modeling results can be used to calculate the die tooling, determine the panel processibility, and control panel rejection in the course of forming.

Solving geosteering inverse problems by stochastic Hybrid Monte Carlo method

DOE PAGES

Shen, Qiuyang; Wu, Xuqing; Chen, Jiefu; ...

2017-11-20

The inverse problems arise in almost all fields of science where the real-world parameters are extracted from a set of measured data. The geosteering inversion plays an essential role in the accurate prediction of oncoming strata as well as a reliable guidance to adjust the borehole position on the fly to reach one or more geological targets. This mathematical treatment is not easy to solve, which requires finding an optimum solution among a large solution space, especially when the problem is non-linear and non-convex. Nowadays, a new generation of logging-while-drilling (LWD) tools has emerged on the market. The so-called azimuthalmore » resistivity LWD tools have azimuthal sensitivity and a large depth of investigation. Hence, the associated inverse problems become much more difficult since the earth model to be inverted will have more detailed structures. The conventional deterministic methods are incapable to solve such a complicated inverse problem, where they suffer from the local minimum trap. Alternatively, stochastic optimizations are in general better at finding global optimal solutions and handling uncertainty quantification. In this article, we investigate the Hybrid Monte Carlo (HMC) based statistical inversion approach and suggest that HMC based inference is more efficient in dealing with the increased complexity and uncertainty faced by the geosteering problems.« less

Group-theoretic models of the inversion process in bacterial genomes.

PubMed

Egri-Nagy, Attila; Gebhardt, Volker; Tanaka, Mark M; Francis, Andrew R

2014-07-01

The variation in genome arrangements among bacterial taxa is largely due to the process of inversion. Recent studies indicate that not all inversions are equally probable, suggesting, for instance, that shorter inversions are more frequent than longer, and those that move the terminus of replication are less probable than those that do not. Current methods for establishing the inversion distance between two bacterial genomes are unable to incorporate such information. In this paper we suggest a group-theoretic framework that in principle can take these constraints into account. In particular, we show that by lifting the problem from circular permutations to the affine symmetric group, the inversion distance can be found in polynomial time for a model in which inversions are restricted to acting on two regions. This requires the proof of new results in group theory, and suggests a vein of new combinatorial problems concerning permutation groups on which group theorists will be needed to collaborate with biologists. We apply the new method to inferring distances and phylogenies for published Yersinia pestis data.

Inverse Problems in Complex Models and Applications to Earth Sciences

NASA Astrophysics Data System (ADS)

Bosch, M. E.

2015-12-01

The inference of the subsurface earth structure and properties requires the integration of different types of data, information and knowledge, by combined processes of analysis and synthesis. To support the process of integrating information, the regular concept of data inversion is evolving to expand its application to models with multiple inner components (properties, scales, structural parameters) that explain multiple data (geophysical survey data, well-logs, core data). The probabilistic inference methods provide the natural framework for the formulation of these problems, considering a posterior probability density function (PDF) that combines the information from a prior information PDF and the new sets of observations. To formulate the posterior PDF in the context of multiple datasets, the data likelihood functions are factorized assuming independence of uncertainties for data originating across different surveys. A realistic description of the earth medium requires modeling several properties and structural parameters, which relate to each other according to dependency and independency notions. Thus, conditional probabilities across model components also factorize. A common setting proceeds by structuring the model parameter space in hierarchical layers. A primary layer (e.g. lithology) conditions a secondary layer (e.g. physical medium properties), which conditions a third layer (e.g. geophysical data). In general, less structured relations within model components and data emerge from the analysis of other inverse problems. They can be described with flexibility via direct acyclic graphs, which are graphs that map dependency relations between the model components. Examples of inverse problems in complex models can be shown at various scales. At local scale, for example, the distribution of gas saturation is inferred from pre-stack seismic data and a calibrated rock-physics model. At regional scale, joint inversion of gravity and magnetic data is applied for the estimation of lithological structure of the crust, with the lithotype body regions conditioning the mass density and magnetic susceptibility fields. At planetary scale, the Earth mantle temperature and element composition is inferred from seismic travel-time and geodetic data.

Decoupled Method for Reconstruction of Surface Conditions From Internal Temperatures On Ablative Materials With Uncertain Recession Model

NASA Technical Reports Server (NTRS)

Oliver, A. Brandon

2017-01-01

Obtaining measurements of flight environments on ablative heat shields is both critical for spacecraft development and extremely challenging due to the harsh heating environment and surface recession. Thermocouples installed several millimeters below the surface are commonly used to measure the heat shield temperature response, but an ill-posed inverse heat conduction problem must be solved to reconstruct the surface heating environment from these measurements. Ablation can contribute substantially to the measurement response making solutions to the inverse problem strongly dependent on the recession model, which is often poorly characterized. To enable efficient surface reconstruction for recession model sensitivity analysis, a method for decoupling the surface recession evaluation from the inverse heat conduction problem is presented. The decoupled method is shown to provide reconstructions of equivalent accuracy to the traditional coupled method but with substantially reduced computational effort. These methods are applied to reconstruct the environments on the Mars Science Laboratory heat shield using diffusion limit and kinetically limited recession models.

Comparing multiple statistical methods for inverse prediction in nuclear forensics applications

DOE PAGES

Lewis, John R.; Zhang, Adah; Anderson-Cook, Christine Michaela

2017-10-29

Forensic science seeks to predict source characteristics using measured observables. Statistically, this objective can be thought of as an inverse problem where interest is in the unknown source characteristics or factors ( X) of some underlying causal model producing the observables or responses (Y = g ( X) + error). Here, this paper reviews several statistical methods for use in inverse problems and demonstrates that comparing results from multiple methods can be used to assess predictive capability. Motivation for assessing inverse predictions comes from the desired application to historical and future experiments involving nuclear material production for forensics research inmore » which inverse predictions, along with an assessment of predictive capability, are desired.« less

Comparing multiple statistical methods for inverse prediction in nuclear forensics applications

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

Lewis, John R.; Zhang, Adah; Anderson-Cook, Christine Michaela

Forensic science seeks to predict source characteristics using measured observables. Statistically, this objective can be thought of as an inverse problem where interest is in the unknown source characteristics or factors ( X) of some underlying causal model producing the observables or responses (Y = g ( X) + error). Here, this paper reviews several statistical methods for use in inverse problems and demonstrates that comparing results from multiple methods can be used to assess predictive capability. Motivation for assessing inverse predictions comes from the desired application to historical and future experiments involving nuclear material production for forensics research inmore » which inverse predictions, along with an assessment of predictive capability, are desired.« less

Approximation of the ruin probability using the scaled Laplace transform inversion

PubMed Central

Mnatsakanov, Robert M.; Sarkisian, Khachatur; Hakobyan, Artak

2015-01-01

The problem of recovering the ruin probability in the classical risk model based on the scaled Laplace transform inversion is studied. It is shown how to overcome the problem of evaluating the ruin probability at large values of an initial surplus process. Comparisons of proposed approximations with the ones based on the Laplace transform inversions using a fixed Talbot algorithm as well as on the ones using the Trefethen–Weideman–Schmelzer and maximum entropy methods are presented via a simulation study. PMID:26752796

3D gravity inversion and uncertainty assessment of basement relief via Particle Swarm Optimization

NASA Astrophysics Data System (ADS)

Pallero, J. L. G.; Fernández-Martínez, J. L.; Bonvalot, S.; Fudym, O.

2017-04-01

Nonlinear gravity inversion in sedimentary basins is a classical problem in applied geophysics. Although a 2D approximation is widely used, 3D models have been also proposed to better take into account the basin geometry. A common nonlinear approach to this 3D problem consists in modeling the basin as a set of right rectangular prisms with prescribed density contrast, whose depths are the unknowns. Then, the problem is iteratively solved via local optimization techniques from an initial model computed using some simplifications or being estimated using prior geophysical models. Nevertheless, this kind of approach is highly dependent on the prior information that is used, and lacks from a correct solution appraisal (nonlinear uncertainty analysis). In this paper, we use the family of global Particle Swarm Optimization (PSO) optimizers for the 3D gravity inversion and model appraisal of the solution that is adopted for basement relief estimation in sedimentary basins. Synthetic and real cases are illustrated, showing that robust results are obtained. Therefore, PSO seems to be a very good alternative for 3D gravity inversion and uncertainty assessment of basement relief when used in a sampling while optimizing approach. That way important geological questions can be answered probabilistically in order to perform risk assessment in the decisions that are made.

PREFACE: First International Congress of the International Association of Inverse Problems (IPIA): Applied Inverse Problems 2007: Theoretical and Computational Aspects

NASA Astrophysics Data System (ADS)

Uhlmann, Gunther

2008-07-01

This volume represents the proceedings of the fourth Applied Inverse Problems (AIP) international conference and the first congress of the Inverse Problems International Association (IPIA) which was held in Vancouver, Canada, June 25 29, 2007. The organizing committee was formed by Uri Ascher, University of British Columbia, Richard Froese, University of British Columbia, Gary Margrave, University of Calgary, and Gunther Uhlmann, University of Washington, chair. The conference was part of the activities of the Pacific Institute of Mathematical Sciences (PIMS) Collaborative Research Group on inverse problems (http://www.pims.math.ca/scientific/collaborative-research-groups/past-crgs). This event was also supported by grants from NSF and MITACS. Inverse Problems (IP) are problems where causes for a desired or an observed effect are to be determined. They lie at the heart of scientific inquiry and technological development. The enormous increase in computing power and the development of powerful algorithms have made it possible to apply the techniques of IP to real-world problems of growing complexity. Applications include a number of medical as well as other imaging techniques, location of oil and mineral deposits in the earth's substructure, creation of astrophysical images from telescope data, finding cracks and interfaces within materials, shape optimization, model identification in growth processes and, more recently, modelling in the life sciences. The series of Applied Inverse Problems (AIP) Conferences aims to provide a primary international forum for academic and industrial researchers working on all aspects of inverse problems, such as mathematical modelling, functional analytic methods, computational approaches, numerical algorithms etc. The steering committee of the AIP conferences consists of Heinz Engl (Johannes Kepler Universität, Austria), Joyce McLaughlin (RPI, USA), William Rundell (Texas A&M, USA), Erkki Somersalo (Helsinki University of Technology, Finland), Masahiro Yamamoto (University of Tokyo, Japan), Gunther Uhlmann (University of Washington) and Jun Zou (Chinese University of Hong Kong). IPIA is a recently formed organization that intends to promote the field of inverse problem at all levels. See http://www.inverse-problems.net/. IPIA awarded the first Calderón prize at the opening of the conference to Matti Lassas (see first article in the Proceedings). There was also a general meeting of IPIA during the workshop. This was probably the largest conference ever on IP with 350 registered participants. The program consisted of 18 invited speakers and the Calderón Prize Lecture given by Matti Lassas. Another integral part of the program was the more than 60 mini-symposia that covered a broad spectrum of the theory and applications of inverse problems, focusing on recent developments in medical imaging, seismic exploration, remote sensing, industrial applications, numerical and regularization methods in inverse problems. Another important related topic was image processing in particular the advances which have allowed for significant enhancement of widely used imaging techniques. For more details on the program see the web page: http://www.pims.math.ca/science/2007/07aip. These proceedings reflect the broad spectrum of topics covered in AIP 2007. The conference and these proceedings would not have happened without the contributions of many people. I thank all my fellow organizers, the invited speakers, the speakers and organizers of mini-symposia for making this an exciting and vibrant event. I also thank PIMS, NSF and MITACS for their generous financial support. I take this opportunity to thank the PIMS staff, particularly Ken Leung, for making the local arrangements. Also thanks are due to Stephen McDowall for his help in preparing the schedule of the conference and Xiaosheng Li for the help in preparing these proceedings. I also would like to thank the contributors of this volume and the referees. Finally, many thanks are due to Graham Douglas and Elaine Longden-Chapman for suggesting publication in Journal of Physics: Conference Series.

Wavelet extractor: A Bayesian well-tie and wavelet extraction program

NASA Astrophysics Data System (ADS)

Gunning, James; Glinsky, Michael E.

2006-06-01

We introduce a new open-source toolkit for the well-tie or wavelet extraction problem of estimating seismic wavelets from seismic data, time-to-depth information, and well-log suites. The wavelet extraction model is formulated as a Bayesian inverse problem, and the software will simultaneously estimate wavelet coefficients, other parameters associated with uncertainty in the time-to-depth mapping, positioning errors in the seismic imaging, and useful amplitude-variation-with-offset (AVO) related parameters in multi-stack extractions. It is capable of multi-well, multi-stack extractions, and uses continuous seismic data-cube interpolation to cope with the problem of arbitrary well paths. Velocity constraints in the form of checkshot data, interpreted markers, and sonic logs are integrated in a natural way. The Bayesian formulation allows computation of full posterior uncertainties of the model parameters, and the important problem of the uncertain wavelet span is addressed uses a multi-model posterior developed from Bayesian model selection theory. The wavelet extraction tool is distributed as part of the Delivery seismic inversion toolkit. A simple log and seismic viewing tool is included in the distribution. The code is written in Java, and thus platform independent, but the Seismic Unix (SU) data model makes the inversion particularly suited to Unix/Linux environments. It is a natural companion piece of software to Delivery, having the capacity to produce maximum likelihood wavelet and noise estimates, but will also be of significant utility to practitioners wanting to produce wavelet estimates for other inversion codes or purposes. The generation of full parameter uncertainties is a crucial function for workers wishing to investigate questions of wavelet stability before proceeding to more advanced inversion studies.

Bayesian Inversion of 2D Models from Airborne Transient EM Data

NASA Astrophysics Data System (ADS)

Blatter, D. B.; Key, K.; Ray, A.

2016-12-01

The inherent non-uniqueness in most geophysical inverse problems leads to an infinite number of Earth models that fit observed data to within an adequate tolerance. To resolve this ambiguity, traditional inversion methods based on optimization techniques such as the Gauss-Newton and conjugate gradient methods rely on an additional regularization constraint on the properties that an acceptable model can possess, such as having minimal roughness. While allowing such an inversion scheme to converge on a solution, regularization makes it difficult to estimate the uncertainty associated with the model parameters. This is because regularization biases the inversion process toward certain models that satisfy the regularization constraint and away from others that don't, even when both may suitably fit the data. By contrast, a Bayesian inversion framework aims to produce not a single `most acceptable' model but an estimate of the posterior likelihood of the model parameters, given the observed data. In this work, we develop a 2D Bayesian framework for the inversion of transient electromagnetic (TEM) data. Our method relies on a reversible-jump Markov Chain Monte Carlo (RJ-MCMC) Bayesian inverse method with parallel tempering. Previous gradient-based inversion work in this area used a spatially constrained scheme wherein individual (1D) soundings were inverted together and non-uniqueness was tackled by using lateral and vertical smoothness constraints. By contrast, our work uses a 2D model space of Voronoi cells whose parameterization (including number of cells) is fully data-driven. To make the problem work practically, we approximate the forward solution for each TEM sounding using a local 1D approximation where the model is obtained from the 2D model by retrieving a vertical profile through the Voronoi cells. The implicit parsimony of the Bayesian inversion process leads to the simplest models that adequately explain the data, obviating the need for explicit smoothness constraints. In addition, credible intervals in model space are directly obtained, resolving some of the uncertainty introduced by regularization. An example application shows how the method can be used to quantify the uncertainty in airborne EM soundings for imaging subglacial brine channels and groundwater systems.

Metamodel-based inverse method for parameter identification: elastic-plastic damage model

NASA Astrophysics Data System (ADS)

Huang, Changwu; El Hami, Abdelkhalak; Radi, Bouchaïb

2017-04-01

This article proposed a metamodel-based inverse method for material parameter identification and applies it to elastic-plastic damage model parameter identification. An elastic-plastic damage model is presented and implemented in numerical simulation. The metamodel-based inverse method is proposed in order to overcome the disadvantage in computational cost of the inverse method. In the metamodel-based inverse method, a Kriging metamodel is constructed based on the experimental design in order to model the relationship between material parameters and the objective function values in the inverse problem, and then the optimization procedure is executed by the use of a metamodel. The applications of the presented material model and proposed parameter identification method in the standard A 2017-T4 tensile test prove that the presented elastic-plastic damage model is adequate to describe the material's mechanical behaviour and that the proposed metamodel-based inverse method not only enhances the efficiency of parameter identification but also gives reliable results.

Prediction and assimilation of surf-zone processes using a Bayesian network: Part II: Inverse models

USGS Publications Warehouse

Plant, Nathaniel G.; Holland, K. Todd

2011-01-01

A Bayesian network model has been developed to simulate a relatively simple problem of wave propagation in the surf zone (detailed in Part I). Here, we demonstrate that this Bayesian model can provide both inverse modeling and data-assimilation solutions for predicting offshore wave heights and depth estimates given limited wave-height and depth information from an onshore location. The inverse method is extended to allow data assimilation using observational inputs that are not compatible with deterministic solutions of the problem. These inputs include sand bar positions (instead of bathymetry) and estimates of the intensity of wave breaking (instead of wave-height observations). Our results indicate that wave breaking information is essential to reduce prediction errors. In many practical situations, this information could be provided from a shore-based observer or from remote-sensing systems. We show that various combinations of the assimilated inputs significantly reduce the uncertainty in the estimates of water depths and wave heights in the model domain. Application of the Bayesian network model to new field data demonstrated significant predictive skill (R2 = 0.7) for the inverse estimate of a month-long time series of offshore wave heights. The Bayesian inverse results include uncertainty estimates that were shown to be most accurate when given uncertainty in the inputs (e.g., depth and tuning parameters). Furthermore, the inverse modeling was extended to directly estimate tuning parameters associated with the underlying wave-process model. The inverse estimates of the model parameters not only showed an offshore wave height dependence consistent with results of previous studies but the uncertainty estimates of the tuning parameters also explain previously reported variations in the model parameters.

Underworld results as a triple (shopping list, posterior, priors)

NASA Astrophysics Data System (ADS)

Quenette, S. M.; Moresi, L. N.; Abramson, D.

2013-12-01

When studying long-term lithosphere deformation and other such large-scale, spatially distinct and behaviour rich problems, there is a natural trade-off between the meaning of a model, the observations used to validate the model and the ability to compute over this space. For example, many models of varying lithologies, rheological properties and underlying physics may reasonably match (or not match) observables. To compound this problem, each realisation is computationally intensive, requiring high resolution, algorithm tuning and code tuning to contemporary computer hardware. It is often intractable to use sampling based assimilation methods, but with better optimisation, the window of tractability becomes wider. The ultimate goal is to find a sweet-spot where a formal assimilation method is used, and where a model affines to observations. Its natural to think of this as an inverse problem, in which the underlying physics may be fixed and the rheological properties and possibly the lithologies themselves are unknown. What happens when we push this approach and treat some portion of the underlying physics as an unknown? At its extreme this is an intractable problem. However, there is an analogy here with how we develop software for these scientific problems. What happens when we treat the changing part of a largely complete code as an unknown, where the changes are working towards this sweet-spot? When posed as a Bayesian inverse problem the result is a triple - the model changes, the real priors and the real posterior. Not only does this give meaning to the process by which a code changes, it forms a mathematical bridge from an inverse problem to compiler optimisations given such changes. As a stepping stone example we show a regional scale heat flow model with constraining observations, and the inverse process including increasingly complexity in the software. The implementation uses Underworld-GT (Underworld plus research extras to import geology and export geothermic measures, etc). Underworld uses StGermain an early (partial) implementation of the theories described here.

Inversion layer MOS solar cells

NASA Technical Reports Server (NTRS)

Ho, Fat Duen

1986-01-01

Inversion layer (IL) Metal Oxide Semiconductor (MOS) solar cells were fabricated. The fabrication technique and problems are discussed. A plan for modeling IL cells is presented. Future work in this area is addressed.

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.

Calibrating the Spatiotemporal Root Density Distribution for Macroscopic Water Uptake Models Using Tikhonov Regularization

NASA Astrophysics Data System (ADS)

Li, N.; Yue, X. Y.

2018-03-01

Macroscopic root water uptake models proportional to a root density distribution function (RDDF) are most commonly used to model water uptake by plants. As the water uptake is difficult and labor intensive to measure, these models are often calibrated by inverse modeling. Most previous inversion studies assume RDDF to be constant with depth and time or dependent on only depth for simplification. However, under field conditions, this function varies with type of soil and root growth and thus changes with both depth and time. This study proposes an inverse method to calibrate both spatially and temporally varying RDDF in unsaturated water flow modeling. To overcome the difficulty imposed by the ill-posedness, the calibration is formulated as an optimization problem in the framework of the Tikhonov regularization theory, adding additional constraint to the objective function. Then the formulated nonlinear optimization problem is numerically solved with an efficient algorithm on the basis of the finite element method. The advantage of our method is that the inverse problem is translated into a Tikhonov regularization functional minimization problem and then solved based on the variational construction, which circumvents the computational complexity in calculating the sensitivity matrix involved in many derivative-based parameter estimation approaches (e.g., Levenberg-Marquardt optimization). Moreover, the proposed method features optimization of RDDF without any prior form, which is applicable to a more general root water uptake model. Numerical examples are performed to illustrate the applicability and effectiveness of the proposed method. Finally, discussions on the stability and extension of this method are presented.

A preprocessing strategy for helioseismic inversions

NASA Astrophysics Data System (ADS)

Christensen-Dalsgaard, J.; Thompson, M. J.

1993-05-01

Helioseismic inversion in general involves considerable computational expense, due to the large number of modes that is typically considered. This is true in particular of the widely used optimally localized averages (OLA) inversion methods, which require the inversion of one or more matrices whose order is the number of modes in the set. However, the number of practically independent pieces of information that a large helioseismic mode set contains is very much less than the number of modes, suggesting that the set might first be reduced before the expensive inversion is performed. We demonstrate with a model problem that by first performing a singular value decomposition the original problem may be transformed into a much smaller one, reducing considerably the cost of the OLA inversion and with no significant loss of information.

A Bayesian method for characterizing distributed micro-releases: II. inference under model uncertainty with short time-series data.

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

Marzouk, Youssef; Fast P.; Kraus, M.

2006-01-01

Terrorist attacks using an aerosolized pathogen preparation have gained credibility as a national security concern after the anthrax attacks of 2001. The ability to characterize such attacks, i.e., to estimate the number of people infected, the time of infection, and the average dose received, is important when planning a medical response. We address this question of characterization by formulating a Bayesian inverse problem predicated on a short time-series of diagnosed patients exhibiting symptoms. To be of relevance to response planning, we limit ourselves to 3-5 days of data. In tests performed with anthrax as the pathogen, we find that thesemore » data are usually sufficient, especially if the model of the outbreak used in the inverse problem is an accurate one. In some cases the scarcity of data may initially support outbreak characterizations at odds with the true one, but with sufficient data the correct inferences are recovered; in other words, the inverse problem posed and its solution methodology are consistent. We also explore the effect of model error-situations for which the model used in the inverse problem is only a partially accurate representation of the outbreak; here, the model predictions and the observations differ by more than a random noise. We find that while there is a consistent discrepancy between the inferred and the true characterizations, they are also close enough to be of relevance when planning a response.« less

Forward and Inverse Modeling of Self-potential. A Tomography of Groundwater Flow and Comparison Between Deterministic and Stochastic Inversion Methods

NASA Astrophysics Data System (ADS)

Quintero-Chavarria, E.; Ochoa Gutierrez, L. H.

2016-12-01

Applications of the Self-potential Method in the fields of Hydrogeology and Environmental Sciences have had significant developments during the last two decades with a strong use on groundwater flows identification. Although only few authors deal with the forward problem's solution -especially in geophysics literature- different inversion procedures are currently being developed but in most cases they are compared with unconventional groundwater velocity fields and restricted to structured meshes. This research solves the forward problem based on the finite element method using the St. Venant's Principle to transform a point dipole, which is the field generated by a single vector, into a distribution of electrical monopoles. Then, two simple aquifer models were generated with specific boundary conditions and head potentials, velocity fields and electric potentials in the medium were computed. With the model's surface electric potential, the inverse problem is solved to retrieve the source of electric potential (vector field associated to groundwater flow) using deterministic and stochastic approaches. The first approach was carried out by implementing a Tikhonov regularization with a stabilized operator adapted to the finite element mesh while for the second a hierarchical Bayesian model based on Markov chain Monte Carlo (McMC) and Markov Random Fields (MRF) was constructed. For all implemented methods, the result between the direct and inverse models was contrasted in two ways: 1) shape and distribution of the vector field, and 2) magnitude's histogram. Finally, it was concluded that inversion procedures are improved when the velocity field's behavior is considered, thus, the deterministic method is more suitable for unconfined aquifers than confined ones. McMC has restricted applications and requires a lot of information (particularly in potentials fields) while MRF has a remarkable response especially when dealing with confined aquifers.

An iterative particle filter approach for coupled hydro-geophysical inversion of a controlled infiltration experiment

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

Manoli, Gabriele, E-mail: manoli@dmsa.unipd.it; Nicholas School of the Environment, Duke University, Durham, NC 27708; Rossi, Matteo

The modeling of unsaturated groundwater flow is affected by a high degree of uncertainty related to both measurement and model errors. Geophysical methods such as Electrical Resistivity Tomography (ERT) can provide useful indirect information on the hydrological processes occurring in the vadose zone. In this paper, we propose and test an iterated particle filter method to solve the coupled hydrogeophysical inverse problem. We focus on an infiltration test monitored by time-lapse ERT and modeled using Richards equation. The goal is to identify hydrological model parameters from ERT electrical potential measurements. Traditional uncoupled inversion relies on the solution of two sequentialmore » inverse problems, the first one applied to the ERT measurements, the second one to Richards equation. This approach does not ensure an accurate quantitative description of the physical state, typically violating mass balance. To avoid one of these two inversions and incorporate in the process more physical simulation constraints, we cast the problem within the framework of a SIR (Sequential Importance Resampling) data assimilation approach that uses a Richards equation solver to model the hydrological dynamics and a forward ERT simulator combined with Archie's law to serve as measurement model. ERT observations are then used to update the state of the system as well as to estimate the model parameters and their posterior distribution. The limitations of the traditional sequential Bayesian approach are investigated and an innovative iterative approach is proposed to estimate the model parameters with high accuracy. The numerical properties of the developed algorithm are verified on both homogeneous and heterogeneous synthetic test cases based on a real-world field experiment.« less

Electrostatic point charge fitting as an inverse problem: Revealing the underlying ill-conditioning

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

Ivanov, Maxim V.; Talipov, Marat R.; Timerghazin, Qadir K., E-mail: qadir.timerghazin@marquette.edu

2015-10-07

Atom-centered point charge (PC) model of the molecular electrostatics—a major workhorse of the atomistic biomolecular simulations—is usually parameterized by least-squares (LS) fitting of the point charge values to a reference electrostatic potential, a procedure that suffers from numerical instabilities due to the ill-conditioned nature of the LS problem. To reveal the origins of this ill-conditioning, we start with a general treatment of the point charge fitting problem as an inverse problem and construct an analytical model with the point charges spherically arranged according to Lebedev quadrature which is naturally suited for the inverse electrostatic problem. This analytical model is contrastedmore » to the atom-centered point-charge model that can be viewed as an irregular quadrature poorly suited for the problem. This analysis shows that the numerical problems of the point charge fitting are due to the decay of the curvatures corresponding to the eigenvectors of LS sum Hessian matrix. In part, this ill-conditioning is intrinsic to the problem and is related to decreasing electrostatic contribution of the higher multipole moments, that are, in the case of Lebedev grid model, directly associated with the Hessian eigenvectors. For the atom-centered model, this association breaks down beyond the first few eigenvectors related to the high-curvature monopole and dipole terms; this leads to even wider spread-out of the Hessian curvature values. Using these insights, it is possible to alleviate the ill-conditioning of the LS point-charge fitting without introducing external restraints and/or constraints. Also, as the analytical Lebedev grid PC model proposed here can reproduce multipole moments up to a given rank, it may provide a promising alternative to including explicit multipole terms in a force field.« less

Joint Model and Parameter Dimension Reduction for Bayesian Inversion Applied to an Ice Sheet Flow Problem

NASA Astrophysics Data System (ADS)

Ghattas, O.; Petra, N.; Cui, T.; Marzouk, Y.; Benjamin, P.; Willcox, K.

2016-12-01

Model-based projections of the dynamics of the polar ice sheets play a central role in anticipating future sea level rise. However, a number of mathematical and computational challenges place significant barriers on improving predictability of these models. One such challenge is caused by the unknown model parameters (e.g., in the basal boundary conditions) that must be inferred from heterogeneous observational data, leading to an ill-posed inverse problem and the need to quantify uncertainties in its solution. In this talk we discuss the problem of estimating the uncertainty in the solution of (large-scale) ice sheet inverse problems within the framework of Bayesian inference. Computing the general solution of the inverse problem--i.e., the posterior probability density--is intractable with current methods on today's computers, due to the expense of solving the forward model (3D full Stokes flow with nonlinear rheology) and the high dimensionality of the uncertain parameters (which are discretizations of the basal sliding coefficient field). To overcome these twin computational challenges, it is essential to exploit problem structure (e.g., sensitivity of the data to parameters, the smoothing property of the forward model, and correlations in the prior). To this end, we present a data-informed approach that identifies low-dimensional structure in both parameter space and the forward model state space. This approach exploits the fact that the observations inform only a low-dimensional parameter space and allows us to construct a parameter-reduced posterior. Sampling this parameter-reduced posterior still requires multiple evaluations of the forward problem, therefore we also aim to identify a low dimensional state space to reduce the computational cost. To this end, we apply a proper orthogonal decomposition (POD) approach to approximate the state using a low-dimensional manifold constructed using ``snapshots'' from the parameter reduced posterior, and the discrete empirical interpolation method (DEIM) to approximate the nonlinearity in the forward problem. We show that using only a limited number of forward solves, the resulting subspaces lead to an efficient method to explore the high-dimensional posterior.

Intelligent inversion method for pre-stack seismic big data based on MapReduce

NASA Astrophysics Data System (ADS)

Yan, Xuesong; Zhu, Zhixin; Wu, Qinghua

2018-01-01

Seismic exploration is a method of oil exploration that uses seismic information; that is, according to the inversion of seismic information, the useful information of the reservoir parameters can be obtained to carry out exploration effectively. Pre-stack data are characterised by a large amount of data, abundant information, and so on, and according to its inversion, the abundant information of the reservoir parameters can be obtained. Owing to the large amount of pre-stack seismic data, existing single-machine environments have not been able to meet the computational needs of the huge amount of data; thus, the development of a method with a high efficiency and the speed to solve the inversion problem of pre-stack seismic data is urgently needed. The optimisation of the elastic parameters by using a genetic algorithm easily falls into a local optimum, which results in a non-obvious inversion effect, especially for the optimisation effect of the density. Therefore, an intelligent optimisation algorithm is proposed in this paper and used for the elastic parameter inversion of pre-stack seismic data. This algorithm improves the population initialisation strategy by using the Gardner formula and the genetic operation of the algorithm, and the improved algorithm obtains better inversion results when carrying out a model test with logging data. All of the elastic parameters obtained by inversion and the logging curve of theoretical model are fitted well, which effectively improves the inversion precision of the density. This algorithm was implemented with a MapReduce model to solve the seismic big data inversion problem. The experimental results show that the parallel model can effectively reduce the running time of the algorithm.

Data fitting and image fine-tuning approach to solve the inverse problem in fluorescence molecular imaging

NASA Astrophysics Data System (ADS)

Gorpas, Dimitris; Politopoulos, Kostas; Yova, Dido; Andersson-Engels, Stefan

2008-02-01

One of the most challenging problems in medical imaging is to "see" a tumour embedded into tissue, which is a turbid medium, by using fluorescent probes for tumour labeling. This problem, despite the efforts made during the last years, has not been fully encountered yet, due to the non-linear nature of the inverse problem and the convergence failures of many optimization techniques. This paper describes a robust solution of the inverse problem, based on data fitting and image fine-tuning techniques. As a forward solver the coupled radiative transfer equation and diffusion approximation model is proposed and compromised via a finite element method, enhanced with adaptive multi-grids for faster and more accurate convergence. A database is constructed by application of the forward model on virtual tumours with known geometry, and thus fluorophore distribution, embedded into simulated tissues. The fitting procedure produces the best matching between the real and virtual data, and thus provides the initial estimation of the fluorophore distribution. Using this information, the coupled radiative transfer equation and diffusion approximation model has the required initial values for a computational reasonable and successful convergence during the image fine-tuning application.

Time-lapse three-dimensional inversion of complex conductivity data using an active time constrained (ATC) approach

USGS Publications Warehouse

Karaoulis, M.; Revil, A.; Werkema, D.D.; Minsley, B.J.; Woodruff, W.F.; Kemna, A.

2011-01-01

Induced polarization (more precisely the magnitude and phase of impedance of the subsurface) is measured using a network of electrodes located at the ground surface or in boreholes. This method yields important information related to the distribution of permeability and contaminants in the shallow subsurface. We propose a new time-lapse 3-D modelling and inversion algorithm to image the evolution of complex conductivity over time. We discretize the subsurface using hexahedron cells. Each cell is assigned a complex resistivity or conductivity value. Using the finite-element approach, we model the in-phase and out-of-phase (quadrature) electrical potentials on the 3-D grid, which are then transformed into apparent complex resistivity. Inhomogeneous Dirichlet boundary conditions are used at the boundary of the domain. The calculation of the Jacobian matrix is based on the principles of reciprocity. The goal of time-lapse inversion is to determine the change in the complex resistivity of each cell of the spatial grid as a function of time. Each model along the time axis is called a 'reference space model'. This approach can be simplified into an inverse problem looking for the optimum of several reference space models using the approximation that the material properties vary linearly in time between two subsequent reference models. Regularizations in both space domain and time domain reduce inversion artefacts and improve the stability of the inversion problem. In addition, the use of the time-lapse equations allows the simultaneous inversion of data obtained at different times in just one inversion step (4-D inversion). The advantages of this new inversion algorithm are demonstrated on synthetic time-lapse data resulting from the simulation of a salt tracer test in a heterogeneous random material described by an anisotropic semi-variogram. ?? 2011 The Authors Geophysical Journal International ?? 2011 RAS.

Mixed linear-nonlinear fault slip inversion: Bayesian inference of model, weighting, and smoothing parameters

NASA Astrophysics Data System (ADS)

Fukuda, J.; Johnson, K. M.

2009-12-01

Studies utilizing inversions of geodetic data for the spatial distribution of coseismic slip on faults typically present the result as a single fault plane and slip distribution. Commonly the geometry of the fault plane is assumed to be known a priori and the data are inverted for slip. However, sometimes there is not strong a priori information on the geometry of the fault that produced the earthquake and the data is not always strong enough to completely resolve the fault geometry. We develop a method to solve for the full posterior probability distribution of fault slip and fault geometry parameters in a Bayesian framework using Monte Carlo methods. The slip inversion problem is particularly challenging because it often involves multiple data sets with unknown relative weights (e.g. InSAR, GPS), model parameters that are related linearly (slip) and nonlinearly (fault geometry) through the theoretical model to surface observations, prior information on model parameters, and a regularization prior to stabilize the inversion. We present the theoretical framework and solution method for a Bayesian inversion that can handle all of these aspects of the problem. The method handles the mixed linear/nonlinear nature of the problem through combination of both analytical least-squares solutions and Monte Carlo methods. We first illustrate and validate the inversion scheme using synthetic data sets. We then apply the method to inversion of geodetic data from the 2003 M6.6 San Simeon, California earthquake. We show that the uncertainty in strike and dip of the fault plane is over 20 degrees. We characterize the uncertainty in the slip estimate with a volume around the mean fault solution in which the slip most likely occurred. Slip likely occurred somewhere in a volume that extends 5-10 km in either direction normal to the fault plane. We implement slip inversions with both traditional, kinematic smoothing constraints on slip and a simple physical condition of uniform stress drop.

Sequential Inverse Problems Bayesian Principles and the Logistic Map Example

NASA Astrophysics Data System (ADS)

Duan, Lian; Farmer, Chris L.; Moroz, Irene M.

2010-09-01

Bayesian statistics provides a general framework for solving inverse problems, but is not without interpretation and implementation problems. This paper discusses difficulties arising from the fact that forward models are always in error to some extent. Using a simple example based on the one-dimensional logistic map, we argue that, when implementation problems are minimal, the Bayesian framework is quite adequate. In this paper the Bayesian Filter is shown to be able to recover excellent state estimates in the perfect model scenario (PMS) and to distinguish the PMS from the imperfect model scenario (IMS). Through a quantitative comparison of the way in which the observations are assimilated in both the PMS and the IMS scenarios, we suggest that one can, sometimes, measure the degree of imperfection.

The 2-D magnetotelluric inverse problem solved with optimization

NASA Astrophysics Data System (ADS)

van Beusekom, Ashley E.; Parker, Robert L.; Bank, Randolph E.; Gill, Philip E.; Constable, Steven

2011-02-01

The practical 2-D magnetotelluric inverse problem seeks to determine the shallow-Earth conductivity structure using finite and uncertain data collected on the ground surface. We present an approach based on using PLTMG (Piecewise Linear Triangular MultiGrid), a special-purpose code for optimization with second-order partial differential equation (PDE) constraints. At each frequency, the electromagnetic field and conductivity are treated as unknowns in an optimization problem in which the data misfit is minimized subject to constraints that include Maxwell's equations and the boundary conditions. Within this framework it is straightforward to accommodate upper and lower bounds or other conditions on the conductivity. In addition, as the underlying inverse problem is ill-posed, constraints may be used to apply various kinds of regularization. We discuss some of the advantages and difficulties associated with using PDE-constrained optimization as the basis for solving large-scale nonlinear geophysical inverse problems. Combined transverse electric and transverse magnetic complex admittances from the COPROD2 data are inverted. First, we invert penalizing size and roughness giving solutions that are similar to those found previously. In a second example, conventional regularization is replaced by a technique that imposes upper and lower bounds on the model. In both examples the data misfit is better than that obtained previously, without any increase in model complexity.

Bayesian parameter estimation in spectral quantitative photoacoustic tomography

NASA Astrophysics Data System (ADS)

Pulkkinen, Aki; Cox, Ben T.; Arridge, Simon R.; Kaipio, Jari P.; Tarvainen, Tanja

2016-03-01

Photoacoustic tomography (PAT) is an imaging technique combining strong contrast of optical imaging to high spatial resolution of ultrasound imaging. These strengths are achieved via photoacoustic effect, where a spatial absorption of light pulse is converted into a measurable propagating ultrasound wave. The method is seen as a potential tool for small animal imaging, pre-clinical investigations, study of blood vessels and vasculature, as well as for cancer imaging. The goal in PAT is to form an image of the absorbed optical energy density field via acoustic inverse problem approaches from the measured ultrasound data. Quantitative PAT (QPAT) proceeds from these images and forms quantitative estimates of the optical properties of the target. This optical inverse problem of QPAT is illposed. To alleviate the issue, spectral QPAT (SQPAT) utilizes PAT data formed at multiple optical wavelengths simultaneously with optical parameter models of tissue to form quantitative estimates of the parameters of interest. In this work, the inverse problem of SQPAT is investigated. Light propagation is modelled using the diffusion equation. Optical absorption is described with chromophore concentration weighted sum of known chromophore absorption spectra. Scattering is described by Mie scattering theory with an exponential power law. In the inverse problem, the spatially varying unknown parameters of interest are the chromophore concentrations, the Mie scattering parameters (power law factor and the exponent), and Gruneisen parameter. The inverse problem is approached with a Bayesian method. It is numerically demonstrated, that estimation of all parameters of interest is possible with the approach.

Collaborative Investigations of Shallow Water Optics Problems

DTIC Science & Technology

2004-12-01

Appendix E. Reprint of Radiative transfer equation inversion: Theory and shape factor models for retrieval of oceanic inherent optical properties, by F ...4829-4834. 5 Hoge, F . E., P. E. Lyon, C. D. Mobley, and L. K. Sundman, 2003. Radiative transfer equation inversion: Theory and shape factor models for...multilinear regression algorithms for the inversion of synthetic ocean colour spectra,, Int. J. Remote Sensing, 25(21), 4829-4834. Hoge, F . E., P. E. Lyon

EDITORIAL: Inverse Problems in Engineering

NASA Astrophysics Data System (ADS)

West, Robert M.; Lesnic, Daniel

2007-01-01

Presented here are 11 noteworthy papers selected from the Fifth International Conference on Inverse Problems in Engineering: Theory and Practice held in Cambridge, UK during 11-15 July 2005. The papers have been peer-reviewed to the usual high standards of this journal and the contributions of reviewers are much appreciated. The conference featured a good balance of the fundamental mathematical concepts of inverse problems with a diverse range of important and interesting applications, which are represented here by the selected papers. Aspects of finite-element modelling and the performance of inverse algorithms are investigated by Autrique et al and Leduc et al. Statistical aspects are considered by Emery et al and Watzenig et al with regard to Bayesian parameter estimation and inversion using particle filters. Electrostatic applications are demonstrated by van Berkel and Lionheart and also Nakatani et al. Contributions to the applications of electrical techniques and specifically electrical tomographies are provided by Wakatsuki and Kagawa, Kim et al and Kortschak et al. Aspects of inversion in optical tomography are investigated by Wright et al and Douiri et al. The authors are representative of the worldwide interest in inverse problems relating to engineering applications and their efforts in producing these excellent papers will be appreciated by many readers of this journal.

Stochastic reduced order models for inverse problems under uncertainty

PubMed Central

Warner, James E.; Aquino, Wilkins; Grigoriu, Mircea D.

2014-01-01

This work presents a novel methodology for solving inverse problems under uncertainty using stochastic reduced order models (SROMs). Given statistical information about an observed state variable in a system, unknown parameters are estimated probabilistically through the solution of a model-constrained, stochastic optimization problem. The point of departure and crux of the proposed framework is the representation of a random quantity using a SROM - a low dimensional, discrete approximation to a continuous random element that permits e cient and non-intrusive stochastic computations. Characterizing the uncertainties with SROMs transforms the stochastic optimization problem into a deterministic one. The non-intrusive nature of SROMs facilitates e cient gradient computations for random vector unknowns and relies entirely on calls to existing deterministic solvers. Furthermore, the method is naturally extended to handle multiple sources of uncertainty in cases where state variable data, system parameters, and boundary conditions are all considered random. The new and widely-applicable SROM framework is formulated for a general stochastic optimization problem in terms of an abstract objective function and constraining model. For demonstration purposes, however, we study its performance in the specific case of inverse identification of random material parameters in elastodynamics. We demonstrate the ability to efficiently recover random shear moduli given material displacement statistics as input data. We also show that the approach remains effective for the case where the loading in the problem is random as well. PMID:25558115

Perturbational and nonperturbational inversion of Rayleigh-wave velocities

USGS Publications Warehouse

Haney, Matt; Tsai, Victor C.

2017-01-01

The inversion of Rayleigh-wave dispersion curves is a classic geophysical inverse problem. We have developed a set of MATLAB codes that performs forward modeling and inversion of Rayleigh-wave phase or group velocity measurements. We describe two different methods of inversion: a perturbational method based on finite elements and a nonperturbational method based on the recently developed Dix-type relation for Rayleigh waves. In practice, the nonperturbational method can be used to provide a good starting model that can be iteratively improved with the perturbational method. Although the perturbational method is well-known, we solve the forward problem using an eigenvalue/eigenvector solver instead of the conventional approach of root finding. Features of the codes include the ability to handle any mix of phase or group velocity measurements, combinations of modes of any order, the presence of a surface water layer, computation of partial derivatives due to changes in material properties and layer boundaries, and the implementation of an automatic grid of layers that is optimally suited for the depth sensitivity of Rayleigh waves.

Inference of emission rates from multiple sources using Bayesian probability theory.

PubMed

Yee, Eugene; Flesch, Thomas K

2010-03-01

The determination of atmospheric emission rates from multiple sources using inversion (regularized least-squares or best-fit technique) is known to be very susceptible to measurement and model errors in the problem, rendering the solution unusable. In this paper, a new perspective is offered for this problem: namely, it is argued that the problem should be addressed as one of inference rather than inversion. Towards this objective, Bayesian probability theory is used to estimate the emission rates from multiple sources. The posterior probability distribution for the emission rates is derived, accounting fully for the measurement errors in the concentration data and the model errors in the dispersion model used to interpret the data. The Bayesian inferential methodology for emission rate recovery is validated against real dispersion data, obtained from a field experiment involving various source-sensor geometries (scenarios) consisting of four synthetic area sources and eight concentration sensors. The recovery of discrete emission rates from three different scenarios obtained using Bayesian inference and singular value decomposition inversion are compared and contrasted.

Using Inverse Problem Methods with Surveillance Data in Pneumococcal Vaccination

PubMed Central

Sutton, Karyn L.; Banks, H. T.; Castillo-Chavez, Carlos

2010-01-01

The design and evaluation of epidemiological control strategies is central to public health policy. While inverse problem methods are routinely used in many applications, this remains an area in which their use is relatively rare, although their potential impact is great. We describe methods particularly relevant to epidemiological modeling at the population level. These methods are then applied to the study of pneumococcal vaccination strategies as a relevant example which poses many challenges common to other infectious diseases. We demonstrate that relevant yet typically unknown parameters may be estimated, and show that a calibrated model may used to assess implemented vaccine policies through the estimation of parameters if vaccine history is recorded along with infection and colonization information. Finally, we show how one might determine an appropriate level of refinement or aggregation in the age-structured model given age-stratified observations. These results illustrate ways in which the collection and analysis of surveillance data can be improved using inverse problem methods. PMID:20209093

Impact of spatio-temporal scale of adjustment on variational assimilation of hydrologic and hydrometeorological data in operational distributed hydrologic models

NASA Astrophysics Data System (ADS)

Lee, H.; Seo, D.; McKee, P.; Corby, R.

2009-12-01

One of the large challenges in data assimilation (DA) into distributed hydrologic models is to reduce the large degrees of freedom involved in the inverse problem to avoid overfitting. To assess the sensitivity of the performance of DA to the dimensionality of the inverse problem, we design and carry out real-world experiments in which the control vector in variational DA (VAR) is solved at different scales in space and time, e.g., lumped, semi-distributed, and fully-distributed in space, and hourly, 6 hourly, etc., in time. The size of the control vector is related to the degrees of freedom in the inverse problem. For the assessment, we use the prototype 4-dimenational variational data assimilator (4DVAR) that assimilates streamflow, precipitation and potential evaporation data into the NWS Hydrology Laboratory’s Research Distributed Hydrologic Model (HL-RDHM). In this talk, we present the initial results for a number of basins in Oklahoma and Texas.

On computational experiments in some inverse problems of heat and mass transfer

NASA Astrophysics Data System (ADS)

Bilchenko, G. G.; Bilchenko, N. G.

2016-11-01

The results of mathematical modeling of effective heat and mass transfer on hypersonic aircraft permeable surfaces are considered. The physic-chemical processes (the dissociation and the ionization) in laminar boundary layer of compressible gas are appreciated. Some algorithms of control restoration are suggested for the interpolation and approximation statements of heat and mass transfer inverse problems. The differences between the methods applied for the problem solutions search for these statements are discussed. Both the algorithms are realized as programs. Many computational experiments were accomplished with the use of these programs. The parameters of boundary layer obtained by means of the A.A.Dorodnicyn's generalized integral relations method from solving the direct problems have been used to obtain the inverse problems solutions. Two types of blowing laws restoration for the inverse problem in interpolation statement are presented as the examples. The influence of the temperature factor on the blowing restoration is investigated. The different character of sensitivity of controllable parameters (the local heat flow and local tangent friction) respectively to step (discrete) changing of control (the blowing) and the switching point position is studied.

On the joint inversion of geophysical data for models of the coupled core-mantle system

NASA Technical Reports Server (NTRS)

Voorhies, Coerte V.

1991-01-01

Joint inversion of magnetic, earth rotation, geoid, and seismic data for a unified model of the coupled core-mantle system is proposed and shown to be possible. A sample objective function is offered and simplified by targeting results from independent inversions and summary travel time residuals instead of original observations. These data are parameterized in terms of a very simple, closed model of the topographically coupled core-mantle system. Minimization of the simplified objective function leads to a nonlinear inverse problem; an iterative method for solution is presented. Parameterization and method are emphasized; numerical results are not presented.

Inverse kinematics of a dual linear actuator pitch/roll heliostat

NASA Astrophysics Data System (ADS)

Freeman, Joshua; Shankar, Balakrishnan; Sundaram, Ganesh

2017-06-01

This work presents a simple, computationally efficient inverse kinematics solution for a pitch/roll heliostat using two linear actuators. The heliostat design and kinematics have been developed, modeled and tested using computer simulation software. A physical heliostat prototype was fabricated to validate the theoretical computations and data. Pitch/roll heliostats have numerous advantages including reduced cost potential and reduced space requirements, with a primary disadvantage being the significantly more complicated kinematics, which are solved here. Novel methods are applied to simplify the inverse kinematics problem which could be applied to other similar problems.

The inverse problem of using the information of historical data to estimate model errors is one of the science frontier research topics. In this study, we investigate such a problem using the classic Lorenz (1963) equation as a prediction model.

NASA Astrophysics Data System (ADS)

Wan, S.; He, W.

2016-12-01

The inverse problem of using the information of historical data to estimate model errors is one of the science frontier research topics. In this study, we investigate such a problem using the classic Lorenz (1963) equation as a prediction model and the Lorenz equation with a periodic evolutionary function as an accurate representation of reality to generate "observational data." On the basis of the intelligent features of evolutionary modeling (EM), including self-organization, self-adaptive and self-learning, the dynamic information contained in the historical data can be identified and extracted by computer automatically. Thereby, a new approach is proposed to estimate model errors based on EM in the present paper. Numerical tests demonstrate the ability of the new approach to correct model structural errors. In fact, it can actualize the combination of the statistics and dynamics to certain extent.

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.

Density-to-Potential Inversions to Guide Development of Exchange-Correlation Approximations at Finite Temperature

NASA Astrophysics Data System (ADS)

Jensen, Daniel; Wasserman, Adam; Baczewski, Andrew

The construction of approximations to the exchange-correlation potential for warm dense matter (WDM) is a topic of significant recent interest. In this work, we study the inverse problem of Kohn-Sham (KS) DFT as a means of guiding functional design at zero temperature and in WDM. Whereas the forward problem solves the KS equations to produce a density from a specified exchange-correlation potential, the inverse problem seeks to construct the exchange-correlation potential from specified densities. These two problems require different computational methods and convergence criteria despite sharing the same mathematical equations. We present two new inversion methods based on constrained variational and PDE-constrained optimization methods. We adapt these methods to finite temperature calculations to reveal the exchange-correlation potential's temperature dependence in WDM-relevant conditions. The different inversion methods presented are applied to both non-interacting and interacting model systems for comparison. Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy's National Security Administration under contract DE-AC04-94.

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.

Inverse problem studies of biochemical systems with structure identification of S-systems by embedding training functions in a genetic algorithm.

PubMed

Sarode, Ketan Dinkar; Kumar, V Ravi; Kulkarni, B D

2016-05-01

An efficient inverse problem approach for parameter estimation, state and structure identification from dynamic data by embedding training functions in a genetic algorithm methodology (ETFGA) is proposed for nonlinear dynamical biosystems using S-system canonical models. Use of multiple shooting and decomposition approach as training functions has been shown for handling of noisy datasets and computational efficiency in studying the inverse problem. The advantages of the methodology are brought out systematically by studying it for three biochemical model systems of interest. By studying a small-scale gene regulatory system described by a S-system model, the first example demonstrates the use of ETFGA for the multifold aims of the inverse problem. The estimation of a large number of parameters with simultaneous state and network identification is shown by training a generalized S-system canonical model with noisy datasets. The results of this study bring out the superior performance of ETFGA on comparison with other metaheuristic approaches. The second example studies the regulation of cAMP oscillations in Dictyostelium cells now assuming limited availability of noisy data. Here, flexibility of the approach to incorporate partial system information in the identification process is shown and its effect on accuracy and predictive ability of the estimated model are studied. The third example studies the phenomenological toy model of the regulation of circadian oscillations in Drosophila that follows rate laws different from S-system power-law. For the limited noisy data, using a priori information about properties of the system, we could estimate an alternate S-system model that showed robust oscillatory behavior with predictive abilities. Copyright © 2016 Elsevier Inc. All rights reserved.

GARCH modelling of covariance in dynamical estimation of inverse solutions

NASA Astrophysics Data System (ADS)

Galka, Andreas; Yamashita, Okito; Ozaki, Tohru

2004-12-01

The problem of estimating unobserved states of spatially extended dynamical systems poses an inverse problem, which can be solved approximately by a recently developed variant of Kalman filtering; in order to provide the model of the dynamics with more flexibility with respect to space and time, we suggest to combine the concept of GARCH modelling of covariance, well known in econometrics, with Kalman filtering. We formulate this algorithm for spatiotemporal systems governed by stochastic diffusion equations and demonstrate its feasibility by presenting a numerical simulation designed to imitate the situation of the generation of electroencephalographic recordings by the human cortex.

An introduction of Markov chain Monte Carlo method to geochemical inverse problems: Reading melting parameters from REE abundances in abyssal peridotites

NASA Astrophysics Data System (ADS)

Liu, Boda; Liang, Yan

2017-04-01

Markov chain Monte Carlo (MCMC) simulation is a powerful statistical method in solving inverse problems that arise from a wide range of applications. In Earth sciences applications of MCMC simulations are primarily in the field of geophysics. The purpose of this study is to introduce MCMC methods to geochemical inverse problems related to trace element fractionation during mantle melting. MCMC methods have several advantages over least squares methods in deciphering melting processes from trace element abundances in basalts and mantle rocks. Here we use an MCMC method to invert for extent of melting, fraction of melt present during melting, and extent of chemical disequilibrium between the melt and residual solid from REE abundances in clinopyroxene in abyssal peridotites from Mid-Atlantic Ridge, Central Indian Ridge, Southwest Indian Ridge, Lena Trough, and American-Antarctic Ridge. We consider two melting models: one with exact analytical solution and the other without. We solve the latter numerically in a chain of melting models according to the Metropolis-Hastings algorithm. The probability distribution of inverted melting parameters depends on assumptions of the physical model, knowledge of mantle source composition, and constraints from the REE data. Results from MCMC inversion are consistent with and provide more reliable uncertainty estimates than results based on nonlinear least squares inversion. We show that chemical disequilibrium is likely to play an important role in fractionating LREE in residual peridotites during partial melting beneath mid-ocean ridge spreading centers. MCMC simulation is well suited for more complicated but physically more realistic melting problems that do not have analytical solutions.

Model based defect characterization in composites

NASA Astrophysics Data System (ADS)

Roberts, R.; Holland, S.

2017-02-01

Work is reported on model-based defect characterization in CFRP composites. The work utilizes computational models of the interaction of NDE probing energy fields (ultrasound and thermography), to determine 1) the measured signal dependence on material and defect properties (forward problem), and 2) an assessment of performance-critical defect properties from analysis of measured NDE signals (inverse problem). Work is reported on model implementation for inspection of CFRP laminates containing multi-ply impact-induced delamination, with application in this paper focusing on ultrasound. A companion paper in these proceedings summarizes corresponding activity in thermography. Inversion of ultrasound data is demonstrated showing the quantitative extraction of damage properties.

Solving inverse problem for Markov chain model of customer lifetime value using flower pollination algorithm

NASA Astrophysics Data System (ADS)

Al-Ma'shumah, Fathimah; Permana, Dony; Sidarto, Kuntjoro Adji

2015-12-01

Customer Lifetime Value is an important and useful concept in marketing. One of its benefits is to help a company for budgeting marketing expenditure for customer acquisition and customer retention. Many mathematical models have been introduced to calculate CLV considering the customer retention/migration classification scheme. A fairly new class of these models which will be described in this paper uses Markov Chain Models (MCM). This class of models has the major advantage for its flexibility to be modified to several different cases/classification schemes. In this model, the probabilities of customer retention and acquisition play an important role. From Pfeifer and Carraway, 2000, the final formula of CLV obtained from MCM usually contains nonlinear form of the transition probability matrix. This nonlinearity makes the inverse problem of CLV difficult to solve. This paper aims to solve this inverse problem, yielding the approximate transition probabilities for the customers, by applying metaheuristic optimization algorithm developed by Yang, 2013, Flower Pollination Algorithm. The major interpretation of obtaining the transition probabilities are to set goals for marketing teams in keeping the relative frequencies of customer acquisition and customer retention.

Approximation Methods for Inverse Problems Governed by Nonlinear Parabolic Systems

DTIC Science & Technology

1999-12-17

We present a rigorous theoretical framework for approximation of nonlinear parabolic systems with delays in the context of inverse least squares...numerical results demonstrating the convergence are given for a model of dioxin uptake and elimination in a distributed liver model that is a special case of the general theoretical framework .

Full-Physics Inverse Learning Machine for Satellite Remote Sensing Retrievals

NASA Astrophysics Data System (ADS)

Loyola, D. G.

2017-12-01

The satellite remote sensing retrievals are usually ill-posed inverse problems that are typically solved by finding a state vector that minimizes the residual between simulated data and real measurements. The classical inversion methods are very time-consuming as they require iterative calls to complex radiative-transfer forward models to simulate radiances and Jacobians, and subsequent inversion of relatively large matrices. In this work we present a novel and extremely fast algorithm for solving inverse problems called full-physics inverse learning machine (FP-ILM). The FP-ILM algorithm consists of a training phase in which machine learning techniques are used to derive an inversion operator based on synthetic data generated using a radiative transfer model (which expresses the "full-physics" component) and the smart sampling technique, and an operational phase in which the inversion operator is applied to real measurements. FP-ILM has been successfully applied to the retrieval of the SO2 plume height during volcanic eruptions and to the retrieval of ozone profile shapes from UV/VIS satellite sensors. Furthermore, FP-ILM will be used for the near-real-time processing of the upcoming generation of European Sentinel sensors with their unprecedented spectral and spatial resolution and associated large increases in the amount of data.

Inversion for the driving forces of plate tectonics

NASA Technical Reports Server (NTRS)

Richardson, R. M.

1983-01-01

Inverse modeling techniques have been applied to the problem of determining the roles of various forces that may drive and resist plate tectonic motions. Separate linear inverse problems have been solved to find the best fitting pole of rotation for finite element grid point velocities and to find the best combination of force models to fit the observed relative plate velocities for the earth's twelve major plates using the generalized inverse operator. Variance-covariance data on plate motion have also been included. Results emphasize the relative importance of ridge push forces in the driving mechanism. Convergent margin forces are smaller by at least a factor of two, and perhaps by as much as a factor of twenty. Slab pull, apparently, is poorly transmitted to the surface plate as a driving force. Drag forces at the base of the plate are smaller than ridge push forces, although the sign of the force remains in question.

Study on Material Parameters Identification of Brain Tissue Considering Uncertainty of Friction Coefficient

NASA Astrophysics Data System (ADS)

Guan, Fengjiao; Zhang, Guanjun; Liu, Jie; Wang, Shujing; Luo, Xu; Zhu, Feng

2017-10-01

Accurate material parameters are critical to construct the high biofidelity finite element (FE) models. However, it is hard to obtain the brain tissue parameters accurately because of the effects of irregular geometry and uncertain boundary conditions. Considering the complexity of material test and the uncertainty of friction coefficient, a computational inverse method for viscoelastic material parameters identification of brain tissue is presented based on the interval analysis method. Firstly, the intervals are used to quantify the friction coefficient in the boundary condition. And then the inverse problem of material parameters identification under uncertain friction coefficient is transformed into two types of deterministic inverse problem. Finally the intelligent optimization algorithm is used to solve the two types of deterministic inverse problems quickly and accurately, and the range of material parameters can be easily acquired with no need of a variety of samples. The efficiency and convergence of this method are demonstrated by the material parameters identification of thalamus. The proposed method provides a potential effective tool for building high biofidelity human finite element model in the study of traffic accident injury.

Wave tilt sounding of multilayered structures. [for probing of stratified planetary surface electrical properties and thickness

NASA Technical Reports Server (NTRS)

Warne, L.; Jaggard, D. L.; Elachi, C.

1979-01-01

The relationship between the wave tilt and the electrical parameters of a multilayered structure is investigated. Particular emphasis is placed on the inverse problem associated with the sounding planetary surfaces. An inversion technique, based on multifrequency wave tilt, is proposed and demonstrated with several computer models. It is determined that there is close agreement between the electrical parameters used in the models and those in the inversion values.

Using field inversion to quantify functional errors in turbulence closures

NASA Astrophysics Data System (ADS)

Singh, Anand Pratap; Duraisamy, Karthik

2016-04-01

A data-informed approach is presented with the objective of quantifying errors and uncertainties in the functional forms of turbulence closure models. The approach creates modeling information from higher-fidelity simulations and experimental data. Specifically, a Bayesian formalism is adopted to infer discrepancies in the source terms of transport equations. A key enabling idea is the transformation of the functional inversion procedure (which is inherently infinite-dimensional) into a finite-dimensional problem in which the distribution of the unknown function is estimated at discrete mesh locations in the computational domain. This allows for the use of an efficient adjoint-driven inversion procedure. The output of the inversion is a full-field of discrepancy that provides hitherto inaccessible modeling information. The utility of the approach is demonstrated by applying it to a number of problems including channel flow, shock-boundary layer interactions, and flows with curvature and separation. In all these cases, the posterior model correlates well with the data. Furthermore, it is shown that even if limited data (such as surface pressures) are used, the accuracy of the inferred solution is improved over the entire computational domain. The results suggest that, by directly addressing the connection between physical data and model discrepancies, the field inversion approach materially enhances the value of computational and experimental data for model improvement. The resulting information can be used by the modeler as a guiding tool to design more accurate model forms, or serve as input to machine learning algorithms to directly replace deficient modeling terms.

3-D minimum-structure inversion of magnetotelluric data using the finite-element method and tetrahedral grids

NASA Astrophysics Data System (ADS)

Jahandari, H.; Farquharson, C. G.

2017-11-01

Unstructured grids enable representing arbitrary structures more accurately and with fewer cells compared to regular structured grids. These grids also allow more efficient refinements compared to rectilinear meshes. In this study, tetrahedral grids are used for the inversion of magnetotelluric (MT) data, which allows for the direct inclusion of topography in the model, for constraining an inversion using a wireframe-based geological model and for local refinement at the observation stations. A minimum-structure method with an iterative model-space Gauss-Newton algorithm for optimization is used. An iterative solver is employed for solving the normal system of equations at each Gauss-Newton step and the sensitivity matrix-vector products that are required by this solver are calculated using pseudo-forward problems. This method alleviates the need to explicitly form the Hessian or Jacobian matrices which significantly reduces the required computation memory. Forward problems are formulated using an edge-based finite-element approach and a sparse direct solver is used for the solutions. This solver allows saving and re-using the factorization of matrices for similar pseudo-forward problems within a Gauss-Newton iteration which greatly minimizes the computation time. Two examples are presented to show the capability of the algorithm: the first example uses a benchmark model while the second example represents a realistic geological setting with topography and a sulphide deposit. The data that are inverted are the full-tensor impedance and the magnetic transfer function vector. The inversions sufficiently recovered the models and reproduced the data, which shows the effectiveness of unstructured grids for complex and realistic MT inversion scenarios. The first example is also used to demonstrate the computational efficiency of the presented model-space method by comparison with its data-space counterpart.

Large-Scale Optimization for Bayesian Inference in Complex Systems

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

Willcox, Karen; Marzouk, Youssef

2013-11-12

The SAGUARO (Scalable Algorithms for Groundwater Uncertainty Analysis and Robust Optimization) Project focused on the development of scalable numerical algorithms for large-scale Bayesian inversion in complex systems that capitalize on advances in large-scale simulation-based optimization and inversion methods. The project was a collaborative effort among MIT, the University of Texas at Austin, Georgia Institute of Technology, and Sandia National Laboratories. The research was directed in three complementary areas: efficient approximations of the Hessian operator, reductions in complexity of forward simulations via stochastic spectral approximations and model reduction, and employing large-scale optimization concepts to accelerate sampling. The MIT--Sandia component of themore » SAGUARO Project addressed the intractability of conventional sampling methods for large-scale statistical inverse problems by devising reduced-order models that are faithful to the full-order model over a wide range of parameter values; sampling then employs the reduced model rather than the full model, resulting in very large computational savings. Results indicate little effect on the computed posterior distribution. On the other hand, in the Texas--Georgia Tech component of the project, we retain the full-order model, but exploit inverse problem structure (adjoint-based gradients and partial Hessian information of the parameter-to-observation map) to implicitly extract lower dimensional information on the posterior distribution; this greatly speeds up sampling methods, so that fewer sampling points are needed. We can think of these two approaches as ``reduce then sample'' and ``sample then reduce.'' In fact, these two approaches are complementary, and can be used in conjunction with each other. Moreover, they both exploit deterministic inverse problem structure, in the form of adjoint-based gradient and Hessian information of the underlying parameter-to-observation map, to achieve their speedups.« less

Acoustic Inversion in Optoacoustic Tomography: A Review

PubMed Central

Rosenthal, Amir; Ntziachristos, Vasilis; Razansky, Daniel

2013-01-01

Optoacoustic tomography enables volumetric imaging with optical contrast in biological tissue at depths beyond the optical mean free path by the use of optical excitation and acoustic detection. The hybrid nature of optoacoustic tomography gives rise to two distinct inverse problems: The optical inverse problem, related to the propagation of the excitation light in tissue, and the acoustic inverse problem, which deals with the propagation and detection of the generated acoustic waves. Since the two inverse problems have different physical underpinnings and are governed by different types of equations, they are often treated independently as unrelated problems. From an imaging standpoint, the acoustic inverse problem relates to forming an image from the measured acoustic data, whereas the optical inverse problem relates to quantifying the formed image. This review focuses on the acoustic aspects of optoacoustic tomography, specifically acoustic reconstruction algorithms and imaging-system practicalities. As these two aspects are intimately linked, and no silver bullet exists in the path towards high-performance imaging, we adopt a holistic approach in our review and discuss the many links between the two aspects. Four classes of reconstruction algorithms are reviewed: time-domain (so called back-projection) formulae, frequency-domain formulae, time-reversal algorithms, and model-based algorithms. These algorithms are discussed in the context of the various acoustic detectors and detection surfaces which are commonly used in experimental studies. We further discuss the effects of non-ideal imaging scenarios on the quality of reconstruction and review methods that can mitigate these effects. Namely, we consider the cases of finite detector aperture, limited-view tomography, spatial under-sampling of the acoustic signals, and acoustic heterogeneities and losses. PMID:24772060

Analysis of the Hessian for Inverse Scattering Problems. Part 3. Inverse Medium Scattering of Electromagnetic Waves in Three Dimensions

DTIC Science & Technology

2012-08-01

small data noise and model error, the discrete Hessian can be approximated by a low-rank matrix. This in turn enables fast solution of an appropriately...implication of the compactness of the Hessian is that for small data noise and model error, the discrete Hessian can be approximated by a low-rank matrix. This...probability distribution is given by the inverse of the Hessian of the negative log likelihood function. For Gaussian data noise and model error, this

Including geological information in the inverse problem of palaeothermal reconstruction

NASA Astrophysics Data System (ADS)

Trautner, S.; Nielsen, S. B.

2003-04-01

A reliable reconstruction of sediment thermal history is of central importance to the assessment of hydrocarbon potential and the understanding of basin evolution. However, only rarely do sedimentation history and borehole data in the form of present day temperatures and vitrinite reflectance constrain the past thermal evolution to a useful level of accuracy (Gallagher and Sambridge,1992; Nielsen,1998; Trautner and Nielsen,2003). This is reflected in the inverse solutions to the problem of determining heat flow history from borehole data: The recent heat flow is constrained by data while older values are governed by the chosen a prior heat flow. In this paper we reduce this problem by including geological information in the inverse problem. Through a careful analysis of geological and geophysical data the timing of the tectonic processes, which may influence heat flow, can be inferred. The heat flow history is then parameterised to allow for the temporal variations characteristic of the different tectonic events. The inversion scheme applies a Markov chain Monte Carlo (MCMC) approach (Nielsen and Gallagher, 1999; Ferrero and Gallagher,2002), which efficiently explores the model space and futhermore samples the posterior probability distribution of the model. The technique is demonstrated on wells in the northern North Sea with emphasis on the stretching event in Late Jurassic. The wells are characterised by maximum sediment temperature at the present day, which is the worst case for resolution of the past thermal history because vitrinite reflectance is determined mainly by the maximum temperature. Including geological information significantly improves the thermal resolution. Ferrero, C. and Gallagher,K.,2002. Stochastic thermal history modelling.1. Constraining heat flow histories and their uncertainty. Marine and Petroleum Geology, 19, 633-648. Gallagher,K. and Sambridge, M., 1992. The resolution of past heat flow in sedimentary basins from non-linear inversion of geochemical data: the smoothest model approach, with synthetic examples. Geophysical Journal International, 109, 78-95. Nielsen, S.B, 1998. Inversion and sensitivity analysis in basin modelling. Geoscience 98. Keele University, UK, Abstract Volume, 56. Nielsen, S.B. and Gallagher, K., 1999. Efficient sampling of 3-D basin modelling scenarios. Extended Abstracts Volume, 1999 AAPG International Conference &Exhibition, Birmingham, England, September 12-15, 1999, p. 369 - 372. Trautner S. and Nielsen, S.B., 2003. 2-D inverse thermal modelling in the Norwegian shelf using Fast Approximate Forward (FAF) solutions. In R. Marzi and Duppenbecker, S. (Ed.), Multi-Dimensional Basin Modeling, AAPG, in press.

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

An ambiguity of information content and error in an ill-posed satellite inversion

NASA Astrophysics Data System (ADS)

Koner, Prabhat

According to Rodgers (2000, stochastic approach), the averaging kernel (AK) is the representational matrix to understand the information content in a scholastic inversion. On the other hand, in deterministic approach this is referred to as model resolution matrix (MRM, Menke 1989). The analysis of AK/MRM can only give some understanding of how much regularization is imposed on the inverse problem. The trace of the AK/MRM matrix, which is the so-called degree of freedom from signal (DFS; stochastic) or degree of freedom in retrieval (DFR; deterministic). There are no physical/mathematical explanations in the literature: why the trace of the matrix is a valid form to calculate this quantity? We will present an ambiguity between information and error using a real life problem of SST retrieval from GOES13. The stochastic information content calculation is based on the linear assumption. The validity of such mathematics in satellite inversion will be questioned because it is based on the nonlinear radiative transfer and ill-conditioned inverse problems. References: Menke, W., 1989: Geophysical data analysis: discrete inverse theory. San Diego academic press. Rodgers, C.D., 2000: Inverse methods for atmospheric soundings: theory and practice. Singapore :World Scientific.

Trinification, the hierarchy problem, and inverse seesaw neutrino masses

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

Cauet, Christophe; Paes, Heinrich; Wiesenfeldt, Soeren

2011-05-01

In minimal trinification models light neutrino masses can be generated via a radiative seesaw mechanism, where the masses of the right-handed neutrinos originate from loops involving Higgs and fermion fields at the unification scale. This mechanism is absent in models aiming at solving or ameliorating the hierarchy problem, such as low-energy supersymmetry, since the large seesaw scale disappears. In this case, neutrino masses need to be generated via a TeV-scale mechanism. In this paper, we investigate an inverse seesaw mechanism and discuss some phenomenological consequences.

Dynamic inverse models in human-cyber-physical systems

NASA Astrophysics Data System (ADS)

Robinson, Ryan M.; Scobee, Dexter R. R.; Burden, Samuel A.; Sastry, S. Shankar

2016-05-01

Human interaction with the physical world is increasingly mediated by automation. This interaction is characterized by dynamic coupling between robotic (i.e. cyber) and neuromechanical (i.e. human) decision-making agents. Guaranteeing performance of such human-cyber-physical systems will require predictive mathematical models of this dynamic coupling. Toward this end, we propose a rapprochement between robotics and neuromechanics premised on the existence of internal forward and inverse models in the human agent. We hypothesize that, in tele-robotic applications of interest, a human operator learns to invert automation dynamics, directly translating from desired task to required control input. By formulating the model inversion problem in the context of a tracking task for a nonlinear control system in control-a_ne form, we derive criteria for exponential tracking and show that the resulting dynamic inverse model generally renders a portion of the physical system state (i.e., the internal dynamics) unobservable from the human operator's perspective. Under stability conditions, we show that the human can achieve exponential tracking without formulating an estimate of the system's state so long as they possess an accurate model of the system's dynamics. These theoretical results are illustrated using a planar quadrotor example. We then demonstrate that the automation can intervene to improve performance of the tracking task by solving an optimal control problem. Performance is guaranteed to improve under the assumption that the human learns and inverts the dynamic model of the altered system. We conclude with a discussion of practical limitations that may hinder exact dynamic model inversion.

Cohesive phase-field fracture and a PDE constrained optimization approach to fracture inverse problems

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

Tupek, Michael R.

2016-06-30

In recent years there has been a proliferation of modeling techniques for forward predictions of crack propagation in brittle materials, including: phase-field/gradient damage models, peridynamics, cohesive-zone models, and G/XFEM enrichment techniques. However, progress on the corresponding inverse problems has been relatively lacking. Taking advantage of key features of existing modeling approaches, we propose a parabolic regularization of Barenblatt cohesive models which borrows extensively from previous phase-field and gradient damage formulations. An efficient explicit time integration strategy for this type of nonlocal fracture model is then proposed and justified. In addition, we present a C++ computational framework for computing in- putmore » parameter sensitivities efficiently for explicit dynamic problems using the adjoint method. This capability allows for solving inverse problems involving crack propagation to answer interesting engineering questions such as: 1) what is the optimal design topology and material placement for a heterogeneous structure to maximize fracture resistance, 2) what loads must have been applied to a structure for it to have failed in an observed way, 3) what are the existing cracks in a structure given various experimental observations, etc. In this work, we focus on the first of these engineering questions and demonstrate a capability to automatically and efficiently compute optimal designs intended to minimize crack propagation in structures.« less

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.

Comparison of Optimal Design Methods in Inverse Problems

PubMed Central

Banks, H. T.; Holm, Kathleen; Kappel, Franz

2011-01-01

Typical optimal design methods for inverse or parameter estimation problems are designed to choose optimal sampling distributions through minimization of a specific cost function related to the resulting error in parameter estimates. It is hoped that the inverse problem will produce parameter estimates with increased accuracy using data collected according to the optimal sampling distribution. Here we formulate the classical optimal design problem in the context of general optimization problems over distributions of sampling times. We present a new Prohorov metric based theoretical framework that permits one to treat succinctly and rigorously any optimal design criteria based on the Fisher Information Matrix (FIM). A fundamental approximation theory is also included in this framework. A new optimal design, SE-optimal design (standard error optimal design), is then introduced in the context of this framework. We compare this new design criteria with the more traditional D-optimal and E-optimal designs. The optimal sampling distributions from each design are used to compute and compare standard errors; the standard errors for parameters are computed using asymptotic theory or bootstrapping and the optimal mesh. We use three examples to illustrate ideas: the Verhulst-Pearl logistic population model [13], the standard harmonic oscillator model [13] and a popular glucose regulation model [16, 19, 29]. PMID:21857762

Exploring equivalence domain in nonlinear inverse problems using Covariance Matrix Adaption Evolution Strategy (CMAES) and random sampling

NASA Astrophysics Data System (ADS)

Grayver, Alexander V.; Kuvshinov, Alexey V.

2016-05-01

This paper presents a methodology to sample equivalence domain (ED) in nonlinear partial differential equation (PDE)-constrained inverse problems. For this purpose, we first applied state-of-the-art stochastic optimization algorithm called Covariance Matrix Adaptation Evolution Strategy (CMAES) to identify low-misfit regions of the model space. These regions were then randomly sampled to create an ensemble of equivalent models and quantify uncertainty. CMAES is aimed at exploring model space globally and is robust on very ill-conditioned problems. We show that the number of iterations required to converge grows at a moderate rate with respect to number of unknowns and the algorithm is embarrassingly parallel. We formulated the problem by using the generalized Gaussian distribution. This enabled us to seamlessly use arbitrary norms for residual and regularization terms. We show that various regularization norms facilitate studying different classes of equivalent solutions. We further show how performance of the standard Metropolis-Hastings Markov chain Monte Carlo algorithm can be substantially improved by using information CMAES provides. This methodology was tested by using individual and joint inversions of magneotelluric, controlled-source electromagnetic (EM) and global EM induction data.

Non-ambiguous recovery of Biot poroelastic parameters of cellular panels using ultrasonicwaves

NASA Astrophysics Data System (ADS)

Ogam, Erick; Fellah, Z. E. A.; Sebaa, Naima; Groby, J.-P.

2011-03-01

The inverse problem of the recovery of the poroelastic parameters of open-cell soft plastic foam panels is solved by employing transmitted ultrasonic waves (USW) and the Biot-Johnson-Koplik-Champoux-Allard (BJKCA) model. It is shown by constructing the objective functional given by the total square of the difference between predictions from the BJKCA interaction model and experimental data obtained with transmitted USW that the inverse problem is ill-posed, since the functional exhibits several local minima and maxima. In order to solve this problem, which is beyond the capability of most off-the-shelf iterative nonlinear least squares optimization algorithms (such as the Levenberg Marquadt or Nelder-Mead simplex methods), simple strategies are developed. The recovered acoustic parameters are compared with those obtained using simpler interaction models and a method employing asymptotic phase velocity of the transmitted USW. The retrieved elastic moduli are validated by solving an inverse vibration spectroscopy problem with data obtained from beam-like specimens cut from the panels using an equivalent solid elastodynamic model as estimator. The phase velocities are reconstructed using computed, measured resonance frequencies and a time-frequency decomposition of transient waves induced in the beam specimen. These confirm that the elastic parameters recovered using vibration are valid over the frequency range ofstudy.

Real-time characterization of partially observed epidemics using surrogate models.

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

Safta, Cosmin; Ray, Jaideep; Lefantzi, Sophia

We present a statistical method, predicated on the use of surrogate models, for the 'real-time' characterization of partially observed epidemics. Observations consist of counts of symptomatic patients, diagnosed with the disease, that may be available in the early epoch of an ongoing outbreak. Characterization, in this context, refers to estimation of epidemiological parameters that can be used to provide short-term forecasts of the ongoing epidemic, as well as to provide gross information on the dynamics of the etiologic agent in the affected population e.g., the time-dependent infection rate. The characterization problem is formulated as a Bayesian inverse problem, and epidemiologicalmore » parameters are estimated as distributions using a Markov chain Monte Carlo (MCMC) method, thus quantifying the uncertainty in the estimates. In some cases, the inverse problem can be computationally expensive, primarily due to the epidemic simulator used inside the inversion algorithm. We present a method, based on replacing the epidemiological model with computationally inexpensive surrogates, that can reduce the computational time to minutes, without a significant loss of accuracy. The surrogates are created by projecting the output of an epidemiological model on a set of polynomial chaos bases; thereafter, computations involving the surrogate model reduce to evaluations of a polynomial. We find that the epidemic characterizations obtained with the surrogate models is very close to that obtained with the original model. We also find that the number of projections required to construct a surrogate model is O(10)-O(10{sup 2}) less than the number of samples required by the MCMC to construct a stationary posterior distribution; thus, depending upon the epidemiological models in question, it may be possible to omit the offline creation and caching of surrogate models, prior to their use in an inverse problem. The technique is demonstrated on synthetic data as well as observations from the 1918 influenza pandemic collected at Camp Custer, Michigan.« less

3-D Magnetotelluric Forward Modeling And Inversion Incorporating Topography By Using Vector Finite-Element Method Combined With Divergence Corrections Based On The Magnetic Field (VFEH++)

NASA Astrophysics Data System (ADS)

Shi, X.; Utada, H.; Jiaying, W.

2009-12-01

The vector finite-element method combined with divergence corrections based on the magnetic field H, referred to as VFEH++ method, is developed to simulate the magnetotelluric (MT) responses of 3-D conductivity models. The advantages of the new VFEH++ method are the use of edge-elements to eliminate the vector parasites and the divergence corrections to explicitly guarantee the divergence-free conditions in the whole modeling domain. 3-D MT topographic responses are modeling using the new VFEH++ method, and are compared with those calculated by other numerical methods. The results show that MT responses can be modeled highly accurate using the VFEH+ +method. The VFEH++ algorithm is also employed for the 3-D MT data inversion incorporating topography. The 3-D MT inverse problem is formulated as a minimization problem of the regularized misfit function. In order to avoid the huge memory requirement and very long time for computing the Jacobian sensitivity matrix for Gauss-Newton method, we employ the conjugate gradient (CG) approach to solve the inversion equation. In each iteration of CG algorithm, the cost computation is the product of the Jacobian sensitivity matrix with a model vector x or its transpose with a data vector y, which can be transformed into two pseudo-forwarding modeling. This avoids the full explicitly Jacobian matrix calculation and storage which leads to considerable savings in the memory required by the inversion program in PC computer. The performance of CG algorithm will be illustrated by several typical 3-D models with horizontal earth surface and topographic surfaces. The results show that the VFEH++ and CG algorithms can be effectively employed to 3-D MT field data inversion.

Regularity Aspects in Inverse Musculoskeletal Biomechanics

NASA Astrophysics Data System (ADS)

Lund, Marie; Stâhl, Fredrik; Gulliksson, Mârten

2008-09-01

Inverse simulations of musculoskeletal models computes the internal forces such as muscle and joint reaction forces, which are hard to measure, using the more easily measured motion and external forces as input data. Because of the difficulties of measuring muscle forces and joint reactions, simulations are hard to validate. One way of reducing errors for the simulations is to ensure that the mathematical problem is well-posed. This paper presents a study of regularity aspects for an inverse simulation method, often called forward dynamics or dynamical optimization, that takes into account both measurement errors and muscle dynamics. Regularity is examined for a test problem around the optimum using the approximated quadratic problem. The results shows improved rank by including a regularization term in the objective that handles the mechanical over-determinancy. Using the 3-element Hill muscle model the chosen regularization term is the norm of the activation. To make the problem full-rank only the excitation bounds should be included in the constraints. However, this results in small negative values of the activation which indicates that muscles are pushing and not pulling, which is unrealistic but the error maybe small enough to be accepted for specific applications. These results are a start to ensure better results of inverse musculoskeletal simulations from a numerical point of view.

The inverse problem of acoustic wave scattering by an air-saturated poroelastic cylinder.

PubMed

Ogam, Erick; Fellah, Z E A; Baki, Paul

2013-03-01

The efficient use of plastic foams in a diverse range of structural applications like in noise reduction, cushioning, and sleeping mattresses requires detailed characterization of their permeability and deformation (load-bearing) behavior. The elastic moduli and airflow resistance properties of foams are often measured using two separate techniques, one employing mechanical vibration methods and the other, flow rates of fluids based on fluid mechanics technology, respectively. A multi-parameter inverse acoustic scattering problem to recover airflow resistivity (AR) and mechanical properties of an air-saturated foam cylinder is solved. A wave-fluid saturated poroelastic structure interaction model based on the modified Biot theory and plane-wave decomposition using orthogonal cylindrical functions is employed to solve the inverse problem. The solutions to the inverse problem are obtained by constructing the objective functional given by the total square of the difference between predictions from the model and scattered acoustic field data acquired in an anechoic chamber. The value of the recovered AR is in good agreement with that of a slab sample cut from the cylinder and characterized using a method employing low frequency transmitted and reflected acoustic waves in a long waveguide developed by Fellah et al. [Rev. Sci. Instrum. 78(11), 114902 (2007)].

Pareto Joint Inversion of Love and Quasi Rayleigh's waves - synthetic study

NASA Astrophysics Data System (ADS)

Bogacz, Adrian; Dalton, David; Danek, Tomasz; Miernik, Katarzyna; Slawinski, Michael A.

2017-04-01

In this contribution the specific application of Pareto joint inversion in solving geophysical problem is presented. Pareto criterion combine with Particle Swarm Optimization were used to solve geophysical inverse problems for Love and Quasi Rayleigh's waves. Basic theory of forward problem calculation for chosen surface waves is described. To avoid computational problems some simplification were made. This operation allowed foster and more straightforward calculation without lost of solution generality. According to the solving scheme restrictions, considered model must have exact two layers, elastic isotropic surface layer and elastic isotropic half space with infinite thickness. The aim of the inversion is to obain elastic parameters and model geometry using dispersion data. In calculations different case were considered, such as different number of modes for different wave types and different frequencies. Created solutions are using OpenMP standard for parallel computing, which help in reduction of computational times. The results of experimental computations are presented and commented. This research was performed in the context of The Geomechanics Project supported by Husky Energy. Also, this research was partially supported by the Natural Sciences and Engineering Research Council of Canada, grant 238416-2013, and by the Polish National Science Center under contract No. DEC-2013/11/B/ST10/0472.

Review: Optimization methods for groundwater modeling and management

NASA Astrophysics Data System (ADS)

Yeh, William W.-G.

2015-09-01

Optimization methods have been used in groundwater modeling as well as for the planning and management of groundwater systems. This paper reviews and evaluates the various optimization methods that have been used for solving the inverse problem of parameter identification (estimation), experimental design, and groundwater planning and management. Various model selection criteria are discussed, as well as criteria used for model discrimination. The inverse problem of parameter identification concerns the optimal determination of model parameters using water-level observations. In general, the optimal experimental design seeks to find sampling strategies for the purpose of estimating the unknown model parameters. A typical objective of optimal conjunctive-use planning of surface water and groundwater is to minimize the operational costs of meeting water demand. The optimization methods include mathematical programming techniques such as linear programming, quadratic programming, dynamic programming, stochastic programming, nonlinear programming, and the global search algorithms such as genetic algorithms, simulated annealing, and tabu search. Emphasis is placed on groundwater flow problems as opposed to contaminant transport problems. A typical two-dimensional groundwater flow problem is used to explain the basic formulations and algorithms that have been used to solve the formulated optimization problems.

A Non-linear Geodetic Data Inversion Using ABIC for Slip Distribution on a Fault With an Unknown dip Angle

NASA Astrophysics Data System (ADS)

Fukahata, Y.; Wright, T. J.

2006-12-01

We developed a method of geodetic data inversion for slip distribution on a fault with an unknown dip angle. When fault geometry is unknown, the problem of geodetic data inversion is non-linear. A common strategy for obtaining slip distribution is to first determine the fault geometry by minimizing the square misfit under the assumption of a uniform slip on a rectangular fault, and then apply the usual linear inversion technique to estimate a slip distribution on the determined fault. It is not guaranteed, however, that the fault determined under the assumption of a uniform slip gives the best fault geometry for a spatially variable slip distribution. In addition, in obtaining a uniform slip fault model, we have to simultaneously determine the values of the nine mutually dependent parameters, which is a highly non-linear, complicated process. Although the inverse problem is non-linear for cases with unknown fault geometries, the non-linearity of the problems is actually weak, when we can assume the fault surface to be flat. In particular, when a clear fault trace is observed on the EarthOs surface after an earthquake, we can precisely estimate the strike and the location of the fault. In this case only the dip angle has large ambiguity. In geodetic data inversion we usually need to introduce smoothness constraints in order to compromise reciprocal requirements for model resolution and estimation errors in a natural way. Strictly speaking, the inverse problem with smoothness constraints is also non-linear, even if the fault geometry is known. The non-linearity has been dissolved by introducing AkaikeOs Bayesian Information Criterion (ABIC), with which the optimal value of the relative weight of observed data to smoothness constraints is objectively determined. In this study, using ABIC in determining the optimal dip angle, we dissolved the non-linearity of the inverse problem. We applied the method to the InSAR data of the 1995 Dinar, Turkey earthquake and obtained a much shallower dip angle than before.

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.

3D CSEM inversion based on goal-oriented adaptive finite element method

NASA Astrophysics Data System (ADS)

Zhang, Y.; Key, K.

2016-12-01

We present a parallel 3D frequency domain controlled-source electromagnetic inversion code name MARE3DEM. Non-linear inversion of observed data is performed with the Occam variant of regularized Gauss-Newton optimization. The forward operator is based on the goal-oriented finite element method that efficiently calculates the responses and sensitivity kernels in parallel using a data decomposition scheme where independent modeling tasks contain different frequencies and subsets of the transmitters and receivers. To accommodate complex 3D conductivity variation with high flexibility and precision, we adopt the dual-grid approach where the forward mesh conforms to the inversion parameter grid and is adaptively refined until the forward solution converges to the desired accuracy. This dual-grid approach is memory efficient, since the inverse parameter grid remains independent from fine meshing generated around the transmitter and receivers by the adaptive finite element method. Besides, the unstructured inverse mesh efficiently handles multiple scale structures and allows for fine-scale model parameters within the region of interest. Our mesh generation engine keeps track of the refinement hierarchy so that the map of conductivity and sensitivity kernel between the forward and inverse mesh is retained. We employ the adjoint-reciprocity method to calculate the sensitivity kernels which establish a linear relationship between changes in the conductivity model and changes in the modeled responses. Our code uses a direcy solver for the linear systems, so the adjoint problem is efficiently computed by re-using the factorization from the primary problem. Further computational efficiency and scalability is obtained in the regularized Gauss-Newton portion of the inversion using parallel dense matrix-matrix multiplication and matrix factorization routines implemented with the ScaLAPACK library. We show the scalability, reliability and the potential of the algorithm to deal with complex geological scenarios by applying it to the inversion of synthetic marine controlled source EM data generated for a complex 3D offshore model with significant seafloor topography.

A Stochastic Inversion Method for Potential Field Data: Ant Colony Optimization

NASA Astrophysics Data System (ADS)

Liu, Shuang; Hu, Xiangyun; Liu, Tianyou

2014-07-01

Simulating natural ants' foraging behavior, the ant colony optimization (ACO) algorithm performs excellently in combinational optimization problems, for example the traveling salesman problem and the quadratic assignment problem. However, the ACO is seldom used to inverted for gravitational and magnetic data. On the basis of the continuous and multi-dimensional objective function for potential field data optimization inversion, we present the node partition strategy ACO (NP-ACO) algorithm for inversion of model variables of fixed shape and recovery of physical property distributions of complicated shape models. We divide the continuous variables into discrete nodes and ants directionally tour the nodes by use of transition probabilities. We update the pheromone trails by use of Gaussian mapping between the objective function value and the quantity of pheromone. It can analyze the search results in real time and promote the rate of convergence and precision of inversion. Traditional mapping, including the ant-cycle system, weaken the differences between ant individuals and lead to premature convergence. We tested our method by use of synthetic data and real data from scenarios involving gravity and magnetic anomalies. The inverted model variables and recovered physical property distributions were in good agreement with the true values. The ACO algorithm for binary representation imaging and full imaging can recover sharper physical property distributions than traditional linear inversion methods. The ACO has good optimization capability and some excellent characteristics, for example robustness, parallel implementation, and portability, compared with other stochastic metaheuristics.

On the optimization of electromagnetic geophysical data: Application of the PSO algorithm

NASA Astrophysics Data System (ADS)

Godio, A.; Santilano, A.

2018-01-01

Particle Swarm optimization (PSO) algorithm resolves constrained multi-parameter problems and is suitable for simultaneous optimization of linear and nonlinear problems, with the assumption that forward modeling is based on good understanding of ill-posed problem for geophysical inversion. We apply PSO for solving the geophysical inverse problem to infer an Earth model, i.e. the electrical resistivity at depth, consistent with the observed geophysical data. The method doesn't require an initial model and can be easily constrained, according to external information for each single sounding. The optimization process to estimate the model parameters from the electromagnetic soundings focuses on the discussion of the objective function to be minimized. We discuss the possibility to introduce in the objective function vertical and lateral constraints, with an Occam-like regularization. A sensitivity analysis allowed us to check the performance of the algorithm. The reliability of the approach is tested on synthetic, real Audio-Magnetotelluric (AMT) and Long Period MT data. The method appears able to solve complex problems and allows us to estimate the a posteriori distribution of the model parameters.

Bayesian inference in geomagnetism

NASA Technical Reports Server (NTRS)

Backus, George E.

1988-01-01

The inverse problem in empirical geomagnetic modeling is investigated, with critical examination of recently published studies. Particular attention is given to the use of Bayesian inference (BI) to select the damping parameter lambda in the uniqueness portion of the inverse problem. The mathematical bases of BI and stochastic inversion are explored, with consideration of bound-softening problems and resolution in linear Gaussian BI. The problem of estimating the radial magnetic field B(r) at the earth core-mantle boundary from surface and satellite measurements is then analyzed in detail, with specific attention to the selection of lambda in the studies of Gubbins (1983) and Gubbins and Bloxham (1985). It is argued that the selection method is inappropriate and leads to lambda values much larger than those that would result if a reasonable bound on the heat flow at the CMB were assumed.

A trade-off solution between model resolution and covariance in surface-wave inversion

USGS Publications Warehouse

Xia, J.; Xu, Y.; Miller, R.D.; Zeng, C.

2010-01-01

Regularization is necessary for inversion of ill-posed geophysical problems. Appraisal of inverse models is essential for meaningful interpretation of these models. Because uncertainties are associated with regularization parameters, extra conditions are usually required to determine proper parameters for assessing inverse models. Commonly used techniques for assessment of a geophysical inverse model derived (generally iteratively) from a linear system are based on calculating the model resolution and the model covariance matrices. Because the model resolution and the model covariance matrices of the regularized solutions are controlled by the regularization parameter, direct assessment of inverse models using only the covariance matrix may provide incorrect results. To assess an inverted model, we use the concept of a trade-off between model resolution and covariance to find a proper regularization parameter with singular values calculated in the last iteration. We plot the singular values from large to small to form a singular value plot. A proper regularization parameter is normally the first singular value that approaches zero in the plot. With this regularization parameter, we obtain a trade-off solution between model resolution and model covariance in the vicinity of a regularized solution. The unit covariance matrix can then be used to calculate error bars of the inverse model at a resolution level determined by the regularization parameter. We demonstrate this approach with both synthetic and real surface-wave data. ?? 2010 Birkh??user / Springer Basel AG.

Proposing integrated Shannon's entropy-inverse data envelopment analysis methods for resource allocation problem under a fuzzy environment

NASA Astrophysics Data System (ADS)

Çakır, Süleyman

2017-10-01

In this study, a two-phase methodology for resource allocation problems under a fuzzy environment is proposed. In the first phase, the imprecise Shannon's entropy method and the acceptability index are suggested, for the first time in the literature, to select input and output variables to be used in the data envelopment analysis (DEA) application. In the second step, an interval inverse DEA model is executed for resource allocation in a short run. In an effort to exemplify the practicality of the proposed fuzzy model, a real case application has been conducted involving 16 cement firms listed in Borsa Istanbul. The results of the case application indicated that the proposed hybrid model is a viable procedure to handle input-output selection and resource allocation problems under fuzzy conditions. The presented methodology can also lend itself to different applications such as multi-criteria decision-making problems.

Parallel three-dimensional magnetotelluric inversion using adaptive finite-element method. Part I: theory and synthetic study

NASA Astrophysics Data System (ADS)

Grayver, Alexander V.

2015-07-01

This paper presents a distributed magnetotelluric inversion scheme based on adaptive finite-element method (FEM). The key novel aspect of the introduced algorithm is the use of automatic mesh refinement techniques for both forward and inverse modelling. These techniques alleviate tedious and subjective procedure of choosing a suitable model parametrization. To avoid overparametrization, meshes for forward and inverse problems were decoupled. For calculation of accurate electromagnetic (EM) responses, automatic mesh refinement algorithm based on a goal-oriented error estimator has been adopted. For further efficiency gain, EM fields for each frequency were calculated using independent meshes in order to account for substantially different spatial behaviour of the fields over a wide range of frequencies. An automatic approach for efficient initial mesh design in inverse problems based on linearized model resolution matrix was developed. To make this algorithm suitable for large-scale problems, it was proposed to use a low-rank approximation of the linearized model resolution matrix. In order to fill a gap between initial and true model complexities and resolve emerging 3-D structures better, an algorithm for adaptive inverse mesh refinement was derived. Within this algorithm, spatial variations of the imaged parameter are calculated and mesh is refined in the neighborhoods of points with the largest variations. A series of numerical tests were performed to demonstrate the utility of the presented algorithms. Adaptive mesh refinement based on the model resolution estimates provides an efficient tool to derive initial meshes which account for arbitrary survey layouts, data types, frequency content and measurement uncertainties. Furthermore, the algorithm is capable to deliver meshes suitable to resolve features on multiple scales while keeping number of unknowns low. However, such meshes exhibit dependency on an initial model guess. Additionally, it is demonstrated that the adaptive mesh refinement can be particularly efficient in resolving complex shapes. The implemented inversion scheme was able to resolve a hemisphere object with sufficient resolution starting from a coarse discretization and refining mesh adaptively in a fully automatic process. The code is able to harness the computational power of modern distributed platforms and is shown to work with models consisting of millions of degrees of freedom. Significant computational savings were achieved by using locally refined decoupled meshes.

Direct Estimation of Optical Parameters From Photoacoustic Time Series in Quantitative Photoacoustic Tomography.

PubMed

Pulkkinen, Aki; Cox, Ben T; Arridge, Simon R; Goh, Hwan; Kaipio, Jari P; Tarvainen, Tanja

2016-11-01

Estimation of optical absorption and scattering of a target is an inverse problem associated with quantitative photoacoustic tomography. Conventionally, the problem is expressed as two folded. First, images of initial pressure distribution created by absorption of a light pulse are formed based on acoustic boundary measurements. Then, the optical properties are determined based on these photoacoustic images. The optical stage of the inverse problem can thus suffer from, for example, artefacts caused by the acoustic stage. These could be caused by imperfections in the acoustic measurement setting, of which an example is a limited view acoustic measurement geometry. In this work, the forward model of quantitative photoacoustic tomography is treated as a coupled acoustic and optical model and the inverse problem is solved by using a Bayesian approach. Spatial distribution of the optical properties of the imaged target are estimated directly from the photoacoustic time series in varying acoustic detection and optical illumination configurations. It is numerically demonstrated, that estimation of optical properties of the imaged target is feasible in limited view acoustic detection setting.

Large-scale 3D inversion of marine controlled source electromagnetic data using the integral equation method

NASA Astrophysics Data System (ADS)

Zhdanov, M. S.; Cuma, M.; Black, N.; Wilson, G. A.

2009-12-01

The marine controlled source electromagnetic (MCSEM) method has become widely used in offshore oil and gas exploration. Interpretation of MCSEM data is still a very challenging problem, especially if one would like to take into account the realistic 3D structure of the subsurface. The inversion of MCSEM data is complicated by the fact that the EM response of a hydrocarbon-bearing reservoir is very weak in comparison with the background EM fields generated by an electric dipole transmitter in complex geoelectrical structures formed by a conductive sea-water layer and the terranes beneath it. In this paper, we present a review of the recent developments in the area of large-scale 3D EM forward modeling and inversion. Our approach is based on using a new integral form of Maxwell’s equations allowing for an inhomogeneous background conductivity, which results in a numerically effective integral representation for 3D EM field. This representation provides an efficient tool for the solution of 3D EM inverse problems. To obtain a robust inverse model of the conductivity distribution, we apply regularization based on a focusing stabilizing functional which allows for the recovery of models with both smooth and sharp geoelectrical boundaries. The method is implemented in a fully parallel computer code, which makes it possible to run large-scale 3D inversions on grids with millions of inversion cells. This new technique can be effectively used for active EM detection and monitoring of the subsurface targets.

On a comparison of two schemes in sequential data assimilation

NASA Astrophysics Data System (ADS)

Grishina, Anastasiia A.; Penenko, Alexey V.

2017-11-01

This paper is focused on variational data assimilation as an approach to mathematical modeling. Realization of the approach requires a sequence of connected inverse problems with different sets of observational data to be solved. Two variational data assimilation schemes, "implicit" and "explicit", are considered in the article. Their equivalence is shown and the numerical results are given on a basis of non-linear Robertson system. To avoid the "inverse problem crime" different schemes were used to produce synthetic measurement and to solve the data assimilation problem.

Inverse problems and computational cell metabolic models: a statistical approach

NASA Astrophysics Data System (ADS)

Calvetti, D.; Somersalo, E.

2008-07-01

In this article, we give an overview of the Bayesian modelling of metabolic systems at the cellular and subcellular level. The models are based on detailed description of key biochemical reactions occurring in tissue, which may in turn be compartmentalized into cytosol and mitochondria, and of transports between the compartments. The classical deterministic approach which models metabolic systems as dynamical systems with Michaelis-Menten kinetics, is replaced by a stochastic extension where the model parameters are interpreted as random variables with an appropriate probability density. The inverse problem of cell metabolism in this setting consists of estimating the density of the model parameters. After discussing some possible approaches to solving the problem, we address the issue of how to assess the reliability of the predictions of a stochastic model by proposing an output analysis in terms of model uncertainties. Visualization modalities for organizing the large amount of information provided by the Bayesian dynamic sensitivity analysis are also illustrated.

3D first-arrival traveltime tomography with modified total variation regularization

NASA Astrophysics Data System (ADS)

Jiang, Wenbin; Zhang, Jie

2018-02-01

Three-dimensional (3D) seismic surveys have become a major tool in the exploration and exploitation of hydrocarbons. 3D seismic first-arrival traveltime tomography is a robust method for near-surface velocity estimation. A common approach for stabilizing the ill-posed inverse problem is to apply Tikhonov regularization to the inversion. However, the Tikhonov regularization method recovers smooth local structures while blurring the sharp features in the model solution. We present a 3D first-arrival traveltime tomography method with modified total variation (MTV) regularization to preserve sharp velocity contrasts and improve the accuracy of velocity inversion. To solve the minimization problem of the new traveltime tomography method, we decouple the original optimization problem into two following subproblems: a standard traveltime tomography problem with the traditional Tikhonov regularization and a L2 total variation problem. We apply the conjugate gradient method and split-Bregman iterative method to solve these two subproblems, respectively. Our synthetic examples show that the new method produces higher resolution models than the conventional traveltime tomography with Tikhonov regularization. We apply the technique to field data. The stacking section shows significant improvements with static corrections from the MTV traveltime tomography.

Structural Damage Detection Using Changes in Natural Frequencies: Theory and Applications

NASA Astrophysics Data System (ADS)

He, K.; Zhu, W. D.

2011-07-01

A vibration-based method that uses changes in natural frequencies of a structure to detect damage has advantages over conventional nondestructive tests in detecting various types of damage, including loosening of bolted joints, using minimum measurement data. Two major challenges associated with applications of the vibration-based damage detection method to engineering structures are addressed: accurate modeling of structures and the development of a robust inverse algorithm to detect damage, which are defined as the forward and inverse problems, respectively. To resolve the forward problem, new physics-based finite element modeling techniques are developed for fillets in thin-walled beams and for bolted joints, so that complex structures can be accurately modeled with a reasonable model size. To resolve the inverse problem, a logistical function transformation is introduced to convert the constrained optimization problem to an unconstrained one, and a robust iterative algorithm using a trust-region method, called the Levenberg-Marquardt method, is developed to accurately detect the locations and extent of damage. The new methodology can ensure global convergence of the iterative algorithm in solving under-determined system equations and deal with damage detection problems with relatively large modeling error and measurement noise. The vibration-based damage detection method is applied to various structures including lightning masts, a space frame structure and one of its components, and a pipeline. The exact locations and extent of damage can be detected in the numerical simulation where there is no modeling error and measurement noise. The locations and extent of damage can be successfully detected in experimental damage detection.

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.

Inverse problem of radiofrequency sounding of ionosphere

NASA Astrophysics Data System (ADS)

Velichko, E. N.; Yu. Grishentsev, A.; Korobeynikov, A. G.

2016-01-01

An algorithm for the solution of the inverse problem of vertical ionosphere sounding and a mathematical model of noise filtering are presented. An automated system for processing and analysis of spectrograms of vertical ionosphere sounding based on our algorithm is described. It is shown that the algorithm we suggest has a rather high efficiency. This is supported by the data obtained at the ionospheric stations of the so-called “AIS-M” type.

A Geophysical Inversion Model Enhancement Technique Based on the Blind Deconvolution

NASA Astrophysics Data System (ADS)

Zuo, B.; Hu, X.; Li, H.

2011-12-01

A model-enhancement technique is proposed to enhance the geophysical inversion model edges and details without introducing any additional information. Firstly, the theoretic correctness of the proposed geophysical inversion model-enhancement technique is discussed. An inversion MRM (model resolution matrix) convolution approximating PSF (Point Spread Function) method is designed to demonstrate the correctness of the deconvolution model enhancement method. Then, a total-variation regularization blind deconvolution geophysical inversion model-enhancement algorithm is proposed. In previous research, Oldenburg et al. demonstrate the connection between the PSF and the geophysical inverse solution. Alumbaugh et al. propose that more information could be provided by the PSF if we return to the idea of it behaving as an averaging or low pass filter. We consider the PSF as a low pass filter to enhance the inversion model basis on the theory of the PSF convolution approximation. Both the 1D linear and the 2D magnetotelluric inversion examples are used to analyze the validity of the theory and the algorithm. To prove the proposed PSF convolution approximation theory, the 1D linear inversion problem is considered. It shows the ratio of convolution approximation error is only 0.15%. The 2D synthetic model enhancement experiment is presented. After the deconvolution enhancement, the edges of the conductive prism and the resistive host become sharper, and the enhancement result is closer to the actual model than the original inversion model according the numerical statistic analysis. Moreover, the artifacts in the inversion model are suppressed. The overall precision of model increases 75%. All of the experiments show that the structure details and the numerical precision of inversion model are significantly improved, especially in the anomalous region. The correlation coefficient between the enhanced inversion model and the actual model are shown in Fig. 1. The figure illustrates that more information and details structure of the actual model are enhanced through the proposed enhancement algorithm. Using the proposed enhancement method can help us gain a clearer insight into the results of the inversions and help make better informed decisions.

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.

Identification procedure for epistemic uncertainties using inverse fuzzy arithmetic

NASA Astrophysics Data System (ADS)

Haag, T.; Herrmann, J.; Hanss, M.

2010-10-01

For the mathematical representation of systems with epistemic uncertainties, arising, for example, from simplifications in the modeling procedure, models with fuzzy-valued parameters prove to be a suitable and promising approach. In practice, however, the determination of these parameters turns out to be a non-trivial problem. The identification procedure to appropriately update these parameters on the basis of a reference output (measurement or output of an advanced model) requires the solution of an inverse problem. Against this background, an inverse method for the computation of the fuzzy-valued parameters of a model with epistemic uncertainties is presented. This method stands out due to the fact that it only uses feedforward simulations of the model, based on the transformation method of fuzzy arithmetic, along with the reference output. An inversion of the system equations is not necessary. The advancement of the method presented in this paper consists of the identification of multiple input parameters based on a single reference output or measurement. An optimization is used to solve the resulting underdetermined problems by minimizing the uncertainty of the identified parameters. Regions where the identification procedure is reliable are determined by the computation of a feasibility criterion which is also based on the output data of the transformation method only. For a frequency response function of a mechanical system, this criterion allows a restriction of the identification process to some special range of frequency where its solution can be guaranteed. Finally, the practicability of the method is demonstrated by covering the measured output of a fluid-filled piping system by the corresponding uncertain FE model in a conservative way.

Inverse problems in 1D hemodynamics on systemic networks: a sequential approach.

PubMed

Lombardi, D

2014-02-01

In this work, a sequential approach based on the unscented Kalman filter is applied to solve inverse problems in 1D hemodynamics, on a systemic network. For instance, the arterial stiffness is estimated by exploiting cross-sectional area and mean speed observations in several locations of the arteries. The results are compared with those ones obtained by estimating the pulse wave velocity and the Moens-Korteweg formula. In the last section, a perspective concerning the identification of the terminal models parameters and peripheral circulation (modeled by a Windkessel circuit) is presented. Copyright © 2013 John Wiley & Sons, Ltd.

2D Inversion of Transient Electromagnetic Method (TEM)

NASA Astrophysics Data System (ADS)

Bortolozo, Cassiano Antonio; Luís Porsani, Jorge; Acácio Monteiro dos Santos, Fernando

2017-04-01

A new methodology was developed for 2D inversion of Transient Electromagnetic Method (TEM). The methodology consists in the elaboration of a set of routines in Matlab code for modeling and inversion of TEM data and the determination of the most efficient field array for the problem. In this research, the 2D TEM modeling uses the finite differences discretization. To solve the inversion problem, were applied an algorithm based on Marquardt technique, also known as Ridge Regression. The algorithm is stable and efficient and it is widely used in geoelectrical inversion problems. The main advantage of 1D survey is the rapid data acquisition in a large area, but in regions with two-dimensional structures or that need more details, is essential to use two-dimensional interpretation methodologies. For an efficient field acquisition we used in an innovative form the fixed-loop array, with a square transmitter loop (200m x 200m) and 25m spacing between the sounding points. The TEM surveys were conducted only inside the transmitter loop, in order to not deal with negative apparent resistivity values. Although it is possible to model the negative values, it makes the inversion convergence more difficult. Therefore the methodology described above has been developed in order to achieve maximum optimization of data acquisition. Since it is necessary only one transmitter loop disposition in the surface for each series of soundings inside the loop. The algorithms were tested with synthetic data and the results were essential to the interpretation of the results with real data and will be useful in future situations. With the inversion of the real data acquired over the Paraná Sedimentary Basin (PSB) was successful realized a 2D TEM inversion. The results indicate a robust geoelectrical characterization for the sedimentary and crystalline aquifers in the PSB. Therefore, using a new and relevant approach for 2D TEM inversion, this research effectively contributed to map the most promising regions for groundwater exploration. In addition, there was the development of new geophysical software that can be applied as an important tool for many geological/hydrogeological applications and educational purposes.

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.

jInv: A Modular and Scalable Framework for Electromagnetic Inverse Problems

NASA Astrophysics Data System (ADS)

Belliveau, P. T.; Haber, E.

2016-12-01

Inversion is a key tool in the interpretation of geophysical electromagnetic (EM) data. Three-dimensional (3D) EM inversion is very computationally expensive and practical software for inverting large 3D EM surveys must be able to take advantage of high performance computing (HPC) resources. It has traditionally been difficult to achieve those goals in a high level dynamic programming environment that allows rapid development and testing of new algorithms, which is important in a research setting. With those goals in mind, we have developed jInv, a framework for PDE constrained parameter estimation problems. jInv provides optimization and regularization routines, a framework for user defined forward problems, and interfaces to several direct and iterative solvers for sparse linear systems. The forward modeling framework provides finite volume discretizations of differential operators on rectangular tensor product meshes and tetrahedral unstructured meshes that can be used to easily construct forward modeling and sensitivity routines for forward problems described by partial differential equations. jInv is written in the emerging programming language Julia. Julia is a dynamic language targeted at the computational science community with a focus on high performance and native support for parallel programming. We have developed frequency and time-domain EM forward modeling and sensitivity routines for jInv. We will illustrate its capabilities and performance with two synthetic time-domain EM inversion examples. First, in airborne surveys, which use many sources, we achieve distributed memory parallelism by decoupling the forward and inverse meshes and performing forward modeling for each source on small, locally refined meshes. Secondly, we invert grounded source time-domain data from a gradient array style induced polarization survey using a novel time-stepping technique that allows us to compute data from different time-steps in parallel. These examples both show that it is possible to invert large scale 3D time-domain EM datasets within a modular, extensible framework written in a high-level, easy to use programming language.

Variational methods to estimate terrestrial ecosystem model parameters

NASA Astrophysics Data System (ADS)

Delahaies, Sylvain; Roulstone, Ian

2016-04-01

Carbon is at the basis of the chemistry of life. Its ubiquity in the Earth system is the result of complex recycling processes. Present in the atmosphere in the form of carbon dioxide it is adsorbed by marine and terrestrial ecosystems and stored within living biomass and decaying organic matter. Then soil chemistry and a non negligible amount of time transform the dead matter into fossil fuels. Throughout this cycle, carbon dioxide is released in the atmosphere through respiration and combustion of fossils fuels. Model-data fusion techniques allow us to combine our understanding of these complex processes with an ever-growing amount of observational data to help improving models and predictions. The data assimilation linked ecosystem carbon (DALEC) model is a simple box model simulating the carbon budget allocation for terrestrial ecosystems. Over the last decade several studies have demonstrated the relative merit of various inverse modelling strategies (MCMC, ENKF, 4DVAR) to estimate model parameters and initial carbon stocks for DALEC and to quantify the uncertainty in the predictions. Despite its simplicity, DALEC represents the basic processes at the heart of more sophisticated models of the carbon cycle. Using adjoint based methods we study inverse problems for DALEC with various data streams (8 days MODIS LAI, monthly MODIS LAI, NEE). The framework of constraint optimization allows us to incorporate ecological common sense into the variational framework. We use resolution matrices to study the nature of the inverse problems and to obtain data importance and information content for the different type of data. We study how varying the time step affect the solutions, and we show how "spin up" naturally improves the conditioning of the inverse problems.

Inverse modeling methods for indoor airborne pollutant tracking: literature review and fundamentals.

PubMed

Liu, X; Zhai, Z

2007-12-01

Reduction in indoor environment quality calls for effective control and improvement measures. Accurate and prompt identification of contaminant sources ensures that they can be quickly removed and contaminated spaces isolated and cleaned. This paper discusses the use of inverse modeling to identify potential indoor pollutant sources with limited pollutant sensor data. The study reviews various inverse modeling methods for advection-dispersion problems and summarizes the methods into three major categories: forward, backward, and probability inverse modeling methods. The adjoint probability inverse modeling method is indicated as an appropriate model for indoor air pollutant tracking because it can quickly find source location, strength and release time without prior information. The paper introduces the principles of the adjoint probability method and establishes the corresponding adjoint equations for both multi-zone airflow models and computational fluid dynamics (CFD) models. The study proposes a two-stage inverse modeling approach integrating both multi-zone and CFD models, which can provide a rapid estimate of indoor pollution status and history for a whole building. Preliminary case study results indicate that the adjoint probability method is feasible for indoor pollutant inverse modeling. The proposed method can help identify contaminant source characteristics (location and release time) with limited sensor outputs. This will ensure an effective and prompt execution of building management strategies and thus achieve a healthy and safe indoor environment. The method can also help design optimal sensor networks.

Cerebellum-inspired neural network solution of the inverse kinematics problem.

PubMed

Asadi-Eydivand, Mitra; Ebadzadeh, Mohammad Mehdi; Solati-Hashjin, Mehran; Darlot, Christian; Abu Osman, Noor Azuan

2015-12-01

The demand today for more complex robots that have manipulators with higher degrees of freedom is increasing because of technological advances. Obtaining the precise movement for a desired trajectory or a sequence of arm and positions requires the computation of the inverse kinematic (IK) function, which is a major problem in robotics. The solution of the IK problem leads robots to the precise position and orientation of their end-effector. We developed a bioinspired solution comparable with the cerebellar anatomy and function to solve the said problem. The proposed model is stable under all conditions merely by parameter determination, in contrast to recursive model-based solutions, which remain stable only under certain conditions. We modified the proposed model for the simple two-segmented arm to prove the feasibility of the model under a basic condition. A fuzzy neural network through its learning method was used to compute the parameters of the system. Simulation results show the practical feasibility and efficiency of the proposed model in robotics. The main advantage of the proposed model is its generalizability and potential use in any robot.

Optimization, Monotonicity and the Determination of Nash Equilibria — An Algorithmic Analysis

NASA Astrophysics Data System (ADS)

Lozovanu, D.; Pickl, S. W.; Weber, G.-W.

2004-08-01

This paper is concerned with the optimization of a nonlinear time-discrete model exploiting the special structure of the underlying cost game and the property of inverse matrices. The costs are interlinked by a system of linear inequalities. It is shown that, if the players cooperate, i.e., minimize the sum of all the costs, they achieve a Nash equilibrium. In order to determine Nash equilibria, the simplex method can be applied with respect to the dual problem. An introduction into the TEM model and its relationship to an economic Joint Implementation program is given. The equivalence problem is presented. The construction of the emission cost game and the allocation problem is explained. The assumption of inverse monotony for the matrices leads to a new result in the area of such allocation problems. A generalization of such problems is presented.

Source localization in electromyography using the inverse potential problem

NASA Astrophysics Data System (ADS)

van den Doel, Kees; Ascher, Uri M.; Pai, Dinesh K.

2011-02-01

We describe an efficient method for reconstructing the activity in human muscles from an array of voltage sensors on the skin surface. MRI is used to obtain morphometric data which are segmented into muscle tissue, fat, bone and skin, from which a finite element model for volume conduction is constructed. The inverse problem of finding the current sources in the muscles is solved using a careful regularization technique which adds a priori information, yielding physically reasonable solutions from among those that satisfy the basic potential problem. Several regularization functionals are considered and numerical experiments on a 2D test model are performed to determine which performs best. The resulting scheme leads to numerical difficulties when applied to large-scale 3D problems. We clarify the nature of these difficulties and provide a method to overcome them, which is shown to perform well in the large-scale problem setting.

Hydrologic characterization of desert soils with varying degrees of pedogenesis: 2. Inverse modeling for eff ective properties

USGS Publications Warehouse

Mirus, B.B.; Perkins, K.S.; Nimmo, J.R.; Singha, K.

2009-01-01

To understand their relation to pedogenic development, soil hydraulic properties in the Mojave Desert were investi- gated for three deposit types: (i) recently deposited sediments in an active wash, (ii) a soil of early Holocene age, and (iii) a highly developed soil of late Pleistocene age. Eff ective parameter values were estimated for a simplifi ed model based on Richards' equation using a fl ow simulator (VS2D), an inverse algorithm (UCODE-2005), and matric pressure and water content data from three ponded infi ltration experiments. The inverse problem framework was designed to account for the eff ects of subsurface lateral spreading of infi ltrated water. Although none of the inverse problems converged on a unique, best-fi t parameter set, a minimum standard error of regression was reached for each deposit type. Parameter sets from the numerous inversions that reached the minimum error were used to develop probability distribu tions for each parameter and deposit type. Electrical resistance imaging obtained for two of the three infi ltration experiments was used to independently test fl ow model performance. Simulations for the active wash and Holocene soil successfully depicted the lateral and vertical fl uxes. Simulations of the more pedogenically developed Pleistocene soil did not adequately replicate the observed fl ow processes, which would require a more complex conceptual model to include smaller scale heterogeneities. The inverse-modeling results, however, indicate that with increasing age, the steep slope of the soil water retention curve shitis toward more negative matric pressures. Assigning eff ective soil hydraulic properties based on soil age provides a promising framework for future development of regional-scale models of soil moisture dynamics in arid environments for land-management applications. ?? Soil Science Society of America.

A sparse reconstruction method for the estimation of multi-resolution emission fields via atmospheric inversion

DOE PAGES

Ray, J.; Lee, J.; Yadav, V.; ...

2015-04-29

Atmospheric inversions are frequently used to estimate fluxes of atmospheric greenhouse gases (e.g., biospheric CO 2 flux fields) at Earth's surface. These inversions typically assume that flux departures from a prior model are spatially smoothly varying, which are then modeled using a multi-variate Gaussian. When the field being estimated is spatially rough, multi-variate Gaussian models are difficult to construct and a wavelet-based field model may be more suitable. Unfortunately, such models are very high dimensional and are most conveniently used when the estimation method can simultaneously perform data-driven model simplification (removal of model parameters that cannot be reliably estimated) andmore » fitting. Such sparse reconstruction methods are typically not used in atmospheric inversions. In this work, we devise a sparse reconstruction method, and illustrate it in an idealized atmospheric inversion problem for the estimation of fossil fuel CO 2 (ffCO 2) emissions in the lower 48 states of the USA. Our new method is based on stagewise orthogonal matching pursuit (StOMP), a method used to reconstruct compressively sensed images. Our adaptations bestow three properties to the sparse reconstruction procedure which are useful in atmospheric inversions. We have modified StOMP to incorporate prior information on the emission field being estimated and to enforce non-negativity on the estimated field. Finally, though based on wavelets, our method allows for the estimation of fields in non-rectangular geometries, e.g., emission fields inside geographical and political boundaries. Our idealized inversions use a recently developed multi-resolution (i.e., wavelet-based) random field model developed for ffCO 2 emissions and synthetic observations of ffCO 2 concentrations from a limited set of measurement sites. We find that our method for limiting the estimated field within an irregularly shaped region is about a factor of 10 faster than conventional approaches. It also reduces the overall computational cost by a factor of 2. Further, the sparse reconstruction scheme imposes non-negativity without introducing strong nonlinearities, such as those introduced by employing log-transformed fields, and thus reaps the benefits of simplicity and computational speed that are characteristic of linear inverse problems.« less

A sparse reconstruction method for the estimation of multi-resolution emission fields via atmospheric inversion

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

Ray, J.; Lee, J.; Yadav, V.

Atmospheric inversions are frequently used to estimate fluxes of atmospheric greenhouse gases (e.g., biospheric CO 2 flux fields) at Earth's surface. These inversions typically assume that flux departures from a prior model are spatially smoothly varying, which are then modeled using a multi-variate Gaussian. When the field being estimated is spatially rough, multi-variate Gaussian models are difficult to construct and a wavelet-based field model may be more suitable. Unfortunately, such models are very high dimensional and are most conveniently used when the estimation method can simultaneously perform data-driven model simplification (removal of model parameters that cannot be reliably estimated) andmore » fitting. Such sparse reconstruction methods are typically not used in atmospheric inversions. In this work, we devise a sparse reconstruction method, and illustrate it in an idealized atmospheric inversion problem for the estimation of fossil fuel CO 2 (ffCO 2) emissions in the lower 48 states of the USA. Our new method is based on stagewise orthogonal matching pursuit (StOMP), a method used to reconstruct compressively sensed images. Our adaptations bestow three properties to the sparse reconstruction procedure which are useful in atmospheric inversions. We have modified StOMP to incorporate prior information on the emission field being estimated and to enforce non-negativity on the estimated field. Finally, though based on wavelets, our method allows for the estimation of fields in non-rectangular geometries, e.g., emission fields inside geographical and political boundaries. Our idealized inversions use a recently developed multi-resolution (i.e., wavelet-based) random field model developed for ffCO 2 emissions and synthetic observations of ffCO 2 concentrations from a limited set of measurement sites. We find that our method for limiting the estimated field within an irregularly shaped region is about a factor of 10 faster than conventional approaches. It also reduces the overall computational cost by a factor of 2. Further, the sparse reconstruction scheme imposes non-negativity without introducing strong nonlinearities, such as those introduced by employing log-transformed fields, and thus reaps the benefits of simplicity and computational speed that are characteristic of linear inverse problems.« less

Using the ARTMO toolbox for automated retrieval of biophysical parameters through radiative transfer model inversion: Optimizing LUT-based inversion

NASA Astrophysics Data System (ADS)

Verrelst, J.; Rivera, J. P.; Leonenko, G.; Alonso, L.; Moreno, J.

2012-04-01

Radiative transfer (RT) modeling plays a key role for earth observation (EO) because it is needed to design EO instruments and to develop and test inversion algorithms. The inversion of a RT model is considered as a successful approach for the retrieval of biophysical parameters because of being physically-based and generally applicable. However, to the broader community this approach is considered as laborious because of its many processing steps and expert knowledge is required to realize precise model parameterization. We have recently developed a radiative transfer toolbox ARTMO (Automated Radiative Transfer Models Operator) with the purpose of providing in a graphical user interface (GUI) essential models and tools required for terrestrial EO applications such as model inversion. In short, the toolbox allows the user: i) to choose between various plant leaf and canopy RT models (e.g. models from the PROSPECT and SAIL family, FLIGHT), ii) to choose between spectral band settings of various air- and space-borne sensors or defining own sensor settings, iii) to simulate a massive amount of spectra based on a look up table (LUT) approach and storing it in a relational database, iv) to plot spectra of multiple models and compare them with measured spectra, and finally, v) to run model inversion against optical imagery given several cost options and accuracy estimates. In this work ARTMO was used to tackle some well-known problems related to model inversion. According to Hadamard conditions, mathematical models of physical phenomena are mathematically invertible if the solution of the inverse problem to be solved exists, is unique and depends continuously on data. This assumption is not always met because of the large number of unknowns and different strategies have been proposed to overcome this problem. Several of these strategies have been implemented in ARTMO and were here analyzed to optimize the inversion performance. Data came from the SPARC-2003 dataset, which was acquired on the agricultural test site Barrax, Spain. LUTs were created using the 4SAIL and FLIGHT models and were inverted against CHRIS data in order to retrieve maps of chlorophyll content (chl) and leaf area index (LAI). The following inversion steps have been optimized: 1. Cost function. The performances of about 50 different cost functions (i.e. minimum distance functions) were compared. Remarkably, in none of the studied cases the widely used root mean square error (RMSE) led to most accurate results. Depending on the retrieved parameter, more successful functions were: 'Sharma and Mittal', 'Shannońs entropy', 'Hellinger distance', 'Pearsońs chi-square'. 2. Gaussian noise. Earth observation data typically encompass a certain degree of noise due to errors related to radiometric and geometric processing. In all cases, adding 5% Gaussian noise to the simulated spectra led to more accurate retrievals as compared to without noise. 3. Average of multiple best solutions. Because multiple parameter combinations may lead to the same spectra, a way to overcome this problem is not searching for the top best match but for a percentage of best matches. Optimized retrievals were encountered when including an average of 7% (Chl) to 10% (LAI) top best matches. 4. Integration of estimates. The option is provided to integrate estimates of biochemical contents at the canopy level (e.g., total chlorophyll: Chl × LAI, or water: Cw × LAI), which can lead to increased robustness and accuracy. 5. Class-based inversion. This option is probably ARTMÓs most powerful feature as it allows model parameterization depending on the imagés land cover classes (e.g. different soil or vegetation types). Class-based inversion can lead to considerably improved accuracies compared to one generic class. Results suggest that 4SAIL and FLIGHT performed alike for Chl but not for LAI. While both models rely on the leaf model PROSPECT for Chl retrieval, their different nature (e.g. numerical vs. ray tracing) may cause that retrieval of structural parameters such as LAI differ. Finally, it should be noted that the whole analysis can be intuitively performed by the toolbox. ARTMO is freely available to the EO community for further development. Expressions of interest are welcome and should be directed to the corresponding author.

A MATLAB implementation of the minimum relative entropy method for linear inverse problems

NASA Astrophysics Data System (ADS)

Neupauer, Roseanna M.; Borchers, Brian

2001-08-01

The minimum relative entropy (MRE) method can be used to solve linear inverse problems of the form Gm= d, where m is a vector of unknown model parameters and d is a vector of measured data. The MRE method treats the elements of m as random variables, and obtains a multivariate probability density function for m. The probability density function is constrained by prior information about the upper and lower bounds of m, a prior expected value of m, and the measured data. The solution of the inverse problem is the expected value of m, based on the derived probability density function. We present a MATLAB implementation of the MRE method. Several numerical issues arise in the implementation of the MRE method and are discussed here. We present the source history reconstruction problem from groundwater hydrology as an example of the MRE implementation.

Full-wave Moment Tensor and Tomographic Inversions Based on 3D Strain Green Tensor

DTIC Science & Technology

2010-01-31

propagation in three-dimensional (3D) earth, linearizes the inverse problem by iteratively updating the earth model , and provides an accurate way to...self-consistent FD-SGT databases constructed from finite-difference simulations of wave propagation in full-wave tomographic models can be used to...determine the moment tensors within minutes after a seismic event, making it possible for real time monitoring using 3D models . 15. SUBJECT TERMS

Distributed micro-releases of bioterror pathogens : threat characterizations and epidemiology from uncertain patient observables.

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

Wolf, Michael M.; Marzouk, Youssef M.; Adams, Brian M.

2008-10-01

Terrorist attacks using an aerosolized pathogen preparation have gained credibility as a national security concern since the anthrax attacks of 2001. The ability to characterize the parameters of such attacks, i.e., to estimate the number of people infected, the time of infection, the average dose received, and the rate of disease spread in contemporary American society (for contagious diseases), is important when planning a medical response. For non-contagious diseases, we address the characterization problem by formulating a Bayesian inverse problem predicated on a short time-series of diagnosed patients exhibiting symptoms. To keep the approach relevant for response planning, we limitmore » ourselves to 3.5 days of data. In computational tests performed for anthrax, we usually find these observation windows sufficient, especially if the outbreak model employed in the inverse problem is accurate. For contagious diseases, we formulated a Bayesian inversion technique to infer both pathogenic transmissibility and the social network from outbreak observations, ensuring that the two determinants of spreading are identified separately. We tested this technique on data collected from a 1967 smallpox epidemic in Abakaliki, Nigeria. We inferred, probabilistically, different transmissibilities in the structured Abakaliki population, the social network, and the chain of transmission. Finally, we developed an individual-based epidemic model to realistically simulate the spread of a rare (or eradicated) disease in a modern society. This model incorporates the mixing patterns observed in an (American) urban setting and accepts, as model input, pathogenic transmissibilities estimated from historical outbreaks that may have occurred in socio-economic environments with little resemblance to contemporary society. Techniques were also developed to simulate disease spread on static and sampled network reductions of the dynamic social networks originally in the individual-based model, yielding faster, though approximate, network-based epidemic models. These reduced-order models are useful in scenario analysis for medical response planning, as well as in computationally intensive inverse problems.« less

Space structures insulating material's thermophysical and radiation properties estimation

NASA Astrophysics Data System (ADS)

Nenarokomov, A. V.; Alifanov, O. M.; Titov, D. M.

2007-11-01

In many practical situations in aerospace technology it is impossible to measure directly such properties of analyzed materials (for example, composites) as thermal and radiation characteristics. The only way that can often be used to overcome these difficulties is indirect measurements. This type of measurement is usually formulated as the solution of inverse heat transfer problems. Such problems are ill-posed in mathematical sense and their main feature shows itself in the solution instabilities. That is why special regularizing methods are needed to solve them. The experimental methods of identification of the mathematical models of heat transfer based on solving the inverse problems are one of the modern effective solving manners. The objective of this paper is to estimate thermal and radiation properties of advanced materials using the approach based on inverse methods.

Analysis of harmonic spline gravity models for Venus and Mars

NASA Technical Reports Server (NTRS)

Bowin, Carl

1986-01-01

Methodology utilizing harmonic splines for determining the true gravity field from Line-Of-Sight (LOS) acceleration data from planetary spacecraft missions was tested. As is well known, the LOS data incorporate errors in the zero reference level that appear to be inherent in the processing procedure used to obtain the LOS vectors. The proposed method offers a solution to this problem. The harmonic spline program was converted from the VAX 11/780 to the Ridge 32C computer. The problem with the matrix inversion routine that improved inversion of the data matrices used in the Optimum Estimate program for global Earth studies was solved. The problem of obtaining a successful matrix inversion for a single rev supplemented by data for the two adjacent revs still remains.

Solving ill-posed inverse problems using iterative deep neural networks

NASA Astrophysics Data System (ADS)

Adler, Jonas; Öktem, Ozan

2017-12-01

We propose a partially learned approach for the solution of ill-posed inverse problems with not necessarily linear forward operators. The method builds on ideas from classical regularisation theory and recent advances in deep learning to perform learning while making use of prior information about the inverse problem encoded in the forward operator, noise model and a regularising functional. The method results in a gradient-like iterative scheme, where the ‘gradient’ component is learned using a convolutional network that includes the gradients of the data discrepancy and regulariser as input in each iteration. We present results of such a partially learned gradient scheme on a non-linear tomographic inversion problem with simulated data from both the Sheep-Logan phantom as well as a head CT. The outcome is compared against filtered backprojection and total variation reconstruction and the proposed method provides a 5.4 dB PSNR improvement over the total variation reconstruction while being significantly faster, giving reconstructions of 512 × 512 pixel images in about 0.4 s using a single graphics processing unit (GPU).

Quo vadis: Hydrologic inverse analyses using high-performance computing and a D-Wave quantum annealer

NASA Astrophysics Data System (ADS)

O'Malley, D.; Vesselinov, V. V.

2017-12-01

Classical microprocessors have had a dramatic impact on hydrology for decades, due largely to the exponential growth in computing power predicted by Moore's law. However, this growth is not expected to continue indefinitely and has already begun to slow. Quantum computing is an emerging alternative to classical microprocessors. Here, we demonstrated cutting edge inverse model analyses utilizing some of the best available resources in both worlds: high-performance classical computing and a D-Wave quantum annealer. The classical high-performance computing resources are utilized to build an advanced numerical model that assimilates data from O(10^5) observations, including water levels, drawdowns, and contaminant concentrations. The developed model accurately reproduces the hydrologic conditions at a Los Alamos National Laboratory contamination site, and can be leveraged to inform decision-making about site remediation. We demonstrate the use of a D-Wave 2X quantum annealer to solve hydrologic inverse problems. This work can be seen as an early step in quantum-computational hydrology. We compare and contrast our results with an early inverse approach in classical-computational hydrology that is comparable to the approach we use with quantum annealing. Our results show that quantum annealing can be useful for identifying regions of high and low permeability within an aquifer. While the problems we consider are small-scale compared to the problems that can be solved with modern classical computers, they are large compared to the problems that could be solved with early classical CPUs. Further, the binary nature of the high/low permeability problem makes it well-suited to quantum annealing, but challenging for classical computers.

Sensitivity computation of the ell1 minimization problem and its application to dictionary design of ill-posed problems

NASA Astrophysics Data System (ADS)

Horesh, L.; Haber, E.

2009-09-01

The ell1 minimization problem has been studied extensively in the past few years. Recently, there has been a growing interest in its application for inverse problems. Most studies have concentrated in devising ways for sparse representation of a solution using a given prototype dictionary. Very few studies have addressed the more challenging problem of optimal dictionary construction, and even these were primarily devoted to the simplistic sparse coding application. In this paper, sensitivity analysis of the inverse solution with respect to the dictionary is presented. This analysis reveals some of the salient features and intrinsic difficulties which are associated with the dictionary design problem. Equipped with these insights, we propose an optimization strategy that alleviates these hurdles while utilizing the derived sensitivity relations for the design of a locally optimal dictionary. Our optimality criterion is based on local minimization of the Bayesian risk, given a set of training models. We present a mathematical formulation and an algorithmic framework to achieve this goal. The proposed framework offers the design of dictionaries for inverse problems that incorporate non-trivial, non-injective observation operators, where the data and the recovered parameters may reside in different spaces. We test our algorithm and show that it yields improved dictionaries for a diverse set of inverse problems in geophysics and medical imaging.

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.

Modification of the nuclear landscape in the inverse problem framework using the generalized Bethe-Weizsäcker mass formula

NASA Astrophysics Data System (ADS)

Mavrodiev, S. Cht.; Deliyergiyev, M. A.

We formalized the nuclear mass problem in the inverse problem framework. This approach allows us to infer the underlying model parameters from experimental observation, rather than to predict the observations from the model parameters. The inverse problem was formulated for the numerically generalized semi-empirical mass formula of Bethe and von Weizsäcker. It was solved in a step-by-step way based on the AME2012 nuclear database. The established parametrization describes the measured nuclear masses of 2564 isotopes with a maximum deviation less than 2.6MeV, starting from the number of protons and number of neutrons equal to 1. The explicit form of unknown functions in the generalized mass formula was discovered in a step-by-step way using the modified least χ2 procedure, that realized in the algorithms which were developed by Lubomir Aleksandrov to solve the nonlinear systems of equations via the Gauss-Newton method, lets us to choose the better one between two functions with same χ2. In the obtained generalized model, the corrections to the binding energy depend on nine proton (2, 8, 14, 20, 28, 50, 82, 108, 124) and ten neutron (2, 8, 14, 20, 28, 50, 82, 124, 152, 202) magic numbers as well on the asymptotic boundaries of their influence. The obtained results were compared with the predictions of other models.

Multimodal, high-dimensional, model-based, Bayesian inverse problems with applications in biomechanics

NASA Astrophysics Data System (ADS)

Franck, I. M.; Koutsourelakis, P. S.

2017-01-01

This paper is concerned with the numerical solution of model-based, Bayesian inverse problems. We are particularly interested in cases where the cost of each likelihood evaluation (forward-model call) is expensive and the number of unknown (latent) variables is high. This is the setting in many problems in computational physics where forward models with nonlinear PDEs are used and the parameters to be calibrated involve spatio-temporarily varying coefficients, which upon discretization give rise to a high-dimensional vector of unknowns. One of the consequences of the well-documented ill-posedness of inverse problems is the possibility of multiple solutions. While such information is contained in the posterior density in Bayesian formulations, the discovery of a single mode, let alone multiple, poses a formidable computational task. The goal of the present paper is two-fold. On one hand, we propose approximate, adaptive inference strategies using mixture densities to capture multi-modal posteriors. On the other, we extend our work in [1] with regard to effective dimensionality reduction techniques that reveal low-dimensional subspaces where the posterior variance is mostly concentrated. We validate the proposed model by employing Importance Sampling which confirms that the bias introduced is small and can be efficiently corrected if the analyst wishes to do so. We demonstrate the performance of the proposed strategy in nonlinear elastography where the identification of the mechanical properties of biological materials can inform non-invasive, medical diagnosis. The discovery of multiple modes (solutions) in such problems is critical in achieving the diagnostic objectives.

Identification of subsurface structures using electromagnetic data and shape priors

NASA Astrophysics Data System (ADS)

Tveit, Svenn; Bakr, Shaaban A.; Lien, Martha; Mannseth, Trond

2015-03-01

We consider the inverse problem of identifying large-scale subsurface structures using the controlled source electromagnetic method. To identify structures in the subsurface where the contrast in electric conductivity can be small, regularization is needed to bias the solution towards preserving structural information. We propose to combine two approaches for regularization of the inverse problem. In the first approach we utilize a model-based, reduced, composite representation of the electric conductivity that is highly flexible, even for a moderate number of degrees of freedom. With a low number of parameters, the inverse problem is efficiently solved using a standard, second-order gradient-based optimization algorithm. Further regularization is obtained using structural prior information, available, e.g., from interpreted seismic data. The reduced conductivity representation is suitable for incorporation of structural prior information. Such prior information cannot, however, be accurately modeled with a gaussian distribution. To alleviate this, we incorporate the structural information using shape priors. The shape prior technique requires the choice of kernel function, which is application dependent. We argue for using the conditionally positive definite kernel which is shown to have computational advantages over the commonly applied gaussian kernel for our problem. Numerical experiments on various test cases show that the methodology is able to identify fairly complex subsurface electric conductivity distributions while preserving structural prior information during the inversion.

Inverse Electrocardiographic Source Localization of Ischemia: An Optimization Framework and Finite Element Solution

PubMed Central

Wang, Dafang; Kirby, Robert M.; MacLeod, Rob S.; Johnson, Chris R.

2013-01-01

With the goal of non-invasively localizing cardiac ischemic disease using body-surface potential recordings, we attempted to reconstruct the transmembrane potential (TMP) throughout the myocardium with the bidomain heart model. The task is an inverse source problem governed by partial differential equations (PDE). Our main contribution is solving the inverse problem within a PDE-constrained optimization framework that enables various physically-based constraints in both equality and inequality forms. We formulated the optimality conditions rigorously in the continuum before deriving finite element discretization, thereby making the optimization independent of discretization choice. Such a formulation was derived for the L2-norm Tikhonov regularization and the total variation minimization. The subsequent numerical optimization was fulfilled by a primal-dual interior-point method tailored to our problem’s specific structure. Our simulations used realistic, fiber-included heart models consisting of up to 18,000 nodes, much finer than any inverse models previously reported. With synthetic ischemia data we localized ischemic regions with roughly a 10% false-negative rate or a 20% false-positive rate under conditions up to 5% input noise. With ischemia data measured from animal experiments, we reconstructed TMPs with roughly 0.9 correlation with the ground truth. While precisely estimating the TMP in general cases remains an open problem, our study shows the feasibility of reconstructing TMP during the ST interval as a means of ischemia localization. PMID:23913980

A modular inverse elastostatics approach to resolve the pressure-induced stress state for in vivo imaging based cardiovascular modeling.

PubMed

Peirlinck, Mathias; De Beule, Matthieu; Segers, Patrick; Rebelo, Nuno

2018-05-28

Patient-specific biomechanical modeling of the cardiovascular system is complicated by the presence of a physiological pressure load given that the imaged tissue is in a pre-stressed and -strained state. Neglect of this prestressed state into solid tissue mechanics models leads to erroneous metrics (e.g. wall deformation, peak stress, wall shear stress) which in their turn are used for device design choices, risk assessment (e.g. procedure, rupture) and surgery planning. It is thus of utmost importance to incorporate this deformed and loaded tissue state into the computational models, which implies solving an inverse problem (calculating an undeformed geometry given the load and the deformed geometry). Methodologies to solve this inverse problem can be categorized into iterative and direct methodologies, both having their inherent advantages and disadvantages. Direct methodologies are typically based on the inverse elastostatics (IE) approach and offer a computationally efficient single shot methodology to compute the in vivo stress state. However, cumbersome and problem-specific derivations of the formulations and non-trivial access to the finite element analysis (FEA) code, especially for commercial products, refrain a broad implementation of these methodologies. For that reason, we developed a novel, modular IE approach and implemented this methodology in a commercial FEA solver with minor user subroutine interventions. The accuracy of this methodology was demonstrated in an arterial tube and porcine biventricular myocardium model. The computational power and efficiency of the methodology was shown by computing the in vivo stress and strain state, and the corresponding unloaded geometry, for two models containing multiple interacting incompressible, anisotropic (fiber-embedded) and hyperelastic material behaviors: a patient-specific abdominal aortic aneurysm and a full 4-chamber heart model. Copyright © 2018 Elsevier Ltd. All rights reserved.

Inversions of the Ledoux discriminant: a closer look at the tachocline

NASA Astrophysics Data System (ADS)

Buldgen, Gaël; Salmon, S. J. A. J.; Godart, M.; Noels, A.; Scuflaire, R.; Dupret, M. A.; Reese, D. R.; Colgan, J.; Fontes, C. J.; Eggenberger, P.; Hakel, P.; Kilcrease, D. P.; Richard, O.

2017-11-01

Modelling the base of the solar convective envelope is a tedious problem. Since the first rotation inversions, solar modellers are confronted with the fact that a region of very limited extent has an enormous physical impact on the Sun. Indeed, it is the transition region from differential to solid body rotation, the tachocline, which furthermore is influenced by turbulence and is also supposed to be the seat of the solar magnetic dynamo. Moreover, solar models show significant disagreement with the sound-speed profile in this region. In this Letter, we show how helioseismology can provide further constraints on this region by carrying out an inversion of the Ledoux discriminant. We compare these inversions for standard solar sodels built using various opacity tables and chemical abundances and discuss the origins of the discrepancies between solar models and the Sun.

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.

Development of a Preventive HIV Vaccine Requires Solving Inverse Problems Which Is Unattainable by Rational Vaccine Design

PubMed Central

Van Regenmortel, Marc H. V.

2018-01-01

Hypotheses and theories are essential constituents of the scientific method. Many vaccinologists are unaware that the problems they try to solve are mostly inverse problems that consist in imagining what could bring about a desired outcome. An inverse problem starts with the result and tries to guess what are the multiple causes that could have produced it. Compared to the usual direct scientific problems that start with the causes and derive or calculate the results using deductive reasoning and known mechanisms, solving an inverse problem uses a less reliable inductive approach and requires the development of a theoretical model that may have different solutions or none at all. Unsuccessful attempts to solve inverse problems in HIV vaccinology by reductionist methods, systems biology and structure-based reverse vaccinology are described. The popular strategy known as rational vaccine design is unable to solve the multiple inverse problems faced by HIV vaccine developers. The term “rational” is derived from “rational drug design” which uses the 3D structure of a biological target for designing molecules that will selectively bind to it and inhibit its biological activity. In vaccine design, however, the word “rational” simply means that the investigator is concentrating on parts of the system for which molecular information is available. The economist and Nobel laureate Herbert Simon introduced the concept of “bounded rationality” to explain why the complexity of the world economic system makes it impossible, for instance, to predict an event like the financial crash of 2007–2008. Humans always operate under unavoidable constraints such as insufficient information, a limited capacity to process huge amounts of data and a limited amount of time available to reach a decision. Such limitations always prevent us from achieving the complete understanding and optimization of a complex system that would be needed to achieve a truly rational design process. This is why the complexity of the human immune system prevents us from rationally designing an HIV vaccine by solving inverse problems. PMID:29387066

Nonlinear inversion of resistivity sounding data for 1-D earth models using the Neighbourhood Algorithm

NASA Astrophysics Data System (ADS)

Ojo, A. O.; Xie, Jun; Olorunfemi, M. O.

2018-01-01

To reduce ambiguity related to nonlinearities in the resistivity model-data relationships, an efficient direct-search scheme employing the Neighbourhood Algorithm (NA) was implemented to solve the 1-D resistivity problem. In addition to finding a range of best-fit models which are more likely to be global minimums, this method investigates the entire multi-dimensional model space and provides additional information about the posterior model covariance matrix, marginal probability density function and an ensemble of acceptable models. This provides new insights into how well the model parameters are constrained and make assessing trade-offs between them possible, thus avoiding some common interpretation pitfalls. The efficacy of the newly developed program is tested by inverting both synthetic (noisy and noise-free) data and field data from other authors employing different inversion methods so as to provide a good base for comparative performance. In all cases, the inverted model parameters were in good agreement with the true and recovered model parameters from other methods and remarkably correlate with the available borehole litho-log and known geology for the field dataset. The NA method has proven to be useful whilst a good starting model is not available and the reduced number of unknowns in the 1-D resistivity inverse problem makes it an attractive alternative to the linearized methods. Hence, it is concluded that the newly developed program offers an excellent complementary tool for the global inversion of the layered resistivity structure.

Query-based learning for aerospace applications.

PubMed

Saad, E W; Choi, J J; Vian, J L; Wunsch, D C Ii

2003-01-01

Models of real-world applications often include a large number of parameters with a wide dynamic range, which contributes to the difficulties of neural network training. Creating the training data set for such applications becomes costly, if not impossible. In order to overcome the challenge, one can employ an active learning technique known as query-based learning (QBL) to add performance-critical data to the training set during the learning phase, thereby efficiently improving the overall learning/generalization. The performance-critical data can be obtained using an inverse mapping called network inversion (discrete network inversion and continuous network inversion) followed by oracle query. This paper investigates the use of both inversion techniques for QBL learning, and introduces an original heuristic to select the inversion target values for continuous network inversion method. Efficiency and generalization was further enhanced by employing node decoupled extended Kalman filter (NDEKF) training and a causality index (CI) as a means to reduce the input search dimensionality. The benefits of the overall QBL approach are experimentally demonstrated in two aerospace applications: a classification problem with large input space and a control distribution problem.

A matched-peak inversion approach for ocean acoustic travel-time tomography

PubMed

Skarsoulis

2000-03-01

A new approach for the inversion of travel-time data is proposed, based on the matching between model arrivals and observed peaks. Using the linearized model relations between sound-speed and arrival-time perturbations about a set of background states, arrival times and associated errors are calculated on a fine grid of model states discretizing the sound-speed parameter space. Each model state can explain (identify) a number of observed peaks in a particular reception lying within the uncertainty intervals of the corresponding predicted arrival times. The model states that explain the maximum number of observed peaks are considered as the more likely parametric descriptions of the reception; these model states can be described in terms of mean values and variances providing a statistical answer (matched-peak solution) to the inversion problem. A basic feature of the matched-peak inversion approach is that each reception can be treated independently, i.e., no constraints are posed from previous-reception identification or inversion results. Accordingly, there is no need for initialization of the inversion procedure and, furthermore, discontinuous travel-time data can be treated. The matched-peak inversion method is demonstrated by application to 9-month-long travel-time data from the Thetis-2 tomography experiment in the western Mediterranean sea.

Escript: Open Source Environment For Solving Large-Scale Geophysical Joint Inversion Problems in Python

NASA Astrophysics Data System (ADS)

Gross, Lutz; Altinay, Cihan; Fenwick, Joel; Smith, Troy

2014-05-01

The program package escript has been designed for solving mathematical modeling problems using python, see Gross et al. (2013). Its development and maintenance has been funded by the Australian Commonwealth to provide open source software infrastructure for the Australian Earth Science community (recent funding by the Australian Geophysical Observing System EIF (AGOS) and the AuScope Collaborative Research Infrastructure Scheme (CRIS)). The key concepts of escript are based on the terminology of spatial functions and partial differential equations (PDEs) - an approach providing abstraction from the underlying spatial discretization method (i.e. the finite element method (FEM)). This feature presents a programming environment to the user which is easy to use even for complex models. Due to the fact that implementations are independent from data structures simulations are easily portable across desktop computers and scalable compute clusters without modifications to the program code. escript has been successfully applied in a variety of applications including modeling mantel convection, melting processes, volcanic flow, earthquakes, faulting, multi-phase flow, block caving and mineralization (see Poulet et al. 2013). The recent escript release (see Gross et al. (2013)) provides an open framework for solving joint inversion problems for geophysical data sets (potential field, seismic and electro-magnetic). The strategy bases on the idea to formulate the inversion problem as an optimization problem with PDE constraints where the cost function is defined by the data defect and the regularization term for the rock properties, see Gross & Kemp (2013). This approach of first-optimize-then-discretize avoids the assemblage of the - in general- dense sensitivity matrix as used in conventional approaches where discrete programming techniques are applied to the discretized problem (first-discretize-then-optimize). In this paper we will discuss the mathematical framework for inversion and appropriate solution schemes in escript. We will also give a brief introduction into escript's open framework for defining and solving geophysical inversion problems. Finally we will show some benchmark results to demonstrate the computational scalability of the inversion method across a large number of cores and compute nodes in a parallel computing environment. References: - L. Gross et al. (2013): Escript Solving Partial Differential Equations in Python Version 3.4, The University of Queensland, https://launchpad.net/escript-finley - 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 - T. Poulet, L. Gross, D. Georgiev, J. Cleverley (2012): escript-RT: Reactive transport simulation in Python using escript, Computers & Geosciences, Volume 45, 168-176. http://dx.doi.org/10.1016/j.cageo.2011.11.005.

A note on convergence of solutions of total variation regularized linear inverse problems

NASA Astrophysics Data System (ADS)

Iglesias, José A.; Mercier, Gwenael; Scherzer, Otmar

2018-05-01

In a recent paper by Chambolle et al (2017 Inverse Problems 33 015002) it was proven that if the subgradient of the total variation at the noise free data is not empty, the level-sets of the total variation denoised solutions converge to the level-sets of the noise free data with respect to the Hausdorff distance. The condition on the subgradient corresponds to the source condition introduced by Burger and Osher (2007 Multiscale Model. Simul. 6 365–95), who proved convergence rates results with respect to the Bregman distance under this condition. We generalize the result of Chambolle et al to total variation regularization of general linear inverse problems under such a source condition. As particular applications we present denoising in bounded and unbounded, convex and non convex domains, deblurring and inversion of the circular Radon transform. In all these examples the convergence result applies. Moreover, we illustrate the convergence behavior through numerical examples.

Simultaneous source and attenuation reconstruction in SPECT using ballistic and single scattering data

NASA Astrophysics Data System (ADS)

Courdurier, M.; Monard, F.; Osses, A.; Romero, F.

2015-09-01

In medical single-photon emission computed tomography (SPECT) imaging, we seek to simultaneously obtain the internal radioactive sources and the attenuation map using not only ballistic measurements but also first-order scattering measurements and assuming a very specific scattering regime. The problem is modeled using the radiative transfer equation by means of an explicit non-linear operator that gives the ballistic and scattering measurements as a function of the radioactive source and attenuation distributions. First, by differentiating this non-linear operator we obtain a linearized inverse problem. Then, under regularity hypothesis for the source distribution and attenuation map and considering small attenuations, we rigorously prove that the linear operator is invertible and we compute its inverse explicitly. This allows proof of local uniqueness for the non-linear inverse problem. Finally, using the previous inversion result for the linear operator, we propose a new type of iterative algorithm for simultaneous source and attenuation recovery for SPECT based on the Neumann series and a Newton-Raphson algorithm.

Inverse dynamics of a 3 degree of freedom spatial flexible manipulator

NASA Technical Reports Server (NTRS)

Bayo, Eduardo; Serna, M.

1989-01-01

A technique is presented for solving the inverse dynamics and kinematics of 3 degree of freedom spatial flexible manipulator. The proposed method finds the joint torques necessary to produce a specified end effector motion. Since the inverse dynamic problem in elastic manipulators is closely coupled to the inverse kinematic problem, the solution of the first also renders the displacements and rotations at any point of the manipulator, including the joints. Furthermore the formulation is complete in the sense that it includes all the nonlinear terms due to the large rotation of the links. The Timoshenko beam theory is used to model the elastic characteristics, and the resulting equations of motion are discretized using the finite element method. An iterative solution scheme is proposed that relies on local linearization of the problem. The solution of each linearization is carried out in the frequency domain. The performance and capabilities of this technique are tested through simulation analysis. Results show the potential use of this method for the smooth motion control of space telerobots.

Bilinear Inverse Problems: Theory, Algorithms, and Applications

NASA Astrophysics Data System (ADS)

Ling, Shuyang

We will discuss how several important real-world signal processing problems, such as self-calibration and blind deconvolution, can be modeled as bilinear inverse problems and solved by convex and nonconvex optimization approaches. In Chapter 2, we bring together three seemingly unrelated concepts, self-calibration, compressive sensing and biconvex optimization. We show how several self-calibration problems can be treated efficiently within the framework of biconvex compressive sensing via a new method called SparseLift. More specifically, we consider a linear system of equations y = DAx, where the diagonal matrix D (which models the calibration error) is unknown and x is an unknown sparse signal. By "lifting" this biconvex inverse problem and exploiting sparsity in this model, we derive explicit theoretical guarantees under which both x and D can be recovered exactly, robustly, and numerically efficiently. In Chapter 3, we study the question of the joint blind deconvolution and blind demixing, i.e., extracting a sequence of functions [special characters omitted] from observing only the sum of their convolutions [special characters omitted]. In particular, for the special case s = 1, it becomes the well-known blind deconvolution problem. We present a non-convex algorithm which guarantees exact recovery under conditions that are competitive with convex optimization methods, with the additional advantage of being computationally much more efficient. We discuss several applications of the proposed framework in image processing and wireless communications in connection with the Internet-of-Things. In Chapter 4, we consider three different self-calibration models of practical relevance. We show how their corresponding bilinear inverse problems can be solved by both the simple linear least squares approach and the SVD-based approach. As a consequence, the proposed algorithms are numerically extremely efficient, thus allowing for real-time deployment. Explicit theoretical guarantees and stability theory are derived and the number of sampling complexity is nearly optimal (up to a poly-log factor). Applications in imaging sciences and signal processing are discussed and numerical simulations are presented to demonstrate the effectiveness and efficiency of our approach.

Modelling night-time ecosystem respiration by a constrained source optimization method

Treesearch

Chun-Tai Lai; Gabriel Katul; John Butnor; David Ellsworth; Ram Oren

2002-01-01

One of the main challenges to quantifying ecosystem carbon budgets is properly quantifying the magnitude of night-time ecosystem respiration. Inverse Lagrangian dispersion analysis provides a promising approach to addressing such a problem when measured mean CO2 concentration profiles and nocturnal velocity statistics are available. An inverse...

Data-resolution matrix and model-resolution matrix for Rayleigh-wave inversion using a damped least-squares method

USGS Publications Warehouse

Xia, J.; Miller, R.D.; Xu, Y.

2008-01-01

Inversion of multimode surface-wave data is of increasing interest in the near-surface geophysics community. For a given near-surface geophysical problem, it is essential to understand how well the data, calculated according to a layered-earth model, might match the observed data. A data-resolution matrix is a function of the data kernel (determined by a geophysical model and a priori information applied to the problem), not the data. A data-resolution matrix of high-frequency (>2 Hz) Rayleigh-wave phase velocities, therefore, offers a quantitative tool for designing field surveys and predicting the match between calculated and observed data. We employed a data-resolution matrix to select data that would be well predicted and we find that there are advantages of incorporating higher modes in inversion. The resulting discussion using the data-resolution matrix provides insight into the process of inverting Rayleigh-wave phase velocities with higher-mode data to estimate S-wave velocity structure. Discussion also suggested that each near-surface geophysical target can only be resolved using Rayleigh-wave phase velocities within specific frequency ranges, and higher-mode data are normally more accurately predicted than fundamental-mode data because of restrictions on the data kernel for the inversion system. We used synthetic and real-world examples to demonstrate that selected data with the data-resolution matrix can provide better inversion results and to explain with the data-resolution matrix why incorporating higher-mode data in inversion can provide better results. We also calculated model-resolution matrices in these examples to show the potential of increasing model resolution with selected surface-wave data. ?? Birkhaueser 2008.

A Solution Framework for Environmental Characterization Problems

EPA Science Inventory

This paper describes experiences developing a grid-enabled framework for solving environmental inverse problems. The solution approach taken here couples environmental simulation models with global search methods and requires readily available computational resources of the grid ...

Numerical methods for the inverse problem of density functional theory

DOE PAGES

Jensen, Daniel S.; Wasserman, Adam

2017-07-17

Here, the inverse problem of Kohn–Sham density functional theory (DFT) is often solved in an effort to benchmark and design approximate exchange-correlation potentials. The forward and inverse problems of DFT rely on the same equations but the numerical methods for solving each problem are substantially different. We examine both problems in this tutorial with a special emphasis on the algorithms and error analysis needed for solving the inverse problem. Two inversion methods based on partial differential equation constrained optimization and constrained variational ideas are introduced. We compare and contrast several different inversion methods applied to one-dimensional finite and periodic modelmore » systems.« less

Numerical methods for the inverse problem of density functional theory

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

Jensen, Daniel S.; Wasserman, Adam

Here, the inverse problem of Kohn–Sham density functional theory (DFT) is often solved in an effort to benchmark and design approximate exchange-correlation potentials. The forward and inverse problems of DFT rely on the same equations but the numerical methods for solving each problem are substantially different. We examine both problems in this tutorial with a special emphasis on the algorithms and error analysis needed for solving the inverse problem. Two inversion methods based on partial differential equation constrained optimization and constrained variational ideas are introduced. We compare and contrast several different inversion methods applied to one-dimensional finite and periodic modelmore » systems.« less

Computing rates of Markov models of voltage-gated ion channels by inverting partial differential equations governing the probability density functions of the conducting and non-conducting states.

PubMed

Tveito, Aslak; Lines, Glenn T; Edwards, Andrew G; McCulloch, Andrew

2016-07-01

Markov models are ubiquitously used to represent the function of single ion channels. However, solving the inverse problem to construct a Markov model of single channel dynamics from bilayer or patch-clamp recordings remains challenging, particularly for channels involving complex gating processes. Methods for solving the inverse problem are generally based on data from voltage clamp measurements. Here, we describe an alternative approach to this problem based on measurements of voltage traces. The voltage traces define probability density functions of the functional states of an ion channel. These probability density functions can also be computed by solving a deterministic system of partial differential equations. The inversion is based on tuning the rates of the Markov models used in the deterministic system of partial differential equations such that the solution mimics the properties of the probability density function gathered from (pseudo) experimental data as well as possible. The optimization is done by defining a cost function to measure the difference between the deterministic solution and the solution based on experimental data. By evoking the properties of this function, it is possible to infer whether the rates of the Markov model are identifiable by our method. We present applications to Markov model well-known from the literature. Copyright © 2016 The Authors. Published by Elsevier Inc. All rights reserved.

Information fusion in regularized inversion of tomographic pumping tests

USGS Publications Warehouse

Bohling, Geoffrey C.; ,

2008-01-01

In this chapter we investigate a simple approach to incorporating geophysical information into the analysis of tomographic pumping tests for characterization of the hydraulic conductivity (K) field in an aquifer. A number of authors have suggested a tomographic approach to the analysis of hydraulic tests in aquifers - essentially simultaneous analysis of multiple tests or stresses on the flow system - in order to improve the resolution of the estimated parameter fields. However, even with a large amount of hydraulic data in hand, the inverse problem is still plagued by non-uniqueness and ill-conditioning and the parameter space for the inversion needs to be constrained in some sensible fashion in order to obtain plausible estimates of aquifer properties. For seismic and radar tomography problems, the parameter space is often constrained through the application of regularization terms that impose penalties on deviations of the estimated parameters from a prior or background model, with the tradeoff between data fit and model norm explored through systematic analysis of results for different levels of weighting on the regularization terms. In this study we apply systematic regularized inversion to analysis of tomographic pumping tests in an alluvial aquifer, taking advantage of the steady-shape flow regime exhibited in these tests to expedite the inversion process. In addition, we explore the possibility of incorporating geophysical information into the inversion through a regularization term relating the estimated K distribution to ground penetrating radar velocity and attenuation distributions through a smoothing spline model. ?? 2008 Springer-Verlag Berlin Heidelberg.

On uncertainty quantification in hydrogeology and hydrogeophysics

NASA Astrophysics Data System (ADS)

Linde, Niklas; Ginsbourger, David; Irving, James; Nobile, Fabio; Doucet, Arnaud

2017-12-01

Recent advances in sensor technologies, field methodologies, numerical modeling, and inversion approaches have contributed to unprecedented imaging of hydrogeological properties and detailed predictions at multiple temporal and spatial scales. Nevertheless, imaging results and predictions will always remain imprecise, which calls for appropriate uncertainty quantification (UQ). In this paper, we outline selected methodological developments together with pioneering UQ applications in hydrogeology and hydrogeophysics. The applied mathematics and statistics literature is not easy to penetrate and this review aims at helping hydrogeologists and hydrogeophysicists to identify suitable approaches for UQ that can be applied and further developed to their specific needs. To bypass the tremendous computational costs associated with forward UQ based on full-physics simulations, we discuss proxy-modeling strategies and multi-resolution (Multi-level Monte Carlo) methods. We consider Bayesian inversion for non-linear and non-Gaussian state-space problems and discuss how Sequential Monte Carlo may become a practical alternative. We also describe strategies to account for forward modeling errors in Bayesian inversion. Finally, we consider hydrogeophysical inversion, where petrophysical uncertainty is often ignored leading to overconfident parameter estimation. The high parameter and data dimensions encountered in hydrogeological and geophysical problems make UQ a complicated and important challenge that has only been partially addressed to date.

The Source Inversion Validation (SIV) Initiative: A Collaborative Study on Uncertainty Quantification in Earthquake Source Inversions

NASA Astrophysics Data System (ADS)

Mai, P. M.; Schorlemmer, D.; Page, M.

2012-04-01

Earthquake source inversions image the spatio-temporal rupture evolution on one or more fault planes using seismic and/or geodetic data. Such studies are critically important for earthquake seismology in general, and for advancing seismic hazard analysis in particular, as they reveal earthquake source complexity and help (i) to investigate earthquake mechanics; (ii) to develop spontaneous dynamic rupture models; (iii) to build models for generating rupture realizations for ground-motion simulations. In applications (i - iii), the underlying finite-fault source models are regarded as "data" (input information), but their uncertainties are essentially unknown. After all, source models are obtained from solving an inherently ill-posed inverse problem to which many a priori assumptions and uncertain observations are applied. The Source Inversion Validation (SIV) project is a collaborative effort to better understand the variability between rupture models for a single earthquake (as manifested in the finite-source rupture model database) and to develop robust uncertainty quantification for earthquake source inversions. The SIV project highlights the need to develop a long-standing and rigorous testing platform to examine the current state-of-the-art in earthquake source inversion, and to develop and test novel source inversion approaches. We will review the current status of the SIV project, and report the findings and conclusions of the recent workshops. We will briefly discuss several source-inversion methods, how they treat uncertainties in data, and assess the posterior model uncertainty. Case studies include initial forward-modeling tests on Green's function calculations, and inversion results for synthetic data from spontaneous dynamic crack-like strike-slip earthquake on steeply dipping fault, embedded in a layered crustal velocity-density structure.

Forward and inverse models of electromagnetic scattering from layered media with rough interfaces

NASA Astrophysics Data System (ADS)

Tabatabaeenejad, Seyed Alireza

This work addresses the problem of electromagnetic scattering from layered dielectric structures with rough boundaries and the associated inverse problem of retrieving the subsurface parameters of the structure using the scattered field. To this end, a forward scattering model based on the Small Perturbation Method (SPM) is developed to calculate the first-order spectral-domain bistatic scattering coefficients of a two-layer rough surface structure. SPM requires the boundaries to be slightly rough compared to the wavelength, but to understand the range of applicability of this method in scattering from two-layer rough surfaces, its region of validity is investigated by comparing its output with that of a first principle solver that does not impose roughness restrictions. The Method of Moments (MoM) is used for this purpose. Finally, for retrieval of the model parameters of the layered structure using scattered field, an inversion scheme based on the Simulated Annealing method is investigated and a strategy is proposed to address convergence to local minimum.

Efficient realization of 3D joint inversion of seismic and magnetotelluric data with cross gradient structure constraint

NASA Astrophysics Data System (ADS)

Luo, H.; Zhang, H.; Gao, J.

2016-12-01

Seismic and magnetotelluric (MT) imaging methods are generally used to characterize subsurface structures at various scales. The two methods are complementary to each other and the integration of them is helpful for more reliably determining the resistivity and velocity models of the target region. Because of the difficulty in finding empirical relationship between resistivity and velocity parameters, Gallardo and Meju [2003] proposed a joint inversion method enforcing resistivity and velocity models consistent in structure, which is realized by minimizing cross gradients between two models. However, it is extremely challenging to combine two different inversion systems together along with the cross gradient constraints. For this reason, Gallardo [2007] proposed a joint inversion scheme that decouples the seismic and MT inversion systems by iteratively performing seismic and MT inversions as well as cross gradient minimization separately. This scheme avoids the complexity of combining two different systems together but it suffers the issue of balancing between data fitting and structure constraint. In this study, we have developed a new joint inversion scheme that avoids the problem encountered by the scheme of Gallardo [2007]. In the new scheme, seismic and MT inversions are still separately performed but the cross gradient minimization is also constrained by model perturbations from separate inversions. In this way, the new scheme still avoids the complexity of combining two different systems together and at the same time the balance between data fitting and structure consistency constraint can be enforced. We have tested our joint inversion algorithm for both 2D and 3D cases. Synthetic tests show that joint inversion better reconstructed the velocity and resistivity models than separate inversions. Compared to separate inversions, joint inversion can remove artifacts in the resistivity model and can improve the resolution for deeper resistivity structures. We will also show results applying the new joint seismic and MT inversion scheme to southwest China, where several MT profiles are available and earthquakes are very active.

Three-dimensional Gravity Inversion with a New Gradient Scheme on Unstructured Grids

NASA Astrophysics Data System (ADS)

Sun, S.; Yin, C.; Gao, X.; Liu, Y.; Zhang, B.

2017-12-01

Stabilized gradient-based methods have been proved to be efficient for inverse problems. Based on these methods, setting gradient close to zero can effectively minimize the objective function. Thus the gradient of objective function determines the inversion results. By analyzing the cause of poor resolution on depth in gradient-based gravity inversion methods, we find that imposing depth weighting functional in conventional gradient can improve the depth resolution to some extent. However, the improvement is affected by the regularization parameter and the effect of the regularization term becomes smaller with increasing depth (shown as Figure 1 (a)). In this paper, we propose a new gradient scheme for gravity inversion by introducing a weighted model vector. The new gradient can improve the depth resolution more efficiently, which is independent of the regularization parameter, and the effect of regularization term will not be weakened when depth increases. Besides, fuzzy c-means clustering method and smooth operator are both used as regularization terms to yield an internal consecutive inverse model with sharp boundaries (Sun and Li, 2015). We have tested our new gradient scheme with unstructured grids on synthetic data to illustrate the effectiveness of the algorithm. Gravity forward modeling with unstructured grids is based on the algorithm proposed by Okbe (1979). We use a linear conjugate gradient inversion scheme to solve the inversion problem. The numerical experiments show a great improvement in depth resolution compared with regular gradient scheme, and the inverse model is compact at all depths (shown as Figure 1 (b)). AcknowledgeThis research is supported by Key Program of National Natural Science Foundation of China (41530320), China Natural Science Foundation for Young Scientists (41404093), and Key National Research Project of China (2016YFC0303100, 2017YFC0601900). ReferencesSun J, Li Y. 2015. Multidomain petrophysically constrained inversion and geology differentiation using guided fuzzy c-means clustering. Geophysics, 80(4): ID1-ID18. Okabe M. 1979. Analytical expressions for gravity anomalies due to homogeneous polyhedral bodies and translations into magnetic anomalies. Geophysics, 44(4), 730-741.

Adjoint-Based Sensitivity Kernels for Glacial Isostatic Adjustment in a Laterally Varying Earth

NASA Astrophysics Data System (ADS)

Crawford, O.; Al-Attar, D.; Tromp, J.; Mitrovica, J. X.; Austermann, J.; Lau, H. C. P.

2017-12-01

We consider a new approach to both the forward and inverse problems in glacial isostatic adjustment. We present a method for forward modelling GIA in compressible and laterally heterogeneous earth models with a variety of linear and non-linear rheologies. Instead of using the so-called sea level equation, which must be solved iteratively, the forward theory we present consists of a number of coupled evolution equations that can be straightforwardly numerically integrated. We also apply the adjoint method to the inverse problem in order to calculate the derivatives of measurements of GIA with respect to the viscosity structure of the Earth. Such derivatives quantify the sensitivity of the measurements to the model. The adjoint method enables efficient calculation of continuous and laterally varying derivatives, allowing us to calculate the sensitivity of measurements of glacial isostatic adjustment to the Earth's three-dimensional viscosity structure. The derivatives have a number of applications within the inverse method. Firstly, they can be used within a gradient-based optimisation method to find a model which minimises some data misfit function. The derivatives can also be used to quantify the uncertainty in such a model and hence to provide understanding of which parts of the model are well constrained. Finally, they enable construction of measurements which provide sensitivity to a particular part of the model space. We illustrate both the forward and inverse aspects with numerical examples in a spherically symmetric earth model.

Chemical Source Inversion using Assimilated Constituent Observations in an Idealized Two-dimensional System

NASA Technical Reports Server (NTRS)

Tangborn, Andrew; Cooper, Robert; Pawson, Steven; Sun, Zhibin

2009-01-01

We present a source inversion technique for chemical constituents that uses assimilated constituent observations rather than directly using the observations. The method is tested with a simple model problem, which is a two-dimensional Fourier-Galerkin transport model combined with a Kalman filter for data assimilation. Inversion is carried out using a Green's function method and observations are simulated from a true state with added Gaussian noise. The forecast state uses the same spectral spectral model, but differs by an unbiased Gaussian model error, and emissions models with constant errors. The numerical experiments employ both simulated in situ and satellite observation networks. Source inversion was carried out by either direct use of synthetically generated observations with added noise, or by first assimilating the observations and using the analyses to extract observations. We have conducted 20 identical twin experiments for each set of source and observation configurations, and find that in the limiting cases of a very few localized observations, or an extremely large observation network there is little advantage to carrying out assimilation first. However, in intermediate observation densities, there decreases in source inversion error standard deviation using the Kalman filter algorithm followed by Green's function inversion by 50% to 95%.

An inverse model for a free-boundary problem with a contact line: Steady case

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

Volkov, Oleg; Protas, Bartosz

2009-07-20

This paper reformulates the two-phase solidification problem (i.e., the Stefan problem) as an inverse problem in which a cost functional is minimized with respect to the position of the interface and subject to PDE constraints. An advantage of this formulation is that it allows for a thermodynamically consistent treatment of the interface conditions in the presence of a contact point involving a third phase. It is argued that such an approach in fact represents a closure model for the original system and some of its key properties are investigated. We describe an efficient iterative solution method for the Stefan problemmore » formulated in this way which uses shape differentiation and adjoint equations to determine the gradient of the cost functional. Performance of the proposed approach is illustrated with sample computations concerning 2D steady solidification phenomena.« less

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.

On determining important aspects of mathematical models: Application to problems in physics and chemistry

NASA Technical Reports Server (NTRS)

Rabitz, Herschel

1987-01-01

The use of parametric and functional gradient sensitivity analysis techniques is considered for models described by partial differential equations. By interchanging appropriate dependent and independent variables, questions of inverse sensitivity may be addressed to gain insight into the inversion of observational data for parameter and function identification in mathematical models. It may be argued that the presence of a subset of dominantly strong coupled dependent variables will result in the overall system sensitivity behavior collapsing into a simple set of scaling and self similarity relations amongst elements of the entire matrix of sensitivity coefficients. These general tools are generic in nature, but herein their application to problems arising in selected areas of physics and chemistry is presented.

Dynamic data integration and stochastic inversion of a confined aquifer

NASA Astrophysics Data System (ADS)

Wang, D.; Zhang, Y.; Irsa, J.; Huang, H.; Wang, L.

2013-12-01

Much work has been done in developing and applying inverse methods to aquifer modeling. The scope of this paper is to investigate the applicability of a new direct method for large inversion problems and to incorporate uncertainty measures in the inversion outcomes (Wang et al., 2013). The problem considered is a two-dimensional inverse model (50×50 grid) of steady-state flow for a heterogeneous ground truth model (500×500 grid) with two hydrofacies. From the ground truth model, decreasing number of wells (12, 6, 3) were sampled for facies types, based on which experimental indicator histograms and directional variograms were computed. These parameters and models were used by Sequential Indicator Simulation to generate 100 realizations of hydrofacies patterns in a 100×100 (geostatistical) grid, which were conditioned to the facies measurements at wells. These realizations were smoothed with Simulated Annealing, coarsened to the 50×50 inverse grid, before they were conditioned with the direct method to the dynamic data, i.e., observed heads and groundwater fluxes at the same sampled wells. A set of realizations of estimated hydraulic conductivities (Ks), flow fields, and boundary conditions were created, which centered on the 'true' solutions from solving the ground truth model. Both hydrofacies conductivities were computed with an estimation accuracy of ×10% (12 wells), ×20% (6 wells), ×35% (3 wells) of the true values. For boundary condition estimation, the accuracy was within × 15% (12 wells), 30% (6 wells), and 50% (3 wells) of the true values. The inversion system of equations was solved with LSQR (Paige et al, 1982), for which coordinate transform and matrix scaling preprocessor were used to improve the condition number (CN) of the coefficient matrix. However, when the inverse grid was refined to 100×100, Gaussian Noise Perturbation was used to limit the growth of the CN before the matrix solve. To scale the inverse problem up (i.e., without smoothing and coarsening and therefore reducing the associated estimation uncertainty), a parallel LSQR solver was written and verified. For the 50×50 grid, the parallel solver sped up the serial solution time by 14X using 4 CPUs (research on parallel performance and scaling is ongoing). A sensitivity analysis was conducted to examine the relation between the observed data and the inversion outcomes, where measurement errors of increasing magnitudes (i.e., ×1, 2, 5, 10% of the total head variation and up to ×2% of the total flux variation) were imposed on the observed data. Inversion results were stable but the accuracy of Ks and boundary estimation degraded with increasing errors, as expected. In particular, quality of the observed heads is critical to hydraulic head recovery, while quality of the observed fluxes plays a dominant role in K estimation. References: Wang, D., Y. Zhang, J. Irsa, H. Huang, and L. Wang (2013), Data integration and stochastic inversion of a confined aquifer with high performance computing, Advances in Water Resources, in preparation. Paige, C. C., and M. A. Saunders (1982), LSQR: an algorithm for sparse linear equations and sparse least squares, ACM Transactions on Mathematical Software, 8(1), 43-71.

Extended fault inversion with random slipmaps: A resolution test for the 2012 Mw 7.6 Nicoya, Costa Rica earthquake from a Popperian inversion strategy.

NASA Astrophysics Data System (ADS)

Ángel López Comino, José; Stich, Daniel; Ferreira, Ana M. G.; Morales Soto, José

2015-04-01

The inversion of seismic data for extended fault slip distributions provides us detailed models of earthquake sources. The validity of the solutions depends on the fit between observed and synthetic seismograms generated with the source model. However, there may exist more than one model that fit the data in a similar way, leading to a multiplicity of solutions. This underdetermined problem has been analyzed and studied by several authors, who agree that inverting for a single best model may become overly dependent on the details of the procedure. We have addressed this resolution problem by using a global search that scans the solutions domain using random slipmaps, applying a Popperian inversion strategy that involves the generation of a representative set of slip distributions. The proposed technique solves the forward problem for a large set of models calculating their corresponding synthetic seismograms. Then, we propose to perform extended fault inversion through falsification, that is, falsify inappropriate trial models that do not reproduce the data within a reasonable level of mismodelling. The remainder of surviving trial models forms our set of coequal solutions. Thereby the ambiguities that might exist can be detected by taking a look at the solutions, allowing for an efficient assessment of the resolution. The solution set may contain only members with similar slip distributions, or else uncover some fundamental ambiguity like, for example, different patterns of main slip patches or different patterns of rupture propagation. For a feasibility study, the proposed resolution test has been evaluated using teleseismic body wave recordings from the September 5th 2012 Nicoya, Costa Rica earthquake. Note that the inversion strategy can be applied to any type of seismic, geodetic or tsunami data for which we can handle the forward problem. A 2D von Karman distribution is used to describe the spectrum of heterogeneity in slipmaps, and we generate possible models by spectral synthesis for random phase, keeping the rake angle, rupture velocity and slip velocity function fixed. The 2012 Nicoya earthquake turns out to be relatively well constrained from 50 teleseismic waveforms. The solution set contains 252 out of 10.000 trial models with normalized L1-fit within 5 percent from the global minimum. The set includes only similar solutions -a single centred slip patch- with minor differences. Uncertainties are related to the details of the slip maximum, including the amount of peak slip (2m to 3.5m), as well as the characteristics of peripheral slip below 1 m. Synthetic tests suggest that slip patterns like Nicoya may be a fortunate case, while it may be more difficult to unambiguously reconstruct more distributed slip from teleseismic data.

Feasibility of waveform inversion of Rayleigh waves for shallow shear-wave velocity using a genetic algorithm

USGS Publications Warehouse

Zeng, C.; Xia, J.; Miller, R.D.; Tsoflias, G.P.

2011-01-01

Conventional surface wave inversion for shallow shear (S)-wave velocity relies on the generation of dispersion curves of Rayleigh waves. This constrains the method to only laterally homogeneous (or very smooth laterally heterogeneous) earth models. Waveform inversion directly fits waveforms on seismograms, hence, does not have such a limitation. Waveforms of Rayleigh waves are highly related to S-wave velocities. By inverting the waveforms of Rayleigh waves on a near-surface seismogram, shallow S-wave velocities can be estimated for earth models with strong lateral heterogeneity. We employ genetic algorithm (GA) to perform waveform inversion of Rayleigh waves for S-wave velocities. The forward problem is solved by finite-difference modeling in the time domain. The model space is updated by generating offspring models using GA. Final solutions can be found through an iterative waveform-fitting scheme. Inversions based on synthetic records show that the S-wave velocities can be recovered successfully with errors no more than 10% for several typical near-surface earth models. For layered earth models, the proposed method can generate one-dimensional S-wave velocity profiles without the knowledge of initial models. For earth models containing lateral heterogeneity in which case conventional dispersion-curve-based inversion methods are challenging, it is feasible to produce high-resolution S-wave velocity sections by GA waveform inversion with appropriate priori information. The synthetic tests indicate that the GA waveform inversion of Rayleigh waves has the great potential for shallow S-wave velocity imaging with the existence of strong lateral heterogeneity. ?? 2011 Elsevier B.V.

The novel high-performance 3-D MT inverse solver

NASA Astrophysics Data System (ADS)

Kruglyakov, Mikhail; Geraskin, Alexey; Kuvshinov, Alexey

2016-04-01

We present novel, robust, scalable, and fast 3-D magnetotelluric (MT) inverse solver. The solver is written in multi-language paradigm to make it as efficient, readable and maintainable as possible. Separation of concerns and single responsibility concepts go through implementation of the solver. As a forward modelling engine a modern scalable solver extrEMe, based on contracting integral equation approach, is used. Iterative gradient-type (quasi-Newton) optimization scheme is invoked to search for (regularized) inverse problem solution, and adjoint source approach is used to calculate efficiently the gradient of the misfit. The inverse solver is able to deal with highly detailed and contrasting models, allows for working (separately or jointly) with any type of MT responses, and supports massive parallelization. Moreover, different parallelization strategies implemented in the code allow optimal usage of available computational resources for a given problem statement. To parameterize an inverse domain the so-called mask parameterization is implemented, which means that one can merge any subset of forward modelling cells in order to account for (usually) irregular distribution of observation sites. We report results of 3-D numerical experiments aimed at analysing the robustness, performance and scalability of the code. In particular, our computational experiments carried out at different platforms ranging from modern laptops to HPC Piz Daint (6th supercomputer in the world) demonstrate practically linear scalability of the code up to thousands of nodes.

Inverse solutions for electrical impedance tomography based on conjugate gradients methods

NASA Astrophysics Data System (ADS)

Wang, M.

2002-01-01

A multistep inverse solution for two-dimensional electric field distribution is developed to deal with the nonlinear inverse problem of electric field distribution in relation to its boundary condition and the problem of divergence due to errors introduced by the ill-conditioned sensitivity matrix and the noise produced by electrode modelling and instruments. This solution is based on a normalized linear approximation method where the change in mutual impedance is derived from the sensitivity theorem and a method of error vector decomposition. This paper presents an algebraic solution of the linear equations at each inverse step, using a generalized conjugate gradients method. Limiting the number of iterations in the generalized conjugate gradients method controls the artificial errors introduced by the assumption of linearity and the ill-conditioned sensitivity matrix. The solution of the nonlinear problem is approached using a multistep inversion. This paper also reviews the mathematical and physical definitions of the sensitivity back-projection algorithm based on the sensitivity theorem. Simulations and discussion based on the multistep algorithm, the sensitivity coefficient back-projection method and the Newton-Raphson method are given. Examples of imaging gas-liquid mixing and a human hand in brine are presented.

Confidence set inference with a prior quadratic bound

NASA Technical Reports Server (NTRS)

Backus, George E.

1989-01-01

In the uniqueness part of a geophysical inverse problem, the observer wants to predict all likely values of P unknown numerical properties z=(z sub 1,...,z sub p) of the earth from measurement of D other numerical properties y (sup 0) = (y (sub 1) (sup 0), ..., y (sub D (sup 0)), using full or partial knowledge of the statistical distribution of the random errors in y (sup 0). The data space Y containing y(sup 0) is D-dimensional, so when the model space X is infinite-dimensional the linear uniqueness problem usually is insoluble without prior information about the correct earth model x. If that information is a quadratic bound on x, Bayesian inference (BI) and stochastic inversion (SI) inject spurious structure into x, implied by neither the data nor the quadratic bound. Confidence set inference (CSI) provides an alternative inversion technique free of this objection. Confidence set inference is illustrated in the problem of estimating the geomagnetic field B at the core-mantle boundary (CMB) from components of B measured on or above the earth's surface.

Feasibility of inverse problem solution for determination of city emission function from night sky radiance measurements

NASA Astrophysics Data System (ADS)

Petržala, Jaromír

2018-07-01

The knowledge of the emission function of a city is crucial for simulation of sky glow in its vicinity. The indirect methods to achieve this function from radiances measured over a part of the sky have been recently developed. In principle, such methods represent an ill-posed inverse problem. This paper deals with the theoretical feasibility study of various approaches to solving of given inverse problem. Particularly, it means testing of fitness of various stabilizing functionals within the Tikhonov's regularization. Further, the L-curve and generalized cross validation methods were investigated as indicators of an optimal regularization parameter. At first, we created the theoretical model for calculation of a sky spectral radiance in the form of a functional of an emission spectral radiance. Consequently, all the mentioned approaches were examined in numerical experiments with synthetical data generated for the fictitious city and loaded by random errors. The results demonstrate that the second order Tikhonov's regularization method together with regularization parameter choice by the L-curve maximum curvature criterion provide solutions which are in good agreement with the supposed model emission functions.

Inverse regression-based uncertainty quantification algorithms for high-dimensional models: Theory and practice

NASA Astrophysics Data System (ADS)

Li, Weixuan; Lin, Guang; Li, Bing

2016-09-01

Many uncertainty quantification (UQ) approaches suffer from the curse of dimensionality, that is, their computational costs become intractable for problems involving a large number of uncertainty parameters. In these situations, the classic Monte Carlo often remains the preferred method of choice because its convergence rate O (n - 1 / 2), where n is the required number of model simulations, does not depend on the dimension of the problem. However, many high-dimensional UQ problems are intrinsically low-dimensional, because the variation of the quantity of interest (QoI) is often caused by only a few latent parameters varying within a low-dimensional subspace, known as the sufficient dimension reduction (SDR) subspace in the statistics literature. Motivated by this observation, we propose two inverse regression-based UQ algorithms (IRUQ) for high-dimensional problems. Both algorithms use inverse regression to convert the original high-dimensional problem to a low-dimensional one, which is then efficiently solved by building a response surface for the reduced model, for example via the polynomial chaos expansion. The first algorithm, which is for the situations where an exact SDR subspace exists, is proved to converge at rate O (n-1), hence much faster than MC. The second algorithm, which doesn't require an exact SDR, employs the reduced model as a control variate to reduce the error of the MC estimate. The accuracy gain could still be significant, depending on how well the reduced model approximates the original high-dimensional one. IRUQ also provides several additional practical advantages: it is non-intrusive; it does not require computing the high-dimensional gradient of the QoI; and it reports an error bar so the user knows how reliable the result is.

Feedback control by online learning an inverse model.

PubMed

Waegeman, Tim; Wyffels, Francis; Schrauwen, Francis

2012-10-01

A model, predictor, or error estimator is often used by a feedback controller to control a plant. Creating such a model is difficult when the plant exhibits nonlinear behavior. In this paper, a novel online learning control framework is proposed that does not require explicit knowledge about the plant. This framework uses two learning modules, one for creating an inverse model, and the other for actually controlling the plant. Except for their inputs, they are identical. The inverse model learns by the exploration performed by the not yet fully trained controller, while the actual controller is based on the currently learned model. The proposed framework allows fast online learning of an accurate controller. The controller can be applied on a broad range of tasks with different dynamic characteristics. We validate this claim by applying our control framework on several control tasks: 1) the heating tank problem (slow nonlinear dynamics); 2) flight pitch control (slow linear dynamics); and 3) the balancing problem of a double inverted pendulum (fast linear and nonlinear dynamics). The results of these experiments show that fast learning and accurate control can be achieved. Furthermore, a comparison is made with some classical control approaches, and observations concerning convergence and stability are made.

Improved resistivity imaging of groundwater solute plumes using POD-based inversion

NASA Astrophysics Data System (ADS)

Oware, E. K.; Moysey, S. M.; Khan, T.

2012-12-01

We propose a new approach for enforcing physics-based regularization in electrical resistivity imaging (ERI) problems. The approach utilizes a basis-constrained inversion where an optimal set of basis vectors is extracted from training data by Proper Orthogonal Decomposition (POD). The key aspect of the approach is that Monte Carlo simulation of flow and transport is used to generate a training dataset, thereby intrinsically capturing the physics of the underlying flow and transport models in a non-parametric form. POD allows for these training data to be projected onto a subspace of the original domain, resulting in the extraction of a basis for the inversion that captures characteristics of the groundwater flow and transport system, while simultaneously allowing for dimensionality reduction of the original problem in the projected space We use two different synthetic transport scenarios in heterogeneous media to illustrate how the POD-based inversion compares with standard Tikhonov and coupled inversion. The first scenario had a single source zone leading to a unimodal solute plume (synthetic #1), whereas, the second scenario had two source zones that produced a bimodal plume (synthetic #2). For both coupled inversion and the POD approach, the conceptual flow and transport model used considered only a single source zone for both scenarios. Results were compared based on multiple metrics (concentration root-mean square error (RMSE), peak concentration, and total solute mass). In addition, results for POD inversion based on 3 different data densities (120, 300, and 560 data points) and varying number of selected basis images (100, 300, and 500) were compared. For synthetic #1, we found that all three methods provided qualitatively reasonable reproduction of the true plume. Quantitatively, the POD inversion performed best overall for each metric considered. Moreover, since synthetic #1 was consistent with the conceptual transport model, a small number of basis vectors (100) contained enough a priori information to constrain the inversion. Increasing the amount of data or number of selected basis images did not translate into significant improvement in imaging results. For synthetic #2, the RMSE and error in total mass were lowest for the POD inversion. However, the peak concentration was significantly overestimated by the POD approach. Regardless, the POD-based inversion was the only technique that could capture the bimodality of the plume in the reconstructed image, thus providing critical information that could be used to reconceptualize the transport problem. We also found that, in the case of synthetic #2, increasing the number of resistivity measurements and the number of selected basis vectors allowed for significant improvements in the reconstructed images.

Spectral-element simulations of wave propagation in complex exploration-industry models: Imaging and adjoint tomography

NASA Astrophysics Data System (ADS)

Luo, Y.; Nissen-Meyer, T.; Morency, C.; Tromp, J.

2008-12-01

Seismic imaging in the exploration industry is often based upon ray-theoretical migration techniques (e.g., Kirchhoff) or other ideas which neglect some fraction of the seismic wavefield (e.g., wavefield continuation for acoustic-wave first arrivals) in the inversion process. In a companion paper we discuss the possibility of solving the full physical forward problem (i.e., including visco- and poroelastic, anisotropic media) using the spectral-element method. With such a tool at hand, we can readily apply the adjoint method to tomographic inversions, i.e., iteratively improving an initial 3D background model to fit the data. In the context of this inversion process, we draw connections between kernels in adjoint tomography and basic imaging principles in migration. We show that the images obtained by migration are nothing but particular kinds of adjoint kernels (mainly density kernels). Migration is basically a first step in the iterative inversion process of adjoint tomography. We apply the approach to basic 2D problems involving layered structures, overthrusting faults, topography, salt domes, and poroelastic regions.

TOMO3D: 3-D joint refraction and reflection traveltime tomography parallel code for active-source seismic data—synthetic test

NASA Astrophysics Data System (ADS)

Meléndez, A.; Korenaga, J.; Sallarès, V.; Miniussi, A.; Ranero, C. R.

2015-10-01

We present a new 3-D traveltime tomography code (TOMO3D) for the modelling of active-source seismic data that uses the arrival times of both refracted and reflected seismic phases to derive the velocity distribution and the geometry of reflecting boundaries in the subsurface. This code is based on its popular 2-D version TOMO2D from which it inherited the methods to solve the forward and inverse problems. The traveltime calculations are done using a hybrid ray-tracing technique combining the graph and bending methods. The LSQR algorithm is used to perform the iterative regularized inversion to improve the initial velocity and depth models. In order to cope with an increased computational demand due to the incorporation of the third dimension, the forward problem solver, which takes most of the run time (˜90 per cent in the test presented here), has been parallelized with a combination of multi-processing and message passing interface standards. This parallelization distributes the ray-tracing and traveltime calculations among available computational resources. The code's performance is illustrated with a realistic synthetic example, including a checkerboard anomaly and two reflectors, which simulates the geometry of a subduction zone. The code is designed to invert for a single reflector at a time. A data-driven layer-stripping strategy is proposed for cases involving multiple reflectors, and it is tested for the successive inversion of the two reflectors. Layers are bound by consecutive reflectors, and an initial velocity model for each inversion step incorporates the results from previous steps. This strategy poses simpler inversion problems at each step, allowing the recovery of strong velocity discontinuities that would otherwise be smoothened.

Spin model for nontrivial types of magnetic order in inverse-perovskite antiferromagnets

NASA Astrophysics Data System (ADS)

Mochizuki, Masahito; Kobayashi, Masaya; Okabe, Reoya; Yamamoto, Daisuke

2018-02-01

Nontrivial magnetic orders in the inverse-perovskite manganese nitrides are theoretically studied by constructing a classical spin model describing the magnetic anisotropy and frustrated exchange interactions inherent in specific crystal and electronic structures of these materials. With a replica-exchange Monte Carlo technique, a theoretical analysis of this model reproduces the experimentally observed triangular Γ5 g and Γ4 g spin-ordered patterns and the systematic evolution of magnetic orders. Our Rapid Communication solves a 40-year-old problem of nontrivial magnetism for the inverse-perovskite manganese nitrides and provides a firm basis for clarifying the magnetism-driven negative thermal expansion phenomenon discovered in this class of materials.

A general approach to regularizing inverse problems with regional data using Slepian wavelets

NASA Astrophysics Data System (ADS)

Michel, Volker; Simons, Frederik J.

2017-12-01

Slepian functions are orthogonal function systems that live on subdomains (for example, geographical regions on the Earth’s surface, or bandlimited portions of the entire spectrum). They have been firmly established as a useful tool for the synthesis and analysis of localized (concentrated or confined) signals, and for the modeling and inversion of noise-contaminated data that are only regionally available or only of regional interest. In this paper, we consider a general abstract setup for inverse problems represented by a linear and compact operator between Hilbert spaces with a known singular-value decomposition (svd). In practice, such an svd is often only given for the case of a global expansion of the data (e.g. on the whole sphere) but not for regional data distributions. We show that, in either case, Slepian functions (associated to an arbitrarily prescribed region and the given compact operator) can be determined and applied to construct a regularization for the ill-posed regional inverse problem. Moreover, we describe an algorithm for constructing the Slepian basis via an algebraic eigenvalue problem. The obtained Slepian functions can be used to derive an svd for the combination of the regionalizing projection and the compact operator. As a result, standard regularization techniques relying on a known svd become applicable also to those inverse problems where the data are regionally given only. In particular, wavelet-based multiscale techniques can be used. An example for the latter case is elaborated theoretically and tested on two synthetic numerical examples.

Final Technical Report for "Applied Mathematics Research: Simulation Based Optimization and Application to Electromagnetic Inverse Problems"

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

Haber, Eldad

2014-03-17

The focus of research was: Developing adaptive mesh for the solution of Maxwell's equations; Developing a parallel framework for time dependent inverse Maxwell's equations; Developing multilevel methods for optimization problems with inequality constraints; A new inversion code for inverse Maxwell's equations in the 0th frequency (DC resistivity); A new inversion code for inverse Maxwell's equations in low frequency regime. Although the research concentrated on electromagnetic forward and in- verse problems the results of the research was applied to the problem of image registration.

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.

Computer modeling of inversion layer MOS solar cells and arrays

NASA Technical Reports Server (NTRS)

Ho, Fat Duen

1991-01-01

A two dimensional numerical model of the inversion layer metal insulator semiconductor (IL/MIS) solar cell is proposed by using the finite element method. The two-dimensional current flow in the device is taken into account in this model. The electrostatic potential distribution, the electron concentration distribution, and the hole concentration distribution for different terminal voltages are simulated. The results of simple calculation are presented. The existing problems for this model are addressed. Future work is proposed. The MIS structures are studied and some of the results are reported.

Effects of adaptive refinement on the inverse EEG solution

NASA Astrophysics Data System (ADS)

Weinstein, David M.; Johnson, Christopher R.; Schmidt, John A.

1995-10-01

One of the fundamental problems in electroencephalography can be characterized by an inverse problem. Given a subset of electrostatic potentials measured on the surface of the scalp and the geometry and conductivity properties within the head, calculate the current vectors and potential fields within the cerebrum. Mathematically the generalized EEG problem can be stated as solving Poisson's equation of electrical conduction for the primary current sources. The resulting problem is mathematically ill-posed i.e., the solution does not depend continuously on the data, such that small errors in the measurement of the voltages on the scalp can yield unbounded errors in the solution, and, for the general treatment of a solution of Poisson's equation, the solution is non-unique. However, if accurate solutions the general treatment of a solution of Poisson's equation, the solution is non-unique. However, if accurate solutions to such problems could be obtained, neurologists would gain noninvasive accesss to patient-specific cortical activity. Access to such data would ultimately increase the number of patients who could be effectively treated for pathological cortical conditions such as temporal lobe epilepsy. In this paper, we present the effects of spatial adaptive refinement on the inverse EEG problem and show that the use of adaptive methods allow for significantly better estimates of electric and potential fileds within the brain through an inverse procedure. To test these methods, we have constructed several finite element head models from magneteic resonance images of a patient. The finite element meshes ranged in size from 2724 nodes and 12,812 elements to 5224 nodes and 29,135 tetrahedral elements, depending on the level of discretization. We show that an adaptive meshing algorithm minimizes the error in the forward problem due to spatial discretization and thus increases the accuracy of the inverse solution.

Discrete Inverse and State Estimation Problems

NASA Astrophysics Data System (ADS)

Wunsch, Carl

2006-06-01

The problems of making inferences about the natural world from noisy observations and imperfect theories occur in almost all scientific disciplines. This book addresses these problems using examples taken from geophysical fluid dynamics. It focuses on discrete formulations, both static and time-varying, known variously as inverse, state estimation or data assimilation problems. Starting with fundamental algebraic and statistical ideas, the book guides the reader through a range of inference tools including the singular value decomposition, Gauss-Markov and minimum variance estimates, Kalman filters and related smoothers, and adjoint (Lagrange multiplier) methods. The final chapters discuss a variety of practical applications to geophysical flow problems. Discrete Inverse and State Estimation Problems is an ideal introduction to the topic for graduate students and researchers in oceanography, meteorology, climate dynamics, and geophysical fluid dynamics. It is also accessible to a wider scientific audience; the only prerequisite is an understanding of linear algebra. Provides a comprehensive introduction to discrete methods of inference from incomplete information Based upon 25 years of practical experience using real data and models Develops sequential and whole-domain analysis methods from simple least-squares Contains many examples and problems, and web-based support through MIT opencourseware

Computed inverse resonance imaging for magnetic susceptibility map reconstruction.

PubMed

Chen, Zikuan; Calhoun, Vince

2012-01-01

This article reports a computed inverse magnetic resonance imaging (CIMRI) model for reconstructing the magnetic susceptibility source from MRI data using a 2-step computational approach. The forward T2*-weighted MRI (T2*MRI) process is broken down into 2 steps: (1) from magnetic susceptibility source to field map establishment via magnetization in the main field and (2) from field map to MR image formation by intravoxel dephasing average. The proposed CIMRI model includes 2 inverse steps to reverse the T2*MRI procedure: field map calculation from MR-phase image and susceptibility source calculation from the field map. The inverse step from field map to susceptibility map is a 3-dimensional ill-posed deconvolution problem, which can be solved with 3 kinds of approaches: the Tikhonov-regularized matrix inverse, inverse filtering with a truncated filter, and total variation (TV) iteration. By numerical simulation, we validate the CIMRI model by comparing the reconstructed susceptibility maps for a predefined susceptibility source. Numerical simulations of CIMRI show that the split Bregman TV iteration solver can reconstruct the susceptibility map from an MR-phase image with high fidelity (spatial correlation ≈ 0.99). The split Bregman TV iteration solver includes noise reduction, edge preservation, and image energy conservation. For applications to brain susceptibility reconstruction, it is important to calibrate the TV iteration program by selecting suitable values of the regularization parameter. The proposed CIMRI model can reconstruct the magnetic susceptibility source of T2*MRI by 2 computational steps: calculating the field map from the phase image and reconstructing the susceptibility map from the field map. The crux of CIMRI lies in an ill-posed 3-dimensional deconvolution problem, which can be effectively solved by the split Bregman TV iteration algorithm.

Computed inverse MRI for magnetic susceptibility map reconstruction

PubMed Central

Chen, Zikuan; Calhoun, Vince

2015-01-01

Objective This paper reports on a computed inverse magnetic resonance imaging (CIMRI) model for reconstructing the magnetic susceptibility source from MRI data using a two-step computational approach. Methods The forward T2*-weighted MRI (T2*MRI) process is decomposed into two steps: 1) from magnetic susceptibility source to fieldmap establishment via magnetization in a main field, and 2) from fieldmap to MR image formation by intravoxel dephasing average. The proposed CIMRI model includes two inverse steps to reverse the T2*MRI procedure: fieldmap calculation from MR phase image and susceptibility source calculation from the fieldmap. The inverse step from fieldmap to susceptibility map is a 3D ill-posed deconvolution problem, which can be solved by three kinds of approaches: Tikhonov-regularized matrix inverse, inverse filtering with a truncated filter, and total variation (TV) iteration. By numerical simulation, we validate the CIMRI model by comparing the reconstructed susceptibility maps for a predefined susceptibility source. Results Numerical simulations of CIMRI show that the split Bregman TV iteration solver can reconstruct the susceptibility map from a MR phase image with high fidelity (spatial correlation≈0.99). The split Bregman TV iteration solver includes noise reduction, edge preservation, and image energy conservation. For applications to brain susceptibility reconstruction, it is important to calibrate the TV iteration program by selecting suitable values of the regularization parameter. Conclusions The proposed CIMRI model can reconstruct the magnetic susceptibility source of T2*MRI by two computational steps: calculating the fieldmap from the phase image and reconstructing the susceptibility map from the fieldmap. The crux of CIMRI lies in an ill-posed 3D deconvolution problem, which can be effectively solved by the split Bregman TV iteration algorithm. PMID:22446372

Constraining Mass Anomalies Using Trans-dimensional Gravity Inversions

NASA Astrophysics Data System (ADS)

Izquierdo, K.; Montesi, L.; Lekic, V.

2016-12-01

The density structure of planetary interiors constitutes a key constraint on their composition, temperature, and dynamics. This has motivated the development of non-invasive methods to infer 3D distribution of density anomalies within a planet's interior using gravity observations made from the surface or orbit. On Earth, this information can be supplemented by seismic and electromagnetic observations, but such data are generally not available on other planets and inferences must be made from gravity observations alone. Unfortunately, inferences of density anomalies from gravity are non-unique and even the dimensionality of the problem - i.e., the number of density anomalies detectable in the planetary interior - is unknown. In this project, we use the Reversible Jump Markov chain Monte Carlo (RJMCMC) algorithm to approach gravity inversions in a trans-dimensional way, that is, considering the magnitude of the mass, the latitude, longitude, depth and number of anomalies itself as unknowns to be constrained by the observed gravity field at the surface of a planet. Our approach builds upon previous work using trans-dimensional gravity inversions in which the density contrast between the anomaly and the surrounding material is known. We validate the algorithm by analyzing a synthetic gravity field produced by a known density structure and comparing the retrieved and input density structures. We find excellent agreement between the input and retrieved structure when working in 1D and 2D domains. However, in 3D domains, comprehensive exploration of the much larger space of possible models makes search efficiency a key ingredient in successful gravity inversion. We find that upon a sufficiently long RJMCMC run, it is possible to use statistical information to recover a predicted model that matches the real model. We argue that even more complex problems, such as those involving real gravity acceleration data of a planet as the constraint, our trans-dimensional gravity inversion algorithm provides a good option to overcome the problem of non-uniqueness while achieving parsimony in gravity inversions.

Adults' understanding of inversion concepts: how does performance on addition and subtraction inversion problems compare to performance on multiplication and division inversion problems?

PubMed

Robinson, Katherine M; Ninowski, Jerilyn E

2003-12-01

Problems of the form a + b - b have been used to assess conceptual understanding of the relationship between addition and subtraction. No study has investigated the same relationship between multiplication and division on problems of the form d x e / e. In both types of inversion problems, no calculation is required if the inverse relationship between the operations is understood. Adult participants solved addition/subtraction and multiplication/division inversion (e.g., 9 x 22 / 22) and standard (e.g., 2 + 27 - 28) problems. Participants started to use the inversion strategy earlier and more frequently on addition/subtraction problems. Participants took longer to solve both types of multiplication/division problems. Overall, conceptual understanding of the relationship between multiplication and division was not as strong as that between addition and subtraction. One explanation for this difference in performance is that the operation of division is more weakly represented and understood than the other operations and that this weakness affects performance on problems of the form d x e / e.

Fast Nonlinear Generalized Inversion of Gravity Data with Application to the Three-Dimensional Crustal Density Structure of Sichuan Basin, Southwest China

NASA Astrophysics Data System (ADS)

Wang, Jun; Meng, Xiaohong; Li, Fang

2017-11-01

Generalized inversion is one of the important steps in the quantitative interpretation of gravity data. With appropriate algorithm and parameters, it gives a view of the subsurface which characterizes different geological bodies. However, generalized inversion of gravity data is time consuming due to the large amount of data points and model cells adopted. Incorporating of various prior information as constraints deteriorates the above situation. In the work discussed in this paper, a method for fast nonlinear generalized inversion of gravity data is proposed. The fast multipole method is employed for forward modelling. The inversion objective function is established with weighted data misfit function along with model objective function. The total objective function is solved by a dataspace algorithm. Moreover, depth weighing factor is used to improve depth resolution of the result, and bound constraint is incorporated by a transfer function to limit the model parameters in a reliable range. The matrix inversion is accomplished by a preconditioned conjugate gradient method. With the above algorithm, equivalent density vectors can be obtained, and interpolation is performed to get the finally density model on the fine mesh in the model domain. Testing on synthetic gravity data demonstrated that the proposed method is faster than conventional generalized inversion algorithm to produce an acceptable solution for gravity inversion problem. The new developed inversion method was also applied for inversion of the gravity data collected over Sichuan basin, southwest China. The established density structure in this study helps understanding the crustal structure of Sichuan basin and provides reference for further oil and gas exploration in this area.

Three-dimensional inversion of multisource array electromagnetic data

NASA Astrophysics Data System (ADS)

Tartaras, Efthimios

Three-dimensional (3-D) inversion is increasingly important for the correct interpretation of geophysical data sets in complex environments. To this effect, several approximate solutions have been developed that allow the construction of relatively fast inversion schemes. One such method that is fast and provides satisfactory accuracy is the quasi-linear (QL) approximation. It has, however, the drawback that it is source-dependent and, therefore, impractical in situations where multiple transmitters in different positions are employed. I have, therefore, developed a localized form of the QL approximation that is source-independent. This so-called localized quasi-linear (LQL) approximation can have a scalar, a diagonal, or a full tensor form. Numerical examples of its comparison with the full integral equation solution, the Born approximation, and the original QL approximation are given. The objective behind developing this approximation is to use it in a fast 3-D inversion scheme appropriate for multisource array data such as those collected in airborne surveys, cross-well logging, and other similar geophysical applications. I have developed such an inversion scheme using the scalar and diagonal LQL approximation. It reduces the original nonlinear inverse electromagnetic (EM) problem to three linear inverse problems. The first of these problems is solved using a weighted regularized linear conjugate gradient method, whereas the last two are solved in the least squares sense. The algorithm I developed provides the option of obtaining either smooth or focused inversion images. I have applied the 3-D LQL inversion to synthetic 3-D EM data that simulate a helicopter-borne survey over different earth models. The results demonstrate the stability and efficiency of the method and show that the LQL approximation can be a practical solution to the problem of 3-D inversion of multisource array frequency-domain EM data. I have also applied the method to helicopter-borne EM data collected by INCO Exploration over the Voisey's Bay area in Labrador, Canada. The results of the 3-D inversion successfully delineate the shallow massive sulfides and show that the method can produce reasonable results even in areas of complex geology and large resistivity contrasts.

Radiative-conductive inverse problem for lumped parameter systems

NASA Astrophysics Data System (ADS)

Alifanov, O. M.; Nenarokomov, A. V.; Gonzalez, V. M.

2008-11-01

The purpose of this paper is to introduce a iterative regularization method in the research of radiative and thermal properties of materials with applications in the design of Thermal Control Systems (TCS) of spacecrafts. In this paper the radiative and thermal properties (emissivity and thermal conductance) of a multilayered thermal-insulating blanket (MLI), which is a screen-vacuum thermal insulation as a part of the (TCS) for perspective spacecrafts, are estimated. Properties of the materials under study are determined in the result of temperature and heat flux measurement data processing based on the solution of the Inverse Heat Transfer Problem (IHTP) technique. Given are physical and mathematical models of heat transfer processes in a specimen of the multilayered thermal-insulating blanket located in the experimental facility. A mathematical formulation of the inverse heat conduction problem is presented too. The practical testing were performed for specimen of the real MLI.

Study of multilayer thermal insulation by inverse problems method

NASA Astrophysics Data System (ADS)

Alifanov, O. M.; Nenarokomov, A. V.; Gonzalez, V. M.

2009-11-01

The purpose of this paper is to introduce a new method in the research of radiative and thermal properties of materials with further applications in the design of thermal control systems (TCS) of spacecrafts. In this paper the radiative and thermal properties (emissivity and thermal conductance) of a multilayered thermal-insulating blanket (MLI), which is a screen-vacuum thermal insulation as a part of the TCS for perspective spacecrafts, are estimated. Properties of the materials under study are determined in the result of temperature and heat flux measurement data processing based on the solution of the inverse heat transfer problem (IHTP) technique. Given are physical and mathematical models of heat transfer processes in a specimen of the multilayered thermal-insulating blanket located in the experimental facility. A mathematical formulation of the inverse heat conduction problem is presented as well. The practical approves were made for specimen of the real MLI.

Use of Genetic Algorithms to solve Inverse Problems in Relativistic Hydrodynamics

NASA Astrophysics Data System (ADS)

Guzmán, F. S.; González, J. A.

2018-04-01

We present the use of Genetic Algorithms (GAs) as a strategy to solve inverse problems associated with models of relativistic hydrodynamics. The signal we consider to emulate an observation is the density of a relativistic gas, measured at a point where a shock is traveling. This shock is generated numerically out of a Riemann problem with mildly relativistic conditions. The inverse problem we propose is the prediction of the initial conditions of density, velocity and pressure of the Riemann problem that gave origin to that signal. For this we use the density, velocity and pressure of the gas at both sides of the discontinuity, as the six genes of an organism, initially with random values within a tolerance. We then prepare an initial population of N of these organisms and evolve them using methods based on GAs. In the end, the organism with the best fitness of each generation is compared to the signal and the process ends when the set of initial conditions of the organisms of a later generation fit the Signal within a tolerance.

Recovery of time-dependent volatility in option pricing model

NASA Astrophysics Data System (ADS)

Deng, Zui-Cha; Hon, Y. C.; Isakov, V.

2016-11-01

In this paper we investigate an inverse problem of determining the time-dependent volatility from observed market prices of options with different strikes. Due to the non linearity and sparsity of observations, an analytical solution to the problem is generally not available. Numerical approximation is also difficult to obtain using most of the existing numerical algorithms. Based on our recent theoretical results, we apply the linearisation technique to convert the problem into an inverse source problem from which recovery of the unknown volatility function can be achieved. Two kinds of strategies, namely, the integral equation method and the Landweber iterations, are adopted to obtain the stable numerical solution to the inverse problem. Both theoretical analysis and numerical examples confirm that the proposed approaches are effective. The work described in this paper was partially supported by a grant from the Research Grant Council of the Hong Kong Special Administrative Region (Project No. CityU 101112) and grants from the NNSF of China (Nos. 11261029, 11461039), and NSF grants DMS 10-08902 and 15-14886 and by Emylou Keith and Betty Dutcher Distinguished Professorship at the Wichita State University (USA).

Towards adjoint-based inversion of time-dependent mantle convection with nonlinear viscosity

NASA Astrophysics Data System (ADS)

Li, Dunzhu; Gurnis, Michael; Stadler, Georg

2017-04-01

We develop and study an adjoint-based inversion method for the simultaneous recovery of initial temperature conditions and viscosity parameters in time-dependent mantle convection from the current mantle temperature and historic plate motion. Based on a realistic rheological model with temperature-dependent and strain-rate-dependent viscosity, we formulate the inversion as a PDE-constrained optimization problem. The objective functional includes the misfit of surface velocity (plate motion) history, the misfit of the current mantle temperature, and a regularization for the uncertain initial condition. The gradient of this functional with respect to the initial temperature and the uncertain viscosity parameters is computed by solving the adjoint of the mantle convection equations. This gradient is used in a pre-conditioned quasi-Newton minimization algorithm. We study the prospects and limitations of the inversion, as well as the computational performance of the method using two synthetic problems, a sinking cylinder and a realistic subduction model. The subduction model is characterized by the migration of a ridge toward a trench whereby both plate motions and subduction evolve. The results demonstrate: (1) for known viscosity parameters, the initial temperature can be well recovered, as in previous initial condition-only inversions where the effective viscosity was given; (2) for known initial temperature, viscosity parameters can be recovered accurately, despite the existence of trade-offs due to ill-conditioning; (3) for the joint inversion of initial condition and viscosity parameters, initial condition and effective viscosity can be reasonably recovered, but the high dimension of the parameter space and the resulting ill-posedness may limit recovery of viscosity parameters.

Sparsity-based acoustic inversion in cross-sectional multiscale optoacoustic imaging.

PubMed

Han, Yiyong; Tzoumas, Stratis; Nunes, Antonio; Ntziachristos, Vasilis; Rosenthal, Amir

2015-09-01

With recent advancement in hardware of optoacoustic imaging systems, highly detailed cross-sectional images may be acquired at a single laser shot, thus eliminating motion artifacts. Nonetheless, other sources of artifacts remain due to signal distortion or out-of-plane signals. The purpose of image reconstruction algorithms is to obtain the most accurate images from noisy, distorted projection data. In this paper, the authors use the model-based approach for acoustic inversion, combined with a sparsity-based inversion procedure. Specifically, a cost function is used that includes the L1 norm of the image in sparse representation and a total variation (TV) term. The optimization problem is solved by a numerically efficient implementation of a nonlinear gradient descent algorithm. TV-L1 model-based inversion is tested in the cross section geometry for numerically generated data as well as for in vivo experimental data from an adult mouse. In all cases, model-based TV-L1 inversion showed a better performance over the conventional Tikhonov regularization, TV inversion, and L1 inversion. In the numerical examples, the images reconstructed with TV-L1 inversion were quantitatively more similar to the originating images. In the experimental examples, TV-L1 inversion yielded sharper images and weaker streak artifact. The results herein show that TV-L1 inversion is capable of improving the quality of highly detailed, multiscale optoacoustic images obtained in vivo using cross-sectional imaging systems. As a result of its high fidelity, model-based TV-L1 inversion may be considered as the new standard for image reconstruction in cross-sectional imaging.

Convergence analysis of surrogate-based methods for Bayesian inverse problems

NASA Astrophysics Data System (ADS)

Yan, Liang; Zhang, Yuan-Xiang

2017-12-01

The major challenges in the Bayesian inverse problems arise from the need for repeated evaluations of the forward model, as required by Markov chain Monte Carlo (MCMC) methods for posterior sampling. Many attempts at accelerating Bayesian inference have relied on surrogates for the forward model, typically constructed through repeated forward simulations that are performed in an offline phase. Although such approaches can be quite effective at reducing computation cost, there has been little analysis of the approximation on posterior inference. In this work, we prove error bounds on the Kullback-Leibler (KL) distance between the true posterior distribution and the approximation based on surrogate models. Our rigorous error analysis show that if the forward model approximation converges at certain rate in the prior-weighted L 2 norm, then the posterior distribution generated by the approximation converges to the true posterior at least two times faster in the KL sense. The error bound on the Hellinger distance is also provided. To provide concrete examples focusing on the use of the surrogate model based methods, we present an efficient technique for constructing stochastic surrogate models to accelerate the Bayesian inference approach. The Christoffel least squares algorithms, based on generalized polynomial chaos, are used to construct a polynomial approximation of the forward solution over the support of the prior distribution. The numerical strategy and the predicted convergence rates are then demonstrated on the nonlinear inverse problems, involving the inference of parameters appearing in partial differential equations.

Mature red blood cells: from optical model to inverse light-scattering problem.

PubMed

Gilev, Konstantin V; Yurkin, Maxim A; Chernyshova, Ekaterina S; Strokotov, Dmitry I; Chernyshev, Andrei V; Maltsev, Valeri P

2016-04-01

We propose a method for characterization of mature red blood cells (RBCs) morphology, based on measurement of light-scattering patterns (LSPs) of individual RBCs with the scanning flow cytometer and on solution of the inverse light-scattering (ILS) problem for each LSP. We considered a RBC shape model, corresponding to the minimal bending energy of the membrane with isotropic elasticity, and constructed an analytical approximation, which allows rapid simulation of the shape, given the diameter and minimal and maximal thicknesses. The ILS problem was solved by the nearest-neighbor interpolation using a preliminary calculated database of 250,000 theoretical LSPs. For each RBC in blood sample we determined three abovementioned shape characteristics and refractive index, which also allows us to calculate volume, surface area, sphericity index, spontaneous curvature, hemoglobin concentration and content.

Mature red blood cells: from optical model to inverse light-scattering problem

PubMed Central

Gilev, Konstantin V.; Yurkin, Maxim A.; Chernyshova, Ekaterina S.; Strokotov, Dmitry I.; Chernyshev, Andrei V.; Maltsev, Valeri P.

2016-01-01

We propose a method for characterization of mature red blood cells (RBCs) morphology, based on measurement of light-scattering patterns (LSPs) of individual RBCs with the scanning flow cytometer and on solution of the inverse light-scattering (ILS) problem for each LSP. We considered a RBC shape model, corresponding to the minimal bending energy of the membrane with isotropic elasticity, and constructed an analytical approximation, which allows rapid simulation of the shape, given the diameter and minimal and maximal thicknesses. The ILS problem was solved by the nearest-neighbor interpolation using a preliminary calculated database of 250,000 theoretical LSPs. For each RBC in blood sample we determined three abovementioned shape characteristics and refractive index, which also allows us to calculate volume, surface area, sphericity index, spontaneous curvature, hemoglobin concentration and content. PMID:27446656

Data matching for free-surface multiple attenuation by multidimensional deconvolution

NASA Astrophysics Data System (ADS)

van der Neut, Joost; Frijlink, Martijn; van Borselen, Roald

2012-09-01

A common strategy for surface-related multiple elimination of seismic data is to predict multiples by a convolutional model and subtract these adaptively from the input gathers. Problems can be posed by interfering multiples and primaries. Removing multiples by multidimensional deconvolution (MDD) (inversion) does not suffer from these problems. However, this approach requires data to be consistent, which is often not the case, especially not at interpolated near-offsets. A novel method is proposed to improve data consistency prior to inversion. This is done by backpropagating first-order multiples with a time-gated reference primary event and matching these with early primaries in the input gather. After data matching, multiple elimination by MDD can be applied with a deterministic inversion scheme.

Variational methods for direct/inverse problems of atmospheric dynamics and chemistry

NASA Astrophysics Data System (ADS)

Penenko, Vladimir; Penenko, Alexey; Tsvetova, Elena

2013-04-01

We present a variational approach for solving direct and inverse problems of atmospheric hydrodynamics and chemistry. It is important that the accurate matching of numerical schemes has to be provided in the chain of objects: direct/adjoint problems - sensitivity relations - inverse problems, including assimilation of all available measurement data. To solve the problems we have developed a new enhanced set of cost-effective algorithms. The matched description of the multi-scale processes is provided by a specific choice of the variational principle functionals for the whole set of integrated models. Then all functionals of variational principle are approximated in space and time by splitting and decomposition methods. Such approach allows us to separately consider, for example, the space-time problems of atmospheric chemistry in the frames of decomposition schemes for the integral identity sum analogs of the variational principle at each time step and in each of 3D finite-volumes. To enhance the realization efficiency, the set of chemical reactions is divided on the subsets related to the operators of production and destruction. Then the idea of the Euler's integrating factors is applied in the frames of the local adjoint problem technique [1]-[3]. The analytical solutions of such adjoint problems play the role of integrating factors for differential equations describing atmospheric chemistry. With their help, the system of differential equations is transformed to the equivalent system of integral equations. As a result we avoid the construction and inversion of preconditioning operators containing the Jacobi matrixes which arise in traditional implicit schemes for ODE solution. This is the main advantage of our schemes. At the same time step but on the different stages of the "global" splitting scheme, the system of atmospheric dynamic equations is solved. For convection - diffusion equations for all state functions in the integrated models we have developed the monotone and stable discrete-analytical numerical schemes [1]-[3] conserving the positivity of the chemical substance concentrations and possessing the properties of energy and mass balance that are postulated in the general variational principle for integrated models. All algorithms for solution of transport, diffusion and transformation problems are direct (without iterations). The work is partially supported by the Programs No 4 of Presidium RAS and No 3 of Mathematical Department of RAS, by RFBR project 11-01-00187 and Integrating projects of SD RAS No 8 and 35. Our studies are in the line with the goals of COST Action ES1004. References Penenko V., Tsvetova E. Discrete-analytical methods for the implementation of variational principles in environmental applications// Journal of computational and applied mathematics, 2009, v. 226, 319-330. Penenko A.V. Discrete-analytic schemes for solving an inverse coefficient heat conduction problem in a layered medium with gradient methods// Numerical Analysis and Applications, 2012, V. 5, pp 326-341. V. Penenko, E. Tsvetova. Variational methods for constructing the monotone approximations for atmospheric chemistry models //Numerical Analysis and Applications, 2013 (in press).

Nonlinear inversion of potential-field data using a hybrid-encoding genetic algorithm

USGS Publications Warehouse

Chen, C.; Xia, J.; Liu, J.; Feng, G.

2006-01-01

Using a genetic algorithm to solve an inverse problem of complex nonlinear geophysical equations is advantageous because it does not require computer gradients of models or "good" initial models. The multi-point search of a genetic algorithm makes it easier to find the globally optimal solution while avoiding falling into a local extremum. As is the case in other optimization approaches, the search efficiency for a genetic algorithm is vital in finding desired solutions successfully in a multi-dimensional model space. A binary-encoding genetic algorithm is hardly ever used to resolve an optimization problem such as a simple geophysical inversion with only three unknowns. The encoding mechanism, genetic operators, and population size of the genetic algorithm greatly affect search processes in the evolution. It is clear that improved operators and proper population size promote the convergence. Nevertheless, not all genetic operations perform perfectly while searching under either a uniform binary or a decimal encoding system. With the binary encoding mechanism, the crossover scheme may produce more new individuals than with the decimal encoding. On the other hand, the mutation scheme in a decimal encoding system will create new genes larger in scope than those in the binary encoding. This paper discusses approaches of exploiting the search potential of genetic operations in the two encoding systems and presents an approach with a hybrid-encoding mechanism, multi-point crossover, and dynamic population size for geophysical inversion. We present a method that is based on the routine in which the mutation operation is conducted in the decimal code and multi-point crossover operation in the binary code. The mix-encoding algorithm is called the hybrid-encoding genetic algorithm (HEGA). HEGA provides better genes with a higher probability by a mutation operator and improves genetic algorithms in resolving complicated geophysical inverse problems. Another significant result is that final solution is determined by the average model derived from multiple trials instead of one computation due to the randomness in a genetic algorithm procedure. These advantages were demonstrated by synthetic and real-world examples of inversion of potential-field data. ?? 2005 Elsevier Ltd. All rights reserved.

Children's Understanding of the Arithmetic Concepts of Inversion and Associativity

ERIC Educational Resources Information Center

Robinson, Katherine M.; Ninowski, Jerilyn E.; Gray, Melissa L.

2006-01-01

Previous studies have shown that even preschoolers can solve inversion problems of the form a + b - b by using the knowledge that addition and subtraction are inverse operations. In this study, a new type of inversion problem of the form d x e [divided by] e was also examined. Grade 6 and 8 students solved inversion problems of both types as well…

Active Subspace Methods for Data-Intensive Inverse Problems

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

Wang, Qiqi

2017-04-27

The project has developed theory and computational tools to exploit active subspaces to reduce the dimension in statistical calibration problems. This dimension reduction enables MCMC methods to calibrate otherwise intractable models. The same theoretical and computational tools can also reduce the measurement dimension for calibration problems that use large stores of data.

Solitons of shallow-water models from energy-dependent spectral problems

NASA Astrophysics Data System (ADS)

Haberlin, Jack; Lyons, Tony

2018-01-01

The current work investigates the soliton solutions of the Kaup-Boussinesq equation using the inverse scattering transform method. We outline the construction of the Riemann-Hilbert problem for a pair of energy-dependent spectral problems for the system, which we then use to construct the solution of this hydrodynamic system.

Collective Socialization and Child Conduct Problems: A Multilevel Analysis with an African American Sample

ERIC Educational Resources Information Center

Simons, Leslie Gordon; Simons, Ronald L.; Conger, Rand D.; Brody, Gene H.

2004-01-01

This article uses hierarchical linear modeling with a sample of African American children and their primary caregivers to examine the association between various community factors and child conduct problems. The analysis revealed a rather strong inverse association between level of collective socialization and conduct problems. This relationship…

Probing flexible thermoplastic thin films on a substrate using ultrasonic waves to retrieve mechanical moduli and density: Inverse problem

NASA Astrophysics Data System (ADS)

Lazri, H.; Ogam, E.; Amar, B.; Fellah, Z. E. A.; Sayoud, N.; Boumaiza, Y.

2018-05-01

Flexible, supple thermoplastic thin films (PVB and PET) placed on elastic substrates were probed using ultrasonic waves to identify their mechanical moduli and density. The composite medium immersed in a fluid host medium (water) was excited using a 50 Mhz transducer operating at normal incidence in reflection mode. Elastic wave propagation data from the stratified medium was captured in the host medium as scattered field. These data were used along with theoretical fluid-solid interaction forward models for stratified-media developed using elasticity theory, to solve an inverse problem for the recovery of the model parameters of the thin films. Two configurations were modeled, one considering the substrate as a semi-infinite elastic medium and the second the substrate having a finite thickness and flanked by a semi-infinite host medium. Transverse slip for the sliding interface between the films and substrate was chosen. This was found to agree with the experiments whereby the thin films were just placed on the substrate without bonding. The inverse problems for the recovery of the mechanical parameters were successful in retrieving the thin films’ parameters under the slip boundary condition. The possible improvements to the new method for the characterization of thin films are discussed.

A 2D forward and inverse code for streaming potential problems

NASA Astrophysics Data System (ADS)

Soueid Ahmed, A.; Jardani, A.; Revil, A.

2013-12-01

The self-potential method corresponds to the passive measurement of the electrical field in response to the occurrence of natural sources of current in the ground. One of these sources corresponds to the streaming current associated with the flow of the groundwater. We can therefore apply the self- potential method to recover non-intrusively some information regarding the groundwater flow. We first solve the forward problem starting with the solution of the groundwater flow problem, then computing the source current density, and finally solving a Poisson equation for the electrical potential. We use the finite-element method to solve the relevant partial differential equations. In order to reduce the number of (petrophysical) model parameters required to solve the forward problem, we introduced an effective charge density tensor of the pore water, which can be determined directly from the permeability tensor for neutral pore waters. The second aspect of our work concerns the inversion of the self-potential data using Tikhonov regularization with smoothness and weighting depth constraints. This approach accounts for the distribution of the electrical resistivity, which can be independently and approximately determined from electrical resistivity tomography. A numerical code, SP2DINV, has been implemented in Matlab to perform both the forward and inverse modeling. Three synthetic case studies are discussed.

Multigrid-based reconstruction algorithm for quantitative photoacoustic tomography

PubMed Central

Li, Shengfu; Montcel, Bruno; Yuan, Zhen; Liu, Wanyu; Vray, Didier

2015-01-01

This paper proposes a multigrid inversion framework for quantitative photoacoustic tomography reconstruction. The forward model of optical fluence distribution and the inverse problem are solved at multiple resolutions. A fixed-point iteration scheme is formulated for each resolution and used as a cost function. The simulated and experimental results for quantitative photoacoustic tomography reconstruction show that the proposed multigrid inversion can dramatically reduce the required number of iterations for the optimization process without loss of reliability in the results. PMID:26203371

Seismic velocity structure of the forearc in northern Cascadia from Bayesian inversion of teleseismic data

NASA Astrophysics Data System (ADS)

Gosselin, J.; Audet, P.; Schaeffer, A. J.

2017-12-01

The seismic velocity structure in the forearc of subduction zones provides important constraints on material properties, with implications for seismogenesis. In Cascadia, previous studies have imaged a downgoing low-velocity zone (LVZ) characterized by an elevated P-to-S velocity ratio (Vp/Vs) down to 45 km depth, near the intersection with the mantle wedge corner, beyond which the signature of the LVZ disappears. These results, combined with the absence of a "normal" continental Moho, indicate that the down-going oceanic crust likely carries large amounts of overpressured free fluids that are released downdip at the onset of crustal eclogitization, and are further stored in the mantle wedge as serpentinite. These overpressured free fluids affect the stability of the plate interface and facilitate slow slip. These results are based on the inversion and migration of scattered teleseismic data for individual layer properties; a methodology which suffers from regularization and smoothing, non-uniqueness, and does not consider model uncertainty. This study instead applies trans-dimensional Bayesian inversion of teleseismic data collected in the forearc of northern Cascadia (the CAFÉ experiment in northern Washington) to provide rigorous, quantitative estimates of local velocity structure, and associated uncertainties (particularly Vp/Vs structure and depth to the plate interface). Trans-dimensional inversion is a generalization of fixed-dimensional inversion that includes the number (and type) of parameters required to describe the velocity model (or data error model) as unknown in the problem. This allows model complexity to be inherently determined by data information content, not by subjective regularization. The inversion is implemented here using the reversible-jump Markov chain Monte Carlo algorithm. The result is an ensemble set of candidate velocity-structure models which approximate the posterior probability density (PPD) of the model parameters. The solution to the inverse problem, and associated uncertainties, are described by properties of the PPD. The results obtained here will eventually be integrated with teleseismic data from OBS stations from the Cascadia Initiative to provide constraints across the entire seismogenic portion of the plate interface.

An adaptive ANOVA-based PCKF for high-dimensional nonlinear inverse modeling

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

Li, Weixuan, E-mail: weixuan.li@usc.edu; Lin, Guang, E-mail: guang.lin@pnnl.gov; Zhang, Dongxiao, E-mail: dxz@pku.edu.cn

2014-02-01

The probabilistic collocation-based Kalman filter (PCKF) is a recently developed approach for solving inverse problems. It resembles the ensemble Kalman filter (EnKF) in every aspect—except that it represents and propagates model uncertainty by polynomial chaos expansion (PCE) instead of an ensemble of model realizations. Previous studies have shown PCKF is a more efficient alternative to EnKF for many data assimilation problems. However, the accuracy and efficiency of PCKF depends on an appropriate truncation of the PCE series. Having more polynomial chaos basis functions in the expansion helps to capture uncertainty more accurately but increases computational cost. Selection of basis functionsmore » is particularly important for high-dimensional stochastic problems because the number of polynomial chaos basis functions required to represent model uncertainty grows dramatically as the number of input parameters (random dimensions) increases. In classic PCKF algorithms, the PCE basis functions are pre-set based on users' experience. Also, for sequential data assimilation problems, the basis functions kept in PCE expression remain unchanged in different Kalman filter loops, which could limit the accuracy and computational efficiency of classic PCKF algorithms. To address this issue, we present a new algorithm that adaptively selects PCE basis functions for different problems and automatically adjusts the number of basis functions in different Kalman filter loops. The algorithm is based on adaptive functional ANOVA (analysis of variance) decomposition, which approximates a high-dimensional function with the summation of a set of low-dimensional functions. Thus, instead of expanding the original model into PCE, we implement the PCE expansion on these low-dimensional functions, which is much less costly. We also propose a new adaptive criterion for ANOVA that is more suited for solving inverse problems. The new algorithm was tested with different examples and demonstrated great effectiveness in comparison with non-adaptive PCKF and EnKF algorithms.« less

An Adaptive ANOVA-based PCKF for High-Dimensional Nonlinear Inverse Modeling

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

LI, Weixuan; Lin, Guang; Zhang, Dongxiao

2014-02-01

The probabilistic collocation-based Kalman filter (PCKF) is a recently developed approach for solving inverse problems. It resembles the ensemble Kalman filter (EnKF) in every aspect—except that it represents and propagates model uncertainty by polynomial chaos expansion (PCE) instead of an ensemble of model realizations. Previous studies have shown PCKF is a more efficient alternative to EnKF for many data assimilation problems. However, the accuracy and efficiency of PCKF depends on an appropriate truncation of the PCE series. Having more polynomial chaos bases in the expansion helps to capture uncertainty more accurately but increases computational cost. Bases selection is particularly importantmore » for high-dimensional stochastic problems because the number of polynomial chaos bases required to represent model uncertainty grows dramatically as the number of input parameters (random dimensions) increases. In classic PCKF algorithms, the PCE bases are pre-set based on users’ experience. Also, for sequential data assimilation problems, the bases kept in PCE expression remain unchanged in different Kalman filter loops, which could limit the accuracy and computational efficiency of classic PCKF algorithms. To address this issue, we present a new algorithm that adaptively selects PCE bases for different problems and automatically adjusts the number of bases in different Kalman filter loops. The algorithm is based on adaptive functional ANOVA (analysis of variance) decomposition, which approximates a high-dimensional function with the summation of a set of low-dimensional functions. Thus, instead of expanding the original model into PCE, we implement the PCE expansion on these low-dimensional functions, which is much less costly. We also propose a new adaptive criterion for ANOVA that is more suited for solving inverse problems. The new algorithm is tested with different examples and demonstrated great effectiveness in comparison with non-adaptive PCKF and EnKF algorithms.« less

Comparison of optimal design methods in inverse problems

NASA Astrophysics Data System (ADS)

Banks, H. T.; Holm, K.; Kappel, F.

2011-07-01

Typical optimal design methods for inverse or parameter estimation problems are designed to choose optimal sampling distributions through minimization of a specific cost function related to the resulting error in parameter estimates. It is hoped that the inverse problem will produce parameter estimates with increased accuracy using data collected according to the optimal sampling distribution. Here we formulate the classical optimal design problem in the context of general optimization problems over distributions of sampling times. We present a new Prohorov metric-based theoretical framework that permits one to treat succinctly and rigorously any optimal design criteria based on the Fisher information matrix. A fundamental approximation theory is also included in this framework. A new optimal design, SE-optimal design (standard error optimal design), is then introduced in the context of this framework. We compare this new design criterion with the more traditional D-optimal and E-optimal designs. The optimal sampling distributions from each design are used to compute and compare standard errors; the standard errors for parameters are computed using asymptotic theory or bootstrapping and the optimal mesh. We use three examples to illustrate ideas: the Verhulst-Pearl logistic population model (Banks H T and Tran H T 2009 Mathematical and Experimental Modeling of Physical and Biological Processes (Boca Raton, FL: Chapman and Hall/CRC)), the standard harmonic oscillator model (Banks H T and Tran H T 2009) and a popular glucose regulation model (Bergman R N, Ider Y Z, Bowden C R and Cobelli C 1979 Am. J. Physiol. 236 E667-77 De Gaetano A and Arino O 2000 J. Math. Biol. 40 136-68 Toffolo G, Bergman R N, Finegood D T, Bowden C R and Cobelli C 1980 Diabetes 29 979-90).

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.

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.

Iterative updating of model error for Bayesian inversion

NASA Astrophysics Data System (ADS)

Calvetti, Daniela; Dunlop, Matthew; Somersalo, Erkki; Stuart, Andrew

2018-02-01

In computational inverse problems, it is common that a detailed and accurate forward model is approximated by a computationally less challenging substitute. The model reduction may be necessary to meet constraints in computing time when optimization algorithms are used to find a single estimate, or to speed up Markov chain Monte Carlo (MCMC) calculations in the Bayesian framework. The use of an approximate model introduces a discrepancy, or modeling error, that may have a detrimental effect on the solution of the ill-posed inverse problem, or it may severely distort the estimate of the posterior distribution. In the Bayesian paradigm, the modeling error can be considered as a random variable, and by using an estimate of the probability distribution of the unknown, one may estimate the probability distribution of the modeling error and incorporate it into the inversion. We introduce an algorithm which iterates this idea to update the distribution of the model error, leading to a sequence of posterior distributions that are demonstrated empirically to capture the underlying truth with increasing accuracy. Since the algorithm is not based on rejections, it requires only limited full model evaluations. We show analytically that, in the linear Gaussian case, the algorithm converges geometrically fast with respect to the number of iterations when the data is finite dimensional. For more general models, we introduce particle approximations of the iteratively generated sequence of distributions; we also prove that each element of the sequence converges in the large particle limit under a simplifying assumption. We show numerically that, as in the linear case, rapid convergence occurs with respect to the number of iterations. Additionally, we show through computed examples that point estimates obtained from this iterative algorithm are superior to those obtained by neglecting the model error.

FOREWORD: 3rd International Workshop on New Computational Methods for Inverse Problems (NCMIP 2013)

NASA Astrophysics Data System (ADS)

Blanc-Féraud, Laure; Joubert, Pierre-Yves

2013-10-01

Conference logo This volume of Journal of Physics: Conference Series is dedicated to the scientific contributions presented during the 3rd International Workshop on New Computational Methods for Inverse Problems, NCMIP 2013 (http://www.farman.ens-cachan.fr/NCMIP_2013.html). This workshop took place at Ecole Normale Supérieure de Cachan, in Cachan, France, on 22 May 2013, at the initiative of Institut Farman. The prior editions of NCMIP also took place in Cachan, France, firstly within the scope of the ValueTools Conference, in May 2011 (http://www.ncmip.org/2011/), and secondly at the initiative of Institut Farman, in May 2012 (http://www.farman.ens-cachan.fr/NCMIP_2012.html). The NCMIP Workshop focused on recent advances in the resolution of inverse problems. Indeed inverse problems appear in numerous scientific areas such as geophysics, biological and medical imaging, material and structure characterization, electrical, mechanical and civil engineering, and finances. The resolution of inverse problems consists of estimating the parameters of the observed system or structure from data collected by an instrumental sensing or imaging device. Its success firstly requires the collection of relevant observation data. It also requires accurate models describing the physical interactions between the instrumental device and the observed system, as well as the intrinsic properties of the solution itself. Finally, it requires the design of robust, accurate and efficient inversion algorithms. Advanced sensor arrays and imaging devices provide high rate and high volume data; in this context, the efficient resolution of the inverse problem requires the joint development of new models and inversion methods, taking computational and implementation aspects into account. During this one-day workshop, researchers had the opportunity to bring to light and share new techniques and results in the field of inverse problems. The topics of the workshop were: algorithms and computational aspects of inversion, Bayesian estimation, kernel methods, learning methods, convex optimization, free discontinuity problems, metamodels, proper orthogonal decomposition, reduced models for the inversion, non-linear inverse scattering, image reconstruction and restoration, and applications (bio-medical imaging, non-destructive evaluation...). NCMIP 2013 was a one-day workshop held in May 2013 which attracted around 60 attendees. Each of the submitted papers has been reviewed by three reviewers. Among the accepted papers, there are seven oral presentations, five posters and one invited poster (On a deconvolution challenge presented by C Vonesch from EPFL, Switzerland). In addition, three international speakers were invited to present a longer talk. The workshop was supported by Institut Farman (ENS Cachan, CNRS) and endorsed by the following French research networks (GDR ISIS, GDR Ondes, GDR MOA, GDR MSPC). The program committee acknowledges the following research laboratories CMLA, LMT, LSV, LURPA, SATIE. Laure Blanc-Féraud and Pierre-Yves Joubert Workshop co-chair Laure Blanc-Féraud, I3S laboratory and INRIA Nice Sophia-Antipolis, France Pierre-Yves Joubert, IEF, Paris-Sud University, CNRS, France Technical program committee Gilles Aubert, J-A Dieudonné Laboratory, CNRS and University of Nice-Sophia Antipolis, France Nabil Anwer, LURPA, ENS Cachan, France Alexandre Baussard, ENSTA Bretagne, Lab-STICC, France Marc Bonnet, ENSTA, ParisTech, France Antonin Chambolle, CMAP, Ecole Polytechnique, CNRS, France Oliver Dorn, School of Mathematics, University of Manchester, UK Cécile Durieu, SATIE, ENS Cachan, CNRS, France Gérard Favier, I3S Laboratory, University of Nice Sophia-Antipolis, France Mário Figueiredo, Instituto Superior Técnico, Lisbon, Portugal Laurent Fribourg, LSV, ENS Cachan, CNRS, France Marc Lambert, L2S Laboratory, CNRS, SupElec, Paris-Sud University, France Dominique Lesselier, L2S Laboratory, CNRS, SupElec, Paris-Sud University, France Matteo Pastorino, DIBE, University of Genoa, Italy Christian Rey, LMT, ENS Cachan, CNRS, France Simon Setzer, Saarland University, Germany Cedric Vonesch, EPFL, Switzerland Local chair Sophie Abriet, SATIE Laboratory, ENS Cachan, France Béatrice Bacquet, SATIE Laboratory, ENS Cachan, France Lydia Matijevic, LMT Laboratory, ENS Cachan France Invited speakers Jérôme Idier, IRCCyN (UMR CNRS 6597), Ecole Centrale de Nantes, France Massimo Fornasier, Faculty of Mathematics, Technical University of Munich, Germany Matthias Fink, Institut Langevin, ESPCI, Université Paris Diderot, France

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.

An alternative empirical likelihood method in missing response problems and causal inference.

PubMed

Ren, Kaili; Drummond, Christopher A; Brewster, Pamela S; Haller, Steven T; Tian, Jiang; Cooper, Christopher J; Zhang, Biao

2016-11-30

Missing responses are common problems in medical, social, and economic studies. When responses are missing at random, a complete case data analysis may result in biases. A popular debias method is inverse probability weighting proposed by Horvitz and Thompson. To improve efficiency, Robins et al. proposed an augmented inverse probability weighting method. The augmented inverse probability weighting estimator has a double-robustness property and achieves the semiparametric efficiency lower bound when the regression model and propensity score model are both correctly specified. In this paper, we introduce an empirical likelihood-based estimator as an alternative to Qin and Zhang (2007). Our proposed estimator is also doubly robust and locally efficient. Simulation results show that the proposed estimator has better performance when the propensity score is correctly modeled. Moreover, the proposed method can be applied in the estimation of average treatment effect in observational causal inferences. Finally, we apply our method to an observational study of smoking, using data from the Cardiovascular Outcomes in Renal Atherosclerotic Lesions clinical trial. Copyright © 2016 John Wiley & Sons, Ltd. Copyright © 2016 John Wiley & Sons, Ltd.

Comparative evolution of the inverse problems (Introduction to an interdisciplinary study of the inverse problems)

NASA Technical Reports Server (NTRS)

Sabatier, P. C.

1972-01-01

The progressive realization of the consequences of nonuniqueness imply an evolution of both the methods and the centers of interest in inverse problems. This evolution is schematically described together with the various mathematical methods used. A comparative description is given of inverse methods in scientific research, with examples taken from mathematics, quantum and classical physics, seismology, transport theory, radiative transfer, electromagnetic scattering, electrocardiology, etc. It is hoped that this paper will pave the way for an interdisciplinary study of inverse problems.

Part-to-itself model inversion in process compensated resonance testing

NASA Astrophysics Data System (ADS)

Mayes, Alexander; Jauriqui, Leanne; Biedermann, Eric; Heffernan, Julieanne; Livings, Richard; Aldrin, John C.; Goodlet, Brent; Mazdiyasni, Siamack

2018-04-01

Process Compensated Resonance Testing (PCRT) is a non-destructive evaluation (NDE) method involving the collection and analysis of a part's resonance spectrum to characterize its material or damage state. Prior work used the finite element method (FEM) to develop forward modeling and model inversion techniques. In many cases, the inversion problem can become confounded by multiple parameters having similar effects on a part's resonance frequencies. To reduce the influence of confounding parameters and isolate the change in a part (e.g., creep), a part-to-itself (PTI) approach can be taken. A PTI approach involves inverting only the change in resonance frequencies from the before and after states of a part. This approach reduces the possible inversion parameters to only those that change in response to in-service loads and damage mechanisms. To evaluate the effectiveness of using a PTI inversion approach, creep strain and material properties were estimated in virtual and real samples using FEM inversion. Virtual and real dog bone samples composed of nickel-based superalloy Mar-M-247 were examined. Virtual samples were modeled with typically observed variations in material properties and dimensions. Creep modeling was verified with the collected resonance spectra from an incrementally crept physical sample. All samples were inverted against a model space that allowed for change in the creep damage state and the material properties but was blind to initial part dimensions. Results quantified the capabilities of PTI inversion in evaluating creep strain and material properties, as well as its sensitivity to confounding initial dimensions.

Bayesian inversion of refraction seismic traveltime data

NASA Astrophysics Data System (ADS)

Ryberg, T.; Haberland, Ch

2018-03-01

We apply a Bayesian Markov chain Monte Carlo (McMC) formalism to the inversion of refraction seismic, traveltime data sets to derive 2-D velocity models below linear arrays (i.e. profiles) of sources and seismic receivers. Typical refraction data sets, especially when using the far-offset observations, are known as having experimental geometries which are very poor, highly ill-posed and far from being ideal. As a consequence, the structural resolution quickly degrades with depth. Conventional inversion techniques, based on regularization, potentially suffer from the choice of appropriate inversion parameters (i.e. number and distribution of cells, starting velocity models, damping and smoothing constraints, data noise level, etc.) and only local model space exploration. McMC techniques are used for exhaustive sampling of the model space without the need of prior knowledge (or assumptions) of inversion parameters, resulting in a large number of models fitting the observations. Statistical analysis of these models allows to derive an average (reference) solution and its standard deviation, thus providing uncertainty estimates of the inversion result. The highly non-linear character of the inversion problem, mainly caused by the experiment geometry, does not allow to derive a reference solution and error map by a simply averaging procedure. We present a modified averaging technique, which excludes parts of the prior distribution in the posterior values due to poor ray coverage, thus providing reliable estimates of inversion model properties even in those parts of the models. The model is discretized by a set of Voronoi polygons (with constant slowness cells) or a triangulated mesh (with interpolation within the triangles). Forward traveltime calculations are performed by a fast, finite-difference-based eikonal solver. The method is applied to a data set from a refraction seismic survey from Northern Namibia and compared to conventional tomography. An inversion test for a synthetic data set from a known model is also presented.

SU-E-J-161: Inverse Problems for Optical Parameters in Laser Induced Thermal Therapy

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

Fahrenholtz, SJ; Stafford, RJ; Fuentes, DT

Purpose: Magnetic resonance-guided laser-induced thermal therapy (MRgLITT) is investigated as a neurosurgical intervention for oncological applications throughout the body by active post market studies. Real-time MR temperature imaging is used to monitor ablative thermal delivery in the clinic. Additionally, brain MRgLITT could improve through effective planning for laser fiber's placement. Mathematical bioheat models have been extensively investigated but require reliable patient specific physical parameter data, e.g. optical parameters. This abstract applies an inverse problem algorithm to characterize optical parameter data obtained from previous MRgLITT interventions. Methods: The implemented inverse problem has three primary components: a parameter-space search algorithm, a physicsmore » model, and training data. First, the parameter-space search algorithm uses a gradient-based quasi-Newton method to optimize the effective optical attenuation coefficient, μ-eff. A parameter reduction reduces the amount of optical parameter-space the algorithm must search. Second, the physics model is a simplified bioheat model for homogeneous tissue where closed-form Green's functions represent the exact solution. Third, the training data was temperature imaging data from 23 MRgLITT oncological brain ablations (980 nm wavelength) from seven different patients. Results: To three significant figures, the descriptive statistics for μ-eff were 1470 m{sup −1} mean, 1360 m{sup −1} median, 369 m{sup −1} standard deviation, 933 m{sup −1} minimum and 2260 m{sup −1} maximum. The standard deviation normalized by the mean was 25.0%. The inverse problem took <30 minutes to optimize all 23 datasets. Conclusion: As expected, the inferred average is biased by underlying physics model. However, the standard deviation normalized by the mean is smaller than literature values and indicates an increased precision in the characterization of the optical parameters needed to plan MRgLITT procedures. This investigation demonstrates the potential for the optimization and validation of more sophisticated bioheat models that incorporate the uncertainty of the data into the predictions, e.g. stochastic finite element methods.« less

A New Paradigm for Satellite Retrieval of Hydrologic Variables: The CDRD Methodology

NASA Astrophysics Data System (ADS)

Smith, E. A.; Mugnai, A.; Tripoli, G. J.

2009-09-01

Historically, retrieval of thermodynamically active geophysical variables in the atmosphere (e.g., temperature, moisture, precipitation) involved some time of inversion scheme - embedded within the retrieval algorithm - to transform radiometric observations (a vector) to the desired geophysical parameter(s) (either a scalar or a vector). Inversion is fundamentally a mathematical operation involving some type of integral-differential radiative transfer equation - often resisting a straightforward algebraic solution - in which the integral side of the equation (typically the right-hand side) contains the desired geophysical vector, while the left-hand side contains the radiative measurement vector often free of operators. Inversion was considered more desirable than forward modeling because the forward model solution had to be selected from a generally unmanageable set of parameter-observation relationships. However, in the classical inversion problem for retrieval of temperature using multiple radiative frequencies along the wing of an absorption band (or line) of a well-mixed radiatively active gas, in either the infrared or microwave spectrums, the inversion equation to be solved consists of a Fredholm integral equation of the 2nd kind - a specific type of transform problem in which there are an infinite number of solutions. This meant that special treatment of the transform process was required in order to obtain a single solution. Inversion had become the method of choice for retrieval in the 1950s because it appealed to the use of mathematical elegance, and because the numerical approaches used to solve the problems (typically some type of relaxation or perturbation scheme) were computationally fast in an age when computers speeds were slow. Like many solution schemes, inversion has lingered on regardless of the fact that computer speeds have increased many orders of magnitude and forward modeling itself has become far more elegant in combination with Bayesian averaging procedures given that the a priori probabilities of occurrence in the true environment of the parameter(s) in question can be approximated (or are actually known). In this presentation, the theory of the more modern retrieval approach using a combination of cloud, radiation and other specialized forward models in conjunction with Bayesian weighted averaging will be reviewed in light of a brief history of inversion. The application of the theory will be cast in the framework of what we call the Cloud-Dynamics-Radiation-Database (CDRD) methodology - which we now use for the retrieval of precipitation from spaceborne passive microwave radiometers. In a companion presentation, we will specifically describe the CDRD methodology and present results for its application within the Mediterranean basin.

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

Oleinikov, A. I., E-mail: a.i.oleinikov@mail.ru; Bormotin, K. S., E-mail: cvmi@knastu.ru

It is shown that inverse problems of steady-state creep bending of plates in both the geometrically linear and nonlinear formulations can be represented in a variational formulation. Steady-state values of the obtained functionals corresponding to the solutions of the problems of inelastic deformation and springback are determined by applying a finite element procedure to the functionals. Optimal laws of creep deformation are formulated using the criterion of minimizing damage in the functionals of the inverse problems. The formulated problems are reduced to the problems solved by the finite element method using MSC.Marc software. Currently, forming of light metals poses tremendousmore » challenges due to their low ductility at room temperature and their unusual deformation characteristics at hot-cold work: strong asymmetry between tensile and compressive behavior, and a very pronounced anisotropy. We used the constitutive models of steady-state creep of initially transverse isotropy structural materials the kind of the stress state has influence. The paper gives basics of the developed computer-aided system of design, modeling, and electronic simulation targeting the processes of manufacture of wing integral panels. The modeling results can be used to calculate the die tooling, determine the panel processibility, and control panel rejection in the course of forming.« less

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).

Joint inversion of seismic refraction and resistivity data using layered models - applications to hydrogeology

NASA Astrophysics Data System (ADS)

Juhojuntti, N. G.; Kamm, J.

2010-12-01

We present a layered-model approach to joint inversion of shallow seismic refraction and resistivity (DC) data, which we believe is a seldom tested method of addressing the problem. This method has been developed as we believe that for shallow sedimentary environments (roughly <100 m depth) a model with a few layers and sharp layer boundaries better represents the subsurface than a smooth minimum-structure (grid) model. Due to the strong assumption our model parameterization implies on the subsurface, only a low number of well resolved model parameters has to be estimated, and provided that this assumptions holds our method can also be applied to other environments. We are using a least-squares inversion, with lateral smoothness constraints, allowing lateral variations in the seismic velocity and the resistivity but no vertical variations. One exception is a positive gradient in the seismic velocity in the uppermost layer in order to get diving rays (the refractions in the deeper layers are modeled as head waves). We assume no connection between seismic velocity and resistivity, and these parameters are allowed to vary individually within the layers. The layer boundaries are, however, common for both parameters. During the inversion lateral smoothing can be applied to the layer boundaries as well as to the seismic velocity and the resistivity. The number of layers is specified before the inversion, and typically we use models with three layers. Depending on the type of environment it is possible to apply smoothing either to the depth of the layer boundaries or to the thickness of the layers, although normally the former is used for shallow sedimentary environments. The smoothing parameters can be chosen independently for each layer. For the DC data we use a finite-difference algorithm to perform the forward modeling and to calculate the Jacobian matrix, while for the seismic data the corresponding entities are retrieved via ray-tracing, using components from the RAYINVR package. The modular layout of the code makes it straightforward to include other types of geophysical data, i.e. gravity. The code has been tested using synthetic examples with fairly simple 2D geometries, mainly for checking the validity of the calculations. The inversion generally converges towards the correct solution, although there could be stability problems if the starting model is too erroneous. We have also applied the code to field data from seismic refraction and multi-electrode resistivity measurements at typical sand-gravel groundwater reservoirs. The tests are promising, as the calculated depths agree fairly well with information from drilling and the velocity and resistivity values appear reasonable. Current work includes better regularization of the inversion as well as defining individual weight factors for the different datasets, as the present algorithm tends to constrain the depths mainly by using the seismic data. More complex synthetic examples will also be tested, including models addressing the seismic hidden-layer problem.

Systematic evaluation of sequential geostatistical resampling within MCMC for posterior sampling of near-surface geophysical inverse problems

NASA Astrophysics Data System (ADS)

Ruggeri, Paolo; Irving, James; Holliger, Klaus

2015-08-01

We critically examine the performance of sequential geostatistical resampling (SGR) as a model proposal mechanism for Bayesian Markov-chain-Monte-Carlo (MCMC) solutions to near-surface geophysical inverse problems. Focusing on a series of simple yet realistic synthetic crosshole georadar tomographic examples characterized by different numbers of data, levels of data error and degrees of model parameter spatial correlation, we investigate the efficiency of three different resampling strategies with regard to their ability to generate statistically independent realizations from the Bayesian posterior distribution. Quite importantly, our results show that, no matter what resampling strategy is employed, many of the examined test cases require an unreasonably high number of forward model runs to produce independent posterior samples, meaning that the SGR approach as currently implemented will not be computationally feasible for a wide range of problems. Although use of a novel gradual-deformation-based proposal method can help to alleviate these issues, it does not offer a full solution. Further, we find that the nature of the SGR is found to strongly influence MCMC performance; however no clear rule exists as to what set of inversion parameters and/or overall proposal acceptance rate will allow for the most efficient implementation. We conclude that although the SGR methodology is highly attractive as it allows for the consideration of complex geostatistical priors as well as conditioning to hard and soft data, further developments are necessary in the context of novel or hybrid MCMC approaches for it to be considered generally suitable for near-surface geophysical inversions.

An adaptive importance sampling algorithm for Bayesian inversion with multimodal distributions

DOE PAGES

Li, Weixuan; Lin, Guang

2015-03-21

Parametric uncertainties are encountered in the simulations of many physical systems, and may be reduced by an inverse modeling procedure that calibrates the simulation results to observations on the real system being simulated. Following Bayes’ rule, a general approach for inverse modeling problems is to sample from the posterior distribution of the uncertain model parameters given the observations. However, the large number of repetitive forward simulations required in the sampling process could pose a prohibitive computational burden. This difficulty is particularly challenging when the posterior is multimodal. We present in this paper an adaptive importance sampling algorithm to tackle thesemore » challenges. Two essential ingredients of the algorithm are: 1) a Gaussian mixture (GM) model adaptively constructed as the proposal distribution to approximate the possibly multimodal target posterior, and 2) a mixture of polynomial chaos (PC) expansions, built according to the GM proposal, as a surrogate model to alleviate the computational burden caused by computational-demanding forward model evaluations. In three illustrative examples, the proposed adaptive importance sampling algorithm demonstrates its capabilities of automatically finding a GM proposal with an appropriate number of modes for the specific problem under study, and obtaining a sample accurately and efficiently representing the posterior with limited number of forward simulations.« less

An adaptive importance sampling algorithm for Bayesian inversion with multimodal distributions

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

Li, Weixuan; Lin, Guang, E-mail: guanglin@purdue.edu

2015-08-01

Parametric uncertainties are encountered in the simulations of many physical systems, and may be reduced by an inverse modeling procedure that calibrates the simulation results to observations on the real system being simulated. Following Bayes' rule, a general approach for inverse modeling problems is to sample from the posterior distribution of the uncertain model parameters given the observations. However, the large number of repetitive forward simulations required in the sampling process could pose a prohibitive computational burden. This difficulty is particularly challenging when the posterior is multimodal. We present in this paper an adaptive importance sampling algorithm to tackle thesemore » challenges. Two essential ingredients of the algorithm are: 1) a Gaussian mixture (GM) model adaptively constructed as the proposal distribution to approximate the possibly multimodal target posterior, and 2) a mixture of polynomial chaos (PC) expansions, built according to the GM proposal, as a surrogate model to alleviate the computational burden caused by computational-demanding forward model evaluations. In three illustrative examples, the proposed adaptive importance sampling algorithm demonstrates its capabilities of automatically finding a GM proposal with an appropriate number of modes for the specific problem under study, and obtaining a sample accurately and efficiently representing the posterior with limited number of forward simulations.« less

Exploring L1 model space in search of conductivity bounds for the MT problem

NASA Astrophysics Data System (ADS)

Wheelock, B. D.; Parker, R. L.

2013-12-01

Geophysical inverse problems of the type encountered in electromagnetic techniques are highly non-unique. As a result, any single inverted model, though feasible, is at best inconclusive and at worst misleading. In this paper, we use modified inversion methods to establish bounds on electrical conductivity within a model of the earth. Our method consists of two steps, each making use of the 1-norm in model regularization. Both 1-norm minimization problems are framed without approximation as non-negative least-squares (NNLS) problems. First, we must identify a parsimonious set of regions within the model for which upper and lower bounds on average conductivity will be sought. This is accomplished by minimizing the 1-norm of spatial variation, which produces a model with a limited number of homogeneous regions; in fact, the number of homogeneous regions will never be greater than the number of data, regardless of the number of free parameters supplied. The second step establishes bounds for each of these regions with pairs of inversions. The new suite of inversions also uses a 1-norm penalty, but applied to the conductivity values themselves, rather than the spatial variation thereof. In the bounding step we use the 1-norm of our model parameters because it is proportional to average conductivity. For a lower bound on average conductivity, the 1-norm within a bounding region is minimized. For an upper bound on average conductivity, the 1-norm everywhere outside a bounding region is minimized. The latter minimization has the effect of concentrating conductance into the bounding region. Taken together, these bounds are a measure of the uncertainty in the associated region of our model. Starting with a blocky inverse solution is key in the selection of the bounding regions. Of course, there is a tradeoff between resolution and uncertainty: an increase in resolution (smaller bounding regions), results in greater uncertainty (wider bounds). Minimization of the 1-norm of spatial variation delivers the fewest possible regions defined by a mean conductivity, the quantity we wish to bound. Thus, these regions present a natural set for which the most narrow and discriminating bounds can be found. For illustration, we apply these techniques to synthetic magnetotelluric (MT) data sets resulting from one-dimensional (1D) earth models. In each case we find that with realistic data coverage, any single inverted model can often stray from the truth, while the computed bounds on an encompassing region contain both the inverted and the true conductivities, indicating that our measure of model uncertainty is robust. Such estimates of uncertainty for conductivity can then be translated to bounds on important petrological parameters such as mineralogy, porosity, saturation, and fluid type.

Simultaneous travel time tomography for updating both velocity and reflector geometry in triangular/tetrahedral cell model

NASA Astrophysics Data System (ADS)

Bai, Chao-ying; He, Lei-yu; Li, Xing-wang; Sun, Jia-yu

2018-05-01

To conduct forward and simultaneous inversion in a complex geological model, including an irregular topography (or irregular reflector or velocity anomaly), we in this paper combined our previous multiphase arrival tracking method (referred as triangular shortest-path method, TSPM) in triangular (2D) or tetrahedral (3D) cell model and a linearized inversion solver (referred to as damped minimum norms and constrained least squares problem solved using the conjugate gradient method, DMNCLS-CG) to formulate a simultaneous travel time inversion method for updating both velocity and reflector geometry by using multiphase arrival times. In the triangular/tetrahedral cells, we deduced the partial derivative of velocity variation with respective to the depth change of reflector. The numerical simulation results show that the computational accuracy can be tuned to a high precision in forward modeling and the irregular velocity anomaly and reflector geometry can be accurately captured in the simultaneous inversion, because the triangular/tetrahedral cell can be easily used to stitch the irregular topography or subsurface interface.

Simultaneous travel time tomography for updating both velocity and reflector geometry in triangular/tetrahedral cell model

NASA Astrophysics Data System (ADS)

Bai, Chao-ying; He, Lei-yu; Li, Xing-wang; Sun, Jia-yu

2017-12-01

To conduct forward and simultaneous inversion in a complex geological model, including an irregular topography (or irregular reflector or velocity anomaly), we in this paper combined our previous multiphase arrival tracking method (referred as triangular shortest-path method, TSPM) in triangular (2D) or tetrahedral (3D) cell model and a linearized inversion solver (referred to as damped minimum norms and constrained least squares problem solved using the conjugate gradient method, DMNCLS-CG) to formulate a simultaneous travel time inversion method for updating both velocity and reflector geometry by using multiphase arrival times. In the triangular/tetrahedral cells, we deduced the partial derivative of velocity variation with respective to the depth change of reflector. The numerical simulation results show that the computational accuracy can be tuned to a high precision in forward modeling and the irregular velocity anomaly and reflector geometry can be accurately captured in the simultaneous inversion, because the triangular/tetrahedral cell can be easily used to stitch the irregular topography or subsurface interface.

A Matlab toolkit for three-dimensional electrical impedance tomography: a contribution to the Electrical Impedance and Diffuse Optical Reconstruction Software project

NASA Astrophysics Data System (ADS)

Polydorides, Nick; Lionheart, William R. B.

2002-12-01

The objective of the Electrical Impedance and Diffuse Optical Reconstruction Software project is to develop freely available software that can be used to reconstruct electrical or optical material properties from boundary measurements. Nonlinear and ill posed problems such as electrical impedance and optical tomography are typically approached using a finite element model for the forward calculations and a regularized nonlinear solver for obtaining a unique and stable inverse solution. Most of the commercially available finite element programs are unsuitable for solving these problems because of their conventional inefficient way of calculating the Jacobian, and their lack of accurate electrode modelling. A complete package for the two-dimensional EIT problem was officially released by Vauhkonen et al at the second half of 2000. However most industrial and medical electrical imaging problems are fundamentally three-dimensional. To assist the development we have developed and released a free toolkit of Matlab routines which can be employed to solve the forward and inverse EIT problems in three dimensions based on the complete electrode model along with some basic visualization utilities, in the hope that it will stimulate further development. We also include a derivation of the formula for the Jacobian (or sensitivity) matrix based on the complete electrode model.

A trade-off between model resolution and variance with selected Rayleigh-wave data

USGS Publications Warehouse

Xia, J.; Miller, R.D.; Xu, Y.

2008-01-01

Inversion of multimode surface-wave data is of increasing interest in the near-surface geophysics community. For a given near-surface geophysical problem, it is essential to understand how well the data, calculated according to a layered-earth model, might match the observed data. A data-resolution matrix is a function of the data kernel (determined by a geophysical model and a priori information applied to the problem), not the data. A data-resolution matrix of high-frequency (??? 2 Hz) Rayleigh-wave phase velocities, therefore, offers a quantitative tool for designing field surveys and predicting the match between calculated and observed data. First, we employed a data-resolution matrix to select data that would be well predicted and to explain advantages of incorporating higher modes in inversion. The resulting discussion using the data-resolution matrix provides insight into the process of inverting Rayleigh-wave phase velocities with higher mode data to estimate S-wave velocity structure. Discussion also suggested that each near-surface geophysical target can only be resolved using Rayleigh-wave phase velocities within specific frequency ranges, and higher mode data are normally more accurately predicted than fundamental mode data because of restrictions on the data kernel for the inversion system. Second, we obtained an optimal damping vector in a vicinity of an inverted model by the singular value decomposition of a trade-off function of model resolution and variance. In the end of the paper, we used a real-world example to demonstrate that selected data with the data-resolution matrix can provide better inversion results and to explain with the data-resolution matrix why incorporating higher mode data in inversion can provide better results. We also calculated model-resolution matrices of these examples to show the potential of increasing model resolution with selected surface-wave data. With the optimal damping vector, we can improve and assess an inverted model obtained by a damped least-square method.

MT+, integrating magnetotellurics to determine earth structure, physical state, and processes

USGS Publications Warehouse

Bedrosian, P.A.

2007-01-01

As one of the few deep-earth imaging techniques, magnetotellurics provides information on both the structure and physical state of the crust and upper mantle. Magnetotellurics is sensitive to electrical conductivity, which varies within the earth by many orders of magnitude and is modified by a range of earth processes. As with all geophysical techniques, magnetotellurics has a non-unique inverse problem and has limitations in resolution and sensitivity. As such, an integrated approach, either via the joint interpretation of independent geophysical models, or through the simultaneous inversion of independent data sets is valuable, and at times essential to an accurate interpretation. Magnetotelluric data and models are increasingly integrated with geological, geophysical and geochemical information. This review considers recent studies that illustrate the ways in which such information is combined, from qualitative comparisons to statistical correlation studies to multi-property inversions. Also emphasized are the range of problems addressed by these integrated approaches, and their value in elucidating earth structure, physical state, and processes. ?? Springer Science+Business Media B.V. 2007.

Reduction of Dynamic Loads in Mine Lifting Installations

NASA Astrophysics Data System (ADS)

Kuznetsov, N. K.; Eliseev, S. V.; Perelygina, A. Yu

2018-01-01

Article is devoted to a problem of decrease in the dynamic loadings arising in transitional operating modes of the mine lifting installations leading to heavy oscillating motions of lifting vessels and decrease in efficiency and reliability of work. The known methods and means of decrease in dynamic loadings and oscillating motions of the similar equipment are analysed. It is shown that an approach based on the concept of the inverse problems of dynamics can be effective method of the solution of this problem. The article describes the design model of a one-ended lifting installation in the form of a two-mass oscillation system, in which the inertial elements are the mass of the lifting vessel and the reduced mass of the engine, reducer, drum and pulley. The simplified mathematical model of this system and results of an efficiency research of an active way of reduction of dynamic loadings of lifting installation on the basis of the concept of the inverse problems of dynamics are given.

Dynamic Shape Reconstruction of Three-Dimensional Frame Structures Using the Inverse Finite Element Method

NASA Technical Reports Server (NTRS)

Gherlone, Marco; Cerracchio, Priscilla; Mattone, Massimiliano; Di Sciuva, Marco; Tessler, Alexander

2011-01-01

A robust and efficient computational method for reconstructing the three-dimensional displacement field of truss, beam, and frame structures, using measured surface-strain data, is presented. Known as shape sensing , this inverse problem has important implications for real-time actuation and control of smart structures, and for monitoring of structural integrity. The present formulation, based on the inverse Finite Element Method (iFEM), uses a least-squares variational principle involving strain measures of Timoshenko theory for stretching, torsion, bending, and transverse shear. Two inverse-frame finite elements are derived using interdependent interpolations whose interior degrees-of-freedom are condensed out at the element level. In addition, relationships between the order of kinematic-element interpolations and the number of required strain gauges are established. As an example problem, a thin-walled, circular cross-section cantilevered beam subjected to harmonic excitations in the presence of structural damping is modeled using iFEM; where, to simulate strain-gauge values and to provide reference displacements, a high-fidelity MSC/NASTRAN shell finite element model is used. Examples of low and high-frequency dynamic motion are analyzed and the solution accuracy examined with respect to various levels of discretization and the number of strain gauges.

Sparsity-based acoustic inversion in cross-sectional multiscale optoacoustic imaging

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

Han, Yiyong; Tzoumas, Stratis; Nunes, Antonio

2015-09-15

Purpose: With recent advancement in hardware of optoacoustic imaging systems, highly detailed cross-sectional images may be acquired at a single laser shot, thus eliminating motion artifacts. Nonetheless, other sources of artifacts remain due to signal distortion or out-of-plane signals. The purpose of image reconstruction algorithms is to obtain the most accurate images from noisy, distorted projection data. Methods: In this paper, the authors use the model-based approach for acoustic inversion, combined with a sparsity-based inversion procedure. Specifically, a cost function is used that includes the L1 norm of the image in sparse representation and a total variation (TV) term. Themore » optimization problem is solved by a numerically efficient implementation of a nonlinear gradient descent algorithm. TV–L1 model-based inversion is tested in the cross section geometry for numerically generated data as well as for in vivo experimental data from an adult mouse. Results: In all cases, model-based TV–L1 inversion showed a better performance over the conventional Tikhonov regularization, TV inversion, and L1 inversion. In the numerical examples, the images reconstructed with TV–L1 inversion were quantitatively more similar to the originating images. In the experimental examples, TV–L1 inversion yielded sharper images and weaker streak artifact. Conclusions: The results herein show that TV–L1 inversion is capable of improving the quality of highly detailed, multiscale optoacoustic images obtained in vivo using cross-sectional imaging systems. As a result of its high fidelity, model-based TV–L1 inversion may be considered as the new standard for image reconstruction in cross-sectional imaging.« less

SaaS Platform for Time Series Data Handling

NASA Astrophysics Data System (ADS)

Oplachko, Ekaterina; Rykunov, Stanislav; Ustinin, Mikhail

2018-02-01

The paper is devoted to the description of MathBrain, a cloud-based resource, which works as a "Software as a Service" model. It is designed to maximize the efficiency of the current technology and to provide a tool for time series data handling. The resource provides access to the following analysis methods: direct and inverse Fourier transforms, Principal component analysis and Independent component analysis decompositions, quantitative analysis, magnetoencephalography inverse problem solution in a single dipole model based on multichannel spectral data.

ITOUGH2(UNIX). Inverse Modeling for TOUGH2 Family of Multiphase Flow Simulators

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

Finsterle, S.

1999-03-01

ITOUGH2 provides inverse modeling capabilities for the TOUGH2 family of numerical simulators for non-isothermal multiphase flows in fractured-porous media. The ITOUGH2 can be used for estimating parameters by automatic modeling calibration, for sensitivity analyses, and for uncertainity propagation analyses (linear and Monte Carlo simulations). Any input parameter to the TOUGH2 simulator can be estimated based on any type of observation for which a corresponding TOUGH2 output is calculated. ITOUGH2 solves a non-linear least-squares problem using direct or gradient-based minimization algorithms. A detailed residual and error analysis is performed, which includes the evaluation of model identification criteria. ITOUGH2 can also bemore » run in forward mode, solving subsurface flow problems related to nuclear waste isolation, oil, gas, and geothermal resevoir engineering, and vadose zone hydrology.« less

BOOK REVIEW: Inverse Problems. Activities for Undergraduates

NASA Astrophysics Data System (ADS)

Yamamoto, Masahiro

2003-06-01

This book is a valuable introduction to inverse problems. In particular, from the educational point of view, the author addresses the questions of what constitutes an inverse problem and how and why we should study them. Such an approach has been eagerly awaited for a long time. Professor Groetsch, of the University of Cincinnati, is a world-renowned specialist in inverse problems, in particular the theory of regularization. Moreover, he has made a remarkable contribution to educational activities in the field of inverse problems, which was the subject of his previous book (Groetsch C W 1993 Inverse Problems in the Mathematical Sciences (Braunschweig: Vieweg)). For this reason, he is one of the most qualified to write an introductory book on inverse problems. Without question, inverse problems are important, necessary and appear in various aspects. So it is crucial to introduce students to exercises in inverse problems. However, there are not many introductory books which are directly accessible by students in the first two undergraduate years. As a consequence, students often encounter diverse concrete inverse problems before becoming aware of their general principles. The main purpose of this book is to present activities to allow first-year undergraduates to learn inverse theory. To my knowledge, this book is a rare attempt to do this and, in my opinion, a great success. The author emphasizes that it is very important to teach inverse theory in the early years. He writes; `If students consider only the direct problem, they are not looking at the problem from all sides .... The habit of always looking at problems from the direct point of view is intellectually limiting ...' (page 21). The book is very carefully organized so that teachers will be able to use it as a textbook. After an introduction in chapter 1, sucessive chapters deal with inverse problems in precalculus, calculus, differential equations and linear algebra. In order to let one gain some insight into the nature of inverse problems and the appropriate mode of thought, chapter 1 offers historical vignettes, most of which have played an essential role in the development of natural science. These vignettes cover the first successful application of `non-destructive testing' by Archimedes (page 4) via Newton's laws of motion up to literary tomography, and readers will be able to enjoy a wide overview of inverse problems. Therefore, as the author asks, the reader should not skip this chapter. This may not be hard to do, since the headings of the sections are quite intriguing (`Archimedes' Bath', `Another World', `Got the Time?', `Head Games', etc). The author embarks on the technical approach to inverse problems in chapter 2. He has elegantly designed each section with a guide specifying course level, objective, mathematical and scientifical background and appropriate technology (e.g. types of calculators required). The guides are designed such that teachers may be able to construct effective and attractive courses by themselves. The book is not intended to offer one rigidly determined course, but should be used flexibly and independently according to the situation. Moreover, every section closes with activities which can be chosen according to the students' interests and levels of ability. Some of these exercises do not have ready solutions, but require long-term study, so readers are not required to solve all of them. After chapter 5, which contains discrete inverse problems such as the algebraic reconstruction technique and the Backus - Gilbert method, there are answers and commentaries to the activities. Finally, scripts in MATLAB are attached, although they can also be downloaded from the author's web page (http://math.uc.edu/~groetsch/). This book is aimed at students but it will be very valuable to researchers wishing to retain a wide overview of inverse problems in the midst of busy research activities. A Japanese version was published in 2002.

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.

Bayesian seismic tomography by parallel interacting Markov chains

NASA Astrophysics Data System (ADS)

Gesret, Alexandrine; Bottero, Alexis; Romary, Thomas; Noble, Mark; Desassis, Nicolas

2014-05-01

The velocity field estimated by first arrival traveltime tomography is commonly used as a starting point for further seismological, mineralogical, tectonic or similar analysis. In order to interpret quantitatively the results, the tomography uncertainty values as well as their spatial distribution are required. The estimated velocity model is obtained through inverse modeling by minimizing an objective function that compares observed and computed traveltimes. This step is often performed by gradient-based optimization algorithms. The major drawback of such local optimization schemes, beyond the possibility of being trapped in a local minimum, is that they do not account for the multiple possible solutions of the inverse problem. They are therefore unable to assess the uncertainties linked to the solution. Within a Bayesian (probabilistic) framework, solving the tomography inverse problem aims at estimating the posterior probability density function of velocity model using a global sampling algorithm. Markov chains Monte-Carlo (MCMC) methods are known to produce samples of virtually any distribution. In such a Bayesian inversion, the total number of simulations we can afford is highly related to the computational cost of the forward model. Although fast algorithms have been recently developed for computing first arrival traveltimes of seismic waves, the complete browsing of the posterior distribution of velocity model is hardly performed, especially when it is high dimensional and/or multimodal. In the latter case, the chain may even stay stuck in one of the modes. In order to improve the mixing properties of classical single MCMC, we propose to make interact several Markov chains at different temperatures. This method can make efficient use of large CPU clusters, without increasing the global computational cost with respect to classical MCMC and is therefore particularly suited for Bayesian inversion. The exchanges between the chains allow a precise sampling of the high probability zones of the model space while avoiding the chains to end stuck in a probability maximum. This approach supplies thus a robust way to analyze the tomography imaging uncertainties. The interacting MCMC approach is illustrated on two synthetic examples of tomography of calibration shots such as encountered in induced microseismic studies. On the second application, a wavelet based model parameterization is presented that allows to significantly reduce the dimension of the problem, making thus the algorithm efficient even for a complex velocity model.

Global Monthly CO2 Flux Inversion Based on Results of Terrestrial Ecosystem Modeling

NASA Astrophysics Data System (ADS)

Deng, F.; Chen, J.; Peters, W.; Krol, M.

2008-12-01

Most of our understanding of the sources and sinks of atmospheric CO2 has come from inverse studies of atmospheric CO2 concentration measurements. However, the number of currently available observation stations and our ability to simulate the diurnal planetary boundary layer evolution over continental regions essentially limit the number of regions that can be reliably inverted globally, especially over continental areas. In order to overcome these restrictions, a nested inverse modeling system was developed based on the Bayesian principle for estimating carbon fluxes of 30 regions in North America and 20 regions for the rest of the globe. Inverse modeling was conducted in monthly steps using CO2 concentration measurements of 5 years (2000 - 2005) with the following two models: (a) An atmospheric transport model (TM5) is used to generate the transport matrix where the diurnal variation n of atmospheric CO2 concentration is considered to enhance the use of the afternoon-hour average CO2 concentration measurements over the continental sites. (b) A process-based terrestrial ecosystem model (BEPS) is used to produce hourly step carbon fluxes, which could minimize the limitation due to our inability to solve the inverse problem in a high resolution, as the background of our inversion. We will present our recent results achieved through a combination of the bottom-up modeling with BEPS and the top-down modeling based on TM5 driven by offline meteorological fields generated by the European Centre for Medium Range Weather Forecast (ECMFW).

Bayesian Approach to the Joint Inversion of Gravity and Magnetic Data, with Application to the Ismenius Area of Mars

NASA Technical Reports Server (NTRS)

Jewell, Jeffrey B.; Raymond, C.; Smrekar, S.; Millbury, C.

2004-01-01

This viewgraph presentation reviews a Bayesian approach to the inversion of gravity and magnetic data with specific application to the Ismenius Area of Mars. Many inverse problems encountered in geophysics and planetary science are well known to be non-unique (i.e. inversion of gravity the density structure of a body). In hopes of reducing the non-uniqueness of solutions, there has been interest in the joint analysis of data. An example is the joint inversion of gravity and magnetic data, with the assumption that the same physical anomalies generate both the observed magnetic and gravitational anomalies. In this talk, we formulate the joint analysis of different types of data in a Bayesian framework and apply the formalism to the inference of the density and remanent magnetization structure for a local region in the Ismenius area of Mars. The Bayesian approach allows prior information or constraints in the solutions to be incorporated in the inversion, with the "best" solutions those whose forward predictions most closely match the data while remaining consistent with assumed constraints. The application of this framework to the inversion of gravity and magnetic data on Mars reveals two typical challenges - the forward predictions of the data have a linear dependence on some of the quantities of interest, and non-linear dependence on others (termed the "linear" and "non-linear" variables, respectively). For observations with Gaussian noise, a Bayesian approach to inversion for "linear" variables reduces to a linear filtering problem, with an explicitly computable "error" matrix. However, for models whose forward predictions have non-linear dependencies, inference is no longer given by such a simple linear problem, and moreover, the uncertainty in the solution is no longer completely specified by a computable "error matrix". It is therefore important to develop methods for sampling from the full Bayesian posterior to provide a complete and statistically consistent picture of model uncertainty, and what has been learned from observations. We will discuss advanced numerical techniques, including Monte Carlo Markov

Applied Mathematics in EM Studies with Special Emphasis on an Uncertainty Quantification and 3-D Integral Equation Modelling

NASA Astrophysics Data System (ADS)

Pankratov, Oleg; Kuvshinov, Alexey

2016-01-01

Despite impressive progress in the development and application of electromagnetic (EM) deterministic inverse schemes to map the 3-D distribution of electrical conductivity within the Earth, there is one question which remains poorly addressed—uncertainty quantification of the recovered conductivity models. Apparently, only an inversion based on a statistical approach provides a systematic framework to quantify such uncertainties. The Metropolis-Hastings (M-H) algorithm is the most popular technique for sampling the posterior probability distribution that describes the solution of the statistical inverse problem. However, all statistical inverse schemes require an enormous amount of forward simulations and thus appear to be extremely demanding computationally, if not prohibitive, if a 3-D set up is invoked. This urges development of fast and scalable 3-D modelling codes which can run large-scale 3-D models of practical interest for fractions of a second on high-performance multi-core platforms. But, even with these codes, the challenge for M-H methods is to construct proposal functions that simultaneously provide a good approximation of the target density function while being inexpensive to be sampled. In this paper we address both of these issues. First we introduce a variant of the M-H method which uses information about the local gradient and Hessian of the penalty function. This, in particular, allows us to exploit adjoint-based machinery that has been instrumental for the fast solution of deterministic inverse problems. We explain why this modification of M-H significantly accelerates sampling of the posterior probability distribution. In addition we show how Hessian handling (inverse, square root) can be made practicable by a low-rank approximation using the Lanczos algorithm. Ultimately we discuss uncertainty analysis based on stochastic inversion results. In addition, we demonstrate how this analysis can be performed within a deterministic approach. In the second part, we summarize modern trends in the development of efficient 3-D EM forward modelling schemes with special emphasis on recent advances in the integral equation approach.

Electromagnetic Inverse Methods and Applications for Inhomogeneous Media Probing and Synthesis.

NASA Astrophysics Data System (ADS)

Xia, Jake Jiqing

The electromagnetic inverse scattering problems concerned in this thesis are to find unknown inhomogeneous permittivity and conductivity profiles in a medium from the scattering data. Both analytical and numerical methods are studied in the thesis. The inverse methods can be applied to geophysical medium probing, non-destructive testing, medical imaging, optical waveguide synthesis and material characterization. An introduction is given in Chapter 1. The first part of the thesis presents inhomogeneous media probing. The Riccati equation approach is discussed in Chapter 2 for a one-dimensional planar profile inversion problem. Two types of the Riccati equations are derived and distinguished. New renormalized formulae based inverting one specific type of the Riccati equation are derived. Relations between the inverse methods of Green's function, the Riccati equation and the Gel'fand-Levitan-Marchenko (GLM) theory are studied. In Chapter 3, the renormalized source-type integral equation (STIE) approach is formulated for inversion of cylindrically inhomogeneous permittivity and conductivity profiles. The advantages of the renormalized STIE approach are demonstrated in numerical examples. The cylindrical profile inversion problem has an application for borehole inversion. In Chapter 4 the renormalized STIE approach is extended to a planar case where the two background media are different. Numerical results have shown fast convergence. This formulation is applied to inversion of the underground soil moisture profiles in remote sensing. The second part of the thesis presents the synthesis problem of inhomogeneous dielectric waveguides using the electromagnetic inverse methods. As a particular example, the rational function representation of reflection coefficients in the GLM theory is used. The GLM method is reviewed in Chapter 5. Relations between modal structures and transverse reflection coefficients of an inhomogeneous medium are established in Chapter 6. A stratified medium model is used to derive the guidance condition and the reflection coefficient. Results obtained in Chapter 6 provide the physical foundation for applying the inverse methods for the waveguide design problem. In Chapter 7, a global guidance condition for continuously varying medium is derived using the Riccati equation. It is further shown that the discrete modes in an inhomogeneous medium have the same wave vectors as the poles of the transverse reflection coefficient. An example of synthesizing an inhomogeneous dielectric waveguide using a rational reflection coefficient is presented. A summary of the thesis is given in Chapter 8. (Copies available exclusively from MIT Libraries, Rm. 14-0551, Cambridge, MA 02139-4307. Ph. 617-253-5668; Fax 617-253-1690.).

An inverse problem for a mathematical model of aquaponic agriculture

NASA Astrophysics Data System (ADS)

Bobak, Carly; Kunze, Herb

2017-01-01

Aquaponic agriculture is a sustainable ecosystem that relies on a symbiotic relationship between fish and macrophytes. While the practice has been growing in popularity, relatively little mathematical models exist which aim to study the system processes. In this paper, we present a system of ODEs which aims to mathematically model the population and concetrations dynamics present in an aquaponic environment. Values of the parameters in the system are estimated from the literature so that simulated results can be presented to illustrate the nature of the solutions to the system. As well, a brief sensitivity analysis is performed in order to identify redundant parameters and highlight those which may need more reliable estimates. Specifically, an inverse problem with manufactured data for fish and plants is presented to demonstrate the ability of the collage theorem to recover parameter estimates.

Models for determining the geometrical properties of halo coronal mass ejections

NASA Astrophysics Data System (ADS)

Zhao, X.; Liu, Y.

2005-12-01

To this day, the prediction of space weather effects near the Earth suffer from a fundamental problem: the necessary condition for determining whether or not and when a part of the huge interplanetary counterpart (ICME) of frontside halo coronal mass ejections (CMEs) is able to hit the Earth and generate goemagnetic storms, i.e., the real angular width, the propagation direction and speed of the CMEs, cannot be measured directly because of the unfavorable geometry. To inverse these geometrical and kinematical properties we have recently developed a few geometrical models, such as the cone model, the ice cream cone model, and the spherical cone model. The inversing solution of the cone model for the 12 may 1997 halo CME has been used as an input to the ENLIL model (a 3D MHD solar wind code) and successfully predicted the ICME near the Earth (Zhao, Plukett & Liu, 2002; Odstrcil, Riley & Zhao, 2004). After briefly describing the geometrical models this presentation will discuss: 1. What kind of halo CMEs can be inversed? 2. How to select the geometrical models given a specific halo CME? 3. Whether or not the inversing solution is unique?

On the possibility of control restoration in some inverse problems of heat and mass transfer

NASA Astrophysics Data System (ADS)

Bilchenko, G. G.; Bilchenko, N. G.

2016-11-01

The hypersonic aircraft permeable surfaces effective heat protection problems are considered. The physic-chemical processes (the dissociation and the ionization) in laminar boundary layer of compressible gas are appreciated in mathematical model. The statements of direct problems of heat and mass transfer are given: according to preset given controls it is necessary to compute the boundary layer mathematical model parameters and determinate the local and total heat flows and friction forces and the power of blowing system. The A.A.Dorodnicyn's generalized integral relations method has been used as calculation basis. The optimal control - the blowing into boundary layer (for continuous functions) was constructed as the solution of direct problem in extreme statement with the use of this approach. The statement of inverse problems are given: the control laws ensuring the preset given local heat flow and local tangent friction are restored. The differences between the interpolation and the approximation statements are discussed. The possibility of unique control restoration is established and proved (in the stagnation point). The computational experiments results are presented.

Geostatistical regularization of inverse models for the retrieval of vegetation biophysical variables

NASA Astrophysics Data System (ADS)

Atzberger, C.; Richter, K.

2009-09-01

The robust and accurate retrieval of vegetation biophysical variables using radiative transfer models (RTM) is seriously hampered by the ill-posedness of the inverse problem. With this research we further develop our previously published (object-based) inversion approach [Atzberger (2004)]. The object-based RTM inversion takes advantage of the geostatistical fact that the biophysical characteristics of nearby pixel are generally more similar than those at a larger distance. A two-step inversion based on PROSPECT+SAIL generated look-up-tables is presented that can be easily implemented and adapted to other radiative transfer models. The approach takes into account the spectral signatures of neighboring pixel and optimizes a common value of the average leaf angle (ALA) for all pixel of a given image object, such as an agricultural field. Using a large set of leaf area index (LAI) measurements (n = 58) acquired over six different crops of the Barrax test site, Spain), we demonstrate that the proposed geostatistical regularization yields in most cases more accurate and spatially consistent results compared to the traditional (pixel-based) inversion. Pros and cons of the approach are discussed and possible future extensions presented.

Towards a new technique to construct a 3D shear-wave velocity model based on converted waves

NASA Astrophysics Data System (ADS)

Hetényi, G.; Colavitti, L.

2017-12-01

A 3D model is essential in all branches of solid Earth sciences because geological structures can be heterogeneous and change significantly in their lateral dimension. The main target of this research is to build a crustal S-wave velocity structure in 3D. The currently popular methodologies to construct 3D shear-wave velocity models are Ambient Noise Tomography (ANT) and Local Earthquake Tomography (LET). Here we propose a new technique to map Earth discontinuities and velocities at depth based on the analysis of receiver functions. The 3D model is obtained by simultaneously inverting P-to-S converted waveforms recorded at a dense array. The individual velocity models corresponding to each trace are extracted from the 3D initial model along ray paths that are calculated using the shooting method, and the velocity model is updated during the inversion. We consider a spherical approximation of ray propagation using a global velocity model (iasp91, Kennett and Engdahl, 1991) for the teleseismic part, while we adopt Cartesian coordinates and a local velocity model for the crust. During the inversion process we work with a multi-layer crustal model for shear-wave velocity, with a flexible mesh for the depth of the interfaces. The RFs inversion represents a complex problem because the amplitude and the arrival time of different phases depend in a non-linear way on the depth of interfaces and the characteristics of the velocity structure. The solution we envisage to manage the inversion problem is the stochastic Neighbourhood Algorithm (NA, Sambridge, 1999), whose goal is to find an ensemble of models that sample the good data-fitting regions of a multidimensional parameter space. Depending on the studied area, this method can accommodate possible independent and complementary geophysical data (gravity, active seismics, LET, ANT, etc.), helping to reduce the non-linearity of the inversion. Our first focus of application is the Central Alps, where a 20-year long dataset of high-quality teleseismic events recorded at 81 stations is available, and we have high-resolution P-wave velocity model available (Diehl et al., 2009). We plan to extend the 3D shear-wave velocity inversion method to the entire Alpine domain in frame of the AlpArray project, and apply it to other areas with a dense network of broadband seismometers.

Improved real-time dynamics from imaginary frequency lattice simulations

NASA Astrophysics Data System (ADS)

Pawlowski, Jan M.; Rothkopf, Alexander

2018-03-01

The computation of real-time properties, such as transport coefficients or bound state spectra of strongly interacting quantum fields in thermal equilibrium is a pressing matter. Since the sign problem prevents a direct evaluation of these quantities, lattice data needs to be analytically continued from the Euclidean domain of the simulation to Minkowski time, in general an ill-posed inverse problem. Here we report on a novel approach to improve the determination of real-time information in the form of spectral functions by setting up a simulation prescription in imaginary frequencies. By carefully distinguishing between initial conditions and quantum dynamics one obtains access to correlation functions also outside the conventional Matsubara frequencies. In particular the range between ω0 and ω1 = 2πT, which is most relevant for the inverse problem may be more highly resolved. In combination with the fact that in imaginary frequencies the kernel of the inverse problem is not an exponential but only a rational function we observe significant improvements in the reconstruction of spectral functions, demonstrated in a simple 0+1 dimensional scalar field theory toy model.

Solving iTOUGH2 simulation and optimization problems using the PEST protocol

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

Finsterle, S.A.; Zhang, Y.

2011-02-01

The PEST protocol has been implemented into the iTOUGH2 code, allowing the user to link any simulation program (with ASCII-based inputs and outputs) to iTOUGH2's sensitivity analysis, inverse modeling, and uncertainty quantification capabilities. These application models can be pre- or post-processors of the TOUGH2 non-isothermal multiphase flow and transport simulator, or programs that are unrelated to the TOUGH suite of codes. PEST-style template and instruction files are used, respectively, to pass input parameters updated by the iTOUGH2 optimization routines to the model, and to retrieve the model-calculated values that correspond to observable variables. We summarize the iTOUGH2 capabilities and demonstratemore » the flexibility added by the PEST protocol for the solution of a variety of simulation-optimization problems. In particular, the combination of loosely coupled and tightly integrated simulation and optimization routines provides both the flexibility and control needed to solve challenging inversion problems for the analysis of multiphase subsurface flow and transport systems.« less

Physics-based Inverse Problem to Deduce Marine Atmospheric Boundary Layer Parameters

DTIC Science & Technology

2017-03-07

please find the Final Technical Report with SF 298 for Dr. Erin E. Hackett’s ONR grant entitled Physics-based Inverse Problem to Deduce Marine...From- To) 07/03/2017 Final Technica l Dec 2012- Dec 2016 4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER Physics-based Inverse Problem to Deduce Marine...SUPPLEMENTARY NOTES 14. ABSTRACT This report describes research results related to the development and implementation of an inverse problem approach for

Joint two dimensional inversion of gravity and magnetotelluric data using correspondence maps

NASA Astrophysics Data System (ADS)

Carrillo Lopez, J.; Gallardo, L. A.

2016-12-01

Inverse problems in Earth sciences are inherently non-unique. To improve models and reduce the number of solutions we need to provide extra information. In geological context, this information could be a priori information, for example, geological information, well log data, smoothness, or actually, information of measures of different kind of data. Joint inversion provides an approach to improve the solution and reduce the errors due to suppositions of each method. To do that, we need a link between two or more models. Some approaches have been explored successfully in recent years. For example, Gallardo and Meju (2003), Gallardo and Meju (2004, 2011), and Gallardo et. al. (2012) used the directions of properties to measure the similarity between models minimizing their cross gradients. In this work, we proposed a joint iterative inversion method that use spatial distribution of properties as a link. Correspondence maps could be better characterizing specific Earth systems due they consider the relation between properties. We implemented a code in Fortran to do a two dimensional inversion of magnetotelluric and gravity data, which are two of the standard methods in geophysical exploration. Synthetic tests show the advantages of joint inversion using correspondence maps against separate inversion. Finally, we applied this technique to magnetotelluric and gravity data in the geothermal zone located in Cerro Prieto, México.

Invisibility problem in acoustics, electromagnetism and heat transfer. Inverse design method

NASA Astrophysics Data System (ADS)

Alekseev, G.; Tokhtina, A.; Soboleva, O.

2017-10-01

Two approaches (direct design and inverse design methods) for solving problems of designing devices providing invisibility of material bodies of detection using different physical fields - electromagnetic, acoustic and static are discussed. The second method is applied for solving problems of designing cloaking devices for the 3D stationary thermal scattering model. Based on this method the design problems under study are reduced to respective control problems. The material parameters (radial and tangential heat conductivities) of the inhomogeneous anisotropic medium filling the thermal cloak and the density of auxiliary heat sources play the role of controls. A unique solvability of direct thermal scattering problem in the Sobolev space is proved and the new estimates of solutions are established. Using these results, the solvability of control problem is proved and the optimality system is derived. Based on analysis of optimality system, the stability estimates of optimal solutions are established and numerical algorithms for solving particular thermal cloaking problem are proposed.

Sub-basalt Imaging of Hydrocarbon-Bearing Mesozoic Sediments Using Ray-Trace Inversion of First-Arrival Seismic Data and Elastic Finite-Difference Full-Wave Modeling Along Sinor-Valod Profile of Deccan Syneclise, India

NASA Astrophysics Data System (ADS)

Talukdar, Karabi; Behera, Laxmidhar

2018-03-01

Imaging below the basalt for hydrocarbon exploration is a global problem because of poor penetration and significant loss of seismic energy due to scattering, attenuation, absorption and mode-conversion when the seismic waves encounter a highly heterogeneous and rugose basalt layer. The conventional (short offset) seismic data acquisition, processing and modeling techniques adopted by the oil industry generally fails to image hydrocarbon-bearing sub-trappean Mesozoic sediments hidden below the basalt and is considered as a serious problem for hydrocarbon exploration in the world. To overcome this difficulty of sub-basalt imaging, we have generated dense synthetic seismic data with the help of elastic finite-difference full-wave modeling using staggered-grid scheme for the model derived from ray-trace inversion using sparse wide-angle seismic data acquired along Sinor-Valod profile in the Deccan Volcanic Province of India. The full-wave synthetic seismic data generated have been processed and imaged using conventional seismic data processing technique with Kirchhoff pre-stack time and depth migrations. The seismic image obtained correlates with all the structural features of the model obtained through ray-trace inversion of wide-angle seismic data, validating the effectiveness of robust elastic finite-difference full-wave modeling approach for imaging below thick basalts. Using the full-wave modeling also allows us to decipher small-scale heterogeneities imposed in the model as a measure of the rugose basalt interfaces, which could not be dealt with ray-trace inversion. Furthermore, we were able to accurately image thin low-velocity hydrocarbon-bearing Mesozoic sediments sandwiched between and hidden below two thick sequences of high-velocity basalt layers lying above the basement.

An interval model updating strategy using interval response surface models

NASA Astrophysics Data System (ADS)

Fang, Sheng-En; Zhang, Qiu-Hu; Ren, Wei-Xin

2015-08-01

Stochastic model updating provides an effective way of handling uncertainties existing in real-world structures. In general, probabilistic theories, fuzzy mathematics or interval analyses are involved in the solution of inverse problems. However in practice, probability distributions or membership functions of structural parameters are often unavailable due to insufficient information of a structure. At this moment an interval model updating procedure shows its superiority in the aspect of problem simplification since only the upper and lower bounds of parameters and responses are sought. To this end, this study develops a new concept of interval response surface models for the purpose of efficiently implementing the interval model updating procedure. The frequent interval overestimation due to the use of interval arithmetic can be maximally avoided leading to accurate estimation of parameter intervals. Meanwhile, the establishment of an interval inverse problem is highly simplified, accompanied by a saving of computational costs. By this means a relatively simple and cost-efficient interval updating process can be achieved. Lastly, the feasibility and reliability of the developed method have been verified against a numerical mass-spring system and also against a set of experimentally tested steel plates.

Inverse scattering approach to improving pattern recognition

NASA Astrophysics Data System (ADS)

Chapline, George; Fu, Chi-Yung

2005-05-01

The Helmholtz machine provides what may be the best existing model for how the mammalian brain recognizes patterns. Based on the observation that the "wake-sleep" algorithm for training a Helmholtz machine is similar to the problem of finding the potential for a multi-channel Schrodinger equation, we propose that the construction of a Schrodinger potential using inverse scattering methods can serve as a model for how the mammalian brain learns to extract essential information from sensory data. In particular, inverse scattering theory provides a conceptual framework for imagining how one might use EEG and MEG observations of brain-waves together with sensory feedback to improve human learning and pattern recognition. Longer term, implementation of inverse scattering algorithms on a digital or optical computer could be a step towards mimicking the seamless information fusion of the mammalian brain.

Inverse Scattering Approach to Improving Pattern Recognition

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

Chapline, G; Fu, C

2005-02-15

The Helmholtz machine provides what may be the best existing model for how the mammalian brain recognizes patterns. Based on the observation that the ''wake-sleep'' algorithm for training a Helmholtz machine is similar to the problem of finding the potential for a multi-channel Schrodinger equation, we propose that the construction of a Schrodinger potential using inverse scattering methods can serve as a model for how the mammalian brain learns to extract essential information from sensory data. In particular, inverse scattering theory provides a conceptual framework for imagining how one might use EEG and MEG observations of brain-waves together with sensorymore » feedback to improve human learning and pattern recognition. Longer term, implementation of inverse scattering algorithms on a digital or optical computer could be a step towards mimicking the seamless information fusion of the mammalian brain.« less

Reconstructing source terms from atmospheric concentration measurements: Optimality analysis of an inversion technique

NASA Astrophysics Data System (ADS)

Turbelin, Grégory; Singh, Sarvesh Kumar; Issartel, Jean-Pierre

2014-12-01

In the event of an accidental or intentional contaminant release in the atmosphere, it is imperative, for managing emergency response, to diagnose the release parameters of the source from measured data. Reconstruction of the source information exploiting measured data is called an inverse problem. To solve such a problem, several techniques are currently being developed. The first part of this paper provides a detailed description of one of them, known as the renormalization method. This technique, proposed by Issartel (2005), has been derived using an approach different from that of standard inversion methods and gives a linear solution to the continuous Source Term Estimation (STE) problem. In the second part of this paper, the discrete counterpart of this method is presented. By using matrix notation, common in data assimilation and suitable for numerical computing, it is shown that the discrete renormalized solution belongs to a family of well-known inverse solutions (minimum weighted norm solutions), which can be computed by using the concept of generalized inverse operator. It is shown that, when the weight matrix satisfies the renormalization condition, this operator satisfies the criteria used in geophysics to define good inverses. Notably, by means of the Model Resolution Matrix (MRM) formalism, we demonstrate that the renormalized solution fulfils optimal properties for the localization of single point sources. Throughout the article, the main concepts are illustrated with data from a wind tunnel experiment conducted at the Environmental Flow Research Centre at the University of Surrey, UK.

Parts-based geophysical inversion with application to water flooding interface detection and geological facies detection

NASA Astrophysics Data System (ADS)

Zhang, Junwei

I built parts-based and manifold based mathematical learning model for the geophysical inverse problem and I applied this approach to two problems. One is related to the detection of the oil-water encroachment front during the water flooding of an oil reservoir. In this application, I propose a new 4D inversion approach based on the Gauss-Newton approach to invert time-lapse cross-well resistance data. The goal of this study is to image the position of the oil-water encroachment front in a heterogeneous clayey sand reservoir. This approach is based on explicitly connecting the change of resistivity to the petrophysical properties controlling the position of the front (porosity and permeability) and to the saturation of the water phase through a petrophysical resistivity model accounting for bulk and surface conductivity contributions and saturation. The distributions of the permeability and porosity are also inverted using the time-lapse resistivity data in order to better reconstruct the position of the oil water encroachment front. In our synthetic test case, we get a better position of the front with the by-products of porosity and permeability inferences near the flow trajectory and close to the wells. The numerical simulations show that the position of the front is recovered well but the distribution of the recovered porosity and permeability is only fair. A comparison with a commercial code based on a classical Gauss-Newton approach with no information provided by the two-phase flow model fails to recover the position of the front. The new approach could be also used for the time-lapse monitoring of various processes in both geothermal fields and oil and gas reservoirs using a combination of geophysical methods. A paper has been published in Geophysical Journal International on this topic and I am the first author of this paper. The second application is related to the detection of geological facies boundaries and their deforation to satisfy to geophysica data and prior distributions. We pose the geophysical inverse problem in terms of Gaussian random fields with mean functions controlled by petrophysical relationships and covariance functions controlled by a prior geological cross-section, including the definition of spatial boundaries for the geological facies. The petrophysical relationship problem is formulated as a regression problem upon each facies. The inversion is performed in a Bayesian framework. We demonstrate the usefulness of this strategy using a first synthetic case study, performing a joint inversion of gravity and galvanometric resistivity data with the stations all located at the ground surface. The joint inversion is used to recover the density and resistivity distributions of the subsurface. In a second step, we consider the possibility that the facies boundaries are deformable and their shapes are inverted as well. We use the level set approach to deform the facies boundaries preserving prior topological properties of the facies throughout the inversion. With the additional help of prior facies petrophysical relationships, topological characteristic of each facies, we make posterior inference about multiple geophysical tomograms based on their corresponding geophysical data misfits. The result of the inversion technique is encouraging when applied to a second synthetic case study, showing that we can recover the heterogeneities inside the facies, the mean values for the petrophysical properties, and, to some extent, the facies boundaries. A paper has been submitted to Geophysics on this topic and I am the first author of this paper. During this thesis, I also worked on the time lapse inversion problem of gravity data in collaboration with Marios Karaoulis and a paper was published in Geophysical Journal international on this topic. I also worked on the time-lapse inversion of cross-well geophysical data (seismic and resistivity) using both a structural approach named the cross-gradient approach and a petrophysical approach. A paper was published in Geophysics on this topic.

Efficient Storage Scheme of Covariance Matrix during Inverse Modeling

NASA Astrophysics Data System (ADS)

Mao, D.; Yeh, T. J.

2013-12-01

During stochastic inverse modeling, the covariance matrix of geostatistical based methods carries the information about the geologic structure. Its update during iterations reflects the decrease of uncertainty with the incorporation of observed data. For large scale problem, its storage and update cost too much memory and computational resources. In this study, we propose a new efficient storage scheme for storage and update. Compressed Sparse Column (CSC) format is utilized to storage the covariance matrix, and users can assign how many data they prefer to store based on correlation scales since the data beyond several correlation scales are usually not very informative for inverse modeling. After every iteration, only the diagonal terms of the covariance matrix are updated. The off diagonal terms are calculated and updated based on shortened correlation scales with a pre-assigned exponential model. The correlation scales are shortened by a coefficient, i.e. 0.95, every iteration to show the decrease of uncertainty. There is no universal coefficient for all the problems and users are encouraged to try several times. This new scheme is tested with 1D examples first. The estimated results and uncertainty are compared with the traditional full storage method. In the end, a large scale numerical model is utilized to validate this new scheme.

Cybernetic group method of data handling (GMDH) statistical learning for hyperspectral remote sensing inverse problems in coastal ocean optics

NASA Astrophysics Data System (ADS)

Filippi, Anthony Matthew

For complex systems, sufficient a priori knowledge is often lacking about the mathematical or empirical relationship between cause and effect or between inputs and outputs of a given system. Automated machine learning may offer a useful solution in such cases. Coastal marine optical environments represent such a case, as the optical remote sensing inverse problem remains largely unsolved. A self-organizing, cybernetic mathematical modeling approach known as the group method of data handling (GMDH), a type of statistical learning network (SLN), was used to generate explicit spectral inversion models for optically shallow coastal waters. Optically shallow water light fields represent a particularly difficult challenge in oceanographic remote sensing. Several algorithm-input data treatment combinations were utilized in multiple experiments to automatically generate inverse solutions for various inherent optical property (IOP), bottom optical property (BOP), constituent concentration, and bottom depth estimations. The objective was to identify the optimal remote-sensing reflectance Rrs(lambda) inversion algorithm. The GMDH also has the potential of inductive discovery of physical hydro-optical laws. Simulated data were used to develop generalized, quasi-universal relationships. The Hydrolight numerical forward model, based on radiative transfer theory, was used to compute simulated above-water remote-sensing reflectance Rrs(lambda) psuedodata, matching the spectral channels and resolution of the experimental Naval Research Laboratory Ocean PHILLS (Portable Hyperspectral Imager for Low-Light Spectroscopy) sensor. The input-output pairs were for GMDH and artificial neural network (ANN) model development, the latter of which was used as a baseline, or control, algorithm. Both types of models were applied to in situ and aircraft data. Also, in situ spectroradiometer-derived Rrs(lambda) were used as input to an optimization-based inversion procedure. Target variables included bottom depth z b, chlorophyll a concentration [chl- a], spectral bottom irradiance reflectance Rb(lambda), and spectral total absorption a(lambda) and spectral total backscattering bb(lambda) coefficients. When applying the cybernetic and neural models to in situ HyperTSRB-derived Rrs, the difference in the means of the absolute error of the inversion estimates for zb was significant (alpha = 0.05). GMDH yielded significantly better zb than the ANN. The ANN model posted a mean absolute error (MAE) of 0.62214 m, compared with 0.55161 m for GMDH.

Joint inversion of geophysical data using petrophysical clustering and facies deformation wth the level set technique

NASA Astrophysics Data System (ADS)

Revil, A.

2015-12-01

Geological expertise and petrophysical relationships can be brought together to provide prior information while inverting multiple geophysical datasets. The merging of such information can result in more realistic solution in the distribution of the model parameters, reducing ipse facto the non-uniqueness of the inverse problem. We consider two level of heterogeneities: facies, described by facies boundaries and heteroegenities inside each facies determined by a correlogram. In this presentation, we pose the geophysical inverse problem in terms of Gaussian random fields with mean functions controlled by petrophysical relationships and covariance functions controlled by a prior geological cross-section, including the definition of spatial boundaries for the geological facies. The petrophysical relationship problem is formulated as a regression problem upon each facies. The inversion of the geophysical data is performed in a Bayesian framework. We demonstrate the usefulness of this strategy using a first synthetic case for which we perform a joint inversion of gravity and galvanometric resistivity data with the stations located at the ground surface. The joint inversion is used to recover the density and resistivity distributions of the subsurface. In a second step, we consider the possibility that the facies boundaries are deformable and their shapes are inverted as well. We use the level set approach to perform such deformation preserving prior topological properties of the facies throughout the inversion. With the help of prior facies petrophysical relationships and topological characteristic of each facies, we make posterior inference about multiple geophysical tomograms based on their corresponding geophysical data misfits. The method is applied to a second synthetic case showing that we can recover the heterogeneities inside the facies, the mean values for the petrophysical properties, and, to some extent, the facies boundaries using the 2D joint inversion of gravity and galvanometric resistivity data. For this 2D synthetic example, we note that the position of the facies are well-recovered except far from the ground surfce where the sensitivity is too low. The figure shows the evolution of the shape of the facies during the inversion itertion by iteration.

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

Fully Nonlinear Modeling and Analysis of Precision Membranes

NASA Technical Reports Server (NTRS)

Pai, P. Frank; Young, Leyland G.

2003-01-01

High precision membranes are used in many current space applications. This paper presents a fully nonlinear membrane theory with forward and inverse analyses of high precision membrane structures. The fully nonlinear membrane theory is derived from Jaumann strains and stresses, exact coordinate transformations, the concept of local relative displacements, and orthogonal virtual rotations. In this theory, energy and Newtonian formulations are fully correlated, and every structural term can be interpreted in terms of vectors. Fully nonlinear ordinary differential equations (ODES) governing the large static deformations of known axisymmetric membranes under known axisymmetric loading (i.e., forward problems) are presented as first-order ODES, and a method for obtaining numerically exact solutions using the multiple shooting procedure is shown. A method for obtaining the undeformed geometry of any axisymmetric membrane with a known inflated geometry and a known internal pressure (i.e., inverse problems) is also derived. Numerical results from forward analysis are verified using results in the literature, and results from inverse analysis are verified using known exact solutions and solutions from the forward analysis. Results show that the membrane theory and the proposed numerical methods for solving nonlinear forward and inverse membrane problems are accurate.

Minimization of model representativity errors in identification of point source emission from atmospheric concentration measurements

NASA Astrophysics Data System (ADS)

Sharan, Maithili; Singh, Amit Kumar; Singh, Sarvesh Kumar

2017-11-01

Estimation of an unknown atmospheric release from a finite set of concentration measurements is considered an ill-posed inverse problem. Besides ill-posedness, the estimation process is influenced by the instrumental errors in the measured concentrations and model representativity errors. The study highlights the effect of minimizing model representativity errors on the source estimation. This is described in an adjoint modelling framework and followed in three steps. First, an estimation of point source parameters (location and intensity) is carried out using an inversion technique. Second, a linear regression relationship is established between the measured concentrations and corresponding predicted using the retrieved source parameters. Third, this relationship is utilized to modify the adjoint functions. Further, source estimation is carried out using these modified adjoint functions to analyse the effect of such modifications. The process is tested for two well known inversion techniques, called renormalization and least-square. The proposed methodology and inversion techniques are evaluated for a real scenario by using concentrations measurements from the Idaho diffusion experiment in low wind stable conditions. With both the inversion techniques, a significant improvement is observed in the retrieval of source estimation after minimizing the representativity errors.

The Prediction Properties of Inverse and Reverse Regression for the Simple Linear Calibration Problem

NASA Technical Reports Server (NTRS)

Parker, Peter A.; Geoffrey, Vining G.; Wilson, Sara R.; Szarka, John L., III; Johnson, Nels G.

2010-01-01

The calibration of measurement systems is a fundamental but under-studied problem within industrial statistics. The origins of this problem go back to basic chemical analysis based on NIST standards. In today's world these issues extend to mechanical, electrical, and materials engineering. Often, these new scenarios do not provide "gold standards" such as the standard weights provided by NIST. This paper considers the classic "forward regression followed by inverse regression" approach. In this approach the initial experiment treats the "standards" as the regressor and the observed values as the response to calibrate the instrument. The analyst then must invert the resulting regression model in order to use the instrument to make actual measurements in practice. This paper compares this classical approach to "reverse regression," which treats the standards as the response and the observed measurements as the regressor in the calibration experiment. Such an approach is intuitively appealing because it avoids the need for the inverse regression. However, it also violates some of the basic regression assumptions.

An Inverse Neural Controller Based on the Applicability Domain of RBF Network Models

PubMed Central

Alexandridis, Alex; Stogiannos, Marios; Papaioannou, Nikolaos; Zois, Elias; Sarimveis, Haralambos

2018-01-01

This paper presents a novel methodology of generic nature for controlling nonlinear systems, using inverse radial basis function neural network models, which may combine diverse data originating from various sources. The algorithm starts by applying the particle swarm optimization-based non-symmetric variant of the fuzzy means (PSO-NSFM) algorithm so that an approximation of the inverse system dynamics is obtained. PSO-NSFM offers models of high accuracy combined with small network structures. Next, the applicability domain concept is suitably tailored and embedded into the proposed control structure in order to ensure that extrapolation is avoided in the controller predictions. Finally, an error correction term, estimating the error produced by the unmodeled dynamics and/or unmeasured external disturbances, is included to the control scheme to increase robustness. The resulting controller guarantees bounded input-bounded state (BIBS) stability for the closed loop system when the open loop system is BIBS stable. The proposed methodology is evaluated on two different control problems, namely, the control of an experimental armature-controlled direct current (DC) motor and the stabilization of a highly nonlinear simulated inverted pendulum. For each one of these problems, appropriate case studies are tested, in which a conventional neural controller employing inverse models and a PID controller are also applied. The results reveal the ability of the proposed control scheme to handle and manipulate diverse data through a data fusion approach and illustrate the superiority of the method in terms of faster and less oscillatory responses. PMID:29361781

Accounting for uncertain fault geometry in earthquake source inversions - I: theory and simplified application

NASA Astrophysics Data System (ADS)

Ragon, Théa; Sladen, Anthony; Simons, Mark

2018-05-01

The ill-posed nature of earthquake source estimation derives from several factors including the quality and quantity of available observations and the fidelity of our forward theory. Observational errors are usually accounted for in the inversion process. Epistemic errors, which stem from our simplified description of the forward problem, are rarely dealt with despite their potential to bias the estimate of a source model. In this study, we explore the impact of uncertainties related to the choice of a fault geometry in source inversion problems. The geometry of a fault structure is generally reduced to a set of parameters, such as position, strike and dip, for one or a few planar fault segments. While some of these parameters can be solved for, more often they are fixed to an uncertain value. We propose a practical framework to address this limitation by following a previously implemented method exploring the impact of uncertainties on the elastic properties of our models. We develop a sensitivity analysis to small perturbations of fault dip and position. The uncertainties in fault geometry are included in the inverse problem under the formulation of the misfit covariance matrix that combines both prediction and observation uncertainties. We validate this approach with the simplified case of a fault that extends infinitely along strike, using both Bayesian and optimization formulations of a static inversion. If epistemic errors are ignored, predictions are overconfident in the data and source parameters are not reliably estimated. In contrast, inclusion of uncertainties in fault geometry allows us to infer a robust posterior source model. Epistemic uncertainties can be many orders of magnitude larger than observational errors for great earthquakes (Mw > 8). Not accounting for uncertainties in fault geometry may partly explain observed shallow slip deficits for continental earthquakes. Similarly, ignoring the impact of epistemic errors can also bias estimates of near surface slip and predictions of tsunamis induced by megathrust earthquakes. (Mw > 8)

A genetic algorithm approach to estimate glacier mass variations from GRACE data

NASA Astrophysics Data System (ADS)

Reimond, Stefan; Klinger, Beate; Krauss, Sandro; Mayer-Gürr, Torsten

2017-04-01

The application of a genetic algorithm (GA) to the inference of glacier mass variations with a point-mass modeling method is described. GRACE K-band ranging data (available since April 2002) processed at the Graz University of Technology serve as input for this study. The reformulation of the point-mass inversion method in terms of an optimization problem is motivated by two reasons: first, an improved choice of the positions of the modeled point-masses (with a particular focus on the depth parameter) is expected to increase the signal-to-noise ratio. Considering these coordinates as additional unknown parameters (besides from the mass change magnitudes) results in a highly non-linear optimization problem. The second reason is that the mass inversion from satellite tracking data is an ill-posed problem, and hence regularization becomes necessary. The main task in this context is the determination of the regularization parameter, which is typically done by means of heuristic selection rules like, e.g., the L-curve criterion. In this study, however, the challenge of selecting a suitable balancing parameter (or even a matrix) is tackled by introducing regularization to the overall optimization problem. Based on this novel approach, estimations of ice-mass changes in various alpine glacier systems (e.g. Svalbard) are presented and compared to existing results and alternative inversion methods.

Minimizing EIT image artefacts from mesh variability in finite element models.

PubMed

Adler, Andy; Lionheart, William R B

2011-07-01

Electrical impedance tomography (EIT) solves an inverse problem to estimate the conductivity distribution within a body from electrical simulation and measurements at the body surface, where the inverse problem is based on a solution of Laplace's equation in the body. Most commonly, a finite element model (FEM) is used, largely because of its ability to describe irregular body shapes. In this paper, we show that simulated variations in the positions of internal nodes within a FEM can result in serious image artefacts in the reconstructed images. Such variations occur when designing FEM meshes to conform to conductivity targets, but the effects may also be seen in other applications of absolute and difference EIT. We explore the hypothesis that these artefacts result from changes in the projection of the anisotropic conductivity tensor onto the FEM system matrix, which introduces anisotropic components into the simulated voltages, which cannot be reconstructed onto an isotropic image, and appear as artefacts. The magnitude of the anisotropic effect is analysed for a small regular FEM, and shown to be proportional to the relative node movement as a fraction of element size. In order to address this problem, we show that it is possible to incorporate a FEM node movement component into the formulation of the inverse problem. These results suggest that it is important to consider artefacts due to FEM mesh geometry in EIT image reconstruction.

Joint inversion of hydraulic head and self-potential data associated with harmonic pumping tests

NASA Astrophysics Data System (ADS)

Soueid Ahmed, A.; Jardani, A.; Revil, A.; Dupont, J. P.

2016-09-01

Harmonic pumping tests consist in stimulating an aquifer by the means of hydraulic stimulations at some discrete frequencies. The inverse problem consisting in retrieving the hydraulic properties is inherently ill posed and is usually underdetermined when considering the number of well head data available in field conditions. To better constrain this inverse problem, we add self-potential data recorded at the ground surface to the head data. The self-potential method is a passive geophysical method. Its signals are generated by the groundwater flow through an electrokinetic coupling. We showed using a 3-D saturated unconfined synthetic aquifer that the self-potential method significantly improves the results of the harmonic hydraulic tomography. The hydroelectric forward problem is obtained by solving first the Richards equation, describing the groundwater flow, and then using the result in an electrical Poisson equation describing the self-potential problem. The joint inversion problem is solved using a reduction model based on the principal component geostatistical approach. In this method, the large prior covariance matrix is truncated and replaced by its low-rank approximation, allowing thus for notable computational time and storage savings. Three test cases are studied, to assess the validity of our approach. In the first test, we show that when the number of harmonic stimulations is low, combining the harmonic hydraulic and self-potential data does not improve the inversion results. In the second test where enough harmonic stimulations are performed, a significant improvement of the hydraulic parameters is observed. In the last synthetic test, we show that the electrical conductivity field required to invert the self-potential data can be determined with enough accuracy using an electrical resistivity tomography survey using the same electrodes configuration as used for the self-potential investigation.

Reducing uncertainties in the velocities determined by inversion of phase velocity dispersion curves using synthetic seismograms

NASA Astrophysics Data System (ADS)

Hosseini, Seyed Mehrdad

Characterizing the near-surface shear-wave velocity structure using Rayleigh-wave phase velocity dispersion curves is widespread in the context of reservoir characterization, exploration seismology, earthquake engineering, and geotechnical engineering. This surface seismic approach provides a feasible and low-cost alternative to the borehole measurements. Phase velocity dispersion curves from Rayleigh surface waves are inverted to yield the vertical shear-wave velocity profile. A significant problem with the surface wave inversion is its intrinsic non-uniqueness, and although this problem is widely recognized, there have not been systematic efforts to develop approaches to reduce the pervasive uncertainty that affects the velocity profiles determined by the inversion. Non-uniqueness cannot be easily studied in a nonlinear inverse problem such as Rayleigh-wave inversion and the only way to understand its nature is by numerical investigation which can get computationally expensive and inevitably time consuming. Regarding the variety of the parameters affecting the surface wave inversion and possible non-uniqueness induced by them, a technique should be established which is not controlled by the non-uniqueness that is already affecting the surface wave inversion. An efficient and repeatable technique is proposed and tested to overcome the non-uniqueness problem; multiple inverted shear-wave velocity profiles are used in a wavenumber integration technique to generate synthetic time series resembling the geophone recordings. The similarity between synthetic and observed time series is used as an additional tool along with the similarity between the theoretical and experimental dispersion curves. The proposed method is proven to be effective through synthetic and real world examples. In these examples, the nature of the non-uniqueness is discussed and its existence is shown. Using the proposed technique, inverted velocity profiles are estimated and effectiveness of this technique is evaluated; in the synthetic example, final inverted velocity profile is compared with the initial target velocity model, and in the real world example, final inverted shear-wave velocity profile is compared with the velocity model from independent measurements in a nearby borehole. Real world example shows that it is possible to overcome the non-uniqueness and distinguish the representative velocity profile for the site that also matches well with the borehole measurements.

Numerical reconstruction of tsunami source using combined seismic, satellite and DART data

NASA Astrophysics Data System (ADS)

Krivorotko, Olga; Kabanikhin, Sergey; Marinin, Igor

2014-05-01

Recent tsunamis, for instance, in Japan (2011), in Sumatra (2004), and at the Indian coast (2004) showed that a system of producing exact and timely information about tsunamis is of a vital importance. Numerical simulation is an effective instrument for providing such information. Bottom relief characteristics and the initial perturbation data (a tsunami source) are required for the direct simulation of tsunamis. The seismic data about the source are usually obtained in a few tens of minutes after an event has occurred (the seismic waves velocity being about five hundred kilometres per minute, while the velocity of tsunami waves is less than twelve kilometres per minute). A difference in the arrival times of seismic and tsunami waves can be used when operationally refining the tsunami source parameters and modelling expected tsunami wave height on the shore. The most suitable physical models related to the tsunamis simulation are based on the shallow water equations. The problem of identification parameters of a tsunami source using additional measurements of a passing wave is called inverse tsunami problem. We investigate three different inverse problems of determining a tsunami source using three different additional data: Deep-ocean Assessment and Reporting of Tsunamis (DART) measurements, satellite wave-form images and seismic data. These problems are severely ill-posed. We apply regularization techniques to control the degree of ill-posedness such as Fourier expansion, truncated singular value decomposition, numerical regularization. The algorithm of selecting the truncated number of singular values of an inverse problem operator which is agreed with the error level in measured data is described and analyzed. In numerical experiment we used gradient methods (Landweber iteration and conjugate gradient method) for solving inverse tsunami problems. Gradient methods are based on minimizing the corresponding misfit function. To calculate the gradient of the misfit function, the adjoint problem is solved. The conservative finite-difference schemes for solving the direct and adjoint problems in the approximation of shallow water are constructed. Results of numerical experiments of the tsunami source reconstruction are presented and discussed. We show that using a combination of three different types of data allows one to increase the stability and efficiency of tsunami source reconstruction. Non-profit organization WAPMERR (World Agency of Planetary Monitoring and Earthquake Risk Reduction) in collaboration with Informap software development department developed the Integrated Tsunami Research and Information System (ITRIS) to simulate tsunami waves and earthquakes, river course changes, coastal zone floods, and risk estimates for coastal constructions at wave run-ups and earthquakes. The special scientific plug-in components are embedded in a specially developed GIS-type graphic shell for easy data retrieval, visualization and processing. This work was supported by the Russian Foundation for Basic Research (project No. 12-01-00773 'Theory and Numerical Methods for Solving Combined Inverse Problems of Mathematical Physics') and interdisciplinary project of SB RAS 14 'Inverse Problems and Applications: Theory, Algorithms, Software'.

Generalized Uncertainty Quantification for Linear Inverse Problems in X-ray Imaging

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

Fowler, Michael James

2014-04-25

In industrial and engineering applications, X-ray radiography has attained wide use as a data collection protocol for the assessment of material properties in cases where direct observation is not possible. The direct measurement of nuclear materials, particularly when they are under explosive or implosive loading, is not feasible, and radiography can serve as a useful tool for obtaining indirect measurements. In such experiments, high energy X-rays are pulsed through a scene containing material of interest, and a detector records a radiograph by measuring the radiation that is not attenuated in the scene. One approach to the analysis of these radiographsmore » is to model the imaging system as an operator that acts upon the object being imaged to produce a radiograph. In this model, the goal is to solve an inverse problem to reconstruct the values of interest in the object, which are typically material properties such as density or areal density. The primary objective in this work is to provide quantitative solutions with uncertainty estimates for three separate applications in X-ray radiography: deconvolution, Abel inversion, and radiation spot shape reconstruction. For each problem, we introduce a new hierarchical Bayesian model for determining a posterior distribution on the unknowns and develop efficient Markov chain Monte Carlo (MCMC) methods for sampling from the posterior. A Poisson likelihood, based on a noise model for photon counts at the detector, is combined with a prior tailored to each application: an edge-localizing prior for deconvolution; a smoothing prior with non-negativity constraints for spot reconstruction; and a full covariance sampling prior based on a Wishart hyperprior for Abel inversion. After developing our methods in a general setting, we demonstrate each model on both synthetically generated datasets, including those from a well known radiation transport code, and real high energy radiographs taken at two U. S. Department of Energy laboratories.« less

Identification of Bouc-Wen hysteretic parameters based on enhanced response sensitivity approach

NASA Astrophysics Data System (ADS)

Wang, Li; Lu, Zhong-Rong

2017-05-01

This paper aims to identify parameters of Bouc-Wen hysteretic model using time-domain measured data. It follows a general inverse identification procedure, that is, identifying model parameters is treated as an optimization problem with the nonlinear least squares objective function. Then, the enhanced response sensitivity approach, which has been shown convergent and proper for such kind of problems, is adopted to solve the optimization problem. Numerical tests are undertaken to verify the proposed identification approach.

The application of the pilot points in groundwater numerical inversion model

NASA Astrophysics Data System (ADS)

Hu, Bin; Teng, Yanguo; Cheng, Lirong

2015-04-01

Numerical inversion simulation of groundwater has been widely applied in groundwater. Compared to traditional forward modeling, inversion model has more space to study. Zones and inversing modeling cell by cell are conventional methods. Pilot points is a method between them. The traditional inverse modeling method often uses software dividing the model into several zones with a few parameters needed to be inversed. However, distribution is usually too simple for modeler and result of simulation deviation. Inverse cell by cell will get the most actual parameter distribution in theory, but it need computational complexity greatly and quantity of survey data for geological statistical simulation areas. Compared to those methods, pilot points distribute a set of points throughout the different model domains for parameter estimation. Property values are assigned to model cells by Kriging to ensure geological units within the parameters of heterogeneity. It will reduce requirements of simulation area geological statistics and offset the gap between above methods. Pilot points can not only save calculation time, increase fitting degree, but also reduce instability of numerical model caused by numbers of parameters and other advantages. In this paper, we use pilot point in a field which structure formation heterogeneity and hydraulics parameter was unknown. We compare inversion modeling results of zones and pilot point methods. With the method of comparative analysis, we explore the characteristic of pilot point in groundwater inversion model. First, modeler generates an initial spatially correlated field given a geostatistical model by the description of the case site with the software named Groundwater Vistas 6. Defining Kriging to obtain the value of the field functions over the model domain on the basis of their values at measurement and pilot point locations (hydraulic conductivity), then we assign pilot points to the interpolated field which have been divided into 4 zones. And add range of disturbance values to inversion targets to calculate the value of hydraulic conductivity. Third, after inversion calculation (PEST), the interpolated field will minimize an objective function measuring the misfit between calculated and measured data. It's an optimization problem to find the optimum value of parameters. After the inversion modeling, the following major conclusion can be found out: (1) In a field structure formation is heterogeneity, the results of pilot point method is more real: better fitting result of parameters, more stable calculation of numerical simulation (stable residual distribution). Compared to zones, it is better of reflecting the heterogeneity of study field. (2) Pilot point method ensures that each parameter is sensitive and not entirely dependent on other parameters. Thus it guarantees the relative independence and authenticity of parameters evaluation results. However, it costs more time to calculate than zones. Key words: groundwater; pilot point; inverse model; heterogeneity; hydraulic conductivity

A Forward Glimpse into Inverse Problems through a Geology Example

ERIC Educational Resources Information Center

Winkel, Brian J.

2012-01-01

This paper describes a forward approach to an inverse problem related to detecting the nature of geological substrata which makes use of optimization techniques in a multivariable calculus setting. The true nature of the related inverse problem is highlighted. (Contains 2 figures.)

Trajectory optimization for the National Aerospace Plane

NASA Technical Reports Server (NTRS)

Lu, Ping

1992-01-01

The primary objective of this research is to develop an efficient and robust trajectory optimization tool for the optimal ascent problem of the National Aerospace Plane (NASP). This report is organized in the following order to summarize the complete work: Section two states the formulation and models of the trajectory optimization problem. An inverse dynamics approach to the problem is introduced in Section three. Optimal trajectories corresponding to various conditions and performance parameters are presented in Section four. A midcourse nonlinear feedback controller is developed in Section five. Section six demonstrates the performance of the inverse dynamics approach and midcourse controller during disturbances. Section seven discusses rocket assisted ascent which may be beneficial when orbital altitude is high. Finally, Section eight recommends areas of future research.

Identification of an internal combustion engine model by nonlinear multi-input multi-output system identification. Ph.D. Thesis

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

Luh, G.C.

1994-01-01

This thesis presents the application of advanced modeling techniques to construct nonlinear forward and inverse models of internal combustion engines for the detection and isolation of incipient faults. The NARMAX (Nonlinear Auto-Regressive Moving Average modeling with eXogenous inputs) technique of system identification proposed by Leontaritis and Billings was used to derive the nonlinear model of a internal combustion engine, over operating conditions corresponding to the I/M240 cycle. The I/M240 cycle is a standard proposed by the United States Environmental Protection Agency to measure tailpipe emissions in inspection and maintenance programs and consists of a driving schedule developed for the purposemore » of testing compliance with federal vehicle emission standards for carbon monoxide, unburned hydrocarbons, and nitrogen oxides. The experimental work for model identification and validation was performed on a 3.0 liter V6 engine installed in an engine test cell at the Center for Automotive Research at The Ohio State University. In this thesis, different types of model structures were proposed to obtain multi-input multi-output (MIMO) nonlinear NARX models. A modification of the algorithm proposed by He and Asada was used to estimate the robust orders of the derived MIMO nonlinear models. A methodology for the analysis of inverse NARX model was developed. Two methods were proposed to derive the inverse NARX model: (1) inversion from the forward NARX model; and (2) direct identification of inverse model from the output-input data set. In this thesis, invertibility, minimum-phase characteristic of zero dynamics, and stability analysis of NARX forward model are also discussed. Stability in the sense of Lyapunov is also investigated to check the stability of the identified forward and inverse models. This application of inverse problem leads to the estimation of unknown inputs and to actuator fault diagnosis.« less

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.

Inverse simulation system for evaluating handling qualities during rendezvous and docking

NASA Astrophysics Data System (ADS)

Zhou, Wanmeng; Wang, Hua; Thomson, Douglas; Tang, Guojin; Zhang, Fan

2017-08-01

The traditional method used for handling qualities assessment of manned space vehicles is too time-consuming to meet the requirements of an increasingly fast design process. In this study, a rendezvous and docking inverse simulation system to assess the handling qualities of spacecraft is proposed using a previously developed model-predictive-control architecture. By considering the fixed discrete force of the thrusters of the system, the inverse model is constructed using the least squares estimation method with a hyper-ellipsoidal restriction, the continuous control outputs of which are subsequently dispersed by pulse width modulation with sensitivity factors introduced. The inputs in every step are deemed constant parameters, and the method could be considered as a general method for solving nominal, redundant, and insufficient inverse problems. The rendezvous and docking inverse simulation is applied to a nine-degrees-of-freedom platform, and a novel handling qualities evaluation scheme is established according to the operation precision and astronauts' workload. Finally, different nominal trajectories are scored by the inverse simulation and an established evaluation scheme. The scores can offer theoretical guidance for astronaut training and more complex operation missions.

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

Kitanidis, Peter

As large-scale, commercial storage projects become operational, the problem of utilizing information from diverse sources becomes more critically important. In this project, we developed, tested, and applied an advanced joint data inversion system for CO 2 storage modeling with large data sets for use in site characterization and real-time monitoring. Emphasis was on the development of advanced and efficient computational algorithms for joint inversion of hydro-geophysical data, coupled with state-of-the-art forward process simulations. The developed system consists of (1) inversion tools using characterization data, such as 3D seismic survey (amplitude images), borehole log and core data, as well as hydraulic,more » tracer and thermal tests before CO 2 injection, (2) joint inversion tools for updating the geologic model with the distribution of rock properties, thus reducing uncertainty, using hydro-geophysical monitoring data, and (3) highly efficient algorithms for directly solving the dense or sparse linear algebra systems derived from the joint inversion. The system combines methods from stochastic analysis, fast linear algebra, and high performance computing. The developed joint inversion tools have been tested through synthetic CO 2 storage examples.« less

Inverse regression-based uncertainty quantification algorithms for high-dimensional models: Theory and practice

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

Li, Weixuan; Lin, Guang; Li, Bing

2016-09-01

A well-known challenge in uncertainty quantification (UQ) is the "curse of dimensionality". However, many high-dimensional UQ problems are essentially low-dimensional, because the randomness of the quantity of interest (QoI) is caused only by uncertain parameters varying within a low-dimensional subspace, known as the sufficient dimension reduction (SDR) subspace. Motivated by this observation, we propose and demonstrate in this paper an inverse regression-based UQ approach (IRUQ) for high-dimensional problems. Specifically, we use an inverse regression procedure to estimate the SDR subspace and then convert the original problem to a low-dimensional one, which can be efficiently solved by building a response surface model such as a polynomial chaos expansion. The novelty and advantages of the proposed approach is seen in its computational efficiency and practicality. Comparing with Monte Carlo, the traditionally preferred approach for high-dimensional UQ, IRUQ with a comparable cost generally gives much more accurate solutions even for high-dimensional problems, and even when the dimension reduction is not exactly sufficient. Theoretically, IRUQ is proved to converge twice as fast as the approach it uses seeking the SDR subspace. For example, while a sliced inverse regression method converges to the SDR subspace at the rate ofmore » $$O(n^{-1/2})$$, the corresponding IRUQ converges at $$O(n^{-1})$$. IRUQ also provides several desired conveniences in practice. It is non-intrusive, requiring only a simulator to generate realizations of the QoI, and there is no need to compute the high-dimensional gradient of the QoI. Finally, error bars can be derived for the estimation results reported by IRUQ.« less

Level-set techniques for facies identification in reservoir modeling

NASA Astrophysics Data System (ADS)

Iglesias, Marco A.; McLaughlin, Dennis

2011-03-01

In this paper we investigate the application of level-set techniques for facies identification in reservoir models. The identification of facies is a geometrical inverse ill-posed problem that we formulate in terms of shape optimization. The goal is to find a region (a geologic facies) that minimizes the misfit between predicted and measured data from an oil-water reservoir. In order to address the shape optimization problem, we present a novel application of the level-set iterative framework developed by Burger in (2002 Interfaces Free Bound. 5 301-29 2004 Inverse Problems 20 259-82) for inverse obstacle problems. The optimization is constrained by (the reservoir model) a nonlinear large-scale system of PDEs that describes the reservoir dynamics. We reformulate this reservoir model in a weak (integral) form whose shape derivative can be formally computed from standard results of shape calculus. At each iteration of the scheme, the current estimate of the shape derivative is utilized to define a velocity in the level-set equation. The proper selection of this velocity ensures that the new shape decreases the cost functional. We present results of facies identification where the velocity is computed with the gradient-based (GB) approach of Burger (2002) and the Levenberg-Marquardt (LM) technique of Burger (2004). While an adjoint formulation allows the straightforward application of the GB approach, the LM technique requires the computation of the large-scale Karush-Kuhn-Tucker system that arises at each iteration of the scheme. We efficiently solve this system by means of the representer method. We present some synthetic experiments to show and compare the capabilities and limitations of the proposed implementations of level-set techniques for the identification of geologic facies.

Ellipsoidal head model for fetal magnetoencephalography: forward and inverse solutions

NASA Astrophysics Data System (ADS)

Gutiérrez, David; Nehorai, Arye; Preissl, Hubert

2005-05-01

Fetal magnetoencephalography (fMEG) is a non-invasive technique where measurements of the magnetic field outside the maternal abdomen are used to infer the source location and signals of the fetus' neural activity. There are a number of aspects related to fMEG modelling that must be addressed, such as the conductor volume, fetal position and orientation, gestation period, etc. We propose a solution to the forward problem of fMEG based on an ellipsoidal head geometry. This model has the advantage of highlighting special characteristics of the field that are inherent to the anisotropy of the human head, such as the spread and orientation of the field in relationship with the localization and position of the fetal head. Our forward solution is presented in the form of a kernel matrix that facilitates the solution of the inverse problem through decoupling of the dipole localization parameters from the source signals. Then, we use this model and the maximum likelihood technique to solve the inverse problem assuming the availability of measurements from multiple trials. The applicability and performance of our methods are illustrated through numerical examples based on a real 151-channel SQUID fMEG measurement system (SARA). SARA is an MEG system especially designed for fetal assessment and is currently used for heart and brain studies. Finally, since our model requires knowledge of the best-fitting ellipsoid's centre location and semiaxes lengths, we propose a method for estimating these parameters through a least-squares fit on anatomical information obtained from three-dimensional ultrasound images.

Using informative priors in facies inversion: The case of C-ISR method

NASA Astrophysics Data System (ADS)

Valakas, G.; Modis, K.

2016-08-01

Inverse problems involving the characterization of hydraulic properties of groundwater flow systems by conditioning on observations of the state variables are mathematically ill-posed because they have multiple solutions and are sensitive to small changes in the data. In the framework of McMC methods for nonlinear optimization and under an iterative spatial resampling transition kernel, we present an algorithm for narrowing the prior and thus producing improved proposal realizations. To achieve this goal, we cosimulate the facies distribution conditionally to facies observations and normal scores transformed hydrologic response measurements, assuming a linear coregionalization model. The approach works by creating an importance sampling effect that steers the process to selected areas of the prior. The effectiveness of our approach is demonstrated by an example application on a synthetic underdetermined inverse problem in aquifer characterization.

Inverse Scattering and Local Observable Algebras in Integrable Quantum Field Theories

NASA Astrophysics Data System (ADS)

Alazzawi, Sabina; Lechner, Gandalf

2017-09-01

We present a solution method for the inverse scattering problem for integrable two-dimensional relativistic quantum field theories, specified in terms of a given massive single particle spectrum and a factorizing S-matrix. An arbitrary number of massive particles transforming under an arbitrary compact global gauge group is allowed, thereby generalizing previous constructions of scalar theories. The two-particle S-matrix S is assumed to be an analytic solution of the Yang-Baxter equation with standard properties, including unitarity, TCP invariance, and crossing symmetry. Using methods from operator algebras and complex analysis, we identify sufficient criteria on S that imply the solution of the inverse scattering problem. These conditions are shown to be satisfied in particular by so-called diagonal S-matrices, but presumably also in other cases such as the O( N)-invariant nonlinear {σ}-models.

Inference of relativistic electron spectra from measurements of inverse Compton radiation

NASA Astrophysics Data System (ADS)

Craig, I. J. D.; Brown, J. C.

1980-07-01

The inference of relativistic electron spectra from spectral measurement of inverse Compton radiation is discussed for the case where the background photon spectrum is a Planck function. The problem is formulated in terms of an integral transform that relates the measured spectrum to the unknown electron distribution. A general inversion formula is used to provide a quantitative assessment of the information content of the spectral data. It is shown that the observations must generally be augmented by additional information if anything other than a rudimentary two or three parameter model of the source function is to be derived. It is also pointed out that since a similar equation governs the continuum spectra emitted by a distribution of black-body radiators, the analysis is relevant to the problem of stellar population synthesis from galactic spectra.

Inverse problems, non-roundness and flat pieces of the effective burning velocity from an inviscid quadratic Hamilton-Jacobi model

NASA Astrophysics Data System (ADS)

Jing, Wenjia; Tran, Hung V.; Yu, Yifeng

2017-05-01

The main goal of this paper is to understand finer properties of the effective burning velocity from a combustion model introduced by Majda and Souganidis (1994 Nonlinearity 7 1-30). Motivated by results in Bangert (1994 Calculus Variations PDE 2 49-63) and applications in turbulent combustion, we show that when the dimension is two and the flow of the ambient fluid is either weak or very strong, the level set of the effective burning velocity has flat pieces. Due to the lack of an applicable Hopf-type rigidity result, we need to identify the exact location of at least one flat piece. Implications on the effective flame front and other related inverse type problems are also discussed.

Reconstruction of the spatial dependence of dielectric and geometrical properties of adhesively bonded structures

NASA Astrophysics Data System (ADS)

Mackay, C.; Hayward, D.; Mulholland, A. J.; McKee, S.; Pethrick, R. A.

2005-06-01

An inverse problem motivated by the nondestructive testing of adhesively bonded structures used in the aircraft industry is studied. Using transmission line theory, a model is developed which, when supplied with electrical and geometrical parameters, accurately predicts the reflection coefficient associated with such structures. Particular attention is paid to modelling the connection between the structures and the equipment used to measure the reflection coefficient. The inverse problem is then studied and an optimization approach employed to recover these electrical and geometrical parameters from experimentally obtained data. In particular the approach focuses on the recovery of spatially varying geometrical parameters as this is paramount to the successful reconstruction of electrical parameters. Reconstructions of structure geometry using this method are found to be in close agreement with experimental observations.

Fundamental Mechanisms of NeuroInformation Processing: Inverse Problems and Spike Processing

DTIC Science & Technology

2016-08-04

platform called Neurokernel for collaborative development of comprehensive models of the brain of the fruit fly Drosophila melanogaster and their execution...example. We investigated the following nonlinear identification problem: given both the input signal u and the time sequence (tk)k2Z at the output of...from a time sequence is to be contrasted with existing methods for rate-based models in neuroscience. In such models the output of the system is taken

The direct and inverse problems of an air-saturated poroelastic cylinder submitted to acoustic radiation

NASA Astrophysics Data System (ADS)

Ogam, Erick; Fellah, Z. E. A.

2011-09-01

A wave-fluid saturated poroelastic structure interaction model based on the modified Biot theory (MBT) and plane-wave decomposition using orthogonal cylindrical functions is developed. The model is employed to recover from real data acquired in an anechoic chamber, the poromechanical properties of a soft cellular melamine cylinder submitted to an audible acoustic radiation. The inverse problem of acoustic diffraction is solved by constructing the objective functional given by the total square of the difference between predictions from the MBT interaction model and diffracted field data from experiment. The faculty of retrieval of the intrinsic poromechanical parameters from the diffracted acoustic fields, indicate that a wave initially propagating in a light fluid (air) medium, is able to carry in the absence of mechanical excitation of the specimen, information on the macroscopic mechanical properties which depend on the microstructural and intrinsic properties of the solid phase.

High effective inverse dynamics modelling for dual-arm robot

NASA Astrophysics Data System (ADS)

Shen, Haoyu; Liu, Yanli; Wu, Hongtao

2018-05-01

To deal with the problem of inverse dynamics modelling for dual arm robot, a recursive inverse dynamics modelling method based on decoupled natural orthogonal complement is presented. In this model, the concepts and methods of Decoupled Natural Orthogonal Complement matrices are used to eliminate the constraint forces in the Newton-Euler kinematic equations, and the screws is used to express the kinematic and dynamics variables. On this basis, the paper has developed a special simulation program with symbol software of Mathematica and conducted a simulation research on the a dual-arm robot. Simulation results show that the proposed method based on decoupled natural orthogonal complement can save an enormous amount of CPU time that was spent in computing compared with the recursive Newton-Euler kinematic equations and the results is correct and reasonable, which can verify the reliability and efficiency of the method.

Bayesian approach to inverse statistical mechanics.

PubMed

Habeck, Michael

2014-05-01

Inverse statistical mechanics aims to determine particle interactions from ensemble properties. This article looks at this inverse problem from a Bayesian perspective and discusses several statistical estimators to solve it. In addition, a sequential Monte Carlo algorithm is proposed that draws the interaction parameters from their posterior probability distribution. The posterior probability involves an intractable partition function that is estimated along with the interactions. The method is illustrated for inverse problems of varying complexity, including the estimation of a temperature, the inverse Ising problem, maximum entropy fitting, and the reconstruction of molecular interaction potentials.

Bayesian approach to inverse statistical mechanics

NASA Astrophysics Data System (ADS)

Habeck, Michael

2014-05-01

Inverse statistical mechanics aims to determine particle interactions from ensemble properties. This article looks at this inverse problem from a Bayesian perspective and discusses several statistical estimators to solve it. In addition, a sequential Monte Carlo algorithm is proposed that draws the interaction parameters from their posterior probability distribution. The posterior probability involves an intractable partition function that is estimated along with the interactions. The method is illustrated for inverse problems of varying complexity, including the estimation of a temperature, the inverse Ising problem, maximum entropy fitting, and the reconstruction of molecular interaction potentials.

Bayesian inversion of marine CSEM data from the Scarborough gas field using a transdimensional 2-D parametrization

NASA Astrophysics Data System (ADS)

Ray, Anandaroop; Key, Kerry; Bodin, Thomas; Myer, David; Constable, Steven

2014-12-01

We apply a reversible-jump Markov chain Monte Carlo method to sample the Bayesian posterior model probability density function of 2-D seafloor resistivity as constrained by marine controlled source electromagnetic data. This density function of earth models conveys information on which parts of the model space are illuminated by the data. Whereas conventional gradient-based inversion approaches require subjective regularization choices to stabilize this highly non-linear and non-unique inverse problem and provide only a single solution with no model uncertainty information, the method we use entirely avoids model regularization. The result of our approach is an ensemble of models that can be visualized and queried to provide meaningful information about the sensitivity of the data to the subsurface, and the level of resolution of model parameters. We represent models in 2-D using a Voronoi cell parametrization. To make the 2-D problem practical, we use a source-receiver common midpoint approximation with 1-D forward modelling. Our algorithm is transdimensional and self-parametrizing where the number of resistivity cells within a 2-D depth section is variable, as are their positions and geometries. Two synthetic studies demonstrate the algorithm's use in the appraisal of a thin, segmented, resistive reservoir which makes for a challenging exploration target. As a demonstration example, we apply our method to survey data collected over the Scarborough gas field on the Northwest Australian shelf.

Imaging metallic samples using electrical capacitance tomography: forward modelling and reconstruction algorithms

NASA Astrophysics Data System (ADS)

Hosani, E. Al; Zhang, M.; Abascal, J. F. P. J.; Soleimani, M.

2016-11-01

Electrical capacitance tomography (ECT) is an imaging technology used to reconstruct the permittivity distribution within the sensing region. So far, ECT has been primarily used to image non-conductive media only, since if the conductivity of the imaged object is high, the capacitance measuring circuit will be almost shortened by the conductivity path and a clear image cannot be produced using the standard image reconstruction approaches. This paper tackles the problem of imaging metallic samples using conventional ECT systems by investigating the two main aspects of image reconstruction algorithms, namely the forward problem and the inverse problem. For the forward problem, two different methods to model the region of high conductivity in ECT is presented. On the other hand, for the inverse problem, three different algorithms to reconstruct the high contrast images are examined. The first two methods are the linear single step Tikhonov method and the iterative total variation regularization method, and use two sets of ECT data to reconstruct the image in time difference mode. The third method, namely the level set method, uses absolute ECT measurements and was developed using a metallic forward model. The results indicate that the applications of conventional ECT systems can be extended to metal samples using the suggested algorithms and forward model, especially using a level set algorithm to find the boundary of the metal.

Modelling and Inverse-Modelling: Experiences with O.D.E. Linear Systems in Engineering Courses

ERIC Educational Resources Information Center

Martinez-Luaces, Victor

2009-01-01

In engineering careers courses, differential equations are widely used to solve problems concerned with modelling. In particular, ordinary differential equations (O.D.E.) linear systems appear regularly in Chemical Engineering, Food Technology Engineering and Environmental Engineering courses, due to the usefulness in modelling chemical kinetics,…

An inverse problem in thermal imaging

NASA Technical Reports Server (NTRS)

Bryan, Kurt; Caudill, Lester F., Jr.

1994-01-01

This paper examines uniqueness and stability results for an inverse problem in thermal imaging. The goal is to identify an unknown boundary of an object by applying a heat flux and measuring the induced temperature on the boundary of the sample. The problem is studied both in the case in which one has data at every point on the boundary of the region and the case in which only finitely many measurements are available. An inversion procedure is developed and used to study the stability of the inverse problem for various experimental configurations.

Inverse problems in quantum chemistry

NASA Astrophysics Data System (ADS)

Karwowski, Jacek

Inverse problems constitute a branch of applied mathematics with well-developed methodology and formalism. A broad family of tasks met in theoretical physics, in civil and mechanical engineering, as well as in various branches of medical and biological sciences has been formulated as specific implementations of the general theory of inverse problems. In this article, it is pointed out that a number of approaches met in quantum chemistry can (and should) be classified as inverse problems. Consequently, the methodology used in these approaches may be enriched by applying ideas and theorems developed within the general field of inverse problems. Several examples, including the RKR method for the construction of potential energy curves, determining parameter values in semiempirical methods, and finding external potentials for which the pertinent Schrödinger equation is exactly solvable, are discussed in detail.

Polarimetric SAR Interferometry Evaluation in Mangroves

NASA Technical Reports Server (NTRS)

Lee, Seung-Kuk; Fatoyinbo,Temilola; Osmanoglu, Batuhan; Sun, Guoqing

2014-01-01

TanDEM-X (TDX) enables to generate an interferometric coherence without temporal decorrelation effect that is the most critical factor for a successful Pol-InSAR inversion, as have recently been used for forest parameter retrieval. This paper presents mangrove forest height estimation only using single-pass/single-baseline/dual-polarization TDX data by means of new dual-Pol-InSAR inversion technique. To overcome a lack of one polarization in a conventional Pol- InSAR inversion (i.e. an underdetermined problem), the ground phase in the Pol-InSAR model is directly estimated from TDX interferograms assuming flat underlying topography in mangrove forest. The inversion result is validated against lidar measurement data (NASA's G-LiHT data).

Angle-domain inverse scattering migration/inversion in isotropic media

NASA Astrophysics Data System (ADS)

Li, Wuqun; Mao, Weijian; Li, Xuelei; Ouyang, Wei; Liang, Quan

2018-07-01

The classical seismic asymptotic inversion can be transformed into a problem of inversion of generalized Radon transform (GRT). In such methods, the combined parameters are linearly attached to the scattered wave-field by Born approximation and recovered by applying an inverse GRT operator to the scattered wave-field data. Typical GRT-style true-amplitude inversion procedure contains an amplitude compensation process after the weighted migration via dividing an illumination associated matrix whose elements are integrals of scattering angles. It is intuitional to some extent that performs the generalized linear inversion and the inversion of GRT together by this process for direct inversion. However, it is imprecise to carry out such operation when the illumination at the image point is limited, which easily leads to the inaccuracy and instability of the matrix. This paper formulates the GRT true-amplitude inversion framework in an angle-domain version, which naturally degrades the external integral term related to the illumination in the conventional case. We solve the linearized integral equation for combined parameters of different fixed scattering angle values. With this step, we obtain high-quality angle-domain common-image gathers (CIGs) in the migration loop which provide correct amplitude-versus-angle (AVA) behavior and reasonable illumination range for subsurface image points. Then we deal with the over-determined problem to solve each parameter in the combination by a standard optimization operation. The angle-domain GRT inversion method keeps away from calculating the inaccurate and unstable illumination matrix. Compared with the conventional method, the angle-domain method can obtain more accurate amplitude information and wider amplitude-preserved range. Several model tests demonstrate the effectiveness and practicability.

A microwave tomography strategy for structural monitoring

NASA Astrophysics Data System (ADS)

Catapano, I.; Crocco, L.; Isernia, T.

2009-04-01

The capability of the electromagnetic waves to penetrate optical dense regions can be conveniently exploited to provide high informative images of the internal status of manmade structures in a non destructive and minimally invasive way. In this framework, as an alternative to the wide adopted radar techniques, Microwave Tomography approaches are worth to be considered. As a matter of fact, they may accurately reconstruct the permittivity and conductivity distributions of a given region from the knowledge of a set of incident fields and measures of the corresponding scattered fields. As far as cultural heritage conservation is concerned, this allow not only to detect the anomalies, which can possibly damage the integrity and the stability of the structure, but also characterize their morphology and electric features, which are useful information to properly address the repair actions. However, since a non linear and ill-posed inverse scattering problem has to be solved, proper regularization strategies and sophisticated data processing tools have to be adopt to assure the reliability of the results. To pursue this aim, in the last years huge attention has been focused on the advantages introduced by diversity in data acquisition (multi-frequency/static/view data) [1,2] as well as on the analysis of the factors affecting the solution of an inverse scattering problem [3]. Moreover, how the degree of non linearity of the relationship between the scattered field and the electromagnetic parameters of the targets can be changed by properly choosing the mathematical model adopt to formulate the scattering problem has been shown in [4]. Exploiting the above results, in this work we propose an imaging procedure in which the inverse scattering problem is formulated as an optimization problem where the mathematical relationship between data and unknowns is expressed by means of a convenient integral equations model and the sought solution is defined as the global minimum of a cost functional. In particular, a local minimization scheme is exploited and a pre-processing step, devoted to preliminary asses the location and the shape of the anomalies, is exploited. The effectiveness of the proposed strategy has been preliminary assessed by means of numerical examples concerning the diagnostic of masonry structures, which will be shown in the Conference. [1] O. M. Bucci, L. Crocco, T. Isernia, and V. Pascazio, Subsurface inverse scattering problems: Quantifying, qualifying and achieving the available information, IEEE Trans. Geosci. Remote Sens., 39(5), 2527-2538, 2001. [2] R. Persico, R. Bernini, and F. Soldovieri, "The role of the measurement configuration in inverse scattering from buried objects under the distorted Born approximation," IEEE Trans. Antennas Propag., vol. 53, no. 6, pp. 1875-1887, Jun. 2005. [3] I. Catapano, L. Crocco, M. D'Urso, T. Isernia, "On the Effect of Support Estimation and of a New Model in 2-D Inverse Scattering Problems," IEEE Trans. Antennas Propagat., vol.55, no.6, pp.1895-1899, 2007. [4] M. D'Urso, I. Catapano, L. Crocco and T. Isernia, Effective solution of 3D scattering problems via series expansions: applicability and a new hybrid scheme, IEEE Trans. On Geosci. Remote Sens., vol.45, no.3, pp. 639-648, 2007.

pyGIMLi: An open-source library for modelling and inversion in geophysics

NASA Astrophysics Data System (ADS)

Rücker, Carsten; Günther, Thomas; Wagner, Florian M.

2017-12-01

Many tasks in applied geosciences cannot be solved by single measurements, but require the integration of geophysical, geotechnical and hydrological methods. Numerical simulation techniques are essential both for planning and interpretation, as well as for the process understanding of modern geophysical methods. These trends encourage open, simple, and modern software architectures aiming at a uniform interface for interdisciplinary and flexible modelling and inversion approaches. We present pyGIMLi (Python Library for Inversion and Modelling in Geophysics), an open-source framework that provides tools for modelling and inversion of various geophysical but also hydrological methods. The modelling component supplies discretization management and the numerical basis for finite-element and finite-volume solvers in 1D, 2D and 3D on arbitrarily structured meshes. The generalized inversion framework solves the minimization problem with a Gauss-Newton algorithm for any physical forward operator and provides opportunities for uncertainty and resolution analyses. More general requirements, such as flexible regularization strategies, time-lapse processing and different sorts of coupling individual methods are provided independently of the actual methods used. The usage of pyGIMLi is first demonstrated by solving the steady-state heat equation, followed by a demonstration of more complex capabilities for the combination of different geophysical data sets. A fully coupled hydrogeophysical inversion of electrical resistivity tomography (ERT) data of a simulated tracer experiment is presented that allows to directly reconstruct the underlying hydraulic conductivity distribution of the aquifer. Another example demonstrates the improvement of jointly inverting ERT and ultrasonic data with respect to saturation by a new approach that incorporates petrophysical relations in the inversion. Potential applications of the presented framework are manifold and include time-lapse, constrained, joint, and coupled inversions of various geophysical and hydrological data sets.

Vibrato in Singing Voice: The Link between Source-Filter and Sinusoidal Models

NASA Astrophysics Data System (ADS)

Arroabarren, Ixone; Carlosena, Alfonso

2004-12-01

The application of inverse filtering techniques for high-quality singing voice analysis/synthesis is discussed. In the context of source-filter models, inverse filtering provides a noninvasive method to extract the voice source, and thus to study voice quality. Although this approach is widely used in speech synthesis, this is not the case in singing voice. Several studies have proved that inverse filtering techniques fail in the case of singing voice, the reasons being unclear. In order to shed light on this problem, we will consider here an additional feature of singing voice, not present in speech: the vibrato. Vibrato has been traditionally studied by sinusoidal modeling. As an alternative, we will introduce here a novel noninteractive source filter model that incorporates the mechanisms of vibrato generation. This model will also allow the comparison of the results produced by inverse filtering techniques and by sinusoidal modeling, as they apply to singing voice and not to speech. In this way, the limitations of these conventional techniques, described in previous literature, will be explained. Both synthetic signals and singer recordings are used to validate and compare the techniques presented in the paper.

Application of a stochastic inverse to the geophysical inverse problem

NASA Technical Reports Server (NTRS)

Jordan, T. H.; Minster, J. B.

1972-01-01

The inverse problem for gross earth data can be reduced to an undertermined linear system of integral equations of the first kind. A theory is discussed for computing particular solutions to this linear system based on the stochastic inverse theory presented by Franklin. The stochastic inverse is derived and related to the generalized inverse of Penrose and Moore. A Backus-Gilbert type tradeoff curve is constructed for the problem of estimating the solution to the linear system in the presence of noise. It is shown that the stochastic inverse represents an optimal point on this tradeoff curve. A useful form of the solution autocorrelation operator as a member of a one-parameter family of smoothing operators is derived.

Analysis of space telescope data collection system

NASA Technical Reports Server (NTRS)

Ingels, F. M.; Schoggen, W. O.

1982-01-01

An analysis of the expected performance for the Multiple Access (MA) system is provided. The analysis covers the expected bit error rate performance, the effects of synchronization loss, the problem of self-interference, and the problem of phase ambiguity. The problem of false acceptance of a command word due to data inversion is discussed. A mathematical determination of the probability of accepting an erroneous command word due to a data inversion is presented. The problem is examined for three cases: (1) a data inversion only, (2) a data inversion and a random error within the same command word, and a block (up to 256 48-bit words) containing both a data inversion and a random error.

EDITORIAL: Introduction to the special issue on electromagnetic inverse problems: emerging methods and novel applications Introduction to the special issue on electromagnetic inverse problems: emerging methods and novel applications

NASA Astrophysics Data System (ADS)

Dorn, O.; Lesselier, D.

2010-07-01

Inverse problems in electromagnetics have a long history and have stimulated exciting research over many decades. New applications and solution methods are still emerging, providing a rich source of challenging topics for further investigation. The purpose of this special issue is to combine descriptions of several such developments that are expected to have the potential to fundamentally fuel new research, and to provide an overview of novel methods and applications for electromagnetic inverse problems. There have been several special sections published in Inverse Problems over the last decade addressing fully, or partly, electromagnetic inverse problems. Examples are: Electromagnetic imaging and inversion of the Earth's subsurface (Guest Editors: D Lesselier and T Habashy) October 2000 Testing inversion algorithms against experimental data (Guest Editors: K Belkebir and M Saillard) December 2001 Electromagnetic and ultrasonic nondestructive evaluation (Guest Editors: D Lesselier and J Bowler) December 2002 Electromagnetic characterization of buried obstacles (Guest Editors: D Lesselier and W C Chew) December 2004 Testing inversion algorithms against experimental data: inhomogeneous targets (Guest Editors: K Belkebir and M Saillard) December 2005 Testing inversion algorithms against experimental data: 3D targets (Guest Editors: A Litman and L Crocco) February 2009 In a certain sense, the current issue can be understood as a continuation of this series of special sections on electromagnetic inverse problems. On the other hand, its focus is intended to be more general than previous ones. Instead of trying to cover a well-defined, somewhat specialized research topic as completely as possible, this issue aims to show the broad range of techniques and applications that are relevant to electromagnetic imaging nowadays, which may serve as a source of inspiration and encouragement for all those entering this active and rapidly developing research area. Also, the construction of this special issue is likely to have been different from preceding ones. In addition to the invitations sent to specific research groups involved in electromagnetic inverse problems, the Guest Editors also solicited recommendations, from a large number of experts, of potential authors who were thereupon encouraged to contribute. Moreover, an open call for contributions was published on the homepage of Inverse Problems in order to attract as wide a scope of contributions as possible. This special issue's attempt at generality might also define its limitations: by no means could this collection of papers be exhaustive or complete, and as Guest Editors we are well aware that many exciting topics and potential contributions will be missing. This, however, also determines its very special flavor: besides addressing electromagnetic inverse problems in a broad sense, there were only a few restrictions on the contributions considered for this section. One requirement was plausible evidence of either novelty or the emergent nature of the technique or application described, judged mainly by the referees, and in some cases by the Guest Editors. The technical quality of the contributions always remained a stringent condition of acceptance, final adjudication (possibly questionable either way, not always positive) being made in most cases once a thorough revision process had been carried out. Therefore, we hope that the final result presented here constitutes an interesting collection of novel ideas and applications, properly refereed and edited, which will find its own readership and which can stimulate significant new research in the topics represented. Overall, as Guest Editors, we feel quite fortunate to have obtained such a strong response to the call for this issue and to have a really wide-ranging collection of high-quality contributions which, indeed, can be read from the first to the last page with sustained enthusiasm. A large number of applications and techniques is represented, overall via 16 contributions with 45 authors in total. This shows, in our opinion, that electromagnetic imaging and inversion remain amongst the most challenging and active research areas in applied inverse problems today. Below, we give a brief overview of the contributions included in this issue, ordered alphabetically by the surname of the leading author. 1. The complexity of handling potential randomness of the source in an inverse scattering problem is not minor, and the literature is far from being replete in this configuration. The contribution by G Bao, S N Chow, P Li and H Zhou, `Numerical solution of an inverse medium scattering problem with a stochastic source', exemplifies how to hybridize Wiener chaos expansion with a recursive linearization method in order to solve the stochastic problem as a set of decoupled deterministic ones. 2. In cases where the forward problem is expensive to evaluate, database methods might become a reliable method of choice, while enabling one to deliver more information on the inversion itself. The contribution by S Bilicz, M Lambert and Sz Gyimóthy, `Kriging-based generation of optimal databases as forward and inverse surrogate models', describes such a technique which uses kriging for constructing an efficient database with the goal of achieving an equidistant distribution of points in the measurement space. 3. Anisotropy remains a considerable challenge in electromagnetic imaging, which is tackled in the contribution by F Cakoni, D Colton, P Monk and J Sun, `The inverse electromagnetic scattering problem for anisotropic media', via the fact that transmission eigenvalues can be retrieved from a far-field scattering pattern, yielding, in particular, lower and upper bounds of the index of refraction of the unknown (dielectric anisotropic) scatterer. 4. So-called subspace optimization methods (SOM) have attracted a lot of interest recently in many fields. The contribution by X Chen, `Subspace-based optimization method for inverse scattering problems with an inhomogeneous background medium', illustrates how to address a realistic situation in which the medium containing the unknown obstacles is not homogeneous, via blending a properly developed SOM with a finite-element approach to the required Green's functions. 5. H Egger, M Hanke, C Schneider, J Schöberl and S Zaglmayr, in their contribution `Adjoint-based sampling methods for electromagnetic scattering', show how to efficiently develop sampling methods without explicit knowledge of the dyadic Green's function once an adjoint problem has been solved at much lower computational cost. This is demonstrated by examples in demanding propagative and diffusive situations. 6. Passive sensor arrays can be employed to image reflectors from ambient noise via proper migration of cross-correlation matrices into their embedding medium. This is investigated, and resolution, in particular, is considered in detail, as a function of the characteristics of the sensor array and those of the noise, in the contribution by J Garnier and G Papanicolaou, `Resolution analysis for imaging with noise'. 7. A direct reconstruction technique based on the conformal mapping theorem is proposed and investigated in depth in the contribution by H Haddar and R Kress, `Conformal mapping and impedance tomography'. This paper expands on previous work, with inclusions in homogeneous media, convergence results, and numerical illustrations. 8. The contribution by T Hohage and S Langer, `Acceleration techniques for regularized Newton methods applied to electromagnetic inverse medium scattering problems', focuses on a spectral preconditioner intended to accelerate regularized Newton methods as employed for the retrieval of a local inhomogeneity in a three-dimensional vector electromagnetic case, while also illustrating the implementation of a Lepskiĭ-type stopping rule outsmarting a traditional discrepancy principle. 9. Geophysical applications are a rich source of practically relevant inverse problems. The contribution by M Li, A Abubakar and T Habashy, `Application of a two-and-a-half dimensional model-based algorithm to crosswell electromagnetic data inversion', deals with a model-based inversion technique for electromagnetic imaging which addresses novel challenges such as multi-physics inversion, and incorporation of prior knowledge, such as in hydrocarbon recovery. 10. Non-stationary inverse problems, considered as a special class of Bayesian inverse problems, are framed via an orthogonal decomposition representation in the contribution by A Lipponen, A Seppänen and J P Kaipio, `Reduced order estimation of nonstationary flows with electrical impedance tomography'. The goal is to simultaneously estimate, from electrical impedance tomography data, certain characteristics of the Navier--Stokes fluid flow model together with time-varying concentration distribution. 11. Non-iterative imaging methods of thin, penetrable cracks, based on asymptotic expansion of the scattering amplitude and analysis of the multi-static response matrix, are discussed in the contribution by W-K Park, `On the imaging of thin dielectric inclusions buried within a half-space', completing, for a shallow burial case at multiple frequencies, the direct imaging of small obstacles (here, along their transverse dimension), MUSIC and non-MUSIC type indicator functions being used for that purpose. 12. The contribution by R Potthast, `A study on orthogonality sampling' envisages quick localization and shaping of obstacles from (portions of) far-field scattering patterns collected at one or more time-harmonic frequencies, via the simple calculation (and summation) of scalar products between those patterns and a test function. This is numerically exemplified for Neumann/Dirichlet boundary conditions and homogeneous/heterogeneous embedding media. 13. The contribution by J D Shea, P Kosmas, B D Van Veen and S C Hagness, `Contrast-enhanced microwave imaging of breast tumors: a computational study using 3D realistic numerical phantoms', aims at microwave medical imaging, namely the early detection of breast cancer. The use of contrast enhancing agents is discussed in detail and a number of reconstructions in three-dimensional geometry of realistic numerical breast phantoms are presented. 14. The contribution by D A Subbarayappa and V Isakov, `Increasing stability of the continuation for the Maxwell system', discusses enhanced log-type stability results for continuation of solutions of the time-harmonic Maxwell system, adding a fresh chapter to the interesting story of the study of the Cauchy problem for PDE. 15. In their contribution, `Recent developments of a monotonicity imaging method for magnetic induction tomography in the small skin-depth regime', A Tamburrino, S Ventre and G Rubinacci extend the recently developed monotonicity method toward the application of magnetic induction tomography in order to map surface-breaking defects affecting a damaged metal component. 16. The contribution by F Viani, P Rocca, M Benedetti, G Oliveri and A Massa, `Electromagnetic passive localization and tracking of moving targets in a WSN-infrastructured environment', contributes to what could still be seen as a niche problem, yet both useful in terms of applications, e.g., security, and challenging in terms of methodologies and experiments, in particular, in view of the complexity of environments in which this endeavor is to take place and the variability of the wireless sensor networks employed. To conclude, we would like to thank the able and tireless work of Kate Watt and Zoë Crossman, as past and present Publishers of the Journal, on what was definitely a long and exciting journey (sometimes a little discouraging when reports were not arriving, or authors were late, or Guest Editors overwhelmed) that started from a thorough discussion at the `Manchester workshop on electromagnetic inverse problems' held mid-June 2009, between Kate Watt and the Guest Editors. We gratefully acknowledge the fact that W W Symes gave us his full backing to carry out this special issue and that A K Louis completed it successfully. Last, but not least, the staff of Inverse Problems should be thanked, since they work together to make it a premier journal.

A linear-encoding model explains the variability of the target morphology in regeneration

PubMed Central

Lobo, Daniel; Solano, Mauricio; Bubenik, George A.; Levin, Michael

2014-01-01

A fundamental assumption of today's molecular genetics paradigm is that complex morphology emerges from the combined activity of low-level processes involving proteins and nucleic acids. An inherent characteristic of such nonlinear encodings is the difficulty of creating the genetic and epigenetic information that will produce a given self-assembling complex morphology. This ‘inverse problem’ is vital not only for understanding the evolution, development and regeneration of bodyplans, but also for synthetic biology efforts that seek to engineer biological shapes. Importantly, the regenerative mechanisms in deer antlers, planarian worms and fiddler crabs can solve an inverse problem: their target morphology can be altered specifically and stably by injuries in particular locations. Here, we discuss the class of models that use pre-specified morphological goal states and propose the existence of a linear encoding of the target morphology, making the inverse problem easy for these organisms to solve. Indeed, many model organisms such as Drosophila, hydra and Xenopus also develop according to nonlinear encodings producing linear encodings of their final morphologies. We propose the development of testable models of regeneration regulation that combine emergence with a top-down specification of shape by linear encodings of target morphology, driving transformative applications in biomedicine and synthetic bioengineering. PMID:24402915

Modeling and forecasting foreign exchange daily closing prices with normal inverse Gaussian

NASA Astrophysics Data System (ADS)

Teneng, Dean

2013-09-01

We fit the normal inverse Gaussian(NIG) distribution to foreign exchange closing prices using the open software package R and select best models by Käärik and Umbleja (2011) proposed strategy. We observe that daily closing prices (12/04/2008 - 07/08/2012) of CHF/JPY, AUD/JPY, GBP/JPY, NZD/USD, QAR/CHF, QAR/EUR, SAR/CHF, SAR/EUR, TND/CHF and TND/EUR are excellent fits while EGP/EUR and EUR/GBP are good fits with a Kolmogorov-Smirnov test p-value of 0.062 and 0.08 respectively. It was impossible to estimate normal inverse Gaussian parameters (by maximum likelihood; computational problem) for JPY/CHF but CHF/JPY was an excellent fit. Thus, while the stochastic properties of an exchange rate can be completely modeled with a probability distribution in one direction, it may be impossible the other way around. We also demonstrate that foreign exchange closing prices can be forecasted with the normal inverse Gaussian (NIG) Lévy process, both in cases where the daily closing prices can and cannot be modeled by NIG distribution.

An Inverse Problem Formulation Methodology for Stochastic Models

DTIC Science & Technology

2010-05-02

form the surveillance data Infection control measures were implemented in the form of health care worker hand - hygiene before and after patients contact...manuscript derives from our interest in understanding the spread of infectious diseases in particular, nosocomial infections , in order to prevent major...given by the inverse of the parameter of the exponential distribution. A hand - hygiene policy applied to health care workers on isolated VRE colonized

Weighted current sheets supported in normal and inverse configurations - A model for prominence observations

NASA Technical Reports Server (NTRS)

Demoulin, P.; Forbes, T. G.

1992-01-01

A technique which incorporates both photospheric and prominence magnetic field observations is used to analyze the magnetic support of solar prominences in two dimensions. The prominence is modeled by a mass-loaded current sheet which is supported against gravity by magnetic fields from a bipolar source in the photosphere and a massless line current in the corona. It is found that prominence support can be achieved in three different kinds of configurations: an arcade topology with a normal polarity; a helical topology with a normal polarity; and a helical topology with an inverse polarity. In all cases the important parameter is the variation of the horizontal component of the prominence field with height. Adding a line current external to the prominence eliminates the nonsupport problem which plagues virtually all previous prominence models with inverse polarity.

Effects of induced stress on seismic forward modelling and inversion

NASA Astrophysics Data System (ADS)

Tromp, Jeroen; Trampert, Jeannot

2018-05-01

We demonstrate how effects of induced stress may be incorporated in seismic modelling and inversion. Our approach is motivated by the accommodation of pre-stress in global seismology. Induced stress modifies both the equation of motion and the constitutive relationship. The theory predicts that induced pressure linearly affects the unstressed isotropic moduli with a slope determined by their adiabatic pressure derivatives. The induced deviatoric stress produces anisotropic compressional and shear wave speeds; the latter result in shear wave splitting. For forward modelling purposes, we determine the weak form of the equation of motion under induced stress. In the context of the inverse problem, we determine induced stress sensitivity kernels, which may be used for adjoint tomography. The theory is illustrated by considering 2-D propagation of SH waves and related Fréchet derivatives based on a spectral-element method.

Kinematic source inversions of teleseismic data based on the QUESO library for uncertainty quantification and prediction

NASA Astrophysics Data System (ADS)

Zielke, O.; McDougall, D.; Mai, P. M.; Babuska, I.

2014-12-01

One fundamental aspect of seismic hazard mitigation is gaining a better understanding of the rupture process. Because direct observation of the relevant parameters and properties is not possible, other means such as kinematic source inversions are used instead. By constraining the spatial and temporal evolution of fault slip during an earthquake, those inversion approaches may enable valuable insights in the physics of the rupture process. However, due to the underdetermined nature of this inversion problem (i.e., inverting a kinematic source model for an extended fault based on seismic data), the provided solutions are generally non-unique. Here we present a statistical (Bayesian) inversion approach based on an open-source library for uncertainty quantification (UQ) called QUESO that was developed at ICES (UT Austin). The approach has advantages with respect to deterministic inversion approaches as it provides not only a single (non-unique) solution but also provides uncertainty bounds with it. Those uncertainty bounds help to qualitatively and quantitatively judge how well constrained an inversion solution is and how much rupture complexity the data reliably resolve. The presented inversion scheme uses only tele-seismically recorded body waves but future developments may lead us towards joint inversion schemes. After giving an insight in the inversion scheme ifself (based on delayed rejection adaptive metropolis, DRAM) we explore the method's resolution potential. For that, we synthetically generate tele-seismic data, add for example different levels of noise and/or change fault plane parameterization and then apply our inversion scheme in the attempt to extract the (known) kinematic rupture model. We conclude with exemplary inverting real tele-seismic data of a recent large earthquake and compare those results with deterministically derived kinematic source models provided by other research groups.

Inverse dynamic substructuring using the direct hybrid assembly in the frequency domain

NASA Astrophysics Data System (ADS)

D'Ambrogio, Walter; Fregolent, Annalisa

2014-04-01

The paper deals with the identification of the dynamic behaviour of a structural subsystem, starting from the known dynamic behaviour of both the coupled system and the remaining part of the structural system (residual subsystem). This topic is also known as decoupling problem, subsystem subtraction or inverse dynamic substructuring. Whenever it is necessary to combine numerical models (e.g. FEM) and test models (e.g. FRFs), one speaks of experimental dynamic substructuring. Substructure decoupling techniques can be classified as inverse coupling or direct decoupling techniques. In inverse coupling, the equations describing the coupling problem are rearranged to isolate the unknown substructure instead of the coupled structure. On the contrary, direct decoupling consists in adding to the coupled system a fictitious subsystem that is the negative of the residual subsystem. Starting from a reduced version of the 3-field formulation (dynamic equilibrium using FRFs, compatibility and equilibrium of interface forces), a direct hybrid assembly is developed by requiring that both compatibility and equilibrium conditions are satisfied exactly, either at coupling DoFs only, or at additional internal DoFs of the residual subsystem. Equilibrium and compatibility DoFs might not be the same: this generates the so-called non-collocated approach. The technique is applied using experimental data from an assembled system made by a plate and a rigid mass.

Sparse deconvolution for the large-scale ill-posed inverse problem of impact force reconstruction

NASA Astrophysics Data System (ADS)

Qiao, Baijie; Zhang, Xingwu; Gao, Jiawei; Liu, Ruonan; Chen, Xuefeng

2017-01-01

Most previous regularization methods for solving the inverse problem of force reconstruction are to minimize the l2-norm of the desired force. However, these traditional regularization methods such as Tikhonov regularization and truncated singular value decomposition, commonly fail to solve the large-scale ill-posed inverse problem in moderate computational cost. In this paper, taking into account the sparse characteristic of impact force, the idea of sparse deconvolution is first introduced to the field of impact force reconstruction and a general sparse deconvolution model of impact force is constructed. Second, a novel impact force reconstruction method based on the primal-dual interior point method (PDIPM) is proposed to solve such a large-scale sparse deconvolution model, where minimizing the l2-norm is replaced by minimizing the l1-norm. Meanwhile, the preconditioned conjugate gradient algorithm is used to compute the search direction of PDIPM with high computational efficiency. Finally, two experiments including the small-scale or medium-scale single impact force reconstruction and the relatively large-scale consecutive impact force reconstruction are conducted on a composite wind turbine blade and a shell structure to illustrate the advantage of PDIPM. Compared with Tikhonov regularization, PDIPM is more efficient, accurate and robust whether in the single impact force reconstruction or in the consecutive impact force reconstruction.

Time-lapse ERT interpretation methodology for leachate injection monitoring based on multiple inversions and a clustering strategy (MICS)

NASA Astrophysics Data System (ADS)

Audebert, M.; Clément, R.; Touze-Foltz, N.; Günther, T.; Moreau, S.; Duquennoi, C.

2014-12-01

Leachate recirculation is a key process in municipal waste landfills functioning as bioreactors. To quantify the water content and to assess the leachate injection system, in-situ methods are required to obtain spatially distributed information, usually electrical resistivity tomography (ERT). This geophysical method is based on the inversion process, which presents two major problems in terms of delimiting the infiltration area. First, it is difficult for ERT users to choose an appropriate inversion parameter set. Indeed, it might not be sufficient to interpret only the optimum model (i.e. the model with the chosen regularisation strength) because it is not necessarily the model which best represents the physical process studied. Second, it is difficult to delineate the infiltration front based on resistivity models because of the smoothness of the inversion results. This paper proposes a new methodology called MICS (multiple inversions and clustering strategy), which allows ERT users to improve the delimitation of the infiltration area in leachate injection monitoring. The MICS methodology is based on (i) a multiple inversion step by varying the inversion parameter values to take a wide range of resistivity models into account and (ii) a clustering strategy to improve the delineation of the infiltration front. In this paper, MICS was assessed on two types of data. First, a numerical assessment allows us to optimise and test MICS for different infiltration area sizes, contrasts and shapes. Second, MICS was applied to a field data set gathered during leachate recirculation on a bioreactor.

3-D linear inversion of gravity data: method and application to Basse-Terre volcanic island, Guadeloupe, Lesser Antilles

NASA Astrophysics Data System (ADS)

Barnoud, Anne; Coutant, Olivier; Bouligand, Claire; Gunawan, Hendra; Deroussi, Sébastien

2016-04-01

We use a Bayesian formalism combined with a grid node discretization for the linear inversion of gravimetric data in terms of 3-D density distribution. The forward modelling and the inversion method are derived from seismological inversion techniques in order to facilitate joint inversion or interpretation of density and seismic velocity models. The Bayesian formulation introduces covariance matrices on model parameters to regularize the ill-posed problem and reduce the non-uniqueness of the solution. This formalism favours smooth solutions and allows us to specify a spatial correlation length and to perform inversions at multiple scales. We also extract resolution parameters from the resolution matrix to discuss how well our density models are resolved. This method is applied to the inversion of data from the volcanic island of Basse-Terre in Guadeloupe, Lesser Antilles. A series of synthetic tests are performed to investigate advantages and limitations of the methodology in this context. This study results in the first 3-D density models of the island of Basse-Terre for which we identify: (i) a southward decrease of densities parallel to the migration of volcanic activity within the island, (ii) three dense anomalies beneath Petite Plaine Valley, Beaugendre Valley and the Grande-Découverte-Carmichaël-Soufrière Complex that may reflect the trace of former major volcanic feeding systems, (iii) shallow low-density anomalies in the southern part of Basse-Terre, especially around La Soufrière active volcano, Piton de Bouillante edifice and along the western coast, reflecting the presence of hydrothermal systems and fractured and altered rocks.

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.

Real-time inversions for finite fault slip models and rupture geometry based on high-rate GPS data

USGS Publications Warehouse

Minson, Sarah E.; Murray, Jessica R.; Langbein, John O.; Gomberg, Joan S.

2015-01-01

We present an inversion strategy capable of using real-time high-rate GPS data to simultaneously solve for a distributed slip model and fault geometry in real time as a rupture unfolds. We employ Bayesian inference to find the optimal fault geometry and the distribution of possible slip models for that geometry using a simple analytical solution. By adopting an analytical Bayesian approach, we can solve this complex inversion problem (including calculating the uncertainties on our results) in real time. Furthermore, since the joint inversion for distributed slip and fault geometry can be computed in real time, the time required to obtain a source model of the earthquake does not depend on the computational cost. Instead, the time required is controlled by the duration of the rupture and the time required for information to propagate from the source to the receivers. We apply our modeling approach, called Bayesian Evidence-based Fault Orientation and Real-time Earthquake Slip, to the 2011 Tohoku-oki earthquake, 2003 Tokachi-oki earthquake, and a simulated Hayward fault earthquake. In all three cases, the inversion recovers the magnitude, spatial distribution of slip, and fault geometry in real time. Since our inversion relies on static offsets estimated from real-time high-rate GPS data, we also present performance tests of various approaches to estimating quasi-static offsets in real time. We find that the raw high-rate time series are the best data to use for determining the moment magnitude of the event, but slightly smoothing the raw time series helps stabilize the inversion for fault geometry.

FOREWORD: 2nd International Workshop on New Computational Methods for Inverse Problems (NCMIP 2012)

NASA Astrophysics Data System (ADS)

Blanc-Féraud, Laure; Joubert, Pierre-Yves

2012-09-01

Conference logo This volume of Journal of Physics: Conference Series is dedicated to the scientific contributions presented during the 2nd International Workshop on New Computational Methods for Inverse Problems, (NCMIP 2012). This workshop took place at Ecole Normale Supérieure de Cachan, in Cachan, France, on 15 May 2012, at the initiative of Institut Farman. The first edition of NCMIP also took place in Cachan, France, within the scope of the ValueTools Conference, in May 2011 (http://www.ncmip.org/2011/). The NCMIP Workshop focused on recent advances in the resolution of inverse problems. Indeed inverse problems appear in numerous scientific areas such as geophysics, biological and medical imaging, material and structure characterization, electrical, mechanical and civil engineering, and finance. The resolution of inverse problems consists of estimating the parameters of the observed system or structure from data collected by an instrumental sensing or imaging device. Its success firstly requires the collection of relevant observation data. It also requires accurate models describing the physical interactions between the instrumental device and the observed system, as well as the intrinsic properties of the solution itself. Finally, it requires the design of robust, accurate and efficient inversion algorithms. Advanced sensor arrays and imaging devices provide high rate and high volume data; in this context, the efficient resolution of the inverse problem requires the joint development of new models and inversion methods, taking computational and implementation aspects into account. During this one-day workshop, researchers had the opportunity to bring to light and share new techniques and results in the field of inverse problems. The topics of the workshop were: algorithms and computational aspects of inversion, Bayesian estimation, kernel methods, learning methods, convex optimization, free discontinuity problems, metamodels, proper orthogonal decomposition, reduced models for the inversion, non-linear inverse scattering, image reconstruction and restoration, applications (bio-medical imaging, non-destructive evaluation etc). NCMIP 2012 was a one-day workshop. Each of the submitted papers was reviewed by 2 to 4 reviewers. Among the accepted papers, there are 8 oral presentations and 5 posters. Three international speakers were invited for a long talk. This second edition attracted 60 registered attendees in May 2012. NCMIP 2012 was supported by Institut Farman (ENS Cachan) and endorsed by the following French research networks (GDR ISIS, GDR Ondes, GDR MOA, GDR MSPC). The program committee acknowledges the following laboratories CMLA, LMT, LSV, LURPA, SATIE, as well as DIGITEO Network. Laure Blanc-Féraud and Pierre-Yves Joubert Workshop Co-chairs Laure Blanc-Féraud, I3S laboratory, CNRS, France Pierre-Yves Joubert, IEF laboratory, Paris-Sud University, CNRS, France Technical Program Committee Alexandre Baussard, ENSTA Bretagne, Lab-STICC, France Marc Bonnet, ENSTA, ParisTech, France Jerôme Darbon, CMLA, ENS Cachan, CNRS, France Oliver Dorn, School of Mathematics, University of Manchester, UK Mário Figueiredo, Instituto Superior Técnico, Lisbon, Portugal Laurent Fribourg, LSV, ENS Cachan, CNRS, France Marc Lambert, L2S Laboratory, CNRS, SupElec, Paris-Sud University, France Anthony Quinn, Trinity College, Dublin, Ireland Christian Rey, LMT, ENS Cachan, CNRS, France Joachim Weickert, Saarland University, Germany Local Chair Alejandro Mottini, Morpheme group I3S-INRIA Sophie Abriet, SATIE, ENS Cachan, CNRS, France Béatrice Bacquet, SATIE, ENS Cachan, CNRS, France Reviewers Gilles Aubert, J-A Dieudonné Laboratory, CNRS and University of Nice-Sophia Antipolis, France Alexandre Baussard, ENSTA Bretagne, Lab-STICC, France Laure Blanc-Féraud, I3S laboratory, CNRS, France Marc Bonnet, ENSTA, ParisTech, France Jerôme Darbon, CMLA, ENS Cachan, CNRS, France Oliver Dorn, School of Mathematics, University of Manchester, UK Gérard Favier, I3S laboratory, CNRS, France Mário Figueiredo, Instituto Superior Técnico, Lisbon, Portugal Laurent Fribourg, LSV, ENS Cachan, CNRS, France Jérôme Idier, IRCCyN, CNRS, France Pierre-Yves Joubert, IEF laboratory, Paris-Sud University, CNRS, France Marc Lambert, L2S Laboratory, CNRS, SupElec, Paris-Sud University, France Dominique Lesselier, L2S Laboratory, CNRS, SupElec, Paris-Sud University, France Anthony Quinn, Trinity College, Dublin, Ireland Christian Rey, LMT, ENS Cachan, CNRS, France Simon Setzer, Saarland University, Germany Joachim Weickert, Saarland University, Germany Invited speakers Antonin Chambolle: CMAP, Ecole Polytechnique, CNRS, France Matteo Pastorino: University of Genoa, Italy Michael Unser: Ecole polytechnique Fédérale de Lausanne, Switzerland

PREFACE: The Second International Conference on Inverse Problems: Recent Theoretical Developments and Numerical Approaches

NASA Astrophysics Data System (ADS)

Cheng, Jin; Hon, Yiu-Chung; Seo, Jin Keun; Yamamoto, Masahiro

2005-01-01

The Second International Conference on Inverse Problems: Recent Theoretical Developments and Numerical Approaches was held at Fudan University, Shanghai from 16-21 June 2004. The first conference in this series was held at the City University of Hong Kong in January 2002 and it was agreed to hold the conference once every two years in a Pan-Pacific Asian country. The next conference is scheduled to be held at Hokkaido University, Sapporo, Japan in July 2006. The purpose of this series of biennial conferences is to establish and develop constant international collaboration, especially among the Pan-Pacific Asian countries. In recent decades, interest in inverse problems has been flourishing all over the globe because of both the theoretical interest and practical requirements. In particular, in Asian countries, one is witnessing remarkable new trends of research in inverse problems as well as the participation of many young talents. Considering these trends, the second conference was organized with the chairperson Professor Li Tat-tsien (Fudan University), in order to provide forums for developing research cooperation and to promote activities in the field of inverse problems. Because solutions to inverse problems are needed in various applied fields, we entertained a total of 92 participants at the second conference and arranged various talks which ranged from mathematical analyses to solutions of concrete inverse problems in the real world. This volume contains 18 selected papers, all of which have undergone peer review. The 18 papers are classified as follows: Surveys: four papers give reviews of specific inverse problems. Theoretical aspects: six papers investigate the uniqueness, stability, and reconstruction schemes. Numerical methods: four papers devise new numerical methods and their applications to inverse problems. Solutions to applied inverse problems: four papers discuss concrete inverse problems such as scattering problems and inverse problems in atmospheric sciences and oceanography. Last but not least is our gratitude. As editors we would like to express our sincere thanks to all the plenary and invited speakers, the members of the International Scientific Committee and the Advisory Board for the success of the conference, which has given rise to this present volume of selected papers. We would also like to thank Mr Wang Yanbo, Miss Wan Xiqiong and the graduate students at Fudan University for their effective work to make this conference a success. The conference was financially supported by the NFS of China, the Mathematical Center of Ministry of Education of China, E-Institutes of Shanghai Municipal Education Commission (No E03004) and Fudan University, Grant 15340027 from the Japan Society for the Promotion of Science, and Grant 15654015 from the Ministry of Education, Cultures, Sports and Technology.

LS-APC v1.0: a tuning-free method for the linear inverse problem and its application to source-term determination

NASA Astrophysics Data System (ADS)

Tichý, Ondřej; Šmídl, Václav; Hofman, Radek; Stohl, Andreas

2016-11-01

Estimation of pollutant releases into the atmosphere is an important problem in the environmental sciences. It is typically formalized as an inverse problem using a linear model that can explain observable quantities (e.g., concentrations or deposition values) as a product of the source-receptor sensitivity (SRS) matrix obtained from an atmospheric transport model multiplied by the unknown source-term vector. Since this problem is typically ill-posed, current state-of-the-art methods are based on regularization of the problem and solution of a formulated optimization problem. This procedure depends on manual settings of uncertainties that are often very poorly quantified, effectively making them tuning parameters. We formulate a probabilistic model, that has the same maximum likelihood solution as the conventional method using pre-specified uncertainties. Replacement of the maximum likelihood solution by full Bayesian estimation also allows estimation of all tuning parameters from the measurements. The estimation procedure is based on the variational Bayes approximation which is evaluated by an iterative algorithm. The resulting method is thus very similar to the conventional approach, but with the possibility to also estimate all tuning parameters from the observations. The proposed algorithm is tested and compared with the standard methods on data from the European Tracer Experiment (ETEX) where advantages of the new method are demonstrated. A MATLAB implementation of the proposed algorithm is available for download.

Applications of the JARS method to study levee sites in southern Texas and southern New Mexico

USGS Publications Warehouse

Ivanov, J.; Miller, R.D.; Xia, J.; Dunbar, J.B.

2007-01-01

We apply the joint analysis of refractions with surface waves (JARS) method to several sites and compare its results to traditional refraction-tomography methods in efforts of finding a more realistic solution to the inverse refraction-traveltime problem. The JARS method uses a reference model, derived from surface-wave shear-wave velocity estimates, as a constraint. In all of the cases JARS estimates appear more realistic than those from the conventional refraction-tomography methods. As a result, we consider, the JARS algorithm as the preferred method for finding solutions to the inverse refraction-tomography problems. ?? 2007 Society of Exploration Geophysicists.

Riemann–Hilbert problem approach for two-dimensional flow inverse scattering

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

Agaltsov, A. D., E-mail: agalets@gmail.com; Novikov, R. G., E-mail: novikov@cmap.polytechnique.fr; IEPT RAS, 117997 Moscow

2014-10-15

We consider inverse scattering for the time-harmonic wave equation with first-order perturbation in two dimensions. This problem arises in particular in the acoustic tomography of moving fluid. We consider linearized and nonlinearized reconstruction algorithms for this problem of inverse scattering. Our nonlinearized reconstruction algorithm is based on the non-local Riemann–Hilbert problem approach. Comparisons with preceding results are given.

Regolith thermal property inversion in the LUNAR-A heat-flow experiment

NASA Astrophysics Data System (ADS)

Hagermann, A.; Tanaka, S.; Yoshida, S.; Fujimura, A.; Mizutani, H.

2001-11-01

In 2003, two penetrators of the LUNAR--A mission of ISAS will investigate the internal structure of the Moon by conducting seismic and heat--flow experiments. Heat-flow is the product of thermal gradient tial T / tial z, and thermal conductivity λ of the lunar regolith. For measuring the thermal conductivity (or dissusivity), each penetrator will carry five thermal property sensors, consisting of small disc heaters. The thermal response Ts(t) of the heater itself to the constant known power supply of approx. 50 mW serves as the data for the subsequent data interpretation. Horai et al. (1991) found a forward analytical solution to the problem of determining the thermal inertia λ ρ c of the regolith for constant thermal properties and a simplyfied geometry. In the inversion, the problem of deriving the unknown thermal properties of a medium from known heat sources and temperatures is an Identification Heat Conduction Problem (IDHCP), an ill--posed inverse problem. Assuming that thermal conductivity λ and heat capacity ρ c are linear functions of temperature (which is reasonable in most cases), one can apply a Kirchhoff transformation to linearize the heat conduction equation, which minimizes computing time. Then the error functional, i.e. the difference between the measured temperature response of the heater and the predicted temperature response, can be minimized, thus solving for thermal dissusivity κ = λ / (ρ c), wich will complete the set of parameters needed for a detailed description of thermal properties of the lunar regolith. Results of model calculations will be presented, in which synthetic data and calibration data are used to invert the unknown thermal diffusivity of the unknown medium by means of a modified Newton Method. Due to the ill-posedness of the problem, the number of parameters to be solved for should be limited. As the model calculations reveal, a homogeneous regolith allows for a fast and accurate inversion.

The Roles of Negative Career Thinking and Career Problem-Solving Self-Efficacy in Career Exploratory Behavior

ERIC Educational Resources Information Center

Bullock-Yowell, Emily; Katz, Sheba P.; Reardon, Robert C.; Peterson, Gary W.

2012-01-01

The respective roles of social cognitive career theory and cognitive information processing in career exploratory behavior were analyzed. A verified path model shows cognitive information processing theory's negative career thoughts inversely predict social cognitive career theory's career problem-solving self-efficacy, which predicts career…

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.

Technical Note: Atmospheric CO2 inversions on the mesoscale using data-driven prior uncertainties: methodology and system evaluation

NASA Astrophysics Data System (ADS)

Kountouris, Panagiotis; Gerbig, Christoph; Rödenbeck, Christian; Karstens, Ute; Koch, Thomas Frank; Heimann, Martin

2018-03-01

Atmospheric inversions are widely used in the optimization of surface carbon fluxes on a regional scale using information from atmospheric CO2 dry mole fractions. In many studies the prior flux uncertainty applied to the inversion schemes does not directly reflect the true flux uncertainties but is used to regularize the inverse problem. Here, we aim to implement an inversion scheme using the Jena inversion system and applying a prior flux error structure derived from a model-data residual analysis using high spatial and temporal resolution over a full year period in the European domain. We analyzed the performance of the inversion system with a synthetic experiment, in which the flux constraint is derived following the same residual analysis but applied to the model-model mismatch. The synthetic study showed a quite good agreement between posterior and true fluxes on European, country, annual and monthly scales. Posterior monthly and country-aggregated fluxes improved their correlation coefficient with the known truth by 7 % compared to the prior estimates when compared to the reference, with a mean correlation of 0.92. The ratio of the SD between the posterior and reference and between the prior and reference was also reduced by 33 % with a mean value of 1.15. We identified temporal and spatial scales on which the inversion system maximizes the derived information; monthly temporal scales at around 200 km spatial resolution seem to maximize the information gain.

Intrinsic nonlinearity and method of disturbed observations in inverse problems of celestial mechanics

NASA Astrophysics Data System (ADS)

Avdyushev, Victor A.

2017-12-01

Orbit determination from a small sample of observations over a very short observed orbital arc is a strongly nonlinear inverse problem. In such problems an evaluation of orbital uncertainty due to random observation errors is greatly complicated, since linear estimations conventionally used are no longer acceptable for describing the uncertainty even as a rough approximation. Nevertheless, if an inverse problem is weakly intrinsically nonlinear, then one can resort to the so-called method of disturbed observations (aka observational Monte Carlo). Previously, we showed that the weaker the intrinsic nonlinearity, the more efficient the method, i.e. the more accurate it enables one to simulate stochastically the orbital uncertainty, while it is strictly exact only when the problem is intrinsically linear. However, as we ascertained experimentally, its efficiency was found to be higher than that of other stochastic methods widely applied in practice. In the present paper we investigate the intrinsic nonlinearity in complicated inverse problems of Celestial Mechanics when orbits are determined from little informative samples of observations, which typically occurs for recently discovered asteroids. To inquire into the question, we introduce an index of intrinsic nonlinearity. In asteroid problems it evinces that the intrinsic nonlinearity can be strong enough to affect appreciably probabilistic estimates, especially at the very short observed orbital arcs that the asteroids travel on for about a hundredth of their orbital periods and less. As it is known from regression analysis, the source of intrinsic nonlinearity is the nonflatness of the estimation subspace specified by a dynamical model in the observation space. Our numerical results indicate that when determining asteroid orbits it is actually very slight. However, in the parametric space the effect of intrinsic nonlinearity is exaggerated mainly by the ill-conditioning of the inverse problem. Even so, as for the method of disturbed observations, we conclude that it practically should be still entirely acceptable to adequately describe the orbital uncertainty since, from a geometrical point of view, the efficiency of the method directly depends only on the nonflatness of the estimation subspace and it gets higher as the nonflatness decreases.

Multiple Fan-Beam Optical Tomography: Modelling Techniques

PubMed Central

Rahim, Ruzairi Abdul; Chen, Leong Lai; San, Chan Kok; Rahiman, Mohd Hafiz Fazalul; Fea, Pang Jon

2009-01-01

This paper explains in detail the solution to the forward and inverse problem faced in this research. In the forward problem section, the projection geometry and the sensor modelling are discussed. The dimensions, distributions and arrangements of the optical fibre sensors are determined based on the real hardware constructed and these are explained in the projection geometry section. The general idea in sensor modelling is to simulate an artificial environment, but with similar system properties, to predict the actual sensor values for various flow models in the hardware system. The sensitivity maps produced from the solution of the forward problems are important in reconstructing the tomographic image. PMID:22291523

Probabilistic Magnetotelluric Inversion with Adaptive Regularisation Using the No-U-Turns Sampler

NASA Astrophysics Data System (ADS)

Conway, Dennis; Simpson, Janelle; Didana, Yohannes; Rugari, Joseph; Heinson, Graham

2018-04-01

We present the first inversion of magnetotelluric (MT) data using a Hamiltonian Monte Carlo algorithm. The inversion of MT data is an underdetermined problem which leads to an ensemble of feasible models for a given dataset. A standard approach in MT inversion is to perform a deterministic search for the single solution which is maximally smooth for a given data-fit threshold. An alternative approach is to use Markov Chain Monte Carlo (MCMC) methods, which have been used in MT inversion to explore the entire solution space and produce a suite of likely models. This approach has the advantage of assigning confidence to resistivity models, leading to better geological interpretations. Recent advances in MCMC techniques include the No-U-Turns Sampler (NUTS), an efficient and rapidly converging method which is based on Hamiltonian Monte Carlo. We have implemented a 1D MT inversion which uses the NUTS algorithm. Our model includes a fixed number of layers of variable thickness and resistivity, as well as probabilistic smoothing constraints which allow sharp and smooth transitions. We present the results of a synthetic study and show the accuracy of the technique, as well as the fast convergence, independence of starting models, and sampling efficiency. Finally, we test our technique on MT data collected from a site in Boulia, Queensland, Australia to show its utility in geological interpretation and ability to provide probabilistic estimates of features such as depth to basement.

Extended fault inversion with random slipmaps: a resolution test for the 2012 Mw 7.6 Nicoya, Costa Rica earthquake

NASA Astrophysics Data System (ADS)

López-Comino, José Ángel; Stich, Daniel; Ferreira, Ana M. G.; Morales, Jose

2015-09-01

Inversions for the full slip distribution of earthquakes provide detailed models of earthquake sources, but stability and non-uniqueness of the inversions is a major concern. The problem is underdetermined in any realistic setting, and significantly different slip distributions may translate to fairly similar seismograms. In such circumstances, inverting for a single best model may become overly dependent on the details of the procedure. Instead, we propose to perform extended fault inversion trough falsification. We generate a representative set of heterogeneous slipmaps, compute their forward predictions, and falsify inappropriate trial models that do not reproduce the data within a reasonable level of mismodelling. The remainder of surviving trial models forms our set of coequal solutions. The solution set may contain only members with similar slip distributions, or else uncover some fundamental ambiguity such as, for example, different patterns of main slip patches. For a feasibility study, we use teleseismic body wave recordings from the 2012 September 5 Nicoya, Costa Rica earthquake, although the inversion strategy can be applied to any type of seismic, geodetic or tsunami data for which we can handle the forward problem. We generate 10 000 pseudo-random, heterogeneous slip distributions assuming a von Karman autocorrelation function, keeping the rake angle, rupture velocity and slip velocity function fixed. The slip distribution of the 2012 Nicoya earthquake turns out to be relatively well constrained from 50 teleseismic waveforms. Two hundred fifty-two slip models with normalized L1-fit within 5 per cent from the global minimum from our solution set. They consistently show a single dominant slip patch around the hypocentre. Uncertainties are related to the details of the slip maximum, including the amount of peak slip (2-3.5 m), as well as the characteristics of peripheral slip below 1 m. Synthetic tests suggest that slip patterns such as Nicoya may be a fortunate case, while it may be more difficult to unambiguously reconstruct more distributed slip from teleseismic data.

Convergence of Chahine's nonlinear relaxation inversion method used for limb viewing remote sensing

NASA Technical Reports Server (NTRS)

Chu, W. P.

1985-01-01

The application of Chahine's (1970) inversion technique to remote sensing problems utilizing the limb viewing geometry is discussed. The problem considered here involves occultation-type measurements and limb radiance-type measurements from either spacecraft or balloon platforms. The kernel matrix of the inversion problem is either an upper or lower triangular matrix. It is demonstrated that the Chahine inversion technique always converges, provided the diagonal elements of the kernel matrix are nonzero.

Object-based inversion of crosswell radar tomography data to monitor vegetable oil injection experiments

USGS Publications Warehouse

Lane, John W.; Day-Lewis, Frederick D.; Versteeg, Roelof J.; Casey, Clifton C.

2004-01-01

Crosswell radar methods can be used to dynamically image ground-water flow and mass transport associated with tracer tests, hydraulic tests, and natural physical processes, for improved characterization of preferential flow paths and complex aquifer heterogeneity. Unfortunately, because the raypath coverage of the interwell region is limited by the borehole geometry, the tomographic inverse problem is typically underdetermined, and tomograms may contain artifacts such as spurious blurring or streaking that confuse interpretation.We implement object-based inversion (using a constrained, non-linear, least-squares algorithm) to improve results from pixel-based inversion approaches that utilize regularization criteria, such as damping or smoothness. Our approach requires pre- and post-injection travel-time data. Parameterization of the image plane comprises a small number of objects rather than a large number of pixels, resulting in an overdetermined problem that reduces the need for prior information. The nature and geometry of the objects are based on hydrologic insight into aquifer characteristics, the nature of the experiment, and the planned use of the geophysical results.The object-based inversion is demonstrated using synthetic and crosswell radar field data acquired during vegetable-oil injection experiments at a site in Fridley, Minnesota. The region where oil has displaced ground water is discretized as a stack of rectangles of variable horizontal extents. The inversion provides the geometry of the affected region and an estimate of the radar slowness change for each rectangle. Applying petrophysical models to these results and porosity from neutron logs, we estimate the vegetable-oil emulsion saturation in various layers.Using synthetic- and field-data examples, object-based inversion is shown to be an effective strategy for inverting crosswell radar tomography data acquired to monitor the emplacement of vegetable-oil emulsions. A principal advantage of object-based inversion is that it yields images that hydrologists and engineers can easily interpret and use for model calibration.

Solution of Inverse Kinematics for 6R Robot Manipulators With Offset Wrist Based on Geometric Algebra.

PubMed

Fu, Zhongtao; Yang, Wenyu; Yang, Zhen

2013-08-01

In this paper, we present an efficient method based on geometric algebra for computing the solutions to the inverse kinematics problem (IKP) of the 6R robot manipulators with offset wrist. Due to the fact that there exist some difficulties to solve the inverse kinematics problem when the kinematics equations are complex, highly nonlinear, coupled and multiple solutions in terms of these robot manipulators stated mathematically, we apply the theory of Geometric Algebra to the kinematic modeling of 6R robot manipulators simply and generate closed-form kinematics equations, reformulate the problem as a generalized eigenvalue problem with symbolic elimination technique, and then yield 16 solutions. Finally, a spray painting robot, which conforms to the type of robot manipulators, is used as an example of implementation for the effectiveness and real-time of this method. The experimental results show that this method has a large advantage over the classical methods on geometric intuition, computation and real-time, and can be directly extended to all serial robot manipulators and completely automatized, which provides a new tool on the analysis and application of general robot manipulators.

IFSM fractal image compression with entropy and sparsity constraints: A sequential quadratic programming approach

NASA Astrophysics Data System (ADS)

Kunze, Herb; La Torre, Davide; Lin, Jianyi

2017-01-01

We consider the inverse problem associated with IFSM: Given a target function f , find an IFSM, such that its fixed point f ¯ is sufficiently close to f in the Lp distance. Forte and Vrscay [1] showed how to reduce this problem to a quadratic optimization model. In this paper, we extend the collage-based method developed by Kunze, La Torre and Vrscay ([2][3][4]), by proposing the minimization of the 1-norm instead of the 0-norm. In fact, optimization problems involving the 0-norm are combinatorial in nature, and hence in general NP-hard. To overcome these difficulties, we introduce the 1-norm and propose a Sequential Quadratic Programming algorithm to solve the corresponding inverse problem. As in Kunze, La Torre and Vrscay [3] in our formulation, the minimization of collage error is treated as a multi-criteria problem that includes three different and conflicting criteria i.e., collage error, entropy and sparsity. This multi-criteria program is solved by means of a scalarization technique which reduces the model to a single-criterion program by combining all objective functions with different trade-off weights. The results of some numerical computations are presented.

Sharp Boundary Inversion of 2D Magnetotelluric Data using Bayesian Method.

NASA Astrophysics Data System (ADS)

Zhou, S.; Huang, Q.

2017-12-01

Normally magnetotelluric(MT) inversion method cannot show the distribution of underground resistivity with clear boundary, even if there are obviously different blocks. Aiming to solve this problem, we develop a Bayesian structure to inverse 2D MT sharp boundary data, using boundary location and inside resistivity as the random variables. Firstly, we use other MT inversion results, like ModEM, to analyze the resistivity distribution roughly. Then, we select the suitable random variables and change its data format to traditional staggered grid parameters, which can be used to do finite difference forward part. Finally, we can shape the posterior probability density(PPD), which contains all the prior information and model-data correlation, by Markov Chain Monte Carlo(MCMC) sampling from prior distribution. The depth, resistivity and their uncertainty can be valued. It also works for sensibility estimation. We applied the method to a synthetic case, which composes two large abnormal blocks in a trivial background. We consider the boundary smooth and the near true model weight constrains that mimic joint inversion or constrained inversion, then we find that the model results a more precise and focused depth distribution. And we also test the inversion without constrains and find that the boundary could also be figured, though not as well. Both inversions have a good valuation of resistivity. The constrained result has a lower root mean square than ModEM inversion result. The data sensibility obtained via PPD shows that the resistivity is the most sensible, center depth comes second and both sides are the worst.

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

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.

Prize for Industrial Applications of Physics Talk: The Inverse Scattering Problem and the role of measurements in its solution

NASA Astrophysics Data System (ADS)

Wyatt, Philip

2009-03-01

The electromagnetic inverse scattering problem suggests that if a homogeneous and non-absorbing object be illuminated with a monochromatic light source and if the far field scattered light intensity is known at sufficient scattering angles, then, in principle, one could derive the dielectric structure of the scattering object. In general, this is an ill-posed problem and methods must be developed to regularize the search for unique solutions. An iterative procedure often begins with a model of the scattering object, solves the forward scattering problem using this model, and then compares these calculated results with the measured values. Key to any such solution is instrumentation capable of providing adequate data. To this end, the development of the first laser based absolute light scattering photometers is described together with their continuing evolution and some of the remarkable discoveries made with them. For particles much smaller than the wavelength of the incident light (e.g. macromolecules), the inverse scattering problems are easily solved. Among the many solutions derived with this instrumentation are the in situ structure of bacterial cells, new drug delivery mechanisms, the development of new vaccines and other biologicals, characterization of wines, the possibility of custom chemotherapy, development of new polymeric materials, identification of protein crystallization conditions, and a variety discoveries concerning protein interactions. A new form of the problem is described to address bioterrorist threats. Over the many years of development and refinement, one element stands out as essential for the successes that followed: the R and D teams were always directed and executed by physics trained theorists and experimentalists. 14 Ph. D. physicists each made his/her unique contribution to the development of these evolving instruments and the interpretation of their results.

A dynamical regularization algorithm for solving inverse source problems of elliptic partial differential equations

NASA Astrophysics Data System (ADS)

Zhang, Ye; Gong, Rongfang; Cheng, Xiaoliang; Gulliksson, Mårten

2018-06-01

This study considers the inverse source problem for elliptic partial differential equations with both Dirichlet and Neumann boundary data. The unknown source term is to be determined by additional boundary conditions. Unlike the existing methods found in the literature, which usually employ the first-order in time gradient-like system (such as the steepest descent methods) for numerically solving the regularized optimization problem with a fixed regularization parameter, we propose a novel method with a second-order in time dissipative gradient-like system and a dynamical selected regularization parameter. A damped symplectic scheme is proposed for the numerical solution. Theoretical analysis is given for both the continuous model and the numerical algorithm. Several numerical examples are provided to show the robustness of the proposed algorithm.

Inferring Spatial Variations of Microstructural Properties from Macroscopic Mechanical Response

PubMed Central

Liu, Tengxiao; Hall, Timothy J.; Barbone, Paul E.; Oberai, Assad A.

2016-01-01

Disease alters tissue microstructure, which in turn affects the macroscopic mechanical properties of tissue. In elasticity imaging, the macroscopic response is measured and is used to infer the spatial distribution of the elastic constitutive parameters. When an empirical constitutive model is used these parameters cannot be linked to the microstructure. However, when the constitutive model is derived from a microstructural representation of the material, it allows for the possibility of inferring the local averages of the spatial distribution of the microstructural parameters. This idea forms the basis of this study. In particular, we first derive a constitutive model by homogenizing the mechanical response of a network of elastic, tortuous fibers. Thereafter, we use this model in an inverse problem to determine the spatial distribution of the microstructural parameters. We solve the inverse problem as a constrained minimization problem, and develop efficient methods for solving it. We apply these methods to displacement fields obtained by deforming gelatin-agar co-gels, and determine the spatial distribution of agar concentration and fiber tortuosity, thereby demonstrating that it is possible to image local averages of microstructural parameters from macroscopic measurements of deformation. PMID:27655420

Elastic-Waveform Inversion with Compressive Sensing for Sparse Seismic Data

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

Lin, Youzuo; Huang, Lianjie

2015-01-28

Accurate velocity models of compressional- and shear-waves are essential for geothermal reservoir characterization and microseismic imaging. Elastic-waveform inversion of multi-component seismic data can provide high-resolution inversion results of subsurface geophysical properties. However, the method requires seismic data acquired using dense source and receiver arrays. In practice, seismic sources and/or geophones are often sparsely distributed on the surface and/or in a borehole, such as 3D vertical seismic profiling (VSP) surveys. We develop a novel elastic-waveform inversion method with compressive sensing for inversion of sparse seismic data. We employ an alternating-minimization algorithm to solve the optimization problem of our new waveform inversionmore » method. We validate our new method using synthetic VSP data for a geophysical model built using geologic features found at the Raft River enhanced-geothermal-system (EGS) field. We apply our method to synthetic VSP data with a sparse source array and compare the results with those obtained with a dense source array. Our numerical results demonstrate that the velocity models produced with our new method using a sparse source array are almost as accurate as those obtained using a dense source array.« less

Active and Passive Hydrologic Tomographic Surveys:A Revolution in Hydrology (Invited)

NASA Astrophysics Data System (ADS)

Yeh, T. J.

2013-12-01

Mathematical forward or inverse problems of flow through geological media always have unique solutions if necessary conditions are givens. Unique mathematical solutions to forward or inverse modeling of field problems are however always uncertain (an infinite number of possibilities) due to many reasons. They include non-representativeness of the governing equations, inaccurate necessary conditions, multi-scale heterogeneity, scale discrepancies between observation and model, noise and others. Conditional stochastic approaches, which derives the unbiased solution and quantifies the solution uncertainty, are therefore most appropriate for forward and inverse modeling of hydrological processes. Conditioning using non-redundant data sets reduces uncertainty. In this presentation, we explain non-redundant data sets in cross-hole aquifer tests, and demonstrate that active hydraulic tomographic survey (using man-made excitations) is a cost-effective approach to collect the same type but non-redundant data sets for reducing uncertainty in the inverse modeling. We subsequently show that including flux measurements (a piece of non-redundant data set) collected in the same well setup as in hydraulic tomography improves the estimated hydraulic conductivity field. We finally conclude with examples and propositions regarding how to collect and analyze data intelligently by exploiting natural recurrent events (river stage fluctuations, earthquakes, lightning, etc.) as energy sources for basin-scale passive tomographic surveys. The development of information fusion technologies that integrate traditional point measurements and active/passive hydrogeophysical tomographic surveys, as well as advances in sensor, computing, and information technologies may ultimately advance our capability of characterizing groundwater basins to achieve resolution far beyond the feat of current science and technology.

An accurate, fast, and scalable solver for high-frequency wave propagation

NASA Astrophysics Data System (ADS)

Zepeda-Núñez, L.; Taus, M.; Hewett, R.; Demanet, L.

2017-12-01

In many science and engineering applications, solving time-harmonic high-frequency wave propagation problems quickly and accurately is of paramount importance. For example, in geophysics, particularly in oil exploration, such problems can be the forward problem in an iterative process for solving the inverse problem of subsurface inversion. It is important to solve these wave propagation problems accurately in order to efficiently obtain meaningful solutions of the inverse problems: low order forward modeling can hinder convergence. Additionally, due to the volume of data and the iterative nature of most optimization algorithms, the forward problem must be solved many times. Therefore, a fast solver is necessary to make solving the inverse problem feasible. For time-harmonic high-frequency wave propagation, obtaining both speed and accuracy is historically challenging. Recently, there have been many advances in the development of fast solvers for such problems, including methods which have linear complexity with respect to the number of degrees of freedom. While most methods scale optimally only in the context of low-order discretizations and smooth wave speed distributions, the method of polarized traces has been shown to retain optimal scaling for high-order discretizations, such as hybridizable discontinuous Galerkin methods and for highly heterogeneous (and even discontinuous) wave speeds. The resulting fast and accurate solver is consequently highly attractive for geophysical applications. To date, this method relies on a layered domain decomposition together with a preconditioner applied in a sweeping fashion, which has limited straight-forward parallelization. In this work, we introduce a new version of the method of polarized traces which reveals more parallel structure than previous versions while preserving all of its other advantages. We achieve this by further decomposing each layer and applying the preconditioner to these new components separately and in parallel. We demonstrate that this produces an even more effective and parallelizable preconditioner for a single right-hand side. As before, additional speed can be gained by pipelining several right-hand-sides.

Using machine learning to accelerate sampling-based inversion

NASA Astrophysics Data System (ADS)

Valentine, A. P.; Sambridge, M.

2017-12-01

In most cases, a complete solution to a geophysical inverse problem (including robust understanding of the uncertainties associated with the result) requires a sampling-based approach. However, the computational burden is high, and proves intractable for many problems of interest. There is therefore considerable value in developing techniques that can accelerate sampling procedures.The main computational cost lies in evaluation of the forward operator (e.g. calculation of synthetic seismograms) for each candidate model. Modern machine learning techniques-such as Gaussian Processes-offer a route for constructing a computationally-cheap approximation to this calculation, which can replace the accurate solution during sampling. Importantly, the accuracy of the approximation can be refined as inversion proceeds, to ensure high-quality results.In this presentation, we describe and demonstrate this approach-which can be seen as an extension of popular current methods, such as the Neighbourhood Algorithm, and bridges the gap between prior- and posterior-sampling frameworks.

Hybrid modeling of spatial continuity for application to numerical inverse problems

USGS Publications Warehouse

Friedel, Michael J.; Iwashita, Fabio

2013-01-01

A novel two-step modeling approach is presented to obtain optimal starting values and geostatistical constraints for numerical inverse problems otherwise characterized by spatially-limited field data. First, a type of unsupervised neural network, called the self-organizing map (SOM), is trained to recognize nonlinear relations among environmental variables (covariates) occurring at various scales. The values of these variables are then estimated at random locations across the model domain by iterative minimization of SOM topographic error vectors. Cross-validation is used to ensure unbiasedness and compute prediction uncertainty for select subsets of the data. Second, analytical functions are fit to experimental variograms derived from original plus resampled SOM estimates producing model variograms. Sequential Gaussian simulation is used to evaluate spatial uncertainty associated with the analytical functions and probable range for constraining variables. The hybrid modeling of spatial continuity is demonstrated using spatially-limited hydrologic measurements at different scales in Brazil: (1) physical soil properties (sand, silt, clay, hydraulic conductivity) in the 42 km2 Vargem de Caldas basin; (2) well yield and electrical conductivity of groundwater in the 132 km2 fractured crystalline aquifer; and (3) specific capacity, hydraulic head, and major ions in a 100,000 km2 transboundary fractured-basalt aquifer. These results illustrate the benefits of exploiting nonlinear relations among sparse and disparate data sets for modeling spatial continuity, but the actual application of these spatial data to improve numerical inverse modeling requires testing.

Training-Image Based Geostatistical Inversion Using a Spatial Generative Adversarial Neural Network

NASA Astrophysics Data System (ADS)

Laloy, Eric; Hérault, Romain; Jacques, Diederik; Linde, Niklas

2018-01-01

Probabilistic inversion within a multiple-point statistics framework is often computationally prohibitive for high-dimensional problems. To partly address this, we introduce and evaluate a new training-image based inversion approach for complex geologic media. Our approach relies on a deep neural network of the generative adversarial network (GAN) type. After training using a training image (TI), our proposed spatial GAN (SGAN) can quickly generate 2-D and 3-D unconditional realizations. A key characteristic of our SGAN is that it defines a (very) low-dimensional parameterization, thereby allowing for efficient probabilistic inversion using state-of-the-art Markov chain Monte Carlo (MCMC) methods. In addition, available direct conditioning data can be incorporated within the inversion. Several 2-D and 3-D categorical TIs are first used to analyze the performance of our SGAN for unconditional geostatistical simulation. Training our deep network can take several hours. After training, realizations containing a few millions of pixels/voxels can be produced in a matter of seconds. This makes it especially useful for simulating many thousands of realizations (e.g., for MCMC inversion) as the relative cost of the training per realization diminishes with the considered number of realizations. Synthetic inversion case studies involving 2-D steady state flow and 3-D transient hydraulic tomography with and without direct conditioning data are used to illustrate the effectiveness of our proposed SGAN-based inversion. For the 2-D case, the inversion rapidly explores the posterior model distribution. For the 3-D case, the inversion recovers model realizations that fit the data close to the target level and visually resemble the true model well.

New Additions to the Toolkit for Forward/Inverse Problems in Electrocardiography within the SCIRun Problem Solving Environment.

PubMed

Coll-Font, Jaume; Burton, Brett M; Tate, Jess D; Erem, Burak; Swenson, Darrel J; Wang, Dafang; Brooks, Dana H; van Dam, Peter; Macleod, Rob S

2014-09-01

Cardiac electrical imaging often requires the examination of different forward and inverse problem formulations based on mathematical and numerical approximations of the underlying source and the intervening volume conductor that can generate the associated voltages on the surface of the body. If the goal is to recover the source on the heart from body surface potentials, the solution strategy must include numerical techniques that can incorporate appropriate constraints and recover useful solutions, even though the problem is badly posed. Creating complete software solutions to such problems is a daunting undertaking. In order to make such tools more accessible to a broad array of researchers, the Center for Integrative Biomedical Computing (CIBC) has made an ECG forward/inverse toolkit available within the open source SCIRun system. Here we report on three new methods added to the inverse suite of the toolkit. These new algorithms, namely a Total Variation method, a non-decreasing TMP inverse and a spline-based inverse, consist of two inverse methods that take advantage of the temporal structure of the heart potentials and one that leverages the spatial characteristics of the transmembrane potentials. These three methods further expand the possibilities of researchers in cardiology to explore and compare solutions to their particular imaging problem.

Inverse Problems in Economic Measurements

NASA Astrophysics Data System (ADS)

Shananin, A. A.

2018-02-01

The problem of economic measurements is discussed. The system of economic indices must reflect the economic relations and mechanisms existing in society. An achievement of the XX century is the development of a system of national accounts and the gross domestic product index. However, the gross domestic product index, which is related to the Hamilton-Pontryagin function in extensive economic growth models, turns out to be inadequate under the conditions of structural changes. New problems of integral geometry related to production models that take into account the substitution of production factors are considered.

Acoustic measurement of bubble size in an inkjet printhead.

PubMed

Jeurissen, Roger; van der Bos, Arjan; Reinten, Hans; van den Berg, Marc; Wijshoff, Herman; de Jong, Jos; Versluis, Michel; Lohse, Detlef

2009-11-01

The volume of a bubble in a piezoinkjet printhead is measured acoustically. The method is based on a numerical model of the investigated system. The piezo not only drives the system but it is also used as a sensor by measuring the current it generates. The numerical model is used to predict this current for a given bubble volume. The inverse problem is to infer the bubble volume from an experimentally obtained piezocurrent. By solving this inverse problem, the size and position of the bubble can thus be measured acoustically. The method is experimentally validated with an inkjet printhead that is augmented with a glass connection channel, through which the bubble was observed optically, while at the same time the piezocurrent was measured. The results from the acoustical measurement method correspond closely to the results from the optical measurement.

Moving from pixel to object scale when inverting radiative transfer models for quantitative estimation of biophysical variables in vegetation (Invited)

NASA Astrophysics Data System (ADS)

Atzberger, C.

2013-12-01

The robust and accurate retrieval of vegetation biophysical variables using RTM is seriously hampered by the ill-posedness of the inverse problem. The contribution presents our object-based inversion approach and evaluate it against measured data. The proposed method takes advantage of the fact that nearby pixels are generally more similar than those at a larger distance. For example, within a given vegetation patch, nearby pixels often share similar leaf angular distributions. This leads to spectral co-variations in the n-dimensional spectral features space, which can be used for regularization purposes. Using a set of leaf area index (LAI) measurements (n=26) acquired over alfalfa, sugar beet and garlic crops of the Barrax test site (Spain), it is demonstrated that the proposed regularization using neighbourhood information yields more accurate results compared to the traditional pixel-based inversion. Principle of the ill-posed inverse problem and the proposed solution illustrated in the red-nIR feature space using (PROSAIL). [A] spectral trajectory ('soil trajectory') obtained for one leaf angle (ALA) and one soil brightness (αsoil), when LAI varies between 0 and 10, [B] 'soil trajectories' for 5 soil brightness values and three leaf angles, [C] ill-posed inverse problem: different combinations of ALA × αsoil yield an identical crossing point, [D] object-based RTM inversion; only one 'soil trajectory' fits all nine pixelswithin a gliding (3×3) window. The black dots (plus the rectangle=central pixel) represent the hypothetical position of nine pixels within a 3×3 (gliding) window. Assuming that over short distances (× 1 pixel) variations in soil brightness can be neglected, the proposed object-based inversion searches for one common set of ALA × αsoil so that the resulting 'soil trajectory' best fits the nine measured pixels. Ground measured vs. retrieved LAI values for three crops. Left: proposed object-based approach. Right: pixel-based inversion

Multi-level damage identification with response reconstruction

NASA Astrophysics Data System (ADS)

Zhang, Chao-Dong; Xu, You-Lin

2017-10-01

Damage identification through finite element (FE) model updating usually forms an inverse problem. Solving the inverse identification problem for complex civil structures is very challenging since the dimension of potential damage parameters in a complex civil structure is often very large. Aside from enormous computation efforts needed in iterative updating, the ill-condition and non-global identifiability features of the inverse problem probably hinder the realization of model updating based damage identification for large civil structures. Following a divide-and-conquer strategy, a multi-level damage identification method is proposed in this paper. The entire structure is decomposed into several manageable substructures and each substructure is further condensed as a macro element using the component mode synthesis (CMS) technique. The damage identification is performed at two levels: the first is at macro element level to locate the potentially damaged region and the second is over the suspicious substructures to further locate as well as quantify the damage severity. In each level's identification, the damage searching space over which model updating is performed is notably narrowed down, not only reducing the computation amount but also increasing the damage identifiability. Besides, the Kalman filter-based response reconstruction is performed at the second level to reconstruct the response of the suspicious substructure for exact damage quantification. Numerical studies and laboratory tests are both conducted on a simply supported overhanging steel beam for conceptual verification. The results demonstrate that the proposed multi-level damage identification via response reconstruction does improve the identification accuracy of damage localization and quantization considerably.

Dynamic electrical impedance imaging with the interacting multiple model scheme.

PubMed

Kim, Kyung Youn; Kim, Bong Seok; Kim, Min Chan; Kim, Sin; Isaacson, David; Newell, Jonathan C

2005-04-01

In this paper, an effective dynamical EIT imaging scheme is presented for on-line monitoring of the abruptly changing resistivity distribution inside the object, based on the interacting multiple model (IMM) algorithm. The inverse problem is treated as a stochastic nonlinear state estimation problem with the time-varying resistivity (state) being estimated on-line with the aid of the IMM algorithm. In the design of the IMM algorithm multiple models with different process noise covariance are incorporated to reduce the modeling uncertainty. Simulations and phantom experiments are provided to illustrate the proposed algorithm.

Children's Understanding of the Inverse Relation between Multiplication and Division

ERIC Educational Resources Information Center

Robinson, Katherine M.; Dube, Adam K.

2009-01-01

Children's understanding of the inversion concept in multiplication and division problems (i.e., that on problems of the form "d multiplied by e/e" no calculations are required) was investigated. Children in Grades 6, 7, and 8 completed an inversion problem-solving task, an assessment of procedures task, and a factual knowledge task of simple…

Model Reduction via Principe Component Analysis and Markov Chain Monte Carlo (MCMC) Methods

NASA Astrophysics Data System (ADS)

Gong, R.; Chen, J.; Hoversten, M. G.; Luo, J.

2011-12-01

Geophysical and hydrogeological inverse problems often include a large number of unknown parameters, ranging from hundreds to millions, depending on parameterization and problems undertaking. This makes inverse estimation and uncertainty quantification very challenging, especially for those problems in two- or three-dimensional spatial domains. Model reduction technique has the potential of mitigating the curse of dimensionality by reducing total numbers of unknowns while describing the complex subsurface systems adequately. In this study, we explore the use of principal component analysis (PCA) and Markov chain Monte Carlo (MCMC) sampling methods for model reduction through the use of synthetic datasets. We compare the performances of three different but closely related model reduction approaches: (1) PCA methods with geometric sampling (referred to as 'Method 1'), (2) PCA methods with MCMC sampling (referred to as 'Method 2'), and (3) PCA methods with MCMC sampling and inclusion of random effects (referred to as 'Method 3'). We consider a simple convolution model with five unknown parameters as our goal is to understand and visualize the advantages and disadvantages of each method by comparing their inversion results with the corresponding analytical solutions. We generated synthetic data with noise added and invert them under two different situations: (1) the noised data and the covariance matrix for PCA analysis are consistent (referred to as the unbiased case), and (2) the noise data and the covariance matrix are inconsistent (referred to as biased case). In the unbiased case, comparison between the analytical solutions and the inversion results show that all three methods provide good estimates of the true values and Method 1 is computationally more efficient. In terms of uncertainty quantification, Method 1 performs poorly because of relatively small number of samples obtained, Method 2 performs best, and Method 3 overestimates uncertainty due to inclusion of random effects. However, in the biased case, only Method 3 correctly estimates all the unknown parameters, and both Methods 1 and 2 provide wrong values for the biased parameters. The synthetic case study demonstrates that if the covariance matrix for PCA analysis is inconsistent with true models, the PCA methods with geometric or MCMC sampling will provide incorrect estimates.

A Volunteer Computing Project for Solving Geoacoustic Inversion Problems

NASA Astrophysics Data System (ADS)

Zaikin, Oleg; Petrov, Pavel; Posypkin, Mikhail; Bulavintsev, Vadim; Kurochkin, Ilya

2017-12-01

A volunteer computing project aimed at solving computationally hard inverse problems in underwater acoustics is described. This project was used to study the possibilities of the sound speed profile reconstruction in a shallow-water waveguide using a dispersion-based geoacoustic inversion scheme. The computational capabilities provided by the project allowed us to investigate the accuracy of the inversion for different mesh sizes of the sound speed profile discretization grid. This problem suits well for volunteer computing because it can be easily decomposed into independent simpler subproblems.

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.

Analysis of the variability in ground-motion synthesis and inversion

USGS Publications Warehouse

Spudich, Paul A.; Cirella, Antonella; Scognamiglio, Laura; Tinti, Elisa

2017-12-07

In almost all past inversions of large-earthquake ground motions for rupture behavior, the goal of the inversion is to find the “best fitting” rupture model that predicts ground motions which optimize some function of the difference between predicted and observed ground motions. This type of inversion was pioneered in the linear-inverse sense by Olson and Apsel (1982), who minimized the square of the difference between observed and simulated motions (“least squares”) while simultaneously minimizing the rupture-model norm (by setting the null-space component of the rupture model to zero), and has been extended in many ways, one of which is the use of nonlinear inversion schemes such as simulated annealing algorithms that optimize some other misfit function. For example, the simulated annealing algorithm of Piatanesi and others (2007) finds the rupture model that minimizes a “cost” function which combines a least-squares and a waveform-correlation measure of misfit.All such inversions that look for a unique “best” model have at least three problems. (1) They have removed the null-space component of the rupture model—that is, an infinite family of rupture models that all fit the data equally well have been narrowed down to a single model. Some property of interest in the rupture model might have been discarded in this winnowing process. (2) Smoothing constraints are commonly used to yield a unique “best” model, in which case spatially rough rupture models will have been discarded, even if they provide a good fit to the data. (3) No estimate of confidence in the resulting rupture models can be given because the effects of unknown errors in the Green’s functions (“theory errors”) have not been assessed. In inversion for rupture behavior, these theory errors are generally larger than the data errors caused by ground noise and instrumental limitations, and so overfitting of the data is probably ubiquitous for such inversions.Recently, attention has turned to the inclusion of theory errors in the inversion process. Yagi and Fukahata (2011) made an important contribution by presenting a method to estimate the uncertainties in predicted large-earthquake ground motions due to uncertainties in the Green’s functions. Here we derive their result and compare it with the results of other recent studies that look at theory errors in a Bayesian inversion context particularly those by Bodin and others (2012), Duputel and others (2012), Dettmer and others (2014), and Minson and others (2014).Notably, in all these studies, the estimates of theory error were obtained from theoretical considerations alone; none of the investigators actually measured Green’s function errors. Large earthquakes typically have aftershocks, which, if their rupture surfaces are physically small enough, can be considered point evaluations of the real Green’s functions of the Earth. Here we simulate smallaftershock ground motions with (erroneous) theoretical Green’s functions. Taking differences between aftershock ground motions and simulated motions to be the “theory error,” we derive a statistical model of the sources of discrepancies between the theoretical and real Green’s functions. We use this model with an extended frequency-domain version of the time-domain theory of Yagi and Fukahata (2011) to determine the expected variance 2 τ caused by Green’s function error in ground motions from a larger (nonpoint) earthquake that we seek to model.We also differ from the above-mentioned Bayesian inversions in our handling of the nonuniqueness problem of seismic inversion. We follow the philosophy of Segall and Du (1993), who, instead of looking for a best-fitting model, looked for slip models that answered specific questions about the earthquakes they studied. In their Bayesian inversions, they inductively derived a posterior probability-density function (PDF) for every model parameter. We instead seek to find two extremal rupture models whose ground motions fit the data within the error bounds given by 2 τ , as quantified by using a chi-squared test described below. So, we can ask questions such as, “What are the rupture models with the highest and lowest average rupture speed consistent with the theory errors?” Having found those models, we can then say with confidence that the true rupture speed is somewhere between those values. Although the Bayesian approach gives a complete solution to the inverse problem, it is computationally demanding: Minson and others (2014) needed 1010 forward kinematic simulations to derive their posterior probability distribution. In our approach, only about107 simulations are needed. Moreover, in practical application, only a small set of rupture models may be needed to answer the relevant questions—for example, determining the maximum likelihood solution (achievable through standard inversion techniques) and the two rupture models bounding some property of interest.The specific property that we wish to investigate is the correlation between various rupturemodel parameters, such as peak slip velocity and rupture velocity, in models of real earthquakes. In some simulations of ground motions for hypothetical large earthquakes, such as those by Aagaard and others (2010) and the Southern California Earthquake Center Broadband Simulation Platform (Graves and Pitarka, 2015), rupture speed is assumed to correlate locally with peak slip, although there is evidence that rupture speed should correlate better with peak slip speed, owing to its dependence on local stress drop. We may be able to determine ways to modify Piatanesi and others’s (2007) inversion’s “cost” function to find rupture models with either high or low degrees of correlation between pairs of rupture parameters. We propose a cost function designed to find these two extremal models.

Conventional and reciprocal approaches to the inverse dipole localization problem for N(20)-P (20) somatosensory evoked potentials.

PubMed

Finke, Stefan; Gulrajani, Ramesh M; Gotman, Jean; Savard, Pierre

2013-01-01

The non-invasive localization of the primary sensory hand area can be achieved by solving the inverse problem of electroencephalography (EEG) for N(20)-P(20) somatosensory evoked potentials (SEPs). This study compares two different mathematical approaches for the computation of transfer matrices used to solve the EEG inverse problem. Forward transfer matrices relating dipole sources to scalp potentials are determined via conventional and reciprocal approaches using individual, realistically shaped head models. The reciprocal approach entails calculating the electric field at the dipole position when scalp electrodes are reciprocally energized with unit current-scalp potentials are obtained from the scalar product of this electric field and the dipole moment. Median nerve stimulation is performed on three healthy subjects and single-dipole inverse solutions for the N(20)-P(20) SEPs are then obtained by simplex minimization and validated against the primary sensory hand area identified on magnetic resonance images. Solutions are presented for different time points, filtering strategies, boundary-element method discretizations, and skull conductivity values. Both approaches produce similarly small position errors for the N(20)-P(20) SEP. Position error for single-dipole inverse solutions is inherently robust to inaccuracies in forward transfer matrices but dependent on the overlapping activity of other neural sources. Significantly smaller time and storage requirements are the principal advantages of the reciprocal approach. Reduced computational requirements and similar dipole position accuracy support the use of reciprocal approaches over conventional approaches for N(20)-P(20) SEP source localization.

Almost but not quite 2D, Non-linear Bayesian Inversion of CSEM Data

NASA Astrophysics Data System (ADS)

Ray, A.; Key, K.; Bodin, T.

2013-12-01

The geophysical inverse problem can be elegantly stated in a Bayesian framework where a probability distribution can be viewed as a statement of information regarding a random variable. After all, the goal of geophysical inversion is to provide information on the random variables of interest - physical properties of the earth's subsurface. However, though it may be simple to postulate, a practical difficulty of fully non-linear Bayesian inversion is the computer time required to adequately sample the model space and extract the information we seek. As a consequence, in geophysical problems where evaluation of a full 2D/3D forward model is computationally expensive, such as marine controlled source electromagnetic (CSEM) mapping of the resistivity of seafloor oil and gas reservoirs, Bayesian studies have largely been conducted with 1D forward models. While the 1D approximation is indeed appropriate for exploration targets with planar geometry and geological stratification, it only provides a limited, site-specific idea of uncertainty in resistivity with depth. In this work, we extend our fully non-linear 1D Bayesian inversion to a 2D model framework, without requiring the usual regularization of model resistivities in the horizontal or vertical directions used to stabilize quasi-2D inversions. In our approach, we use the reversible jump Markov-chain Monte-Carlo (RJ-MCMC) or trans-dimensional method and parameterize the subsurface in a 2D plane with Voronoi cells. The method is trans-dimensional in that the number of cells required to parameterize the subsurface is variable, and the cells dynamically move around and multiply or combine as demanded by the data being inverted. This approach allows us to expand our uncertainty analysis of resistivity at depth to more than a single site location, allowing for interactions between model resistivities at different horizontal locations along a traverse over an exploration target. While the model is parameterized in 2D, we efficiently evaluate the forward response using 1D profiles extracted from the model at the common-midpoints of the EM source-receiver pairs. Since the 1D approximation is locally valid at different midpoint locations, the computation time is far lower than is required by a full 2D or 3D simulation. We have applied this method to both synthetic and real CSEM survey data from the Scarborough gas field on the Northwest shelf of Australia, resulting in a spatially variable quantification of resistivity and its uncertainty in 2D. This Bayesian approach results in a large database of 2D models that comprise a posterior probability distribution, which we can subset to test various hypotheses about the range of model structures compatible with the data. For example, we can subset the model distributions to examine the hypothesis that a resistive reservoir extends overs a certain spatial extent. Depending on how this conditions other parts of the model space, light can be shed on the geological viability of the hypothesis. Since tackling spatially variable uncertainty and trade-offs in 2D and 3D is a challenging research problem, the insights gained from this work may prove valuable for subsequent full 2D and 3D Bayesian inversions.

Coupled Hydrogeophysical Inversion and Hydrogeological Data Fusion

NASA Astrophysics Data System (ADS)

Cirpka, O. A.; Schwede, R. L.; Li, W.

2012-12-01

Tomographic geophysical monitoring methods give the opportunity to observe hydrogeological tests at higher spatial resolution than is possible with classical hydraulic monitoring tools. This has been demonstrated in a substantial number of studies in which electrical resistivity tomography (ERT) has been used to monitor salt-tracer experiments. It is now accepted that inversion of such data sets requires a fully coupled framework, explicitly accounting for the hydraulic processes (groundwater flow and solute transport), the relationship between solute and geophysical properties (petrophysical relationship such as Archie's law), and the governing equations of the geophysical surveying techniques (e.g., the Poisson equation) as consistent coupled system. These data sets can be amended with data from other - more direct - hydrogeological tests to infer the distribution of hydraulic aquifer parameters. In the inversion framework, meaningful condensation of data does not only contribute to inversion efficiency but also increases the stability of the inversion. In particular, transient concentration data themselves only weakly depend on hydraulic conductivity, and model improvement using gradient-based methods is only possible when a substantial agreement between measurements and model output already exists. The latter also holds when concentrations are monitored by ERT. Tracer arrival times, by contrast, show high sensitivity and a more monotonic dependence on hydraulic conductivity than concentrations themselves. Thus, even without using temporal-moment generating equations, inverting travel times rather than concentrations or related geoelectrical signals themselves is advantageous. We have applied this approach to concentrations measured directly or via ERT, and to heat-tracer data. We present a consistent inversion framework including temporal moments of concentrations, geoelectrical signals obtained during salt-tracer tests, drawdown data from hydraulic tomography and flowmeter measurements to identify mainly the hydraulic-conductivity distribution. By stating the inversion as geostatistical conditioning problem, we obtain parameter sets together with their correlated uncertainty. While we have applied the quasi-linear geostatistical approach as inverse kernel, other methods - such as ensemble Kalman methods - may suit the same purpose, particularly when many data points are to be included. In order to identify 3-D fields, discretized by about 50 million grid points, we use the high-performance-computing framework DUNE to solve the involved partial differential equations on midrange computer cluster. We have quantified the worth of different data types in these inference problems. In practical applications, the constitutive relationships between geophysical, thermal, and hydraulic properties can pose a problem, requiring additional inversion. However, not well constrained transient boundary conditions may put inversion efforts on larger (e.g. regional) scales even more into question. We envision that future hydrogeophysical inversion efforts will target boundary conditions, such as groundwater recharge rates, in conjunction with - or instead of - aquifer parameters. By this, the distinction between data assimilation and parameter estimation will gradually vanish.

Diagnostic methods for atmospheric inversions of long-lived greenhouse gases

NASA Astrophysics Data System (ADS)

Michalak, Anna M.; Randazzo, Nina A.; Chevallier, Frédéric

2017-06-01

The ability to predict the trajectory of climate change requires a clear understanding of the emissions and uptake (i.e., surface fluxes) of long-lived greenhouse gases (GHGs). Furthermore, the development of climate policies is driving a need to constrain the budgets of anthropogenic GHG emissions. Inverse problems that couple atmospheric observations of GHG concentrations with an atmospheric chemistry and transport model have increasingly been used to gain insights into surface fluxes. Given the inherent technical challenges associated with their solution, it is imperative that objective approaches exist for the evaluation of such inverse problems. Because direct observation of fluxes at compatible spatiotemporal scales is rarely possible, diagnostics tools must rely on indirect measures. Here we review diagnostics that have been implemented in recent studies and discuss their use in informing adjustments to model setup. We group the diagnostics along a continuum starting with those that are most closely related to the scientific question being targeted, and ending with those most closely tied to the statistical and computational setup of the inversion. We thus begin with diagnostics based on assessments against independent information (e.g., unused atmospheric observations, large-scale scientific constraints), followed by statistical diagnostics of inversion results, diagnostics based on sensitivity tests, and analyses of robustness (e.g., tests focusing on the chemistry and transport model, the atmospheric observations, or the statistical and computational framework), and close with the use of synthetic data experiments (i.e., observing system simulation experiments, OSSEs). We find that existing diagnostics provide a crucial toolbox for evaluating and improving flux estimates but, not surprisingly, cannot overcome the fundamental challenges associated with limited atmospheric observations or the lack of direct flux measurements at compatible scales. As atmospheric inversions are increasingly expected to contribute to national reporting of GHG emissions, the need for developing and implementing robust and transparent evaluation approaches will only grow.

Simultaneous stochastic inversion for geomagnetic main field and secular variation. I - A large-scale inverse problem

NASA Technical Reports Server (NTRS)

Bloxham, Jeremy

1987-01-01

The method of stochastic inversion is extended to the simultaneous inversion of both main field and secular variation. In the present method, the time dependency is represented by an expansion in Legendre polynomials, resulting in a simple diagonal form for the a priori covariance matrix. The efficient preconditioned Broyden-Fletcher-Goldfarb-Shanno algorithm is used to solve the large system of equations resulting from expansion of the field spatially to spherical harmonic degree 14 and temporally to degree 8. Application of the method to observatory data spanning the 1900-1980 period results in a data fit of better than 30 nT, while providing temporally and spatially smoothly varying models of the magnetic field at the core-mantle boundary.

New Global 3D Upper to Mid-mantle Electrical Conductivity Model Based on Observatory Data with Realistic Auroral Sources

NASA Astrophysics Data System (ADS)

Kelbert, A.; Egbert, G. D.; Sun, J.

2011-12-01

Poleward of 45-50 degrees (geomagnetic) observatory data are influenced significantly by auroral ionospheric current systems, invalidating the simplifying zonal dipole source assumption traditionally used for long period (T > 2 days) geomagnetic induction studies. Previous efforts to use these data to obtain the global electrical conductivity distribution in Earth's mantle have omitted high-latitude sites (further thinning an already sparse dataset) and/or corrected the affected transfer functions using a highly simplified model of auroral source currents. Although these strategies are partly effective, there remain clear suggestions of source contamination in most recent 3D inverse solutions - specifically, bands of conductive features are found near auroral latitudes. We report on a new approach to this problem, based on adjusting both external field structure and 3D Earth conductivity to fit observatory data. As an initial step towards full joint inversion we are using a two step procedure. In the first stage, we adopt a simplified conductivity model, with a thin-sheet of variable conductance (to represent the oceans) overlying a 1D Earth, to invert observed magnetic fields for external source spatial structure. Input data for this inversion are obtained from frequency domain principal components (PC) analysis of geomagnetic observatory hourly mean values. To make this (essentially linear) inverse problem well-posed we regularize using covariances for source field structure that are consistent with well-established properties of auroral ionospheric (and magnetospheric) current systems, and basic physics of the EM fields. In the second stage, we use a 3D finite difference inversion code, with source fields estimated from the first stage, to further fit the observatory PC modes. We incorporate higher latitude data into the inversion, and maximize the amount of available information by directly inverting the magnetic field components of the PC modes, instead of transfer functions such as C-responses used previously. Recent improvements in accuracy and speed of the forward and inverse finite difference codes (a secondary field formulation and parallelization over frequencies) allow us to use finer computational grid for inversion, and thus to model finer scale features, making full use of the expanded data set. Overall, our approach presents an improvement over earlier observatory data interpretation techniques, making better use of the available data, and allowing to explore the trade-offs between complications in source structure, and heterogeneities in mantle conductivity. We will also report on progress towards applying the same approach to simultaneous source/conductivity inversion of shorter period observatory data, focusing especially on the daily variation band.

Evaluation of various procedures transposing global tilted irradiance to horizontal surface irradiance

NASA Astrophysics Data System (ADS)

Housmans, Caroline; Bertrand, Cédric

2017-02-01

Many transposition models have been proposed in the literature to convert solar irradiance on the horizontal plane to that on a tilted plane. The inverse process, i.e. the conversion from tilted to horizontal is investigated here based upon seven months of in-plane global solar irradiance measurements recorded on the roof of the Royal Meteorological Institute of Belgium's radiation tower in Uccle (Longitude 4.35° E, Latitude 50.79° N). Up to three pyranometers mounted on inclined planes of different tilts and orientations were involved in the inverse transposition process. Our results indicate that (1) the tilt to horizontal irradiance conversion is improved when measurements from more than one tilted pyranometer are considered (i.e. by using a multi-pyranometer approach) and (2) the improvement from using an isotropic model to anisotropic models in the inverse transposition problem is not significant.

A Generalized Approach for the Interpretation of Geophysical Well Logs in Ground-Water Studies - Theory and Application

USGS Publications Warehouse

Paillet, Frederick L.; Crowder, R.E.

1996-01-01

Quantitative analysis of geophysical logs in ground-water studies often involves at least as broad a range of applications and variation in lithology as is typically encountered in petroleum exploration, making such logs difficult to calibrate and complicating inversion problem formulation. At the same time, data inversion and analysis depend on inversion model formulation and refinement, so that log interpretation cannot be deferred to a geophysical log specialist unless active involvement with interpretation can be maintained by such an expert over the lifetime of the project. We propose a generalized log-interpretation procedure designed to guide hydrogeologists in the interpretation of geophysical logs, and in the integration of log data into ground-water models that may be systematically refined and improved in an iterative way. The procedure is designed to maximize the effective use of three primary contributions from geophysical logs: (1) The continuous depth scale of the measurements along the well bore; (2) The in situ measurement of lithologic properties and the correlation with hydraulic properties of the formations over a finite sample volume; and (3) Multiple independent measurements that can potentially be inverted for multiple physical or hydraulic properties of interest. The approach is formulated in the context of geophysical inversion theory, and is designed to be interfaced with surface geophysical soundings and conventional hydraulic testing. The step-by-step procedures given in our generalized interpretation and inversion technique are based on both qualitative analysis designed to assist formulation of the interpretation model, and quantitative analysis used to assign numerical values to model parameters. The approach bases a decision as to whether quantitative inversion is statistically warranted by formulating an over-determined inversion. If no such inversion is consistent with the inversion model, quantitative inversion is judged not possible with the given data set. Additional statistical criteria such as the statistical significance of regressions are used to guide the subsequent calibration of geophysical data in terms of hydraulic variables in those situations where quantitative data inversion is considered appropriate.

The ZH ratio method for long-period seismic data: inversion for S-wave velocity structure

NASA Astrophysics Data System (ADS)

Yano, Tomoko; Tanimoto, T.; Rivera, L.

2009-10-01

The particle motion of surface waves, in addition to phase and group velocities, can provide useful information for S-wave velocity structure in the crust and upper mantle. In this study, we applied a new method to retrieve velocity structure using the ZH ratio, the ratio between vertical and horizontal surface amplitudes of Rayleigh waves. Analysing data from the GEOSCOPE network, we measured the ZH ratios for frequencies between 0.004 and 0.05 Hz (period between 20 and 250s) and inverted them for S-wave velocity structure beneath each station. Our analysis showed that the resolving power of the ZH ratio is limited and final solutions display dependence on starting models; in particular, the depth of the Moho in the starting model is important in order to get reliable results. Thus, initial models for the inversion need to be carefully constructed. We chose PREM and CRUST2.0 in this study as a starting model for all but one station (ECH). The eigenvalue analysis of the least-squares problem that arises for each step of the iterative process shows a few dominant eigenvalues which explains the cause of the inversion's initial-model dependence. However, the ZH ratio is unique in having high sensitivity to near-surface structure and thus provides complementary information to phase and group velocities. Application of this method to GEOSCOPE data suggest that low velocity zones may exist beneath some stations near hotspots. Our tests with different starting models show that the models with low-velocity anomalies fit better to the ZH ratio data. Such low velocity zones are seen near Hawaii (station KIP), Crozet Island (CRZF) and Djibuti (ATD) but not near Reunion Island (RER). It is also found near Echery (ECH) which is in a geothermal area. However, this method has a tendency to produce spurious low velocity zones and resolution of the low velocity zones requires further careful study. We also performed simultaneous inversions for volumetric perturbation and discontinuity-depth perturbation. While its formulation and inversion were straightforward, there seemed to be a difficult trade-off problem between volumetric perturbation and discontinuity-depth perturbation.

Gender-Related Quality of Parent-Child Interactions and Early Adolescent Problem Behaviors: Exploratory Study with Midwestern Samples

ERIC Educational Resources Information Center

Spoth, Richard; Neppl, Tricia; Goldberg-Lillehoj, Catherine; Jung, Tony; Ramisetty-Mikler, Suhasini

2006-01-01

This article reports two exploratory studies testing a model guided by a social interactional perspective, positing an inverse relation between the quality of parent-child interactions and adolescent problem behaviors. It addresses mixed findings in the literature related to gender differences. Study 1 uses cross-sectional survey data from…

Nonlinear functional approximation with networks using adaptive neurons

NASA Technical Reports Server (NTRS)

Tawel, Raoul

1992-01-01

A novel mathematical framework for the rapid learning of nonlinear mappings and topological transformations is presented. It is based on allowing the neuron's parameters to adapt as a function of learning. This fully recurrent adaptive neuron model (ANM) has been successfully applied to complex nonlinear function approximation problems such as the highly degenerate inverse kinematics problem in robotics.

Comment on 'A reporting protocol for thermochronologic modeling illustrated with data from the Grand Canyon' by Flowers, Farley and Ketcham

NASA Astrophysics Data System (ADS)

Gallagher, Kerry

2016-05-01

Flowers et al. (2015) propose a framework for reporting modeling results for thermochronological data problems, particularly when using inversion approaches. In the final paragraph, they state 'we hope that the suggested reporting table template will stimulate additional community discussion about modeling philosophies and reporting formats'. In this spirit the purpose of this comment is to suggest that they have underplayed the importance of presenting a comparison of the model predictions with the observations. An inversion-based modeling approach aims to identify those models which makes predictions consistent, perhaps to varying degrees, with the observed data. The concluding section includes the phrase 'clear documentation of the model inputs and outputs', but their example from the Grand Canyon shows only the observed data.

Identification of dynamic characteristics of flexible rotors as dynamic inverse problem

NASA Technical Reports Server (NTRS)

Roisman, W. P.; Vajingortin, L. D.

1991-01-01

The problem of dynamic and balancing of flexible rotors were considered, which were set and solved as the problem of the identification of flexible rotor systems, which is the same as the inverse problem of the oscillation theory dealing with the task of the identifying the outside influences and system parameters on the basis of the known laws of motion. This approach to the problem allows the disclosure the picture of disbalances throughout the rotor-under-test (which traditional methods of flexible rotor balancing, based on natural oscillations, could not provide), and identify dynamic characteristics of the system, which correspond to a selected mathematical model. Eventually, various methods of balancing were developed depending on the special features of the machines as to their design, technology, and operation specifications. Also, theoretical and practical methods are given for the flexible rotor balancing at far from critical rotation frequencies, which does not necessarily require the knowledge forms of oscillation, dissipation, and elasticity and inertia characteristics, and to use testing masses.

Inverse Scattering Problem For The Schrödinger Equation With An Additional Quadratic Potential On The Entire Axis

NASA Astrophysics Data System (ADS)

Guseinov, I. M.; Khanmamedov, A. Kh.; Mamedova, A. F.

2018-04-01

We consider the Schrödinger equation with an additional quadratic potential on the entire axis and use the transformation operator method to study the direct and inverse problems of the scattering theory. We obtain the main integral equations of the inverse problem and prove that the basic equations are uniquely solvable.

Assessing non-uniqueness: An algebraic approach

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

Vasco, Don W.

Geophysical inverse problems are endowed with a rich mathematical structure. When discretized, most differential and integral equations of interest are algebraic (polynomial) in form. Techniques from algebraic geometry and computational algebra provide a means to address questions of existence and uniqueness for both linear and non-linear inverse problem. In a sense, the methods extend ideas which have proven fruitful in treating linear inverse problems.

INTRODUCTION Introduction to the conference proceeding of the Workshop on Electromagnetic Inverse ProblemsThe University of Manchester, UK, 15-18 June, 2009

NASA Astrophysics Data System (ADS)

Dorn, Oliver; Lionheart, Bill

2010-11-01

This proceeding combines selected contributions from participants of the Workshop on Electromagnetic Inverse Problems which was hosted by the University of Manchester in June 2009. The workshop was organized by the two guest editors of this conference proceeding and ran in parallel to the 10th International Conference on Electrical Impedance Tomography, which was guided by Bill Lionheart, Richard Bayford, and Eung Je Woo. Both events shared plenary talks and several selected sessions. One reason for combining these two events was the goal of bringing together scientists from various related disciplines who normally might not attend the same conferences, and to enhance discussions between these different groups. So, for example, one day of the workshop was dedicated to the broader area of geophysical inverse problems (including inverse problems in petroleum engineering), where participants from the EIT community and from the medical imaging community were also encouraged to participate, with great success. Other sessions concentrated on microwave medical imaging, on inverse scattering, or on eddy current imaging, with active feedback also from geophysically oriented scientists. Furthermore, several talks addressed such diverse topics as optical tomography, photoacoustic tomography, time reversal, or electrosensing fish. As a result of the workshop, speakers were invited to contribute extended papers to this conference proceeding. All submissions were thoroughly reviewed and, after a thoughtful revision by the authors, combined in this proceeding. The resulting set of six papers presenting the work of in total 22 authors from 5 different countries provides a very interesting overview of several of the themes which were represented at the workshop. These can be divided into two important categories, namely (i) modelling and (ii) data inversion. The first three papers of this selection, as outlined below, focus more on modelling aspects, being an essential component of any successful inversion, whereas the other three papers discuss novel inversion techniques for specific applications. In the first contribution, with the title A Novel Simplified Mathematical Model for Antennas used in Medical Imaging Applications, the authors M J Fernando, M Elsdon, K Busawon and D Smith discuss a new technique for modelling the current across a monopole antenna from which the radiation fields of the antenna can be calculated very efficiently in specific medical imaging applications. This new technique is then tested on two examples, a quarter wavelength and a three quarter wavelength monopole antenna. The next contribution, with the title An investigation into the use of a mixture model for simulating the electrical properties of soil with varying effective saturation levels for sub-soil imaging using ECT by R R Hayes, P A Newill, F J W Podd, T A York, B D Grieve and O Dorn, considers the development of a new visualization tool for monitoring soil moisture content surrounding certain seed breeder plants. An electrical capacitance tomography technique is employed for verifying how efficiently each plant utilises the water and nutrients available in the surrounding soil. The goal of this study is to help in developing and identifying new drought tolerant food crops. In the third contribution Combination of Maximin and Kriging Prediction Methods for Eddy-Current Testing Database Generation by S Bilicz, M Lambert, E Vazquez and S Gyimóthy, a novel database generation technique is proposed for its use in solving inverse eddy-current testing problems. For avoiding expensive repeated forward simulations during the creation of this database, a kriging interpolation technique is employed for filling uniformly the data output space with sample points. Mathematically this is achieved by using a maximin formalism. The paper 2.5D inversion of CSEM data in a vertically anisotropic earth by C Ramananjaona and L MacGregor considers controlled-source electromagnetic techniques for imaging the earth in a marine environment. It focuses in particular on taking into account anisotropy effects in the inversion. Results of this technique are demonstrated from simulated and from real field data. Furthermore, in the contribution Multiple level-sets for elliptic Cauchy problems in three-dimensional domains by A Leitão and M Marques Alves the authors consider a TV-H1regularization technique for multiple level-set inversion of elliptic Cauchy problems. Generalized minimizers are defined and convergence and stability results are provided for this method, in addition to several numerical experiments. Finally, in the paper Development of in-vivo fluorescence imaging with the matrix-free method, the authors A Zacharopoulos, A Garofalakis, J Ripoll and S Arridge address a recently developed non-contact fluorescence molecular tomography technique where the use of non-contact acquisition systems poses new challenges on computational efficiency during data processing. The matrix-free method is designed to reduce computational cost and memory requirements during the inversion. Reconstructions from a simulated mouse phantom are provided for demonstrating the performance of the proposed technique in realistic scenarios. We hope that this selection of strong and thought-provoking papers will help stimulating further cross-disciplinary research in the spirit of the workshop. We thank all authors for providing us with this excellent set of high-quality contributions. We also thank EPSRC for having provided funding for the workshop under grant EP/G065047/1. Oliver Dorn, Bill Lionheart School of Mathematics, University of Manchester, Alan Turing Building, Oxford Rd Manchester, M13 9PL, UK E-mail: oliver.dorn@manchester.ac.uk, bill.lionheart@manchester.ac.uk Guest Editors

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.

Review on solving the forward problem in EEG source analysis

PubMed Central

Hallez, Hans; Vanrumste, Bart; Grech, Roberta; Muscat, Joseph; De Clercq, Wim; Vergult, Anneleen; D'Asseler, Yves; Camilleri, Kenneth P; Fabri, Simon G; Van Huffel, Sabine; Lemahieu, Ignace

2007-01-01

Background The aim of electroencephalogram (EEG) source localization is to find the brain areas responsible for EEG waves of interest. It consists of solving forward and inverse problems. The forward problem is solved by starting from a given electrical source and calculating the potentials at the electrodes. These evaluations are necessary to solve the inverse problem which is defined as finding brain sources which are responsible for the measured potentials at the EEG electrodes. Methods While other reviews give an extensive summary of the both forward and inverse problem, this review article focuses on different aspects of solving the forward problem and it is intended for newcomers in this research field. Results It starts with focusing on the generators of the EEG: the post-synaptic potentials in the apical dendrites of pyramidal neurons. These cells generate an extracellular current which can be modeled by Poisson's differential equation, and Neumann and Dirichlet boundary conditions. The compartments in which these currents flow can be anisotropic (e.g. skull and white matter). In a three-shell spherical head model an analytical expression exists to solve the forward problem. During the last two decades researchers have tried to solve Poisson's equation in a realistically shaped head model obtained from 3D medical images, which requires numerical methods. The following methods are compared with each other: the boundary element method (BEM), the finite element method (FEM) and the finite difference method (FDM). In the last two methods anisotropic conducting compartments can conveniently be introduced. Then the focus will be set on the use of reciprocity in EEG source localization. It is introduced to speed up the forward calculations which are here performed for each electrode position rather than for each dipole position. Solving Poisson's equation utilizing FEM and FDM corresponds to solving a large sparse linear system. Iterative methods are required to solve these sparse linear systems. The following iterative methods are discussed: successive over-relaxation, conjugate gradients method and algebraic multigrid method. Conclusion Solving the forward problem has been well documented in the past decades. In the past simplified spherical head models are used, whereas nowadays a combination of imaging modalities are used to accurately describe the geometry of the head model. Efforts have been done on realistically describing the shape of the head model, as well as the heterogenity of the tissue types and realistically determining the conductivity. However, the determination and validation of the in vivo conductivity values is still an important topic in this field. In addition, more studies have to be done on the influence of all the parameters of the head model and of the numerical techniques on the solution of the forward problem. PMID:18053144

Porosity Estimation By Artificial Neural Networks Inversion . Application to Algerian South Field

NASA Astrophysics Data System (ADS)

Eladj, Said; Aliouane, Leila; Ouadfeul, Sid-Ali

2017-04-01

One of the main geophysicist's current challenge is the discovery and the study of stratigraphic traps, this last is a difficult task and requires a very fine analysis of the seismic data. The seismic data inversion allows obtaining lithological and stratigraphic information for the reservoir characterization . However, when solving the inverse problem we encounter difficult problems such as: Non-existence and non-uniqueness of the solution add to this the instability of the processing algorithm. Therefore, uncertainties in the data and the non-linearity of the relationship between the data and the parameters must be taken seriously. In this case, the artificial intelligence techniques such as Artificial Neural Networks(ANN) is used to resolve this ambiguity, this can be done by integrating different physical properties data which requires a supervised learning methods. In this work, we invert the acoustic impedance 3D seismic cube using the colored inversion method, then, the introduction of the acoustic impedance volume resulting from the first step as an input of based model inversion method allows to calculate the Porosity volume using the Multilayer Perceptron Artificial Neural Network. Application to an Algerian South hydrocarbon field clearly demonstrate the power of the proposed processing technique to predict the porosity for seismic data, obtained results can be used for reserves estimation, permeability prediction, recovery factor and reservoir monitoring. Keywords: Artificial Neural Networks, inversion, non-uniqueness , nonlinear, 3D porosity volume, reservoir characterization .

Anisotropic three-dimensional inversion of CSEM data using finite-element techniques on unstructured grids

NASA Astrophysics Data System (ADS)

Wang, Feiyan; Morten, Jan Petter; Spitzer, Klaus

2018-05-01

In this paper, we present a recently developed anisotropic 3-D inversion framework for interpreting controlled-source electromagnetic (CSEM) data in the frequency domain. The framework integrates a high-order finite-element forward operator and a Gauss-Newton inversion algorithm. Conductivity constraints are applied using a parameter transformation. We discretize the continuous forward and inverse problems on unstructured grids for a flexible treatment of arbitrarily complex geometries. Moreover, an unstructured mesh is more desirable in comparison to a single rectilinear mesh for multisource problems because local grid refinement will not significantly influence the mesh density outside the region of interest. The non-uniform spatial discretization facilitates parametrization of the inversion domain at a suitable scale. For a rapid simulation of multisource EM data, we opt to use a parallel direct solver. We further accelerate the inversion process by decomposing the entire data set into subsets with respect to frequencies (and transmitters if memory requirement is affordable). The computational tasks associated with each data subset are distributed to different processes and run in parallel. We validate the scheme using a synthetic marine CSEM model with rough bathymetry, and finally, apply it to an industrial-size 3-D data set from the Troll field oil province in the North Sea acquired in 2008 to examine its robustness and practical applicability.

The investigation of advanced remote sensing techniques for the measurement of aerosol characteristics

NASA Technical Reports Server (NTRS)

Deepak, A.; Becher, J.

1979-01-01

Advanced remote sensing techniques and inversion methods for the measurement of characteristics of aerosol and gaseous species in the atmosphere were investigated. Of particular interest were the physical and chemical properties of aerosols, such as their size distribution, number concentration, and complex refractive index, and the vertical distribution of these properties on a local as well as global scale. Remote sensing techniques for monitoring of tropospheric aerosols were developed as well as satellite monitoring of upper tropospheric and stratospheric aerosols. Computer programs were developed for solving multiple scattering and radiative transfer problems, as well as inversion/retrieval problems. A necessary aspect of these efforts was to develop models of aerosol properties.

Methods for the identification of material parameters in distributed models for flexible structures

NASA Technical Reports Server (NTRS)

Banks, H. T.; Crowley, J. M.; Rosen, I. G.

1986-01-01

Theoretical and numerical results are presented for inverse problems involving estimation of spatially varying parameters such as stiffness and damping in distributed models for elastic structures such as Euler-Bernoulli beams. An outline of algorithms used and a summary of computational experiences are presented.

Numerical convergence and validation of the DIMP inverse particle transport model

DOE PAGES

Nelson, Noel; Azmy, Yousry

2017-09-01

The data integration with modeled predictions (DIMP) model is a promising inverse radiation transport method for solving the special nuclear material (SNM) holdup problem. Unlike previous methods, DIMP is a completely passive nondestructive assay technique that requires no initial assumptions regarding the source distribution or active measurement time. DIMP predicts the most probable source location and distribution through Bayesian inference and quasi-Newtonian optimization of predicted detector re-sponses (using the adjoint transport solution) with measured responses. DIMP performs well with for-ward hemispherical collimation and unshielded measurements, but several considerations are required when using narrow-view collimated detectors. DIMP converged well to themore » correct source distribution as the number of synthetic responses increased. DIMP also performed well for the first experimental validation exercise after applying a collimation factor, and sufficiently reducing the source search vol-ume's extent to prevent the optimizer from getting stuck in local minima. DIMP's simple point detector response function (DRF) is being improved to address coplanar false positive/negative responses, and an angular DRF is being considered for integration with the next version of DIMP to account for highly collimated responses. Overall, DIMP shows promise for solving the SNM holdup inverse problem, especially once an improved optimization algorithm is implemented.« less

Quantifying the Uncertainties and Multi-parameter Trade-offs in Joint Inversion of Receiver Functions and Surface Wave Velocity and Ellipticity

NASA Astrophysics Data System (ADS)

Gao, C.; Lekic, V.

2016-12-01

When constraining the structure of the Earth's continental lithosphere, multiple seismic observables are often combined due to their complementary sensitivities.The transdimensional Bayesian (TB) approach in seismic inversion allows model parameter uncertainties and trade-offs to be quantified with few assumptions. TB sampling yields an adaptive parameterization that enables simultaneous inversion for different model parameters (Vp, Vs, density, radial anisotropy), without the need for strong prior information or regularization. We use a reversible jump Markov chain Monte Carlo (rjMcMC) algorithm to incorporate different seismic observables - surface wave dispersion (SWD), Rayleigh wave ellipticity (ZH ratio), and receiver functions - into the inversion for the profiles of shear velocity (Vs), compressional velocity (Vp), density (ρ), and radial anisotropy (ξ) beneath a seismic station. By analyzing all three data types individually and together, we show that TB sampling can eliminate the need for a fixed parameterization based on prior information, and reduce trade-offs in model estimates. We then explore the effect of different types of misfit functions for receiver function inversion, which is a highly non-unique problem. We compare the synthetic inversion results using the L2 norm, cross-correlation type and integral type misfit function by their convergence rates and retrieved seismic structures. In inversions in which only one type of model parameter (Vs for the case of SWD) is inverted, assumed scaling relationships are often applied to account for sensitivity to other model parameters (e.g. Vp, ρ, ξ). Here we show that under a TB framework, we can eliminate scaling assumptions, while simultaneously constraining multiple model parameters to varying degrees. Furthermore, we compare the performance of TB inversion when different types of model parameters either share the same or use independent parameterizations. We show that different parameterizations can lead to differences in retrieved model parameters, consistent with limited data constraints. We then quantitatively examine the model parameter trade-offs and find that trade-offs between Vp and radial anisotropy might limit our ability to constrain shallow-layer radial anisotropy using current seismic observables.

The Earth isn't flat: The (large) influence of topography on geodetic fault slip imaging.

NASA Astrophysics Data System (ADS)

Thompson, T. B.; Meade, B. J.

2017-12-01

While earthquakes both occur near and generate steep topography, most geodetic slip inversions assume that the Earth's surface is flat. We have developed a new boundary element tool, Tectosaur, with the capability to study fault and earthquake problems including complex fault system geometries, topography, material property contrasts, and millions of elements. Using Tectosaur, we study the model error induced by neglecting topography in both idealized synthetic fault models and for the cases of the MW=7.3 Landers and MW=8.0 Wenchuan earthquakes. Near the steepest topography, we find the use of flat Earth dislocation models may induce errors of more than 100% in the inferred slip magnitude and rake. In particular, neglecting topographic effects leads to an inferred shallow slip deficit. Thus, we propose that the shallow slip deficit observed in several earthquakes may be an artefact resulting from the systematic use of elastic dislocation models assuming a flat Earth. Finally, using this study as an example, we emphasize the dangerous potential for forward model errors to be amplified by an order of magnitude in inverse problems.

Joint inversion of multiple geophysical and petrophysical data using generalized fuzzy clustering algorithms

NASA Astrophysics Data System (ADS)

Sun, Jiajia; Li, Yaoguo

2017-02-01

Joint inversion that simultaneously inverts multiple geophysical data sets to recover a common Earth model is increasingly being applied to exploration problems. Petrophysical data can serve as an effective constraint to link different physical property models in such inversions. There are two challenges, among others, associated with the petrophysical approach to joint inversion. One is related to the multimodality of petrophysical data because there often exist more than one relationship between different physical properties in a region of study. The other challenge arises from the fact that petrophysical relationships have different characteristics and can exhibit point, linear, quadratic, or exponential forms in a crossplot. The fuzzy c-means (FCM) clustering technique is effective in tackling the first challenge and has been applied successfully. We focus on the second challenge in this paper and develop a joint inversion method based on variations of the FCM clustering technique. To account for the specific shapes of petrophysical relationships, we introduce several different fuzzy clustering algorithms that are capable of handling different shapes of petrophysical relationships. We present two synthetic and one field data examples and demonstrate that, by choosing appropriate distance measures for the clustering component in the joint inversion algorithm, the proposed joint inversion method provides an effective means of handling common petrophysical situations we encounter in practice. The jointly inverted models have both enhanced structural similarity and increased petrophysical correlation, and better represent the subsurface in the spatial domain and the parameter domain of physical properties.

Blocky inversion of multichannel elastic impedance for elastic parameters

NASA Astrophysics Data System (ADS)

Mozayan, Davoud Karami; Gholami, Ali; Siahkoohi, Hamid Reza

2018-04-01

Petrophysical description of reservoirs requires proper knowledge of elastic parameters like P- and S-wave velocities (Vp and Vs) and density (ρ), which can be retrieved from pre-stack seismic data using the concept of elastic impedance (EI). We propose an inversion algorithm which recovers elastic parameters from pre-stack seismic data in two sequential steps. In the first step, using the multichannel blind seismic inversion method (exploited recently for recovering acoustic impedance from post-stack seismic data), high-resolution blocky EI models are obtained directly from partial angle-stacks. Using an efficient total-variation (TV) regularization, each angle-stack is inverted independently in a multichannel form without prior knowledge of the corresponding wavelet. The second step involves inversion of the resulting EI models for elastic parameters. Mathematically, under some assumptions, the EI's are linearly described by the elastic parameters in the logarithm domain. Thus a linear weighted least squares inversion is employed to perform this step. Accuracy of the concept of elastic impedance in predicting reflection coefficients at low and high angles of incidence is compared with that of exact Zoeppritz elastic impedance and the role of low frequency content in the problem is discussed. The performance of the proposed inversion method is tested using synthetic 2D data sets obtained from the Marmousi model and also 2D field data sets. The results confirm the efficiency and accuracy of the proposed method for inversion of pre-stack seismic data.

A fixed energy fixed angle inverse scattering in interior transmission problem

NASA Astrophysics Data System (ADS)

Chen, Lung-Hui

2017-06-01

We study the inverse acoustic scattering problem in mathematical physics. The problem is to recover the index of refraction in an inhomogeneous medium by measuring the scattered wave fields in the far field. We transform the problem to the interior transmission problem in the study of the Helmholtz equation. We find an inverse uniqueness on the scatterer with a knowledge of a fixed interior transmission eigenvalue. By examining the solution in a series of spherical harmonics in the far field, we can determine uniquely the perturbation source for the radially symmetric perturbations.

Numerical simulations of induction and MWD logging tools and data inversion method with X-window interface on a UNIX workstation

NASA Astrophysics Data System (ADS)

Tian, Xiang-Dong

The purpose of this research is to simulate induction and measuring-while-drilling (MWD) logs. In simulation of logs, there are two tasks. The first task, the forward modeling procedure, is to compute the logs from known formation. The second task, the inversion procedure, is to determine the unknown properties of the formation from the measured field logs. In general, the inversion procedure requires the solution of a forward model. In this study, a stable numerical method to simulate induction and MWD logs is presented. The proposed algorithm is based on a horizontal eigenmode expansion method. Vertical propagation of modes is modeled by a three-layer module. The multilayer cases are treated as a cascade of these modules. The mode tracing algorithm possesses stable characteristics that are superior to other methods. This method is applied to simulate the logs in the formations with both vertical and horizontal layers, and also used to study the groove effects of the MWD tool. The results are very good. Two-dimensional inversion of induction logs is an nonlinear problem. Nonlinear functions of the apparent conductivity are expanded into a Taylor series. After truncating the high order terms in this Taylor series, the nonlinear functions are linearized. An iterative procedure is then devised to solve the inversion problem. In each iteration, the Jacobian matrix is calculated, and a small variation computed using the least-squares method is used to modify the background medium. Finally, the inverted medium is obtained. The horizontal eigenstate method is used to solve the forward problem. It is found that a good inverted formation can be obtained by using measurements. In order to help the user simulate the induction logs conveniently, a Wellog Simulator, based on the X-window system, is developed. The application software (FORTRAN codes) embedded in the Simulator is designed to simulate the responses of the induction tools in the layered formation with dipping beds. The graphic user-interface part of the Wellog Simulator is implemented with C and Motif. Through the user interface, the user can prepare the simulation data, select the tools, simulate the logs and plot the results.

DAMIT: a database of asteroid models

NASA Astrophysics Data System (ADS)

Durech, J.; Sidorin, V.; Kaasalainen, M.

2010-04-01

Context. Apart from a few targets that were directly imaged by spacecraft, remote sensing techniques are the main source of information about the basic physical properties of asteroids, such as the size, the spin state, or the spectral type. The most widely used observing technique - time-resolved photometry - provides us with data that can be used for deriving asteroid shapes and spin states. In the past decade, inversion of asteroid lightcurves has led to more than a hundred asteroid models. In the next decade, when data from all-sky surveys are available, the number of asteroid models will increase. Combining photometry with, e.g., adaptive optics data produces more detailed models. Aims: We created the Database of Asteroid Models from Inversion Techniques (DAMIT) with the aim of providing the astronomical community access to reliable and up-to-date physical models of asteroids - i.e., their shapes, rotation periods, and spin axis directions. Models from DAMIT can be used for further detailed studies of individual objects, as well as for statistical studies of the whole set. Methods: Most DAMIT models were derived from photometric data by the lightcurve inversion method. Some of them have been further refined or scaled using adaptive optics images, infrared observations, or occultation data. A substantial number of the models were derived also using sparse photometric data from astrometric databases. Results: At present, the database contains models of more than one hundred asteroids. For each asteroid, DAMIT provides the polyhedral shape model, the sidereal rotation period, the spin axis direction, and the photometric data used for the inversion. The database is updated when new models are available or when already published models are updated or refined. We have also released the C source code for the lightcurve inversion and for the direct problem (updates and extensions will follow).

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.

Application of a Numerical Inverse Laplace Integration Method to Surface Loading on a Viscoelastic Compressible Earth Model

NASA Astrophysics Data System (ADS)

Tanaka, Yoshiyuki; Klemann, Volker; Okuno, Jun'ichi

2009-09-01

Normal mode approaches for calculating viscoelastic responses of self-gravitating and compressible spherical earth models have an intrinsic problem of determining the roots of the secular equation and the associated residues in the Laplace domain. To bypass this problem, a method based on numerical inverse Laplace integration was developed by T anaka et al. (2006, 2007) for computations of viscoelastic deformation caused by an internal dislocation. The advantage of this approach is that the root-finding problem is avoided without imposing additional constraints on the governing equations and earth models. In this study, we apply the same algorithm to computations of viscoelastic responses to a surface load and show that the results obtained by this approach agree well with those obtained by a time-domain approach that does not need determinations of the normal modes in the Laplace domain. Using the elastic earth model PREM and a convex viscosity profile, we calculate viscoelastic load Love numbers ( h, l, k) for compressible and incompressible models. Comparisons between the results show that effects due to compressibility are consistent with results obtained by previous studies and that the rate differences between the two models total 10-40%. This will serve as an independent method to confirm results obtained by time-domain approaches and will usefully increase the reliability when modeling postglacial rebound.

Visualizing the ill-posedness of the inversion of a canopy radiative transfer model: A case study for Sentinel-2

NASA Astrophysics Data System (ADS)

Zurita-Milla, R.; Laurent, V. C. E.; van Gijsel, J. A. E.

2015-12-01

Monitoring biophysical and biochemical vegetation variables in space and time is key to understand the earth system. Operational approaches using remote sensing imagery rely on the inversion of radiative transfer models, which describe the interactions between light and vegetation canopies. The inversion required to estimate vegetation variables is, however, an ill-posed problem because of variable compensation effects that can cause different combinations of soil and canopy variables to yield extremely similar spectral responses. In this contribution, we present a novel approach to visualise the ill-posed problem using self-organizing maps (SOM), which are a type of unsupervised neural network. The approach is demonstrated with simulations for Sentinel-2 data (13 bands) made with the Soil-Leaf-Canopy (SLC) radiative transfer model. A look-up table of 100,000 entries was built by randomly sampling 14 SLC model input variables between their minimum and maximum allowed values while using both a dark and a bright soil. The Sentinel-2 spectral simulations were used to train a SOM of 200 × 125 neurons. The training projected similar spectral signatures onto either the same, or contiguous, neuron(s). Tracing back the inputs that generated each spectral signature, we created a 200 × 125 map for each of the SLC variables. The lack of spatial patterns and the variability in these maps indicate ill-posed situations, where similar spectral signatures correspond to different canopy variables. For Sentinel-2, our results showed that leaf area index, crown cover and leaf chlorophyll, water and brown pigment content are less confused in the inversion than variables with noisier maps like fraction of brown canopy area, leaf dry matter content and the PROSPECT mesophyll parameter. This study supports both educational and on-going research activities on inversion algorithms and might be useful to evaluate the uncertainties of retrieved canopy biophysical and biochemical state variables.

Total-variation based velocity inversion with Bregmanized operator splitting algorithm

NASA Astrophysics Data System (ADS)

Zand, Toktam; Gholami, Ali

2018-04-01

Many problems in applied geophysics can be formulated as a linear inverse problem. The associated problems, however, are large-scale and ill-conditioned. Therefore, regularization techniques are needed to be employed for solving them and generating a stable and acceptable solution. We consider numerical methods for solving such problems in this paper. In order to tackle the ill-conditioning of the problem we use blockiness as a prior information of the subsurface parameters and formulate the problem as a constrained total variation (TV) regularization. The Bregmanized operator splitting (BOS) algorithm as a combination of the Bregman iteration and the proximal forward backward operator splitting method is developed to solve the arranged problem. Two main advantages of this new algorithm are that no matrix inversion is required and that a discrepancy stopping criterion is used to stop the iterations, which allow efficient solution of large-scale problems. The high performance of the proposed TV regularization method is demonstrated using two different experiments: 1) velocity inversion from (synthetic) seismic data which is based on Born approximation, 2) computing interval velocities from RMS velocities via Dix formula. Numerical examples are presented to verify the feasibility of the proposed method for high-resolution velocity inversion.