Forecasting experiments of a dynamical-statistical model of the sea surface temperature anomaly field based on the improved self-memorization principle

NASA Astrophysics Data System (ADS)

Hong, Mei; Chen, Xi; Zhang, Ren; Wang, Dong; Shen, Shuanghe; Singh, Vijay P.

2018-04-01

With the objective of tackling the problem of inaccurate long-term El Niño-Southern Oscillation (ENSO) forecasts, this paper develops a new dynamical-statistical forecast model of the sea surface temperature anomaly (SSTA) field. To avoid single initial prediction values, a self-memorization principle is introduced to improve the dynamical reconstruction model, thus making the model more appropriate for describing such chaotic systems as ENSO events. The improved dynamical-statistical model of the SSTA field is used to predict SSTA in the equatorial eastern Pacific and during El Niño and La Niña events. The long-term step-by-step forecast results and cross-validated retroactive hindcast results of time series T1 and T2 are found to be satisfactory, with a Pearson correlation coefficient of approximately 0.80 and a mean absolute percentage error (MAPE) of less than 15 %. The corresponding forecast SSTA field is accurate in that not only is the forecast shape similar to the actual field but also the contour lines are essentially the same. This model can also be used to forecast the ENSO index. The temporal correlation coefficient is 0.8062, and the MAPE value of 19.55 % is small. The difference between forecast results in spring and those in autumn is not high, indicating that the improved model can overcome the spring predictability barrier to some extent. Compared with six mature models published previously, the present model has an advantage in prediction precision and length, and is a novel exploration of the ENSO forecast method.

A Diagnostics Tool to detect ensemble forecast system anomaly and guide operational decisions

NASA Astrophysics Data System (ADS)

Park, G. H.; Srivastava, A.; Shrestha, E.; Thiemann, M.; Day, G. N.; Draijer, S.

2017-12-01

The hydrologic community is moving toward using ensemble forecasts to take uncertainty into account during the decision-making process. The New York City Department of Environmental Protection (DEP) implements several types of ensemble forecasts in their decision-making process: ensemble products for a statistical model (Hirsch and enhanced Hirsch); the National Weather Service (NWS) Advanced Hydrologic Prediction Service (AHPS) forecasts based on the classical Ensemble Streamflow Prediction (ESP) technique; and the new NWS Hydrologic Ensemble Forecasting Service (HEFS) forecasts. To remove structural error and apply the forecasts to additional forecast points, the DEP post processes both the AHPS and the HEFS forecasts. These ensemble forecasts provide mass quantities of complex data, and drawing conclusions from these forecasts is time-consuming and difficult. The complexity of these forecasts also makes it difficult to identify system failures resulting from poor data, missing forecasts, and server breakdowns. To address these issues, we developed a diagnostic tool that summarizes ensemble forecasts and provides additional information such as historical forecast statistics, forecast skill, and model forcing statistics. This additional information highlights the key information that enables operators to evaluate the forecast in real-time, dynamically interact with the data, and review additional statistics, if needed, to make better decisions. We used Bokeh, a Python interactive visualization library, and a multi-database management system to create this interactive tool. This tool compiles and stores data into HTML pages that allows operators to readily analyze the data with built-in user interaction features. This paper will present a brief description of the ensemble forecasts, forecast verification results, and the intended applications for the diagnostic tool.

Evaluation of statistical models for forecast errors from the HBV model

NASA Astrophysics Data System (ADS)

Engeland, Kolbjørn; Renard, Benjamin; Steinsland, Ingelin; Kolberg, Sjur

2010-04-01

SummaryThree statistical models for the forecast errors for inflow into the Langvatn reservoir in Northern Norway have been constructed and tested according to the agreement between (i) the forecast distribution and the observations and (ii) median values of the forecast distribution and the observations. For the first model observed and forecasted inflows were transformed by the Box-Cox transformation before a first order auto-regressive model was constructed for the forecast errors. The parameters were conditioned on weather classes. In the second model the Normal Quantile Transformation (NQT) was applied on observed and forecasted inflows before a similar first order auto-regressive model was constructed for the forecast errors. For the third model positive and negative errors were modeled separately. The errors were first NQT-transformed before conditioning the mean error values on climate, forecasted inflow and yesterday's error. To test the three models we applied three criterions: we wanted (a) the forecast distribution to be reliable; (b) the forecast intervals to be narrow; (c) the median values of the forecast distribution to be close to the observed values. Models 1 and 2 gave almost identical results. The median values improved the forecast with Nash-Sutcliffe R eff increasing from 0.77 for the original forecast to 0.87 for the corrected forecasts. Models 1 and 2 over-estimated the forecast intervals but gave the narrowest intervals. Their main drawback was that the distributions are less reliable than Model 3. For Model 3 the median values did not fit well since the auto-correlation was not accounted for. Since Model 3 did not benefit from the potential variance reduction that lies in bias estimation and removal it gave on average wider forecasts intervals than the two other models. At the same time Model 3 on average slightly under-estimated the forecast intervals, probably explained by the use of average measures to evaluate the fit.

The Wind Forecast Improvement Project (WFIP). A Public/Private Partnership for Improving Short Term Wind Energy Forecasts and Quantifying the Benefits of Utility Operations -- the Northern Study Area

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

Finley, Cathy

2014-04-30

This report contains the results from research aimed at improving short-range (0-6 hour) hub-height wind forecasts in the NOAA weather forecast models through additional data assimilation and model physics improvements for use in wind energy forecasting. Additional meteorological observing platforms including wind profilers, sodars, and surface stations were deployed for this study by NOAA and DOE, and additional meteorological data at or near wind turbine hub height were provided by South Dakota State University and WindLogics/NextEra Energy Resources over a large geographical area in the U.S. Northern Plains for assimilation into NOAA research weather forecast models. The resulting improvements inmore » wind energy forecasts based on the research weather forecast models (with the additional data assimilation and model physics improvements) were examined in many different ways and compared with wind energy forecasts based on the current operational weather forecast models to quantify the forecast improvements important to power grid system operators and wind plant owners/operators participating in energy markets. Two operational weather forecast models (OP_RUC, OP_RAP) and two research weather forecast models (ESRL_RAP, HRRR) were used as the base wind forecasts for generating several different wind power forecasts for the NextEra Energy wind plants in the study area. Power forecasts were generated from the wind forecasts in a variety of ways, from very simple to quite sophisticated, as they might be used by a wide range of both general users and commercial wind energy forecast vendors. The error characteristics of each of these types of forecasts were examined and quantified using bulk error statistics for both the local wind plant and the system aggregate forecasts. The wind power forecast accuracy was also evaluated separately for high-impact wind energy ramp events. The overall bulk error statistics calculated over the first six hours of the forecasts at both the individual wind plant and at the system-wide aggregate level over the one year study period showed that the research weather model-based power forecasts (all types) had lower overall error rates than the current operational weather model-based power forecasts, both at the individual wind plant level and at the system aggregate level. The bulk error statistics of the various model-based power forecasts were also calculated by season and model runtime/forecast hour as power system operations are more sensitive to wind energy forecast errors during certain times of year and certain times of day. The results showed that there were significant differences in seasonal forecast errors between the various model-based power forecasts. The results from the analysis of the various wind power forecast errors by model runtime and forecast hour showed that the forecast errors were largest during the times of day that have increased significance to power system operators (the overnight hours and the morning/evening boundary layer transition periods), but the research weather model-based power forecasts showed improvement over the operational weather model-based power forecasts at these times.« less

A Comparison of the Forecast Skills among Three Numerical Models

NASA Astrophysics Data System (ADS)

Lu, D.; Reddy, S. R.; White, L. J.

2003-12-01

Three numerical weather forecast models, MM5, COAMPS and WRF, operating with a joint effort of NOAA HU-NCAS and Jackson State University (JSU) during summer 2003 have been chosen to study their forecast skills against observations. The models forecast over the same region with the same initialization, boundary condition, forecast length and spatial resolution. AVN global dataset have been ingested as initial conditions. Grib resolution of 27 km is chosen to represent the current mesoscale model. The forecasts with the length of 36h are performed to output the result with 12h interval. The key parameters used to evaluate the forecast skill include 12h accumulated precipitation, sea level pressure, wind, surface temperature and dew point. Precipitation is evaluated statistically using conventional skill scores, Threat Score (TS) and Bias Score (BS), for different threshold values based on 12h rainfall observations whereas other statistical methods such as Mean Error (ME), Mean Absolute Error(MAE) and Root Mean Square Error (RMSE) are applied to other forecast parameters.

National Centers for Environmental Prediction

Science.gov Websites

Statistics Observational Data Processing Data Assimilation Monsoon Desk Model Transition Seminars Seminar Hurricane Weather Research and Forecast System ANALYSIS FORECAST MODEL GSI Gridpoint Statistical Weather and Climate Prediction (NCWCP) 5830 University Research Court College Park, MD 20740 Page Author

Evaluation Of Statistical Models For Forecast Errors From The HBV-Model

NASA Astrophysics Data System (ADS)

Engeland, K.; Kolberg, S.; Renard, B.; Stensland, I.

2009-04-01

Three statistical models for the forecast errors for inflow to the Langvatn reservoir in Northern Norway have been constructed and tested according to how well the distribution and median values of the forecasts errors fit to the observations. For the first model observed and forecasted inflows were transformed by the Box-Cox transformation before a first order autoregressive model was constructed for the forecast errors. The parameters were conditioned on climatic conditions. In the second model the Normal Quantile Transformation (NQT) was applied on observed and forecasted inflows before a similar first order autoregressive model was constructed for the forecast errors. For the last model positive and negative errors were modeled separately. The errors were first NQT-transformed before a model where the mean values were conditioned on climate, forecasted inflow and yesterday's error. To test the three models we applied three criterions: We wanted a) the median values to be close to the observed values; b) the forecast intervals to be narrow; c) the distribution to be correct. The results showed that it is difficult to obtain a correct model for the forecast errors, and that the main challenge is to account for the auto-correlation in the errors. Model 1 and 2 gave similar results, and the main drawback is that the distributions are not correct. The 95% forecast intervals were well identified, but smaller forecast intervals were over-estimated, and larger intervals were under-estimated. Model 3 gave a distribution that fits better, but the median values do not fit well since the auto-correlation is not properly accounted for. If the 95% forecast interval is of interest, Model 2 is recommended. If the whole distribution is of interest, Model 3 is recommended.

A Comparison of the Performance of Advanced Statistical Techniques for the Refinement of Day-ahead and Longer NWP-based Wind Power Forecasts

NASA Astrophysics Data System (ADS)

Zack, J. W.

2015-12-01

Predictions from Numerical Weather Prediction (NWP) models are the foundation for wind power forecasts for day-ahead and longer forecast horizons. The NWP models directly produce three-dimensional wind forecasts on their respective computational grids. These can be interpolated to the location and time of interest. However, these direct predictions typically contain significant systematic errors ("biases"). This is due to a variety of factors including the limited space-time resolution of the NWP models and shortcomings in the model's representation of physical processes. It has become common practice to attempt to improve the raw NWP forecasts by statistically adjusting them through a procedure that is widely known as Model Output Statistics (MOS). The challenge is to identify complex patterns of systematic errors and then use this knowledge to adjust the NWP predictions. The MOS-based improvements are the basis for much of the value added by commercial wind power forecast providers. There are an enormous number of statistical approaches that can be used to generate the MOS adjustments to the raw NWP forecasts. In order to obtain insight into the potential value of some of the newer and more sophisticated statistical techniques often referred to as "machine learning methods" a MOS-method comparison experiment has been performed for wind power generation facilities in 6 wind resource areas of California. The underlying NWP models that provided the raw forecasts were the two primary operational models of the US National Weather Service: the GFS and NAM models. The focus was on 1- and 2-day ahead forecasts of the hourly wind-based generation. The statistical methods evaluated included: (1) screening multiple linear regression, which served as a baseline method, (2) artificial neural networks, (3) a decision-tree approach called random forests, (4) gradient boosted regression based upon an decision-tree algorithm, (5) support vector regression and (6) analog ensemble, which is a case-matching scheme. The presentation will provide (1) an overview of each method and the experimental design, (2) performance comparisons based on standard metrics such as bias, MAE and RMSE, (3) a summary of the performance characteristics of each approach and (4) a preview of further experiments to be conducted.

A probabilistic drought forecasting framework: A combined dynamical and statistical approach

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

Yan, Hongxiang; Moradkhani, Hamid; Zarekarizi, Mahkameh

In order to improve drought forecasting skill, this study develops a probabilistic drought forecasting framework comprised of dynamical and statistical modeling components. The novelty of this study is to seek the use of data assimilation to quantify initial condition uncertainty with the Monte Carlo ensemble members, rather than relying entirely on the hydrologic model or land surface model to generate a single deterministic initial condition, as currently implemented in the operational drought forecasting systems. Next, the initial condition uncertainty is quantified through data assimilation and coupled with a newly developed probabilistic drought forecasting model using a copula function. The initialmore » condition at each forecast start date are sampled from the data assimilation ensembles for forecast initialization. Finally, seasonal drought forecasting products are generated with the updated initial conditions. This study introduces the theory behind the proposed drought forecasting system, with an application in Columbia River Basin, Pacific Northwest, United States. Results from both synthetic and real case studies suggest that the proposed drought forecasting system significantly improves the seasonal drought forecasting skills and can facilitate the state drought preparation and declaration, at least three months before the official state drought declaration.« less

Efficient ensemble forecasting of marine ecology with clustered 1D models and statistical lateral exchange: application to the Red Sea

NASA Astrophysics Data System (ADS)

Dreano, Denis; Tsiaras, Kostas; Triantafyllou, George; Hoteit, Ibrahim

2017-07-01

Forecasting the state of large marine ecosystems is important for many economic and public health applications. However, advanced three-dimensional (3D) ecosystem models, such as the European Regional Seas Ecosystem Model (ERSEM), are computationally expensive, especially when implemented within an ensemble data assimilation system requiring several parallel integrations. As an alternative to 3D ecological forecasting systems, we propose to implement a set of regional one-dimensional (1D) water-column ecological models that run at a fraction of the computational cost. The 1D model domains are determined using a Gaussian mixture model (GMM)-based clustering method and satellite chlorophyll-a (Chl-a) data. Regionally averaged Chl-a data is assimilated into the 1D models using the singular evolutive interpolated Kalman (SEIK) filter. To laterally exchange information between subregions and improve the forecasting skills, we introduce a new correction step to the assimilation scheme, in which we assimilate a statistical forecast of future Chl-a observations based on information from neighbouring regions. We apply this approach to the Red Sea and show that the assimilative 1D ecological models can forecast surface Chl-a concentration with high accuracy. The statistical assimilation step further improves the forecasting skill by as much as 50%. This general approach of clustering large marine areas and running several interacting 1D ecological models is very flexible. It allows many combinations of clustering, filtering and regression technics to be used and can be applied to build efficient forecasting systems in other large marine ecosystems.

Can we use Earth Observations to improve monthly water level forecasts?

NASA Astrophysics Data System (ADS)

Slater, L. J.; Villarini, G.

2017-12-01

Dynamical-statistical hydrologic forecasting approaches benefit from different strengths in comparison with traditional hydrologic forecasting systems: they are computationally efficient, can integrate and `learn' from a broad selection of input data (e.g., General Circulation Model (GCM) forecasts, Earth Observation time series, teleconnection patterns), and can take advantage of recent progress in machine learning (e.g. multi-model blending, post-processing and ensembling techniques). Recent efforts to develop a dynamical-statistical ensemble approach for forecasting seasonal streamflow using both GCM forecasts and changing land cover have shown promising results over the U.S. Midwest. Here, we use climate forecasts from several GCMs of the North American Multi Model Ensemble (NMME) alongside 15-minute stage time series from the National River Flow Archive (NRFA) and land cover classes extracted from the European Space Agency's Climate Change Initiative 300 m annual Global Land Cover time series. With these data, we conduct systematic long-range probabilistic forecasting of monthly water levels in UK catchments over timescales ranging from one to twelve months ahead. We evaluate the improvement in model fit and model forecasting skill that comes from using land cover classes as predictors in the models. This work opens up new possibilities for combining Earth Observation time series with GCM forecasts to predict a variety of hazards from space using data science techniques.

Comparisons of Three-Dimensional Variational Data Assimilation and Model Output Statistics in Improving Atmospheric Chemistry Forecasts

NASA Astrophysics Data System (ADS)

Ma, Chaoqun; Wang, Tijian; Zang, Zengliang; Li, Zhijin

2018-07-01

Atmospheric chemistry models usually perform badly in forecasting wintertime air pollution because of their uncertainties. Generally, such uncertainties can be decreased effectively by techniques such as data assimilation (DA) and model output statistics (MOS). However, the relative importance and combined effects of the two techniques have not been clarified. Here, a one-month air quality forecast with the Weather Research and Forecasting-Chemistry (WRF-Chem) model was carried out in a virtually operational setup focusing on Hebei Province, China. Meanwhile, three-dimensional variational (3DVar) DA and MOS based on one-dimensional Kalman filtering were implemented separately and simultaneously to investigate their performance in improving the model forecast. Comparison with observations shows that the chemistry forecast with MOS outperforms that with 3DVar DA, which could be seen in all the species tested over the whole 72 forecast hours. Combined use of both techniques does not guarantee a better forecast than MOS only, with the improvements and degradations being small and appearing rather randomly. Results indicate that the implementation of MOS is more suitable than 3DVar DA in improving the operational forecasting ability of WRF-Chem.

Ensemble Solar Forecasting Statistical Quantification and Sensitivity Analysis: Preprint

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

Cheung, WanYin; Zhang, Jie; Florita, Anthony

2015-12-08

Uncertainties associated with solar forecasts present challenges to maintain grid reliability, especially at high solar penetrations. This study aims to quantify the errors associated with the day-ahead solar forecast parameters and the theoretical solar power output for a 51-kW solar power plant in a utility area in the state of Vermont, U.S. Forecasts were generated by three numerical weather prediction (NWP) models, including the Rapid Refresh, the High Resolution Rapid Refresh, and the North American Model, and a machine-learning ensemble model. A photovoltaic (PV) performance model was adopted to calculate theoretical solar power generation using the forecast parameters (e.g., irradiance,more » cell temperature, and wind speed). Errors of the power outputs were quantified using statistical moments and a suite of metrics, such as the normalized root mean squared error (NRMSE). In addition, the PV model's sensitivity to different forecast parameters was quantified and analyzed. Results showed that the ensemble model yielded forecasts in all parameters with the smallest NRMSE. The NRMSE of solar irradiance forecasts of the ensemble NWP model was reduced by 28.10% compared to the best of the three NWP models. Further, the sensitivity analysis indicated that the errors of the forecasted cell temperature attributed only approximately 0.12% to the NRMSE of the power output as opposed to 7.44% from the forecasted solar irradiance.« less

Bayesian analyses of seasonal runoff forecasts

NASA Astrophysics Data System (ADS)

Krzysztofowicz, R.; Reese, S.

1991-12-01

Forecasts of seasonal snowmelt runoff volume provide indispensable information for rational decision making by water project operators, irrigation district managers, and farmers in the western United States. Bayesian statistical models and communication frames have been researched in order to enhance the forecast information disseminated to the users, and to characterize forecast skill from the decision maker's point of view. Four products are presented: (i) a Bayesian Processor of Forecasts, which provides a statistical filter for calibrating the forecasts, and a procedure for estimating the posterior probability distribution of the seasonal runoff; (ii) the Bayesian Correlation Score, a new measure of forecast skill, which is related monotonically to the ex ante economic value of forecasts for decision making; (iii) a statistical predictor of monthly cumulative runoffs within the snowmelt season, conditional on the total seasonal runoff forecast; and (iv) a framing of the forecast message that conveys the uncertainty associated with the forecast estimates to the users. All analyses are illustrated with numerical examples of forecasts for six gauging stations from the period 1971 1988.

Hyperparameterization of soil moisture statistical models for North America with Ensemble Learning Models (Elm)

NASA Astrophysics Data System (ADS)

Steinberg, P. D.; Brener, G.; Duffy, D.; Nearing, G. S.; Pelissier, C.

2017-12-01

Hyperparameterization, of statistical models, i.e. automated model scoring and selection, such as evolutionary algorithms, grid searches, and randomized searches, can improve forecast model skill by reducing errors associated with model parameterization, model structure, and statistical properties of training data. Ensemble Learning Models (Elm), and the related Earthio package, provide a flexible interface for automating the selection of parameters and model structure for machine learning models common in climate science and land cover classification, offering convenient tools for loading NetCDF, HDF, Grib, or GeoTiff files, decomposition methods like PCA and manifold learning, and parallel training and prediction with unsupervised and supervised classification, clustering, and regression estimators. Continuum Analytics is using Elm to experiment with statistical soil moisture forecasting based on meteorological forcing data from NASA's North American Land Data Assimilation System (NLDAS). There Elm is using the NSGA-2 multiobjective optimization algorithm for optimizing statistical preprocessing of forcing data to improve goodness-of-fit for statistical models (i.e. feature engineering). This presentation will discuss Elm and its components, including dask (distributed task scheduling), xarray (data structures for n-dimensional arrays), and scikit-learn (statistical preprocessing, clustering, classification, regression), and it will show how NSGA-2 is being used for automate selection of soil moisture forecast statistical models for North America.

Forecasting of Radiation Belts: Results From the PROGRESS Project.

NASA Astrophysics Data System (ADS)

Balikhin, M. A.; Arber, T. D.; Ganushkina, N. Y.; Walker, S. N.

2017-12-01

Forecasting of Radiation Belts: Results from the PROGRESS Project. The overall goal of the PROGRESS project, funded in frame of EU Horizon2020 programme, is to combine first principles based models with the systems science methodologies to achieve reliable forecasts of the geo-space particle radiation environment.The PROGRESS incorporates three themes : The propagation of the solar wind to L1, Forecast of geomagnetic indices, and forecast of fluxes of energetic electrons within the magnetosphere. One of the important aspects of the PROGRESS project is the development of statistical wave models for magnetospheric waves that affect the dynamics of energetic electrons such as lower band chorus, hiss and equatorial noise. The error reduction ratio (ERR) concept has been used to optimise the set of solar wind and geomagnetic parameters for organisation of statistical wave models for these emissions. The resulting sets of parameters and statistical wave models will be presented and discussed. However the ERR analysis also indicates that the combination of solar wind and geomagnetic parameters accounts for only part of the variance of the emissions under investigation (lower band chorus, hiss and equatorial noise). In addition, advances in the forecast of fluxes of energetic electrons, exploiting empirical models and the first principles IMPTAM model achieved by the PROGRESS project is presented.

Mixture EMOS model for calibrating ensemble forecasts of wind speed.

PubMed

Baran, S; Lerch, S

2016-03-01

Ensemble model output statistics (EMOS) is a statistical tool for post-processing forecast ensembles of weather variables obtained from multiple runs of numerical weather prediction models in order to produce calibrated predictive probability density functions. The EMOS predictive probability density function is given by a parametric distribution with parameters depending on the ensemble forecasts. We propose an EMOS model for calibrating wind speed forecasts based on weighted mixtures of truncated normal (TN) and log-normal (LN) distributions where model parameters and component weights are estimated by optimizing the values of proper scoring rules over a rolling training period. The new model is tested on wind speed forecasts of the 50 member European Centre for Medium-range Weather Forecasts ensemble, the 11 member Aire Limitée Adaptation dynamique Développement International-Hungary Ensemble Prediction System ensemble of the Hungarian Meteorological Service, and the eight-member University of Washington mesoscale ensemble, and its predictive performance is compared with that of various benchmark EMOS models based on single parametric families and combinations thereof. The results indicate improved calibration of probabilistic and accuracy of point forecasts in comparison with the raw ensemble and climatological forecasts. The mixture EMOS model significantly outperforms the TN and LN EMOS methods; moreover, it provides better calibrated forecasts than the TN-LN combination model and offers an increased flexibility while avoiding covariate selection problems. © 2016 The Authors Environmetrics Published by JohnWiley & Sons Ltd.

Statistical Analysis of Atmospheric Forecast Model Accuracy - A Focus on Multiple Atmospheric Variables and Location-Based Analysis

DTIC Science & Technology

2014-04-01

WRF ) model is a numerical weather prediction system designed for operational forecasting and atmospheric research. This report examined WRF model... WRF , weather research and forecasting, atmospheric effects 16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF ABSTRACT SAR 18. NUMBER OF...and Forecasting ( WRF ) model. The authors would also like to thank Ms. Sherry Larson, STS Systems Integration, LLC, ARL Technical Publishing Branch

Air Quality Forecasting through Different Statistical and Artificial Intelligence Techniques

NASA Astrophysics Data System (ADS)

Mishra, D.; Goyal, P.

2014-12-01

Urban air pollution forecasting has emerged as an acute problem in recent years because there are sever environmental degradation due to increase in harmful air pollutants in the ambient atmosphere. In this study, there are different types of statistical as well as artificial intelligence techniques are used for forecasting and analysis of air pollution over Delhi urban area. These techniques are principle component analysis (PCA), multiple linear regression (MLR) and artificial neural network (ANN) and the forecasting are observed in good agreement with the observed concentrations through Central Pollution Control Board (CPCB) at different locations in Delhi. But such methods suffers from disadvantages like they provide limited accuracy as they are unable to predict the extreme points i.e. the pollution maximum and minimum cut-offs cannot be determined using such approach. Also, such methods are inefficient approach for better output forecasting. But with the advancement in technology and research, an alternative to the above traditional methods has been proposed i.e. the coupling of statistical techniques with artificial Intelligence (AI) can be used for forecasting purposes. The coupling of PCA, ANN and fuzzy logic is used for forecasting of air pollutant over Delhi urban area. The statistical measures e.g., correlation coefficient (R), normalized mean square error (NMSE), fractional bias (FB) and index of agreement (IOA) of the proposed model are observed in better agreement with the all other models. Hence, the coupling of statistical and artificial intelligence can be use for the forecasting of air pollutant over urban area.

A Bayesian modelling method for post-processing daily sub-seasonal to seasonal rainfall forecasts from global climate models and evaluation for 12 Australian catchments

NASA Astrophysics Data System (ADS)

Schepen, Andrew; Zhao, Tongtiegang; Wang, Quan J.; Robertson, David E.

2018-03-01

Rainfall forecasts are an integral part of hydrological forecasting systems at sub-seasonal to seasonal timescales. In seasonal forecasting, global climate models (GCMs) are now the go-to source for rainfall forecasts. For hydrological applications however, GCM forecasts are often biased and unreliable in uncertainty spread, and calibration is therefore required before use. There are sophisticated statistical techniques for calibrating monthly and seasonal aggregations of the forecasts. However, calibration of seasonal forecasts at the daily time step typically uses very simple statistical methods or climate analogue methods. These methods generally lack the sophistication to achieve unbiased, reliable and coherent forecasts of daily amounts and seasonal accumulated totals. In this study, we propose and evaluate a Rainfall Post-Processing method for Seasonal forecasts (RPP-S), which is based on the Bayesian joint probability modelling approach for calibrating daily forecasts and the Schaake Shuffle for connecting the daily ensemble members of different lead times. We apply the method to post-process ACCESS-S forecasts for 12 perennial and ephemeral catchments across Australia and for 12 initialisation dates. RPP-S significantly reduces bias in raw forecasts and improves both skill and reliability. RPP-S forecasts are also more skilful and reliable than forecasts derived from ACCESS-S forecasts that have been post-processed using quantile mapping, especially for monthly and seasonal accumulations. Several opportunities to improve the robustness and skill of RPP-S are identified. The new RPP-S post-processed forecasts will be used in ensemble sub-seasonal to seasonal streamflow applications.

A hybrid ARIMA and neural network model applied to forecast catch volumes of Selar crumenophthalmus

NASA Astrophysics Data System (ADS)

Aquino, Ronald L.; Alcantara, Nialle Loui Mar T.; Addawe, Rizavel C.

2017-11-01

The Selar crumenophthalmus with the English name big-eyed scad fish, locally known as matang-baka, is one of the fishes commonly caught along the waters of La Union, Philippines. The study deals with the forecasting of catch volumes of big-eyed scad fish for commercial consumption. The data used are quarterly caught volumes of big-eyed scad fish from 2002 to first quarter of 2017. This actual data is available from the open stat database published by the Philippine Statistics Authority (PSA)whose task is to collect, compiles, analyzes and publish information concerning different aspects of the Philippine setting. Autoregressive Integrated Moving Average (ARIMA) models, Artificial Neural Network (ANN) model and the Hybrid model consisting of ARIMA and ANN were developed to forecast catch volumes of big-eyed scad fish. Statistical errors such as Mean Absolute Errors (MAE) and Root Mean Square Errors (RMSE) were computed and compared to choose the most suitable model for forecasting the catch volume for the next few quarters. A comparison of the results of each model and corresponding statistical errors reveals that the hybrid model, ARIMA-ANN (2,1,2)(6:3:1), is the most suitable model to forecast the catch volumes of the big-eyed scad fish for the next few quarters.

Statistical Correction of Air Temperature Forecasts for City and Road Weather Applications

NASA Astrophysics Data System (ADS)

Mahura, Alexander; Petersen, Claus; Sass, Bent; Gilet, Nicolas

2014-05-01

The method for statistical correction of air /road surface temperatures forecasts was developed based on analysis of long-term time-series of meteorological observations and forecasts (from HIgh Resolution Limited Area Model & Road Conditions Model; 3 km horizontal resolution). It has been tested for May-Aug 2012 & Oct 2012 - Mar 2013, respectively. The developed method is based mostly on forecasted meteorological parameters with a minimal inclusion of observations (covering only a pre-history period). Although the st iteration correction is based taking into account relevant temperature observations, but the further adjustment of air and road temperature forecasts is based purely on forecasted meteorological parameters. The method is model independent, e.g. it can be applied for temperature correction with other types of models having different horizontal resolutions. It is relatively fast due to application of the singular value decomposition method for matrix solution to find coefficients. Moreover, there is always a possibility for additional improvement due to extra tuning of the temperature forecasts for some locations (stations), and in particular, where for example, the MAEs are generally higher compared with others (see Gilet et al., 2014). For the city weather applications, new operationalized procedure for statistical correction of the air temperature forecasts has been elaborated and implemented for the HIRLAM-SKA model runs at 00, 06, 12, and 18 UTCs covering forecast lengths up to 48 hours. The procedure includes segments for extraction of observations and forecast data, assigning these to forecast lengths, statistical correction of temperature, one-&multi-days statistical evaluation of model performance, decision-making on using corrections by stations, interpolation, visualisation and storage/backup. Pre-operational air temperature correction runs were performed for the mainland Denmark since mid-April 2013 and shown good results. Tests also showed that the CPU time required for the operational procedure is relatively short (less than 15 minutes including a large time spent for interpolation). These also showed that in order to start correction of forecasts there is no need to have a long-term pre-historical data (containing forecasts and observations) and, at least, a couple of weeks will be sufficient when a new observational station is included and added to the forecast point. Note for the road weather application, the operationalization of the statistical correction of the road surface temperature forecasts (for the RWM system daily hourly runs covering forecast length up to 5 hours ahead) for the Danish road network (for about 400 road stations) was also implemented, and it is running in a test mode since Sep 2013. The method can also be applied for correction of the dew point temperature and wind speed (as a part of observations/ forecasts at synoptical stations), where these both meteorological parameters are parts of the proposed system of equations. The evaluation of the method performance for improvement of the wind speed forecasts is planned as well, with considering possibilities for the wind direction improvements (which is more complex due to multi-modal types of such data distribution). The method worked for the entire domain of mainland Denmark (tested for 60 synoptical and 395 road stations), and hence, it can be also applied for any geographical point within this domain, as through interpolation to about 100 cities' locations (for Danish national byvejr forecasts). Moreover, we can assume that the same method can be used in other geographical areas. The evaluation for other domains (with a focus on Greenland and Nordic countries) is planned. In addition, a similar approach might be also tested for statistical correction of concentrations of chemical species, but such approach will require additional elaboration and evaluation.

The GEOS Ozone Data Assimilation System: Specification of Error Statistics

NASA Technical Reports Server (NTRS)

Stajner, Ivanka; Riishojgaard, Lars Peter; Rood, Richard B.

2000-01-01

A global three-dimensional ozone data assimilation system has been developed at the Data Assimilation Office of the NASA/Goddard Space Flight Center. The Total Ozone Mapping Spectrometer (TOMS) total ozone and the Solar Backscatter Ultraviolet (SBUV) or (SBUV/2) partial ozone profile observations are assimilated. The assimilation, into an off-line ozone transport model, is done using the global Physical-space Statistical Analysis Scheme (PSAS). This system became operational in December 1999. A detailed description of the statistical analysis scheme, and in particular, the forecast and observation error covariance models is given. A new global anisotropic horizontal forecast error correlation model accounts for a varying distribution of observations with latitude. Correlations are largest in the zonal direction in the tropics where data is sparse. Forecast error variance model is proportional to the ozone field. The forecast error covariance parameters were determined by maximum likelihood estimation. The error covariance models are validated using x squared statistics. The analyzed ozone fields in the winter 1992 are validated against independent observations from ozone sondes and HALOE. There is better than 10% agreement between mean Halogen Occultation Experiment (HALOE) and analysis fields between 70 and 0.2 hPa. The global root-mean-square (RMS) difference between TOMS observed and forecast values is less than 4%. The global RMS difference between SBUV observed and analyzed ozone between 50 and 3 hPa is less than 15%.

A Comparison of Forecast Error Generators for Modeling Wind and Load Uncertainty

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

Lu, Ning; Diao, Ruisheng; Hafen, Ryan P.

2013-12-18

This paper presents four algorithms to generate random forecast error time series, including a truncated-normal distribution model, a state-space based Markov model, a seasonal autoregressive moving average (ARMA) model, and a stochastic-optimization based model. The error time series are used to create real-time (RT), hour-ahead (HA), and day-ahead (DA) wind and load forecast time series that statistically match historically observed forecasting data sets, used for variable generation integration studies. A comparison is made using historical DA load forecast and actual load values to generate new sets of DA forecasts with similar stoical forecast error characteristics. This paper discusses and comparesmore » the capabilities of each algorithm to preserve the characteristics of the historical forecast data sets.« less

Three-Month Real-Time Dengue Forecast Models: An Early Warning System for Outbreak Alerts and Policy Decision Support in Singapore.

PubMed

Shi, Yuan; Liu, Xu; Kok, Suet-Yheng; Rajarethinam, Jayanthi; Liang, Shaohong; Yap, Grace; Chong, Chee-Seng; Lee, Kim-Sung; Tan, Sharon S Y; Chin, Christopher Kuan Yew; Lo, Andrew; Kong, Waiming; Ng, Lee Ching; Cook, Alex R

2016-09-01

With its tropical rainforest climate, rapid urbanization, and changing demography and ecology, Singapore experiences endemic dengue; the last large outbreak in 2013 culminated in 22,170 cases. In the absence of a vaccine on the market, vector control is the key approach for prevention. We sought to forecast the evolution of dengue epidemics in Singapore to provide early warning of outbreaks and to facilitate the public health response to moderate an impending outbreak. We developed a set of statistical models using least absolute shrinkage and selection operator (LASSO) methods to forecast the weekly incidence of dengue notifications over a 3-month time horizon. This forecasting tool used a variety of data streams and was updated weekly, including recent case data, meteorological data, vector surveillance data, and population-based national statistics. The forecasting methodology was compared with alternative approaches that have been proposed to model dengue case data (seasonal autoregressive integrated moving average and step-down linear regression) by fielding them on the 2013 dengue epidemic, the largest on record in Singapore. Operationally useful forecasts were obtained at a 3-month lag using the LASSO-derived models. Based on the mean average percentage error, the LASSO approach provided more accurate forecasts than the other methods we assessed. We demonstrate its utility in Singapore's dengue control program by providing a forecast of the 2013 outbreak for advance preparation of outbreak response. Statistical models built using machine learning methods such as LASSO have the potential to markedly improve forecasting techniques for recurrent infectious disease outbreaks such as dengue. Shi Y, Liu X, Kok SY, Rajarethinam J, Liang S, Yap G, Chong CS, Lee KS, Tan SS, Chin CK, Lo A, Kong W, Ng LC, Cook AR. 2016. Three-month real-time dengue forecast models: an early warning system for outbreak alerts and policy decision support in Singapore. Environ Health Perspect 124:1369-1375; http://dx.doi.org/10.1289/ehp.1509981.

Improving the long-lead predictability of El Niño using a novel forecasting scheme based on a dynamic components model

NASA Astrophysics Data System (ADS)

Petrova, Desislava; Koopman, Siem Jan; Ballester, Joan; Rodó, Xavier

2017-02-01

El Niño (EN) is a dominant feature of climate variability on inter-annual time scales driving changes in the climate throughout the globe, and having wide-spread natural and socio-economic consequences. In this sense, its forecast is an important task, and predictions are issued on a regular basis by a wide array of prediction schemes and climate centres around the world. This study explores a novel method for EN forecasting. In the state-of-the-art the advantageous statistical technique of unobserved components time series modeling, also known as structural time series modeling, has not been applied. Therefore, we have developed such a model where the statistical analysis, including parameter estimation and forecasting, is based on state space methods, and includes the celebrated Kalman filter. The distinguishing feature of this dynamic model is the decomposition of a time series into a range of stochastically time-varying components such as level (or trend), seasonal, cycles of different frequencies, irregular, and regression effects incorporated as explanatory covariates. These components are modeled separately and ultimately combined in a single forecasting scheme. Customary statistical models for EN prediction essentially use SST and wind stress in the equatorial Pacific. In addition to these, we introduce a new domain of regression variables accounting for the state of the subsurface ocean temperature in the western and central equatorial Pacific, motivated by our analysis, as well as by recent and classical research, showing that subsurface processes and heat accumulation there are fundamental for the genesis of EN. An important feature of the scheme is that different regression predictors are used at different lead months, thus capturing the dynamical evolution of the system and rendering more efficient forecasts. The new model has been tested with the prediction of all warm events that occurred in the period 1996-2015. Retrospective forecasts of these events were made for long lead times of at least two and a half years. Hence, the present study demonstrates that the theoretical limit of ENSO prediction should be sought much longer than the commonly accepted "Spring Barrier". The high correspondence between the forecasts and observations indicates that the proposed model outperforms all current operational statistical models, and behaves comparably to the best dynamical models used for EN prediction. Thus, the novel way in which the modeling scheme has been structured could also be used for improving other statistical and dynamical modeling systems.

Spatial Ensemble Postprocessing of Precipitation Forecasts Using High Resolution Analyses

NASA Astrophysics Data System (ADS)

Lang, Moritz N.; Schicker, Irene; Kann, Alexander; Wang, Yong

2017-04-01

Ensemble prediction systems are designed to account for errors or uncertainties in the initial and boundary conditions, imperfect parameterizations, etc. However, due to sampling errors and underestimation of the model errors, these ensemble forecasts tend to be underdispersive, and to lack both reliability and sharpness. To overcome such limitations, statistical postprocessing methods are commonly applied to these forecasts. In this study, a full-distributional spatial post-processing method is applied to short-range precipitation forecasts over Austria using Standardized Anomaly Model Output Statistics (SAMOS). Following Stauffer et al. (2016), observation and forecast fields are transformed into standardized anomalies by subtracting a site-specific climatological mean and dividing by the climatological standard deviation. Due to the need of fitting only a single regression model for the whole domain, the SAMOS framework provides a computationally inexpensive method to create operationally calibrated probabilistic forecasts for any arbitrary location or for all grid points in the domain simultaneously. Taking advantage of the INCA system (Integrated Nowcasting through Comprehensive Analysis), high resolution analyses are used for the computation of the observed climatology and for model training. The INCA system operationally combines station measurements and remote sensing data into real-time objective analysis fields at 1 km-horizontal resolution and 1 h-temporal resolution. The precipitation forecast used in this study is obtained from a limited area model ensemble prediction system also operated by ZAMG. The so called ALADIN-LAEF provides, by applying a multi-physics approach, a 17-member forecast at a horizontal resolution of 10.9 km and a temporal resolution of 1 hour. The performed SAMOS approach statistically combines the in-house developed high resolution analysis and ensemble prediction system. The station-based validation of 6 hour precipitation sums shows a mean improvement of more than 40% in CRPS when compared to bilinearly interpolated uncalibrated ensemble forecasts. The validation on randomly selected grid points, representing the true height distribution over Austria, still indicates a mean improvement of 35%. The applied statistical model is currently set up for 6-hourly and daily accumulation periods, but will be extended to a temporal resolution of 1-3 hours within a new probabilistic nowcasting system operated by ZAMG.

Multivariate Statistical Postprocessing of Ensemble Forcasts of Precipitation and Temperature over four River Basins in California

NASA Astrophysics Data System (ADS)

Scheuerer, Michael; Hamill, Thomas M.; Whitin, Brett; He, Minxue; Henkel, Arthur

2017-04-01

Hydrological forecasts strongly rely on predictions of precipitation amounts and temperature as meteorological inputs to hydrological models. Ensemble weather predictions provide a number of different scenarios that reflect the uncertainty about these meteorological inputs, but are often biased and underdispersive, and therefore require statistical postprocessing. In hydrological applications it is crucial that spatial and temporal (i.e. between different forecast lead times) dependencies as well as dependence between the two weather variables is adequately represented by the recalibrated forecasts. We present a study with temperature and precipitation forecasts over four river basins over California that are postprocessed with a variant of the nonhomogeneous Gaussian regression method (Gneiting et al., 2005) and the censored, shifted gamma distribution approach (Scheuerer and Hamill, 2015) respectively. For modelling spatial, temporal and inter-variable dependence we propose a variant of the Schaake Shuffle (Clark et al., 2005) that uses spatio-temporal trajectories of observed temperture and precipitation as a dependence template, and chooses the historic dates in such a way that the divergence between the marginal distributions of these trajectories and the univariate forecast distributions is minimized. For the four river basins considered in our study, this new multivariate modelling technique consistently improves upon the Schaake Shuffle and yields reliable spatio-temporal forecast trajectories of temperature and precipitation that can be used to force hydrological forecast systems. References: Clark, M., Gangopadhyay, S., Hay, L., Rajagopalan, B., Wilby, R., 2004. The Schaake Shuffle: A method for reconstructing space-time variability in forecasted precipitation and temperature fields. Journal of Hydrometeorology, 5, pp.243-262. Gneiting, T., Raftery, A.E., Westveld, A.H., Goldman, T., 2005. Calibrated probabilistic forecasting using ensemble model output statistics and minimum CRPS. Monthly Weather Review, 133, pp.1098-1118. Scheuerer, M., Hamill, T.M., 2015. Statistical postprocessing of ensemble precipitation forecasts by fitting censored, shifted gamma distributions. Monthly Weather Review, 143, pp.4578-4596. Scheuerer, M., Hamill, T.M., Whitin, B., He, M., and Henkel, A., 2016: A method for preferential selection of dates in the Schaake shuffle approach to constructing spatio-temporal forecast fields of temperature and precipitation. Water Resources Research, submitted.

An Objective Verification of the North American Mesoscale Model for Kennedy Space Center and Cape Canaveral Air Force Station

NASA Technical Reports Server (NTRS)

Bauman, William H., III

2010-01-01

The 45th Weather Squadron (45 WS) Launch Weather Officers use the 12-km resolution North American Mesoscale (NAM) model (MesoNAM) text and graphical product forecasts extensively to support launch weather operations. However, the actual performance of the model at Kennedy Space Center (KSC) and Cape Canaveral Air Force Station (CCAFS) has not been measured objectively. In order to have tangible evidence of model performance, the 45 WS tasked the Applied Meteorology Unit to conduct a detailed statistical analysis of model output compared to observed values. The model products are provided to the 45 WS by ACTA, Inc. and include hourly forecasts from 0 to 84 hours based on model initialization times of 00, 06, 12 and 18 UTC. The objective analysis compared the MesoNAM forecast winds, temperature and dew point, as well as the changes in these parameters over time, to the observed values from the sensors in the KSC/CCAFS wind tower network. Objective statistics will give the forecasters knowledge of the model's strength and weaknesses, which will result in improved forecasts for operations.

Glossary - NOAA's National Weather Service

Science.gov Websites

Organization Search NWS All NOAA Go Local forecast by "City, St" Search by city. Press enter or Text Bulletins By State By Message Type National Forecast Models Numerical Models Statistical Models

Nowcasting and Forecasting the Monthly Food Stamps Data in the US Using Online Search Data

PubMed Central

Fantazzini, Dean

2014-01-01

We propose the use of Google online search data for nowcasting and forecasting the number of food stamps recipients. We perform a large out-of-sample forecasting exercise with almost 3000 competing models with forecast horizons up to 2 years ahead, and we show that models including Google search data statistically outperform the competing models at all considered horizons. These results hold also with several robustness checks, considering alternative keywords, a falsification test, different out-of-samples, directional accuracy and forecasts at the state-level. PMID:25369315

Simultaneous calibration of ensemble river flow predictions over an entire range of lead times

NASA Astrophysics Data System (ADS)

Hemri, S.; Fundel, F.; Zappa, M.

2013-10-01

Probabilistic estimates of future water levels and river discharge are usually simulated with hydrologic models using ensemble weather forecasts as main inputs. As hydrologic models are imperfect and the meteorological ensembles tend to be biased and underdispersed, the ensemble forecasts for river runoff typically are biased and underdispersed, too. Thus, in order to achieve both reliable and sharp predictions statistical postprocessing is required. In this work Bayesian model averaging (BMA) is applied to statistically postprocess ensemble runoff raw forecasts for a catchment in Switzerland, at lead times ranging from 1 to 240 h. The raw forecasts have been obtained using deterministic and ensemble forcing meteorological models with different forecast lead time ranges. First, BMA is applied based on mixtures of univariate normal distributions, subject to the assumption of independence between distinct lead times. Then, the independence assumption is relaxed in order to estimate multivariate runoff forecasts over the entire range of lead times simultaneously, based on a BMA version that uses multivariate normal distributions. Since river runoff is a highly skewed variable, Box-Cox transformations are applied in order to achieve approximate normality. Both univariate and multivariate BMA approaches are able to generate well calibrated probabilistic forecasts that are considerably sharper than climatological forecasts. Additionally, multivariate BMA provides a promising approach for incorporating temporal dependencies into the postprocessed forecasts. Its major advantage against univariate BMA is an increase in reliability when the forecast system is changing due to model availability.

Forecasting Responses of a Northern Peatland Carbon Cycle to Elevated CO2 and a Gradient of Experimental Warming

NASA Astrophysics Data System (ADS)

Jiang, Jiang; Huang, Yuanyuan; Ma, Shuang; Stacy, Mark; Shi, Zheng; Ricciuto, Daniel M.; Hanson, Paul J.; Luo, Yiqi

2018-03-01

The ability to forecast ecological carbon cycling is imperative to land management in a world where past carbon fluxes are no longer a clear guide in the Anthropocene. However, carbon-flux forecasting has not been practiced routinely like numerical weather prediction. This study explored (1) the relative contributions of model forcing data and parameters to uncertainty in forecasting flux- versus pool-based carbon cycle variables and (2) the time points when temperature and CO2 treatments may cause statistically detectable differences in those variables. We developed an online forecasting workflow (Ecological Platform for Assimilation of Data (EcoPAD)), which facilitates iterative data-model integration. EcoPAD automates data transfer from sensor networks, data assimilation, and ecological forecasting. We used the Spruce and Peatland Responses Under Changing Experiments data collected from 2011 to 2014 to constrain the parameters in the Terrestrial Ecosystem Model, forecast carbon cycle responses to elevated CO2 and a gradient of warming from 2015 to 2024, and specify uncertainties in the model output. Our results showed that data assimilation substantially reduces forecasting uncertainties. Interestingly, we found that the stochasticity of future external forcing contributed more to the uncertainty of forecasting future dynamics of C flux-related variables than model parameters. However, the parameter uncertainty primarily contributes to the uncertainty in forecasting C pool-related response variables. Given the uncertainties in forecasting carbon fluxes and pools, our analysis showed that statistically different responses of fast-turnover pools to various CO2 and warming treatments were observed sooner than slow-turnover pools. Our study has identified the sources of uncertainties in model prediction and thus leads to improve ecological carbon cycling forecasts in the future.

EVALUATION OF SEVERAL PM 2.5 FORECAST MODELS USING DATA COLLECTED DURING THE ICARTT/NEAQS 2004 FIELD STUDY

EPA Science Inventory

Real-time forecasts of PM_2.5 aerosol mass from seven air-quality forecast models (AQFMs) are statistically evaluated against observations collected in the northeastern U.S. and southeastern Canada from two surface networks and aircraft data during the summer of 2004 IC...

A Comparison of Conventional Linear Regression Methods and Neural Networks for Forecasting Educational Spending.

ERIC Educational Resources Information Center

Baker, Bruce D.; Richards, Craig E.

1999-01-01

Applies neural network methods for forecasting 1991-95 per-pupil expenditures in U.S. public elementary and secondary schools. Forecasting models included the National Center for Education Statistics' multivariate regression model and three neural architectures. Regarding prediction accuracy, neural network results were comparable or superior to…

Dynamical-statistical seasonal prediction for western North Pacific typhoons based on APCC multi-models

NASA Astrophysics Data System (ADS)

Kim, Ok-Yeon; Kim, Hye-Mi; Lee, Myong-In; Min, Young-Mi

2017-01-01

This study aims at predicting the seasonal number of typhoons (TY) over the western North Pacific with an Asia-Pacific Climate Center (APCC) multi-model ensemble (MME)-based dynamical-statistical hybrid model. The hybrid model uses the statistical relationship between the number of TY during the typhoon season (July-October) and the large-scale key predictors forecasted by APCC MME for the same season. The cross validation result from the MME hybrid model demonstrates high prediction skill, with a correlation of 0.67 between the hindcasts and observation for 1982-2008. The cross validation from the hybrid model with individual models participating in MME indicates that there is no single model which consistently outperforms the other models in predicting typhoon number. Although the forecast skill of MME is not always the highest compared to that of each individual model, the skill of MME presents rather higher averaged correlations and small variance of correlations. Given large set of ensemble members from multi-models, a relative operating characteristic score reveals an 82 % (above-) and 78 % (below-normal) improvement for the probabilistic prediction of the number of TY. It implies that there is 82 % (78 %) probability that the forecasts can successfully discriminate between above normal (below-normal) from other years. The forecast skill of the hybrid model for the past 7 years (2002-2008) is more skillful than the forecast from the Tropical Storm Risk consortium. Using large set of ensemble members from multi-models, the APCC MME could provide useful deterministic and probabilistic seasonal typhoon forecasts to the end-users in particular, the residents of tropical cyclone-prone areas in the Asia-Pacific region.

Does money matter in inflation forecasting?

NASA Astrophysics Data System (ADS)

Binner, J. M.; Tino, P.; Tepper, J.; Anderson, R.; Jones, B.; Kendall, G.

2010-11-01

This paper provides the most fully comprehensive evidence to date on whether or not monetary aggregates are valuable for forecasting US inflation in the early to mid 2000s. We explore a wide range of different definitions of money, including different methods of aggregation and different collections of included monetary assets. In our forecasting experiment we use two nonlinear techniques, namely, recurrent neural networks and kernel recursive least squares regression-techniques that are new to macroeconomics. Recurrent neural networks operate with potentially unbounded input memory, while the kernel regression technique is a finite memory predictor. The two methodologies compete to find the best fitting US inflation forecasting models and are then compared to forecasts from a naïve random walk model. The best models were nonlinear autoregressive models based on kernel methods. Our findings do not provide much support for the usefulness of monetary aggregates in forecasting inflation. Beyond its economic findings, our study is in the tradition of physicists’ long-standing interest in the interconnections among statistical mechanics, neural networks, and related nonparametric statistical methods, and suggests potential avenues of extension for such studies.

A synoptic view of the Third Uniform California Earthquake Rupture Forecast (UCERF3)

USGS Publications Warehouse

Field, Edward; Jordan, Thomas H.; Page, Morgan T.; Milner, Kevin R.; Shaw, Bruce E.; Dawson, Timothy E.; Biasi, Glenn; Parsons, Thomas E.; Hardebeck, Jeanne L.; Michael, Andrew J.; Weldon, Ray; Powers, Peter; Johnson, Kaj M.; Zeng, Yuehua; Bird, Peter; Felzer, Karen; van der Elst, Nicholas; Madden, Christopher; Arrowsmith, Ramon; Werner, Maximillan J.; Thatcher, Wayne R.

2017-01-01

Probabilistic forecasting of earthquake‐producing fault ruptures informs all major decisions aimed at reducing seismic risk and improving earthquake resilience. Earthquake forecasting models rely on two scales of hazard evolution: long‐term (decades to centuries) probabilities of fault rupture, constrained by stress renewal statistics, and short‐term (hours to years) probabilities of distributed seismicity, constrained by earthquake‐clustering statistics. Comprehensive datasets on both hazard scales have been integrated into the Uniform California Earthquake Rupture Forecast, Version 3 (UCERF3). UCERF3 is the first model to provide self‐consistent rupture probabilities over forecasting intervals from less than an hour to more than a century, and it is the first capable of evaluating the short‐term hazards that result from multievent sequences of complex faulting. This article gives an overview of UCERF3, illustrates the short‐term probabilities with aftershock scenarios, and draws some valuable scientific conclusions from the modeling results. In particular, seismic, geologic, and geodetic data, when combined in the UCERF3 framework, reject two types of fault‐based models: long‐term forecasts constrained to have local Gutenberg–Richter scaling, and short‐term forecasts that lack stress relaxation by elastic rebound.

Operational planning using Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS)

NASA Astrophysics Data System (ADS)

O'Connor, Alison; Kirtman, Benjamin; Harrison, Scott; Gorman, Joe

2016-05-01

The US Navy faces several limitations when planning operations in regard to forecasting environmental conditions. Currently, mission analysis and planning tools rely heavily on short-term (less than a week) forecasts or long-term statistical climate products. However, newly available data in the form of weather forecast ensembles provides dynamical and statistical extended-range predictions that can produce more accurate predictions if ensemble members can be combined correctly. Charles River Analytics is designing the Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS), which performs data fusion over extended-range multi-model ensembles, such as the North American Multi-Model Ensemble (NMME), to produce a unified forecast for several weeks to several seasons in the future. We evaluated thirty years of forecasts using machine learning to select predictions for an all-encompassing and superior forecast that can be used to inform the Navy's decision planning process.

Forecasting in foodservice: model development, testing, and evaluation.

PubMed

Miller, J L; Thompson, P A; Orabella, M M

1991-05-01

This study was designed to develop, test, and evaluate mathematical models appropriate for forecasting menu-item production demand in foodservice. Data were collected from residence and dining hall foodservices at Ohio State University. Objectives of the study were to collect, code, and analyze the data; develop and test models using actual operation data; and compare forecasting results with current methods in use. Customer count was forecast using deseasonalized simple exponential smoothing. Menu-item demand was forecast by multiplying the count forecast by a predicted preference statistic. Forecasting models were evaluated using mean squared error, mean absolute deviation, and mean absolute percentage error techniques. All models were more accurate than current methods. A broad spectrum of forecasting techniques could be used by foodservice managers with access to a personal computer and spread-sheet and database-management software. The findings indicate that mathematical forecasting techniques may be effective in foodservice operations to control costs, increase productivity, and maximize profits.

The probability forecast evaluation of hazard and storm wind over the territories of Russia and Europe

NASA Astrophysics Data System (ADS)

Perekhodtseva, E. V.

2012-04-01

The results of the probability forecast methods of summer storm and hazard wind over territories of Russia and Europe are submitted at this paper. These methods use the hydrodynamic-statistical model of these phenomena. The statistical model was developed for the recognition of the situation involving these phenomena. For this perhaps the samples of the values of atmospheric parameters (n=40) for the presence and for the absence of these phenomena of storm and hazard wind were accumulated. The compressing of the predictors space without the information losses was obtained by special algorithm (k=7<19m/s, the values of 65%24m/s, the values of 75%29m/s or the area of the tornado and strong squalls. The evaluation of this probability forecast was provided by criterion of Brayer. The estimation was successful and was equal for the European part of Russia B=0,37. The application of the probability forecast of storm and hazard winds allows to mitigate the economic losses when the errors of the first and second kinds of storm wind categorical forecast are not so small. A lot of examples of the storm wind probability forecast are submitted at this report.

Evaluation of model-based seasonal streamflow and water allocation forecasts for the Elqui Valley, Chile

NASA Astrophysics Data System (ADS)

Delorit, Justin; Cristian Gonzalez Ortuya, Edmundo; Block, Paul

2017-09-01

In many semi-arid regions, multisectoral demands often stress available water supplies. Such is the case in the Elqui River valley of northern Chile, which draws on a limited-capacity reservoir to allocate 25 000 water rights. Delayed infrastructure investment forces water managers to address demand-based allocation strategies, particularly in dry years, which are realized through reductions in the volume associated with each water right. Skillful season-ahead streamflow forecasts have the potential to inform managers with an indication of future conditions to guide reservoir allocations. This work evaluates season-ahead statistical prediction models of October-January (growing season) streamflow at multiple lead times associated with manager and user decision points, and links predictions with a reservoir allocation tool. Skillful results (streamflow forecasts outperform climatology) are produced for short lead times (1 September: ranked probability skill score (RPSS) of 0.31, categorical hit skill score of 61 %). At longer lead times, climatological skill exceeds forecast skill due to fewer observations of precipitation. However, coupling the 1 September statistical forecast model with a sea surface temperature phase and strength statistical model allows for equally skillful categorical streamflow forecasts to be produced for a 1 May lead, triggered for 60 % of years (1950-2015), suggesting forecasts need not be strictly deterministic to be useful for water rights holders. An early (1 May) categorical indication of expected conditions is reinforced with a deterministic forecast (1 September) as more observations of local variables become available. The reservoir allocation model is skillful at the 1 September lead (categorical hit skill score of 53 %); skill improves to 79 % when categorical allocation prediction certainty exceeds 80 %. This result implies that allocation efficiency may improve when forecasts are integrated into reservoir decision frameworks. The methods applied here advance the understanding of the mechanisms and timing responsible for moisture transport to the Elqui Valley and provide a unique application of streamflow forecasting in the prediction of water right allocations.

Testing the Predictive Power of Coulomb Stress on Aftershock Sequences

NASA Astrophysics Data System (ADS)

Woessner, J.; Lombardi, A.; Werner, M. J.; Marzocchi, W.

2009-12-01

Empirical and statistical models of clustered seismicity are usually strongly stochastic and perceived to be uninformative in their forecasts, since only marginal distributions are used, such as the Omori-Utsu and Gutenberg-Richter laws. In contrast, so-called physics-based aftershock models, based on seismic rate changes calculated from Coulomb stress changes and rate-and-state friction, make more specific predictions: anisotropic stress shadows and multiplicative rate changes. We test the predictive power of models based on Coulomb stress changes against statistical models, including the popular Short Term Earthquake Probabilities and Epidemic-Type Aftershock Sequences models: We score and compare retrospective forecasts on the aftershock sequences of the 1992 Landers, USA, the 1997 Colfiorito, Italy, and the 2008 Selfoss, Iceland, earthquakes. To quantify predictability, we use likelihood-based metrics that test the consistency of the forecasts with the data, including modified and existing tests used in prospective forecast experiments within the Collaboratory for the Study of Earthquake Predictability (CSEP). Our results indicate that a statistical model performs best. Moreover, two Coulomb model classes seem unable to compete: Models based on deterministic Coulomb stress changes calculated from a given fault-slip model, and those based on fixed receiver faults. One model of Coulomb stress changes does perform well and sometimes outperforms the statistical models, but its predictive information is diluted, because of uncertainties included in the fault-slip model. Our results suggest that models based on Coulomb stress changes need to incorporate stochastic features that represent model and data uncertainty.

Increased Accuracy in Statistical Seasonal Hurricane Forecasting

NASA Astrophysics Data System (ADS)

Nateghi, R.; Quiring, S. M.; Guikema, S. D.

2012-12-01

Hurricanes are among the costliest and most destructive natural hazards in the U.S. Accurate hurricane forecasts are crucial to optimal preparedness and mitigation decisions in the U.S. where 50 percent of the population lives within 50 miles of the coast. We developed a flexible statistical approach to forecast annual number of hurricanes in the Atlantic region during the hurricane season. Our model is based on the method of Random Forest and captures the complex relationship between hurricane activity and climatic conditions through careful variable selection, model testing and validation. We used the National Hurricane Center's Best Track hurricane data from 1949-2011 and sixty-one candidate climate descriptors to develop our model. The model includes information prior to the hurricane season, i.e., from the last three months of the previous year (Oct. through Dec.) and the first five months of the current year (January through May). Our forecast errors are substantially lower than other leading forecasts such as that of the National Oceanic and Atmospheric Administration (NOAA).

The MSFC Solar Activity Future Estimation (MSAFE) Model

NASA Technical Reports Server (NTRS)

Suggs, Ron

2017-01-01

The Natural Environments Branch of the Engineering Directorate at Marshall Space Flight Center (MSFC) provides solar cycle forecasts for NASA space flight programs and the aerospace community. These forecasts provide future statistical estimates of sunspot number, solar radio 10.7 cm flux (F10.7), and the geomagnetic planetary index, Ap, for input to various space environment models. For example, many thermosphere density computer models used in spacecraft operations, orbital lifetime analysis, and the planning of future spacecraft missions require as inputs the F10.7 and Ap. The solar forecast is updated each month by executing MSAFE using historical and the latest month's observed solar indices to provide estimates for the balance of the current solar cycle. The forecasted solar indices represent the 13-month smoothed values consisting of a best estimate value stated as a 50 percentile value along with approximate +/- 2 sigma values stated as 95 and 5 percentile statistical values. This presentation will give an overview of the MSAFE model and the forecast for the current solar cycle.

Development of visibility forecasting modeling framework for the Lower Fraser Valley of British Columbia using Canada's Regional Air Quality Deterministic Prediction System.

PubMed

So, Rita; Teakles, Andrew; Baik, Jonathan; Vingarzan, Roxanne; Jones, Keith

2018-05-01

Visibility degradation, one of the most noticeable indicators of poor air quality, can occur despite relatively low levels of particulate matter when the risk to human health is low. The availability of timely and reliable visibility forecasts can provide a more comprehensive understanding of the anticipated air quality conditions to better inform local jurisdictions and the public. This paper describes the development of a visibility forecasting modeling framework, which leverages the existing air quality and meteorological forecasts from Canada's operational Regional Air Quality Deterministic Prediction System (RAQDPS) for the Lower Fraser Valley of British Columbia. A baseline model (GM-IMPROVE) was constructed using the revised IMPROVE algorithm based on unprocessed forecasts from the RAQDPS. Three additional prototypes (UMOS-HYB, GM-MLR, GM-RF) were also developed and assessed for forecast performance of up to 48 hr lead time during various air quality and meteorological conditions. Forecast performance was assessed by examining their ability to provide both numerical and categorical forecasts in the form of 1-hr total extinction and Visual Air Quality Ratings (VAQR), respectively. While GM-IMPROVE generally overestimated extinction more than twofold, it had skill in forecasting the relative species contribution to visibility impairment, including ammonium sulfate and ammonium nitrate. Both statistical prototypes, GM-MLR and GM-RF, performed well in forecasting 1-hr extinction during daylight hours, with correlation coefficients (R) ranging from 0.59 to 0.77. UMOS-HYB, a prototype based on postprocessed air quality forecasts without additional statistical modeling, provided reasonable forecasts during most daylight hours. In terms of categorical forecasts, the best prototype was approximately 75 to 87% correct, when forecasting for a condensed three-category VAQR. A case study, focusing on a poor visual air quality yet low Air Quality Health Index episode, illustrated that the statistical prototypes were able to provide timely and skillful visibility forecasts with lead time up to 48 hr. This study describes the development of a visibility forecasting modeling framework, which leverages the existing air quality and meteorological forecasts from Canada's operational Regional Air Quality Deterministic Prediction System. The main applications include tourism and recreation planning, input into air quality management programs, and educational outreach. Visibility forecasts, when supplemented with the existing air quality and health based forecasts, can assist jurisdictions to anticipate the visual air quality impacts as perceived by the public, which can potentially assist in formulating the appropriate air quality bulletins and recommendations.

Statistical Forecasting of Current and Future Circum-Arctic Ground Temperatures and Active Layer Thickness

NASA Astrophysics Data System (ADS)

Aalto, J.; Karjalainen, O.; Hjort, J.; Luoto, M.

2018-05-01

Mean annual ground temperature (MAGT) and active layer thickness (ALT) are key to understanding the evolution of the ground thermal state across the Arctic under climate change. Here a statistical modeling approach is presented to forecast current and future circum-Arctic MAGT and ALT in relation to climatic and local environmental factors, at spatial scales unreachable with contemporary transient modeling. After deploying an ensemble of multiple statistical techniques, distance-blocked cross validation between observations and predictions suggested excellent and reasonable transferability of the MAGT and ALT models, respectively. The MAGT forecasts indicated currently suitable conditions for permafrost to prevail over an area of 15.1 ± 2.8 × 106 km2. This extent is likely to dramatically contract in the future, as the results showed consistent, but region-specific, changes in ground thermal regime due to climate change. The forecasts provide new opportunities to assess future Arctic changes in ground thermal state and biogeochemical feedback.

Influenza forecasting with Google Flu Trends.

PubMed

Dugas, Andrea Freyer; Jalalpour, Mehdi; Gel, Yulia; Levin, Scott; Torcaso, Fred; Igusa, Takeru; Rothman, Richard E

2013-01-01

We developed a practical influenza forecast model based on real-time, geographically focused, and easy to access data, designed to provide individual medical centers with advanced warning of the expected number of influenza cases, thus allowing for sufficient time to implement interventions. Secondly, we evaluated the effects of incorporating a real-time influenza surveillance system, Google Flu Trends, and meteorological and temporal information on forecast accuracy. Forecast models designed to predict one week in advance were developed from weekly counts of confirmed influenza cases over seven seasons (2004-2011) divided into seven training and out-of-sample verification sets. Forecasting procedures using classical Box-Jenkins, generalized linear models (GLM), and generalized linear autoregressive moving average (GARMA) methods were employed to develop the final model and assess the relative contribution of external variables such as, Google Flu Trends, meteorological data, and temporal information. A GARMA(3,0) forecast model with Negative Binomial distribution integrating Google Flu Trends information provided the most accurate influenza case predictions. The model, on the average, predicts weekly influenza cases during 7 out-of-sample outbreaks within 7 cases for 83% of estimates. Google Flu Trend data was the only source of external information to provide statistically significant forecast improvements over the base model in four of the seven out-of-sample verification sets. Overall, the p-value of adding this external information to the model is 0.0005. The other exogenous variables did not yield a statistically significant improvement in any of the verification sets. Integer-valued autoregression of influenza cases provides a strong base forecast model, which is enhanced by the addition of Google Flu Trends confirming the predictive capabilities of search query based syndromic surveillance. This accessible and flexible forecast model can be used by individual medical centers to provide advanced warning of future influenza cases.

Spatial Pattern Classification for More Accurate Forecasting of Variable Energy Resources

NASA Astrophysics Data System (ADS)

Novakovskaia, E.; Hayes, C.; Collier, C.

2014-12-01

The accuracy of solar and wind forecasts is becoming increasingly essential as grid operators continue to integrate additional renewable generation onto the electric grid. Forecast errors affect rate payers, grid operators, wind and solar plant maintenance crews and energy traders through increases in prices, project down time or lost revenue. While extensive and beneficial efforts were undertaken in recent years to improve physical weather models for a broad spectrum of applications these improvements have generally not been sufficient to meet the accuracy demands of system planners. For renewables, these models are often used in conjunction with additional statistical models utilizing both meteorological observations and the power generation data. Forecast accuracy can be dependent on specific weather regimes for a given location. To account for these dependencies it is important that parameterizations used in statistical models change as the regime changes. An automated tool, based on an artificial neural network model, has been developed to identify different weather regimes as they impact power output forecast accuracy at wind or solar farms. In this study, improvements in forecast accuracy were analyzed for varying time horizons for wind farms and utility-scale PV plants located in different geographical regions.

Assessment of the long-lead probabilistic prediction for the Asian summer monsoon precipitation (1983-2011) based on the APCC multimodel system and a statistical model

NASA Astrophysics Data System (ADS)

Sohn, Soo-Jin; Min, Young-Mi; Lee, June-Yi; Tam, Chi-Yung; Kang, In-Sik; Wang, Bin; Ahn, Joong-Bae; Yamagata, Toshio

2012-02-01

The performance of the probabilistic multimodel prediction (PMMP) system of the APEC Climate Center (APCC) in predicting the Asian summer monsoon (ASM) precipitation at a four-month lead (with February initial condition) was compared with that of a statistical model using hindcast data for 1983-2005 and real-time forecasts for 2006-2011. Particular attention was paid to probabilistic precipitation forecasts for the boreal summer after the mature phase of El Niño and Southern Oscillation (ENSO). Taking into account the fact that coupled models' skill for boreal spring and summer precipitation mainly comes from their ability to capture ENSO teleconnection, we developed the statistical model using linear regression with the preceding winter ENSO condition as the predictor. Our results reveal several advantages and disadvantages in both forecast systems. First, the PMMP appears to have higher skills for both above- and below-normal categories in the six-year real-time forecast period, whereas the cross-validated statistical model has higher skills during the 23-year hindcast period. This implies that the cross-validated statistical skill may be overestimated. Second, the PMMP is the better tool for capturing atypical ENSO (or non-canonical ENSO related) teleconnection, which has affected the ASM precipitation during the early 1990s and in the recent decade. Third, the statistical model is more sensitive to the ENSO phase and has an advantage in predicting the ASM precipitation after the mature phase of La Niña.

Providing peak river flow statistics and forecasting in the Niger River basin

NASA Astrophysics Data System (ADS)

Andersson, Jafet C. M.; Ali, Abdou; Arheimer, Berit; Gustafsson, David; Minoungou, Bernard

2017-08-01

Flooding is a growing concern in West Africa. Improved quantification of discharge extremes and associated uncertainties is needed to improve infrastructure design, and operational forecasting is needed to provide timely warnings. In this study, we use discharge observations, a hydrological model (Niger-HYPE) and extreme value analysis to estimate peak river flow statistics (e.g. the discharge magnitude with a 100-year return period) across the Niger River basin. To test the model's capacity of predicting peak flows, we compared 30-year maximum discharge and peak flow statistics derived from the model vs. derived from nine observation stations. The results indicate that the model simulates peak discharge reasonably well (on average + 20%). However, the peak flow statistics have a large uncertainty range, which ought to be considered in infrastructure design. We then applied the methodology to derive basin-wide maps of peak flow statistics and their associated uncertainty. The results indicate that the method is applicable across the hydrologically active part of the river basin, and that the uncertainty varies substantially depending on location. Subsequently, we used the most recent bias-corrected climate projections to analyze potential changes in peak flow statistics in a changed climate. The results are generally ambiguous, with consistent changes only in very few areas. To test the forecasting capacity, we ran Niger-HYPE with a combination of meteorological data sets for the 2008 high-flow season and compared with observations. The results indicate reasonable forecasting capacity (on average 17% deviation), but additional years should also be evaluated. We finish by presenting a strategy and pilot project which will develop an operational flood monitoring and forecasting system based in-situ data, earth observations, modelling, and extreme statistics. In this way we aim to build capacity to ultimately improve resilience toward floods, protecting lives and infrastructure in the region.

Assessment of summer rainfall forecast skill in the Intra-Americas in GFDL high and low-resolution models

NASA Astrophysics Data System (ADS)

Krishnamurthy, Lakshmi; Muñoz, Ángel G.; Vecchi, Gabriel A.; Msadek, Rym; Wittenberg, Andrew T.; Stern, Bill; Gudgel, Rich; Zeng, Fanrong

2018-05-01

The Caribbean low-level jet (CLLJ) is an important component of the atmospheric circulation over the Intra-Americas Sea (IAS) which impacts the weather and climate both locally and remotely. It influences the rainfall variability in the Caribbean, Central America, northern South America, the tropical Pacific and the continental Unites States through the transport of moisture. We make use of high-resolution coupled and uncoupled models from the Geophysical Fluid Dynamics Laboratory (GFDL) to investigate the simulation of the CLLJ and its teleconnections and further compare with low-resolution models. The high-resolution coupled model FLOR shows improvements in the simulation of the CLLJ and its teleconnections with rainfall and SST over the IAS compared to the low-resolution coupled model CM2.1. The CLLJ is better represented in uncoupled models (AM2.1 and AM2.5) forced with observed sea-surface temperatures (SSTs), emphasizing the role of SSTs in the simulation of the CLLJ. Further, we determine the forecast skill for observed rainfall using both high- and low-resolution predictions of rainfall and SSTs for the July-August-September season. We determine the role of statistical correction of model biases, coupling and horizontal resolution on the forecast skill. Statistical correction dramatically improves area-averaged forecast skill. But the analysis of spatial distribution in skill indicates that the improvement in skill after statistical correction is region dependent. Forecast skill is sensitive to coupling in parts of the Caribbean, Central and northern South America, and it is mostly insensitive over North America. Comparison of forecast skill between high and low-resolution coupled models does not show any dramatic difference. However, uncoupled models show improvement in the area-averaged skill in the high-resolution atmospheric model compared to lower resolution model. Understanding and improving the forecast skill over the IAS has important implications for highly vulnerable nations in the region.

A comparative verification of high resolution precipitation forecasts using model output statistics

NASA Astrophysics Data System (ADS)

van der Plas, Emiel; Schmeits, Maurice; Hooijman, Nicolien; Kok, Kees

2017-04-01

Verification of localized events such as precipitation has become even more challenging with the advent of high-resolution meso-scale numerical weather prediction (NWP). The realism of a forecast suggests that it should compare well against precipitation radar imagery with similar resolution, both spatially and temporally. Spatial verification methods solve some of the representativity issues that point verification gives rise to. In this study a verification strategy based on model output statistics is applied that aims to address both double penalty and resolution effects that are inherent to comparisons of NWP models with different resolutions. Using predictors based on spatial precipitation patterns around a set of stations, an extended logistic regression (ELR) equation is deduced, leading to a probability forecast distribution of precipitation for each NWP model, analysis and lead time. The ELR equations are derived for predictands based on areal calibrated radar precipitation and SYNOP observations. The aim is to extract maximum information from a series of precipitation forecasts, like a trained forecaster would. The method is applied to the non-hydrostatic model Harmonie (2.5 km resolution), Hirlam (11 km resolution) and the ECMWF model (16 km resolution), overall yielding similar Brier skill scores for the 3 post-processed models, but larger differences for individual lead times. Besides, the Fractions Skill Score is computed using the 3 deterministic forecasts, showing somewhat better skill for the Harmonie model. In other words, despite the realism of Harmonie precipitation forecasts, they only perform similarly or somewhat better than precipitation forecasts from the 2 lower resolution models, at least in the Netherlands.

Post-processing method for wind speed ensemble forecast using wind speed and direction

NASA Astrophysics Data System (ADS)

Sofie Eide, Siri; Bjørnar Bremnes, John; Steinsland, Ingelin

2017-04-01

Statistical methods are widely applied to enhance the quality of both deterministic and ensemble NWP forecasts. In many situations, like wind speed forecasting, most of the predictive information is contained in one variable in the NWP models. However, in statistical calibration of deterministic forecasts it is often seen that including more variables can further improve forecast skill. For ensembles this is rarely taken advantage of, mainly due to that it is generally not straightforward how to include multiple variables. In this study, it is demonstrated how multiple variables can be included in Bayesian model averaging (BMA) by using a flexible regression method for estimating the conditional means. The method is applied to wind speed forecasting at 204 Norwegian stations based on wind speed and direction forecasts from the ECMWF ensemble system. At about 85 % of the sites the ensemble forecasts were improved in terms of CRPS by adding wind direction as predictor compared to only using wind speed. On average the improvements were about 5 %, but mainly for moderate to strong wind situations. For weak wind speeds adding wind direction had more or less neutral impact.

Incorporating Yearly Derived Winter Wheat Maps Into Winter Wheat Yield Forecasting Model

NASA Technical Reports Server (NTRS)

Skakun, S.; Franch, B.; Roger, J.-C.; Vermote, E.; Becker-Reshef, I.; Justice, C.; Santamaría-Artigas, A.

2016-01-01

Wheat is one of the most important cereal crops in the world. Timely and accurate forecast of wheat yield and production at global scale is vital in implementing food security policy. Becker-Reshef et al. (2010) developed a generalized empirical model for forecasting winter wheat production using remote sensing data and official statistics. This model was implemented using static wheat maps. In this paper, we analyze the impact of incorporating yearly wheat masks into the forecasting model. We propose a new approach of producing in season winter wheat maps exploiting satellite data and official statistics on crop area only. Validation on independent data showed that the proposed approach reached 6% to 23% of omission error and 10% to 16% of commission error when mapping winter wheat 2-3 months before harvest. In general, we found a limited impact of using yearly winter wheat masks over a static mask for the study regions.

Joint Seasonal ARMA Approach for Modeling of Load Forecast Errors in Planning Studies

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

Hafen, Ryan P.; Samaan, Nader A.; Makarov, Yuri V.

2014-04-14

To make informed and robust decisions in the probabilistic power system operation and planning process, it is critical to conduct multiple simulations of the generated combinations of wind and load parameters and their forecast errors to handle the variability and uncertainty of these time series. In order for the simulation results to be trustworthy, the simulated series must preserve the salient statistical characteristics of the real series. In this paper, we analyze day-ahead load forecast error data from multiple balancing authority locations and characterize statistical properties such as mean, standard deviation, autocorrelation, correlation between series, time-of-day bias, and time-of-day autocorrelation.more » We then construct and validate a seasonal autoregressive moving average (ARMA) model to model these characteristics, and use the model to jointly simulate day-ahead load forecast error series for all BAs.« less

Assessing North American multimodel ensemble (NMME) seasonal forecast skill to assist in the early warning of hydrometeorological extremes over East Africa

USGS Publications Warehouse

Shukla, Shraddhanand; Roberts, Jason B.; Hoell. Andrew,; Funk, Chris; Robertson, Franklin R.; Kirtmann, Benjamin

2016-01-01

The skill of North American multimodel ensemble (NMME) seasonal forecasts in East Africa (EA), which encompasses one of the most food and water insecure areas of the world, is evaluated using deterministic, categorical, and probabilistic evaluation methods. The skill is estimated for all three primary growing seasons: March–May (MAM), July–September (JAS), and October–December (OND). It is found that the precipitation forecast skill in this region is generally limited and statistically significant over only a small part of the domain. In the case of MAM (JAS) [OND] season it exceeds the skill of climatological forecasts in parts of equatorial EA (Northern Ethiopia) [equatorial EA] for up to 2 (5) [5] months lead. Temperature forecast skill is generally much higher than precipitation forecast skill (in terms of deterministic and probabilistic skill scores) and statistically significant over a majority of the region. Over the region as a whole, temperature forecasts also exhibit greater reliability than the precipitation forecasts. The NMME ensemble forecasts are found to be more skillful and reliable than the forecast from any individual model. The results also demonstrate that for some seasons (e.g. JAS), the predictability of precipitation signals varies and is higher during certain climate events (e.g. ENSO). Finally, potential room for improvement in forecast skill is identified in some models by comparing homogeneous predictability in individual NMME models with their respective forecast skill.

An Objective Verification of the North American Mesoscale Model for Kennedy Space Center and Cape Canaveral Air Force Station

NASA Technical Reports Server (NTRS)

Bauman, William H., III

2010-01-01

The 45th Weather Squadron (45 WS) Launch Weather Officers (LWO's) use the 12-km resolution North American Mesoscale (NAM) model (MesoNAM) text and graphical product forecasts extensively to support launch weather operations. However, the actual performance of the model at Kennedy Space Center (KSC) and Cape Canaveral Air Force Station (CCAFS) has not been measured objectively. In order to have tangible evidence of model performance, the 45 WS tasked the Applied Meteorology Unit (AMU; Bauman et ai, 2004) to conduct a detailed statistical analysis of model output compared to observed values. The model products are provided to the 45 WS by ACTA, Inc. and include hourly forecasts from 0 to 84 hours based on model initialization times of 00, 06, 12 and 18 UTC. The objective analysis compared the MesoNAM forecast winds, temperature (T) and dew pOint (T d), as well as the changes in these parameters over time, to the observed values from the sensors in the KSC/CCAFS wind tower network shown in Table 1. These objective statistics give the forecasters knowledge of the model's strengths and weaknesses, which will result in improved forecasts for operations.

Seasonal forecasting of discharge for the Raccoon River, Iowa

NASA Astrophysics Data System (ADS)

Slater, Louise; Villarini, Gabriele; Bradley, Allen; Vecchi, Gabriel

2016-04-01

The state of Iowa (central United States) is regularly afflicted by severe natural hazards such as the 2008/2013 floods and the 2012 drought. To improve preparedness for these catastrophic events and allow Iowans to make more informed decisions about the most suitable water management strategies, we have developed a framework for medium to long range probabilistic seasonal streamflow forecasting for the Raccoon River at Van Meter, a 8900-km2 catchment located in central-western Iowa. Our flow forecasts use statistical models to predict seasonal discharge for low to high flows, with lead forecasting times ranging from one to ten months. Historical measurements of daily discharge are obtained from the U.S. Geological Survey (USGS) at the Van Meter stream gage, and used to compute quantile time series from minimum to maximum seasonal flow. The model is forced with basin-averaged total seasonal precipitation records from the PRISM Climate Group and annual row crop production acreage from the U.S. Department of Agriculture's National Agricultural Statistics Services database. For the forecasts, we use corn and soybean production from the previous year (persistence forecast) as a proxy for the impacts of agricultural practices on streamflow. The monthly precipitation forecasts are provided by eight Global Climate Models (GCMs) from the North American Multi-Model Ensemble (NMME), with lead times ranging from 0.5 to 11.5 months, and a resolution of 1 decimal degree. Additionally, precipitation from the month preceding each season is used to characterize antecedent soil moisture conditions. The accuracy of our modelled (1927-2015) and forecasted (2001-2015) discharge values is assessed by comparison with the observed USGS data. We explore the sensitivity of forecast skill over the full range of lead times, flow quantiles, forecast seasons, and with each GCM. Forecast skill is also examined using different formulations of the statistical models, as well as NMME forecast weighting procedures based on the computed potential skill (historical forecast accuracy) of the different GCMs. We find that the models describe the year-to-year variability in streamflow accurately, as well as the overall tendency towards increasing (and more variable) discharge over time. Surprisingly, forecast skill does not decrease markedly with lead time, and high flows tend to be well predicted, suggesting that these forecasts may have considerable practical applications. Further, the seasonal flow forecast accuracy is substantially improved by weighting the contribution of individual GCMs to the forecasts, and also by the inclusion of antecedent precipitation. Our results can provide critical information for adaptation strategies aiming to mitigate the costs and disruptions arising from flood and drought conditions, and allow us to determine how far in advance skillful forecasts can be issued. The availability of these discharge forecasts would have major societal and economic benefits for hydrology and water resources management, agriculture, disaster forecasts and prevention, energy, finance and insurance, food security, policy-making and public authorities, and transportation.

The total probabilities from high-resolution ensemble forecasting of floods

NASA Astrophysics Data System (ADS)

Olav Skøien, Jon; Bogner, Konrad; Salamon, Peter; Smith, Paul; Pappenberger, Florian

2015-04-01

Ensemble forecasting has for a long time been used in meteorological modelling, to give an indication of the uncertainty of the forecasts. As meteorological ensemble forecasts often show some bias and dispersion errors, there is a need for calibration and post-processing of the ensembles. Typical methods for this are Bayesian Model Averaging (Raftery et al., 2005) and Ensemble Model Output Statistics (EMOS) (Gneiting et al., 2005). There are also methods for regionalizing these methods (Berrocal et al., 2007) and for incorporating the correlation between lead times (Hemri et al., 2013). To make optimal predictions of floods along the stream network in hydrology, we can easily use the ensemble members as input to the hydrological models. However, some of the post-processing methods will need modifications when regionalizing the forecasts outside the calibration locations, as done by Hemri et al. (2013). We present a method for spatial regionalization of the post-processed forecasts based on EMOS and top-kriging (Skøien et al., 2006). We will also look into different methods for handling the non-normality of runoff and the effect on forecasts skills in general and for floods in particular. Berrocal, V. J., Raftery, A. E. and Gneiting, T.: Combining Spatial Statistical and Ensemble Information in Probabilistic Weather Forecasts, Mon. Weather Rev., 135(4), 1386-1402, doi:10.1175/MWR3341.1, 2007. Gneiting, T., Raftery, A. E., Westveld, A. H. and Goldman, T.: Calibrated Probabilistic Forecasting Using Ensemble Model Output Statistics and Minimum CRPS Estimation, Mon. Weather Rev., 133(5), 1098-1118, doi:10.1175/MWR2904.1, 2005. Hemri, S., Fundel, F. and Zappa, M.: Simultaneous calibration of ensemble river flow predictions over an entire range of lead times, Water Resour. Res., 49(10), 6744-6755, doi:10.1002/wrcr.20542, 2013. Raftery, A. E., Gneiting, T., Balabdaoui, F. and Polakowski, M.: Using Bayesian Model Averaging to Calibrate Forecast Ensembles, Mon. Weather Rev., 133(5), 1155-1174, doi:10.1175/MWR2906.1, 2005. Skøien, J. O., Merz, R. and Blöschl, G.: Top-kriging - Geostatistics on stream networks, Hydrol. Earth Syst. Sci., 10(2), 277-287, 2006.

Three-Month Real-Time Dengue Forecast Models: An Early Warning System for Outbreak Alerts and Policy Decision Support in Singapore

PubMed Central

Shi, Yuan; Liu, Xu; Kok, Suet-Yheng; Rajarethinam, Jayanthi; Liang, Shaohong; Yap, Grace; Chong, Chee-Seng; Lee, Kim-Sung; Tan, Sharon S.Y.; Chin, Christopher Kuan Yew; Lo, Andrew; Kong, Waiming; Ng, Lee Ching; Cook, Alex R.

2015-01-01

Background: With its tropical rainforest climate, rapid urbanization, and changing demography and ecology, Singapore experiences endemic dengue; the last large outbreak in 2013 culminated in 22,170 cases. In the absence of a vaccine on the market, vector control is the key approach for prevention. Objectives: We sought to forecast the evolution of dengue epidemics in Singapore to provide early warning of outbreaks and to facilitate the public health response to moderate an impending outbreak. Methods: We developed a set of statistical models using least absolute shrinkage and selection operator (LASSO) methods to forecast the weekly incidence of dengue notifications over a 3-month time horizon. This forecasting tool used a variety of data streams and was updated weekly, including recent case data, meteorological data, vector surveillance data, and population-based national statistics. The forecasting methodology was compared with alternative approaches that have been proposed to model dengue case data (seasonal autoregressive integrated moving average and step-down linear regression) by fielding them on the 2013 dengue epidemic, the largest on record in Singapore. Results: Operationally useful forecasts were obtained at a 3-month lag using the LASSO-derived models. Based on the mean average percentage error, the LASSO approach provided more accurate forecasts than the other methods we assessed. We demonstrate its utility in Singapore’s dengue control program by providing a forecast of the 2013 outbreak for advance preparation of outbreak response. Conclusions: Statistical models built using machine learning methods such as LASSO have the potential to markedly improve forecasting techniques for recurrent infectious disease outbreaks such as dengue. Citation: Shi Y, Liu X, Kok SY, Rajarethinam J, Liang S, Yap G, Chong CS, Lee KS, Tan SS, Chin CK, Lo A, Kong W, Ng LC, Cook AR. 2016. Three-month real-time dengue forecast models: an early warning system for outbreak alerts and policy decision support in Singapore. Environ Health Perspect 124:1369–1375; http://dx.doi.org/10.1289/ehp.1509981 PMID:26662617

Error models for official mortality forecasts.

PubMed

Alho, J M; Spencer, B D

1990-09-01

"The Office of the Actuary, U.S. Social Security Administration, produces alternative forecasts of mortality to reflect uncertainty about the future.... In this article we identify the components and assumptions of the official forecasts and approximate them by stochastic parametric models. We estimate parameters of the models from past data, derive statistical intervals for the forecasts, and compare them with the official high-low intervals. We use the models to evaluate the forecasts rather than to develop different predictions of the future. Analysis of data from 1972 to 1985 shows that the official intervals for mortality forecasts for males or females aged 45-70 have approximately a 95% chance of including the true mortality rate in any year. For other ages the chances are much less than 95%." excerpt

Forecasting runout of rock and debris avalanches

USGS Publications Warehouse

Iverson, Richard M.; Evans, S.G.; Mugnozza, G.S.; Strom, A.; Hermanns, R.L.

2006-01-01

Physically based mathematical models and statistically based empirical equations each may provide useful means of forecasting runout of rock and debris avalanches. This paper compares the foundations, strengths, and limitations of a physically based model and a statistically based forecasting method, both of which were developed to predict runout across three-dimensional topography. The chief advantage of the physically based model results from its ties to physical conservation laws and well-tested axioms of soil and rock mechanics, such as the Coulomb friction rule and effective-stress principle. The output of this model provides detailed information about the dynamics of avalanche runout, at the expense of high demands for accurate input data, numerical computation, and experimental testing. In comparison, the statistical method requires relatively modest computation and no input data except identification of prospective avalanche source areas and a range of postulated avalanche volumes. Like the physically based model, the statistical method yields maps of predicted runout, but it provides no information on runout dynamics. Although the two methods differ significantly in their structure and objectives, insights gained from one method can aid refinement of the other.

Evaluation of the 29-km Eta Model. Part 1; Objective Verification at Three Selected Stations

NASA Technical Reports Server (NTRS)

Nutter, Paul A.; Manobianco, John; Merceret, Francis J. (Technical Monitor)

1998-01-01

This paper describes an objective verification of the National Centers for Environmental Prediction (NCEP) 29-km eta model from May 1996 through January 1998. The evaluation was designed to assess the model's surface and upper-air point forecast accuracy at three selected locations during separate warm (May - August) and cool (October - January) season periods. In order to enhance sample sizes available for statistical calculations, the objective verification includes two consecutive warm and cool season periods. Systematic model deficiencies comprise the larger portion of the total error in most of the surface forecast variables that were evaluated. The error characteristics for both surface and upper-air forecasts vary widely by parameter, season, and station location. At upper levels, a few characteristic biases are identified. Overall however, the upper-level errors are more nonsystematic in nature and could be explained partly by observational measurement uncertainty. With a few exceptions, the upper-air results also indicate that 24-h model error growth is not statistically significant. In February and August 1997, NCEP implemented upgrades to the eta model's physical parameterizations that were designed to change some of the model's error characteristics near the surface. The results shown in this paper indicate that these upgrades led to identifiable and statistically significant changes in forecast accuracy for selected surface parameters. While some of the changes were expected, others were not consistent with the intent of the model updates and further emphasize the need for ongoing sensitivity studies and localized statistical verification efforts. Objective verification of point forecasts is a stringent measure of model performance, but when used alone, is not enough to quantify the overall value that model guidance may add to the forecast process. Therefore, results from a subjective verification of the meso-eta model over the Florida peninsula are discussed in the companion paper by Manobianco and Nutter. Overall verification results presented here and in part two should establish a reasonable benchmark from which model users and developers may pursue the ongoing eta model verification strategies in the future.

Assimilating the Future for Better Forecasts and Earlier Warnings

NASA Astrophysics Data System (ADS)

Du, H.; Wheatcroft, E.; Smith, L. A.

2016-12-01

Multi-model ensembles have become popular tools to account for some of the uncertainty due to model inadequacy in weather and climate simulation-based predictions. The current multi-model forecasts focus on combining single model ensemble forecasts by means of statistical post-processing. Assuming each model is developed independently or with different primary target variables, each is likely to contain different dynamical strengths and weaknesses. Using statistical post-processing, such information is only carried by the simulations under a single model ensemble: no advantage is taken to influence simulations under the other models. A novel methodology, named Multi-model Cross Pollination in Time, is proposed for multi-model ensemble scheme with the aim of integrating the dynamical information regarding the future from each individual model operationally. The proposed approach generates model states in time via applying data assimilation scheme(s) to yield truly "multi-model trajectories". It is demonstrated to outperform traditional statistical post-processing in the 40-dimensional Lorenz96 flow. Data assimilation approaches are originally designed to improve state estimation from the past to the current time. The aim of this talk is to introduce a framework that uses data assimilation to improve model forecasts at future time (not to argue for any one particular data assimilation scheme). Illustration of applying data assimilation "in the future" to provide early warning of future high-impact events is also presented.

The meta-Gaussian Bayesian Processor of forecasts and associated preliminary experiments

NASA Astrophysics Data System (ADS)

Chen, Fajing; Jiao, Meiyan; Chen, Jing

2013-04-01

Public weather services are trending toward providing users with probabilistic weather forecasts, in place of traditional deterministic forecasts. Probabilistic forecasting techniques are continually being improved to optimize available forecasting information. The Bayesian Processor of Forecast (BPF), a new statistical method for probabilistic forecast, can transform a deterministic forecast into a probabilistic forecast according to the historical statistical relationship between observations and forecasts generated by that forecasting system. This technique accounts for the typical forecasting performance of a deterministic forecasting system in quantifying the forecast uncertainty. The meta-Gaussian likelihood model is suitable for a variety of stochastic dependence structures with monotone likelihood ratios. The meta-Gaussian BPF adopting this kind of likelihood model can therefore be applied across many fields, including meteorology and hydrology. The Bayes theorem with two continuous random variables and the normal-linear BPF are briefly introduced. The meta-Gaussian BPF for a continuous predictand using a single predictor is then presented and discussed. The performance of the meta-Gaussian BPF is tested in a preliminary experiment. Control forecasts of daily surface temperature at 0000 UTC at Changsha and Wuhan stations are used as the deterministic forecast data. These control forecasts are taken from ensemble predictions with a 96-h lead time generated by the National Meteorological Center of the China Meteorological Administration, the European Centre for Medium-Range Weather Forecasts, and the US National Centers for Environmental Prediction during January 2008. The results of the experiment show that the meta-Gaussian BPF can transform a deterministic control forecast of surface temperature from any one of the three ensemble predictions into a useful probabilistic forecast of surface temperature. These probabilistic forecasts quantify the uncertainty of the control forecast; accordingly, the performance of the probabilistic forecasts differs based on the source of the underlying deterministic control forecasts.

Automated Statistical Forecast Method to 36-48H ahead of Storm Wind and Dangerous Precipitation at the Mediterranean Region

NASA Astrophysics Data System (ADS)

Perekhodtseva, E. V.

2009-09-01

Development of successful method of forecast of storm winds, including squalls and tornadoes and heavy rainfalls, that often result in human and material losses, could allow one to take proper measures against destruction of buildings and to protect people. Well-in-advance successful forecast (from 12 hours to 48 hour) makes possible to reduce the losses. Prediction of the phenomena involved is a very difficult problem for synoptic till recently. The existing graphic and calculation methods still depend on subjective decision of an operator. Nowadays in Russia there is no hydrodynamic model for forecast of the maximal precipitation and wind velocity V> 25m/c, hence the main tools of objective forecast are statistical methods using the dependence of the phenomena involved on a number of atmospheric parameters (predictors). Statistical decisive rule of the alternative and probability forecast of these events was obtained in accordance with the concept of "perfect prognosis" using the data of objective analysis. For this purpose the different teaching samples of present and absent of this storm wind and rainfalls were automatically arranged that include the values of forty physically substantiated potential predictors. Then the empirical statistical method was used that involved diagonalization of the mean correlation matrix R of the predictors and extraction of diagonal blocks of strongly correlated predictors. Thus for these phenomena the most informative predictors were selected without loosing information. The statistical decisive rules for diagnosis and prognosis of the phenomena involved U(X) were calculated for choosing informative vector-predictor. We used the criterion of distance of Mahalanobis and criterion of minimum of entropy by Vapnik-Chervonenkis for the selection predictors. Successful development of hydrodynamic models for short-term forecast and improvement of 36-48h forecasts of pressure, temperature and others parameters allowed us to use the prognostic fields of those models for calculations of the discriminant functions in the nodes of the grid 150x150km and the values of probabilities P of dangerous wind and thus to get fully automated forecasts. In order to change to the alternative forecast the author proposes the empirical threshold values specified for this phenomenon and advance period 36 hours. In the accordance to the Pirsey-Obukhov criterion (T), the success of these automated statistical methods of forecast of squalls and tornadoes to 36 -48 hours ahead and heavy rainfalls in the warm season for the territory of Italy, Spain and Balkan countries is T = 1-a-b=0,54: 0,78 after author experiments. A lot of examples of very successful forecasts of summer storm wind and heavy rainfalls over the Italy and Spain territory are submitted at this report. The same decisive rules were applied to the forecast of these phenomena during cold period in this year too. This winter heavy snowfalls in Spain and in Italy and storm wind at this territory were observed very often. And our forecasts are successful.

Statistical and Machine Learning forecasting methods: Concerns and ways forward

PubMed Central

Makridakis, Spyros; Assimakopoulos, Vassilios

2018-01-01

Machine Learning (ML) methods have been proposed in the academic literature as alternatives to statistical ones for time series forecasting. Yet, scant evidence is available about their relative performance in terms of accuracy and computational requirements. The purpose of this paper is to evaluate such performance across multiple forecasting horizons using a large subset of 1045 monthly time series used in the M3 Competition. After comparing the post-sample accuracy of popular ML methods with that of eight traditional statistical ones, we found that the former are dominated across both accuracy measures used and for all forecasting horizons examined. Moreover, we observed that their computational requirements are considerably greater than those of statistical methods. The paper discusses the results, explains why the accuracy of ML models is below that of statistical ones and proposes some possible ways forward. The empirical results found in our research stress the need for objective and unbiased ways to test the performance of forecasting methods that can be achieved through sizable and open competitions allowing meaningful comparisons and definite conclusions. PMID:29584784

Short-term Wind Forecasting at Wind Farms using WRF-LES and Actuator Disk Model

NASA Astrophysics Data System (ADS)

Kirkil, Gokhan

2017-04-01

Short-term wind forecasts are obtained for a wind farm on a mountainous terrain using WRF-LES. Multi-scale simulations are also performed using different PBL parameterizations. Turbines are parameterized using Actuator Disc Model. LES models improved the forecasts. Statistical error analysis is performed and ramp events are analyzed. Complex topography of the study area affects model performance, especially the accuracy of wind forecasts were poor for cross valley-mountain flows. By means of LES, we gain new knowledge about the sources of spatial and temporal variability of wind fluctuations such as the configuration of wind turbines.

Modeling and forecasting the distribution of Vibrio vulnificus in Chesapeake Bay

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

Jacobs, John M.; Rhodes, M.; Brown, C. W.

The aim is to construct statistical models to predict the presence, abundance and potential virulence of Vibrio vulnificus in surface waters. A variety of statistical techniques were used in concert to identify water quality parameters associated with V. vulnificus presence, abundance and virulence markers in the interest of developing strong predictive models for use in regional oceanographic modeling systems. A suite of models are provided to represent the best model fit and alternatives using environmental variables that allow them to be put to immediate use in current ecological forecasting efforts. Conclusions: Environmental parameters such as temperature, salinity and turbidity aremore » capable of accurately predicting abundance and distribution of V. vulnificus in Chesapeake Bay. Forcing these empirical models with output from ocean modeling systems allows for spatially explicit forecasts for up to 48 h in the future. This study uses one of the largest data sets compiled to model Vibrio in an estuary, enhances our understanding of environmental correlates with abundance, distribution and presence of potentially virulent strains and offers a method to forecast these pathogens that may be replicated in other regions.« less

Verification of Ensemble Forecasts for the New York City Operations Support Tool

NASA Astrophysics Data System (ADS)

Day, G.; Schaake, J. C.; Thiemann, M.; Draijer, S.; Wang, L.

2012-12-01

The New York City water supply system operated by the Department of Environmental Protection (DEP) serves nine million people. It covers 2,000 square miles of portions of the Catskill, Delaware, and Croton watersheds, and it includes nineteen reservoirs and three controlled lakes. DEP is developing an Operations Support Tool (OST) to support its water supply operations and planning activities. OST includes historical and real-time data, a model of the water supply system complete with operating rules, and lake water quality models developed to evaluate alternatives for managing turbidity in the New York City Catskill reservoirs. OST will enable DEP to manage turbidity in its unfiltered system while satisfying its primary objective of meeting the City's water supply needs, in addition to considering secondary objectives of maintaining ecological flows, supporting fishery and recreation releases, and mitigating downstream flood peaks. The current version of OST relies on statistical forecasts of flows in the system based on recent observed flows. To improve short-term decision making, plans are being made to transition to National Weather Service (NWS) ensemble forecasts based on hydrologic models that account for short-term weather forecast skill, longer-term climate information, as well as the hydrologic state of the watersheds and recent observed flows. To ensure that the ensemble forecasts are unbiased and that the ensemble spread reflects the actual uncertainty of the forecasts, a statistical model has been developed to post-process the NWS ensemble forecasts to account for hydrologic model error as well as any inherent bias and uncertainty in initial model states, meteorological data and forecasts. The post-processor is designed to produce adjusted ensemble forecasts that are consistent with the DEP historical flow sequences that were used to develop the system operating rules. A set of historical hindcasts that is representative of the real-time ensemble forecasts is needed to verify that the post-processed forecasts are unbiased, statistically reliable, and preserve the skill inherent in the "raw" NWS ensemble forecasts. A verification procedure and set of metrics will be presented that provide an objective assessment of ensemble forecasts. The procedure will be applied to both raw ensemble hindcasts and to post-processed ensemble hindcasts. The verification metrics will be used to validate proper functioning of the post-processor and to provide a benchmark for comparison of different types of forecasts. For example, current NWS ensemble forecasts are based on climatology, using each historical year to generate a forecast trace. The NWS Hydrologic Ensemble Forecast System (HEFS) under development will utilize output from both the National Oceanic Atmospheric Administration (NOAA) Global Ensemble Forecast System (GEFS) and the Climate Forecast System (CFS). Incorporating short-term meteorological forecasts and longer-term climate forecast information should provide sharper, more accurate forecasts. Hindcasts from HEFS will enable New York City to generate verification results to validate the new forecasts and further fine-tune system operating rules. Project verification results will be presented for different watersheds across a range of seasons, lead times, and flow levels to assess the quality of the current ensemble forecasts.

Eruption Forecasting in Alaska: A Retrospective and Test of the Distal VT Model

NASA Astrophysics Data System (ADS)

Prejean, S. G.; Pesicek, J. D.; Wellik, J.; Cameron, C.; White, R. A.; McCausland, W. A.; Buurman, H.

2015-12-01

United States volcano observatories have successfully forecast most significant US eruptions in the past decade. However, eruptions of some volcanoes remain stubbornly difficult to forecast effectively using seismic data alone. The Alaska Volcano Observatory (AVO) has responded to 28 eruptions from 10 volcanoes since 2005. Eruptions that were not forecast include those of frequently active volcanoes with basaltic-andesite magmas, like Pavlof, Veniaminof, and Okmok volcanoes. In this study we quantify the success rate of eruption forecasting in Alaska and explore common characteristics of eruptions not forecast. In an effort to improve future forecasts, we re-examine seismic data from eruptions and known intrusive episodes in Alaska to test the effectiveness of the distal VT model commonly employed by the USGS-USAID Volcano Disaster Assistance Program (VDAP). In the distal VT model, anomalous brittle failure or volcano-tectonic (VT) earthquake swarms in the shallow crust surrounding the volcano occur as a secondary response to crustal strain induced by magma intrusion. Because the Aleutian volcanic arc is among the most seismically active regions on Earth, distinguishing distal VT earthquake swarms for eruption forecasting purposes from tectonic seismicity unrelated to volcanic processes poses a distinct challenge. In this study, we use a modified beta-statistic to identify pre-eruptive distal VT swarms and establish their statistical significance with respect to long-term background seismicity. This analysis allows us to explore the general applicability of the distal VT model and quantify the likelihood of encountering false positives in eruption forecasting using this model alone.

Forecasting approaches to the Mekong River

NASA Astrophysics Data System (ADS)

Plate, E. J.

2009-04-01

Hydrologists distinguish between flood forecasts, which are concerned with events of the immediate future, and flood predictions, which are concerned with events that are possible, but whose date of occurrence is not determined. Although in principle both involve the determination of runoff from rainfall, the analytical approaches differ because of different objectives. The differences between the two approaches will be discussed, starting with an analysis of the forecasting process. The Mekong River in south-east Asia is used as an example. Prediction is defined as forecast for a hypothetical event, such as the 100-year flood, which is usually sufficiently specified by its magnitude and its probability of occurrence. It forms the basis for designing flood protection structures and risk management activities. The method for determining these quantities is hydrological modeling combined with extreme value statistics, today usually applied both to rainfall events and to observed river discharges. A rainfall-runoff model converts extreme rainfall events into extreme discharges, which at certain gage points along a river are calibrated against observed discharges. The quality of the model output is assessed against the mean value by means of the Nash-Sutcliffe quality criterion. The result of this procedure is a design hydrograph (or a family of design hydrographs) which are used as inputs into a hydraulic model, which converts the hydrograph into design water levels according to the hydraulic situation of the location. The accuracy of making a prediction in this sense is not particularly high: hydrologists know that the 100-year flood is a statistical quantity which can be estimated only within comparatively wide error bounds, and the hydraulics of a river site, in particular under conditions of heavy sediment loads has many uncertainties. Safety margins, such as additional freeboards are arranged to compensate for the uncertainty of the prediction. Forecasts, on the other hand, have as objective to obtain an accurate hydrograph of the near future. The method by means of which this is done is not as important as the accuracy of the forecast. A mathematical rainfall-runoff model is not necessarily a good forecast model. It has to be very carefully designed, and in many cases statistical models are found to give better results than mathematical models. Forecasters have the advantage of knowing the course of the hydrographs up to the point in time where forecasts have to be made. Therefore, models can be calibrated on line against the hydrograph of the immediate past. To assess the quality of a forecast, the quality criterion should not be based on the mean value, as does the Nash-Sutcliffe criterion, but should be based on the best forecast given the information up to the forecast time. Without any additional information, the best forecast when only the present day value is known is to assume a no-change scenario, i.e. to assume that the present value does not change in the immediate future. For the Mekong there exists a forecasting system which is based on a rainfall-runoff model operated by the Mekong River Commission. This model is found not to be adequate for forecasting for periods longer than one or two days ahead. Improvements are sought through two approaches: a strictly deterministic rainfall-runoff model, and a strictly statistical model based on regression with upstream stations. The two approaches are com-pared, and suggestions are made how to best combine the advantages of both approaches. This requires that due consideration is given to critical hydraulic conditions of the river at and in between the gauging stations. Critical situations occur in two ways: when the river overtops, in which case the rainfall-runoff model is incomplete unless overflow losses are considered, and at the confluence with tributaries. Of particular importance is the role of the large Tonle Sap Lake, which dampens the hydrograph downstream of Phnom Penh. The effect of these components of river hydraulics on forecasting accuracy will be assessed.

Using High Resolution Model Data to Improve Lightning Forecasts across Southern California

NASA Astrophysics Data System (ADS)

Capps, S. B.; Rolinski, T.

2014-12-01

Dry lightning often results in a significant amount of fire starts in areas where the vegetation is dry and continuous. Meteorologists from the USDA Forest Service Predictive Services' program in Riverside, California are tasked to provide southern and central California's fire agencies with fire potential outlooks. Logistic regression equations were developed by these meteorologists several years ago, which forecast probabilities of lightning as well as lightning amounts, out to seven days across southern California. These regression equations were developed using ten years of historical gridded data from the Global Forecast System (GFS) model on a coarse scale (0.5 degree resolution), correlated with historical lightning strike data. These equations do a reasonably good job of capturing a lightning episode (3-5 consecutive days or greater of lightning), but perform poorly regarding more detailed information such as exact location and amounts. It is postulated that the inadequacies in resolving the finer details of episodic lightning events is due to the coarse resolution of the GFS data, along with limited predictors. Stability parameters, such as the Lifted Index (LI), the Total Totals index (TT), Convective Available Potential Energy (CAPE), along with Precipitable Water (PW) are the only parameters being considered as predictors. It is hypothesized that the statistical forecasts will benefit from higher resolution data both in training and implementing the statistical model. We have dynamically downscaled NCEP FNL (Final) reanalysis data using the Weather Research and Forecasting model (WRF) to 3km spatial and hourly temporal resolution across a decade. This dataset will be used to evaluate the contribution to the success of the statistical model of additional predictors in higher vertical, spatial and temporal resolution. If successful, we will implement an operational dynamically downscaled GFS forecast product to generate predictors for the resulting statistical lightning model. This data will help fire agencies be better prepared to pre-deploy resources in advance of these events. Specific information regarding duration, amount, and location will be especially valuable.

Development of a Statistical Model for Forecasting Episodes of Visibility Degradation in the Denver Metropolitan Area.

NASA Astrophysics Data System (ADS)

Reddy, P. J.; Barbarick, D. E.; Osterburg, R. D.

1995-03-01

In 1990, the State of Colorado implemented a visibility standard of 0.076 km1 of beta extinction for the Denver metropolitan area. Meteorologists with Colorado's Air Pollution Control Division forecast high pollution days associated with visibility impairment as well as those due to high levels of the federal criteria pollutants. Visibility forecasts are made from a few hours up to about 26 h in advance of the period of interest. Here we discuss the key microscale, mesoscale, and synoptic-scale features associated with episodes of visibility impairment. Data from special studies, case studies, and the 22 NOAA Program for Regional Observing and Forecasting Services mesonet sites have been invaluable in identifying patterns associated with extremes in visibility conditions. A preliminary statistical forecast model has been developed using variables that represent many of these patterns. Six variables were selected from an initial pool of 27 to be used in a model based on linear logistic regression. These six variables include forecast measures of snow cover, surface pressures and a surface pressure gradient in eastern Colorado, relative humidity, and 500-mb ridge position. The initial testing of the model has been encouraging. The model correctly predicted 76% of the good visibility days and 67% of the poor visibility days for a test set of 171 days.

Forecasting incidence of dengue in Rajasthan, using time series analyses.

PubMed

Bhatnagar, Sunil; Lal, Vivek; Gupta, Shiv D; Gupta, Om P

2012-01-01

To develop a prediction model for dengue fever/dengue haemorrhagic fever (DF/DHF) using time series data over the past decade in Rajasthan and to forecast monthly DF/DHF incidence for 2011. Seasonal autoregressive integrated moving average (SARIMA) model was used for statistical modeling. During January 2001 to December 2010, the reported DF/DHF cases showed a cyclical pattern with seasonal variation. SARIMA (0,0,1) (0,1,1) 12 model had the lowest normalized Bayesian information criteria (BIC) of 9.426 and mean absolute percentage error (MAPE) of 263.361 and appeared to be the best model. The proportion of variance explained by the model was 54.3%. Adequacy of the model was established through Ljung-Box test (Q statistic 4.910 and P-value 0.996), which showed no significant correlation between residuals at different lag times. The forecast for the year 2011 showed a seasonal peak in the month of October with an estimated 546 cases. Application of SARIMA model may be useful for forecast of cases and impending outbreaks of DF/DHF and other infectious diseases, which exhibit seasonal pattern.

Science and Engineering of an Operational Tsunami Forecasting System

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

Gonzalez, Frank

2009-04-06

After a review of tsunami statistics and the destruction caused by tsunamis, a means of forecasting tsunamis is discussed as part of an overall program of reducing fatalities through hazard assessment, education, training, mitigation, and a tsunami warning system. The forecast is accomplished via a concept called Deep Ocean Assessment and Reporting of Tsunamis (DART). Small changes of pressure at the sea floor are measured and relayed to warning centers. Under development is an international modeling network to transfer, maintain, and improve tsunami forecast models.

Science and Engineering of an Operational Tsunami Forecasting System

ScienceCinema

Gonzalez, Frank

2017-12-09

After a review of tsunami statistics and the destruction caused by tsunamis, a means of forecasting tsunamis is discussed as part of an overall program of reducing fatalities through hazard assessment, education, training, mitigation, and a tsunami warning system. The forecast is accomplished via a concept called Deep Ocean Assessment and Reporting of Tsunamis (DART). Small changes of pressure at the sea floor are measured and relayed to warning centers. Under development is an international modeling network to transfer, maintain, and improve tsunami forecast models.

Avoiding the ensemble decorrelation problem using member-by-member post-processing

NASA Astrophysics Data System (ADS)

Van Schaeybroeck, Bert; Vannitsem, Stéphane

2014-05-01

Forecast calibration or post-processing has become a standard tool in atmospheric and climatological science due to the presence of systematic initial condition and model errors. For ensemble forecasts the most competitive methods derive from the assumption of a fixed ensemble distribution. However, when independently applying such 'statistical' methods at different locations, lead times or for multiple variables the correlation structure for individual ensemble members is destroyed. Instead of reastablishing the correlation structure as in Schefzik et al. (2013) we instead propose a calibration method that avoids such problem by correcting each ensemble member individually. Moreover, we analyse the fundamental mechanisms by which the probabilistic ensemble skill can be enhanced. In terms of continuous ranked probability score, our member-by-member approach amounts to skill gain that extends for lead times far beyond the error doubling time and which is as good as the one of the most competitive statistical approach, non-homogeneous Gaussian regression (Gneiting et al. 2005). Besides the conservation of correlation structure, additional benefits arise including the fact that higher-order ensemble moments like kurtosis and skewness are inherited from the uncorrected forecasts. Our detailed analysis is performed in the context of the Kuramoto-Sivashinsky equation and different simple models but the results extent succesfully to the ensemble forecast of the European Centre for Medium-Range Weather Forecasts (Van Schaeybroeck and Vannitsem, 2013, 2014) . References [1] Gneiting, T., Raftery, A. E., Westveld, A., Goldman, T., 2005: Calibrated probabilistic forecasting using ensemble model output statistics and minimum CRPS estimation. Mon. Weather Rev. 133, 1098-1118. [2] Schefzik, R., T.L. Thorarinsdottir, and T. Gneiting, 2013: Uncertainty Quantification in Complex Simulation Models Using Ensemble Copula Coupling. To appear in Statistical Science 28. [3] Van Schaeybroeck, B., and S. Vannitsem, 2013: Reliable probabilities through statistical post-processing of ensemble forecasts. Proceedings of the European Conference on Complex Systems 2012, Springer proceedings on complexity, XVI, p. 347-352. [4] Van Schaeybroeck, B., and S. Vannitsem, 2014: Ensemble post-processing using member-by-member approaches: theoretical aspects, under review.

National Centers for Environmental Prediction

Science.gov Websites

Statistics Observational Data Processing Data Assimilation Monsoon Desk Model Transition Seminars Seminar WEATHER RESEARCH and FORECASTING HMON HMON - OPERATIONAL HURRICANE FORECASTING WAVEWATCH III WAVEWATCH III Modeling Center NOAA Center for Weather and Climate Prediction (NCWCP) 5830 University Research Court

Forecast and virtual weather driven plant disease risk modeling system

USDA-ARS?s Scientific Manuscript database

We describe a system in use and development that leverages public weather station data, several spatialized weather forecast types, leaf wetness estimation, generic plant disease models, and online statistical evaluation. Convergent technological developments in all these areas allow, with funding f...

Satellite Sounder Data Assimilation for Improving Alaska Region Weather Forecast

NASA Technical Reports Server (NTRS)

Zhu, Jiang; Stevens, E.; Zhang, X.; Zavodsky, B. T.; Heinrichs, T.; Broderson, D.

2014-01-01

A case study and monthly statistical analysis using sounder data assimilation to improve the Alaska regional weather forecast model are presented. Weather forecast in Alaska faces challenges as well as opportunities. Alaska has a large land with multiple types of topography and coastal area. Weather forecast models must be finely tuned in order to accurately predict weather in Alaska. Being in the high-latitudes provides Alaska greater coverage of polar orbiting satellites for integration into forecasting models than the lower 48. Forecasting marine low stratus clouds is critical to the Alaska aviation and oil industry and is the current focus of the case study. NASA AIRS/CrIS sounder profiles data are used to do data assimilation for the Alaska regional weather forecast model to improve Arctic marine stratus clouds forecast. Choosing physical options for the WRF model is discussed. Preprocess of AIRS/CrIS sounder data for data assimilation is described. Local observation data, satellite data, and global data assimilation data are used to verify and/or evaluate the forecast results by the MET tools Model Evaluation Tools (MET).

Relative effects of statistical preprocessing and postprocessing on a regional hydrological ensemble prediction system

NASA Astrophysics Data System (ADS)

Sharma, Sanjib; Siddique, Ridwan; Reed, Seann; Ahnert, Peter; Mendoza, Pablo; Mejia, Alfonso

2018-03-01

The relative roles of statistical weather preprocessing and streamflow postprocessing in hydrological ensemble forecasting at short- to medium-range forecast lead times (day 1-7) are investigated. For this purpose, a regional hydrologic ensemble prediction system (RHEPS) is developed and implemented. The RHEPS is comprised of the following components: (i) hydrometeorological observations (multisensor precipitation estimates, gridded surface temperature, and gauged streamflow); (ii) weather ensemble forecasts (precipitation and near-surface temperature) from the National Centers for Environmental Prediction 11-member Global Ensemble Forecast System Reforecast version 2 (GEFSRv2); (iii) NOAA's Hydrology Laboratory-Research Distributed Hydrologic Model (HL-RDHM); (iv) heteroscedastic censored logistic regression (HCLR) as the statistical preprocessor; (v) two statistical postprocessors, an autoregressive model with a single exogenous variable (ARX(1,1)) and quantile regression (QR); and (vi) a comprehensive verification strategy. To implement the RHEPS, 1 to 7 days weather forecasts from the GEFSRv2 are used to force HL-RDHM and generate raw ensemble streamflow forecasts. Forecasting experiments are conducted in four nested basins in the US Middle Atlantic region, ranging in size from 381 to 12 362 km2. Results show that the HCLR preprocessed ensemble precipitation forecasts have greater skill than the raw forecasts. These improvements are more noticeable in the warm season at the longer lead times (> 3 days). Both postprocessors, ARX(1,1) and QR, show gains in skill relative to the raw ensemble streamflow forecasts, particularly in the cool season, but QR outperforms ARX(1,1). The scenarios that implement preprocessing and postprocessing separately tend to perform similarly, although the postprocessing-alone scenario is often more effective. The scenario involving both preprocessing and postprocessing consistently outperforms the other scenarios. In some cases, however, the differences between this scenario and the scenario with postprocessing alone are not as significant. We conclude that implementing both preprocessing and postprocessing ensures the most skill improvements, but postprocessing alone can often be a competitive alternative.

Value of Forecaster in the Loop

DTIC Science & Technology

2014-09-01

forecast system IFR instrument flight rules IMC instrument meteorological conditions LAMP Localized Aviation Model Output Statistics Program METOC...obtaining valuable experience. Additional factors have impacted the Navy weather forecast process. There has been a the realignment of the meteorology...forecasts that are assessed, it may be a relatively small number that have direct impact on the decision-making process. Whether the value is minimal or

Synoptic scale forecast skill and systematic errors in the MASS 2.0 model. [Mesoscale Atmospheric Simulation System

NASA Technical Reports Server (NTRS)

Koch, S. E.; Skillman, W. C.; Kocin, P. J.; Wetzel, P. J.; Brill, K. F.

1985-01-01

The synoptic scale performance characteristics of MASS 2.0 are determined by comparing filtered 12-24 hr model forecasts to same-case forecasts made by the National Meteorological Center's synoptic-scale Limited-area Fine Mesh model. Characteristics of the two systems are contrasted, and the analysis methodology used to determine statistical skill scores and systematic errors is described. The overall relative performance of the two models in the sample is documented, and important systematic errors uncovered are presented.

A hybrid spatiotemporal drought forecasting model for operational use

NASA Astrophysics Data System (ADS)

Vasiliades, L.; Loukas, A.

2010-09-01

Drought forecasting plays an important role in the planning and management of natural resources and water resource systems in a river basin. Early and timelines forecasting of a drought event can help to take proactive measures and set out drought mitigation strategies to alleviate the impacts of drought. Spatiotemporal data mining is the extraction of unknown and implicit knowledge, structures, spatiotemporal relationships, or patterns not explicitly stored in spatiotemporal databases. As one of data mining techniques, forecasting is widely used to predict the unknown future based upon the patterns hidden in the current and past data. This study develops a hybrid spatiotemporal scheme for integrated spatial and temporal forecasting. Temporal forecasting is achieved using feed-forward neural networks and the temporal forecasts are extended to the spatial dimension using a spatial recurrent neural network model. The methodology is demonstrated for an operational meteorological drought index the Standardized Precipitation Index (SPI) calculated at multiple timescales. 48 precipitation stations and 18 independent precipitation stations, located at Pinios river basin in Thessaly region, Greece, were used for the development and spatiotemporal validation of the hybrid spatiotemporal scheme. Several quantitative temporal and spatial statistical indices were considered for the performance evaluation of the models. Furthermore, qualitative statistical criteria based on contingency tables between observed and forecasted drought episodes were calculated. The results show that the lead time of forecasting for operational use depends on the SPI timescale. The hybrid spatiotemporal drought forecasting model could be operationally used for forecasting up to three months ahead for SPI short timescales (e.g. 3-6 months) up to six months ahead for large SPI timescales (e.g. 24 months). The above findings could be useful in developing a drought preparedness plan in the region.

Improvement of PM concentration predictability using WRF-CMAQ-DLM coupled system and its applications

NASA Astrophysics Data System (ADS)

Lee, Soon Hwan; Kim, Ji Sun; Lee, Kang Yeol; Shon, Keon Tae

2017-04-01

Air quality due to increasing Particulate Matter(PM) in Korea in Asia is getting worse. At present, the PM forecast is announced based on the PM concentration predicted from the air quality prediction numerical model. However, forecast accuracy is not as high as expected due to various uncertainties for PM physical and chemical characteristics. The purpose of this study was to develop a numerical-statistically ensemble models to improve the accuracy of prediction of PM10 concentration. Numerical models used in this study are the three dimensional atmospheric model Weather Research and Forecasting(WRF) and the community multiscale air quality model (CMAQ). The target areas for the PM forecast are Seoul, Busan, Daegu, and Daejeon metropolitan areas in Korea. The data used in the model development are PM concentration and CMAQ predictions and the data period is 3 months (March 1 - May 31, 2014). The dynamic-statistical technics for reducing the systematic error of the CMAQ predictions was applied to the dynamic linear model(DLM) based on the Baysian Kalman filter technic. As a result of applying the metrics generated from the dynamic linear model to the forecasting of PM concentrations accuracy was improved. Especially, at the high PM concentration where the damage is relatively large, excellent improvement results are shown.

Evaluation of probabilistic forecasts with the scoringRules package

NASA Astrophysics Data System (ADS)

Jordan, Alexander; Krüger, Fabian; Lerch, Sebastian

2017-04-01

Over the last decades probabilistic forecasts in the form of predictive distributions have become popular in many scientific disciplines. With the proliferation of probabilistic models arises the need for decision-theoretically principled tools to evaluate the appropriateness of models and forecasts in a generalized way in order to better understand sources of prediction errors and to improve the models. Proper scoring rules are functions S(F,y) which evaluate the accuracy of a forecast distribution F , given that an outcome y was observed. In coherence with decision-theoretical principles they allow to compare alternative models, a crucial ability given the variety of theories, data sources and statistical specifications that is available in many situations. This contribution presents the software package scoringRules for the statistical programming language R, which provides functions to compute popular scoring rules such as the continuous ranked probability score for a variety of distributions F that come up in applied work. For univariate variables, two main classes are parametric distributions like normal, t, or gamma distributions, and distributions that are not known analytically, but are indirectly described through a sample of simulation draws. For example, ensemble weather forecasts take this form. The scoringRules package aims to be a convenient dictionary-like reference for computing scoring rules. We offer state of the art implementations of several known (but not routinely applied) formulas, and implement closed-form expressions that were previously unavailable. Whenever more than one implementation variant exists, we offer statistically principled default choices. Recent developments include the addition of scoring rules to evaluate multivariate forecast distributions. The use of the scoringRules package is illustrated in an example on post-processing ensemble forecasts of temperature.

Using Temperature Forecasts to Improve Seasonal Streamflow Forecasts in the Colorado and Rio Grande Basins

NASA Astrophysics Data System (ADS)

Lehner, F.; Wood, A.; Llewellyn, D.; Blatchford, D. B.; Goodbody, A. G.; Pappenberger, F.

2017-12-01

Recent studies have documented the influence of increasing temperature on streamflow across the American West, including snow-melt driven rivers such as the Colorado or Rio Grande. At the same time, some basins are reporting decreasing skill in seasonal streamflow forecasts, termed water supply forecasts (WSFs), over the recent decade. While the skill in seasonal precipitation forecasts from dynamical models remains low, their skill in predicting seasonal temperature variations could potentially be harvested for WSFs to account for non-stationarity in regional temperatures. Here, we investigate whether WSF skill can be improved by incorporating seasonal temperature forecasts from dynamical forecasting models (from the North American Multi Model Ensemble and the European Centre for Medium-Range Weather Forecast System 4) into traditional statistical forecast models. We find improved streamflow forecast skill relative to traditional WSF approaches in a majority of headwater locations in the Colorado and Rio Grande basins. Incorporation of temperature into WSFs thus provides a promising avenue to increase the robustness of current forecasting techniques in the face of continued regional warming.

Exploring the calibration of a wind forecast ensemble for energy applications

NASA Astrophysics Data System (ADS)

Heppelmann, Tobias; Ben Bouallegue, Zied; Theis, Susanne

2015-04-01

In the German research project EWeLiNE, Deutscher Wetterdienst (DWD) and Fraunhofer Institute for Wind Energy and Energy System Technology (IWES) are collaborating with three German Transmission System Operators (TSO) in order to provide the TSOs with improved probabilistic power forecasts. Probabilistic power forecasts are derived from probabilistic weather forecasts, themselves derived from ensemble prediction systems (EPS). Since the considered raw ensemble wind forecasts suffer from underdispersiveness and bias, calibration methods are developed for the correction of the model bias and the ensemble spread bias. The overall aim is to improve the ensemble forecasts such that the uncertainty of the possible weather deployment is depicted by the ensemble spread from the first forecast hours. Additionally, the ensemble members after calibration should remain physically consistent scenarios. We focus on probabilistic hourly wind forecasts with horizon of 21 h delivered by the convection permitting high-resolution ensemble system COSMO-DE-EPS which has become operational in 2012 at DWD. The ensemble consists of 20 ensemble members driven by four different global models. The model area includes whole Germany and parts of Central Europe with a horizontal resolution of 2.8 km and a vertical resolution of 50 model levels. For verification we use wind mast measurements around 100 m height that corresponds to the hub height of wind energy plants that belong to wind farms within the model area. Calibration of the ensemble forecasts can be performed by different statistical methods applied to the raw ensemble output. Here, we explore local bivariate Ensemble Model Output Statistics at individual sites and quantile regression with different predictors. Applying different methods, we already show an improvement of ensemble wind forecasts from COSMO-DE-EPS for energy applications. In addition, an ensemble copula coupling approach transfers the time-dependencies of the raw ensemble to the calibrated ensemble. The calibrated wind forecasts are evaluated first with univariate probabilistic scores and additionally with diagnostics of wind ramps in order to assess the time-consistency of the calibrated ensemble members.

Stochastic demographic forecasting.

PubMed

Lee, R D

1992-11-01

"This paper describes a particular approach to stochastic population forecasting, which is implemented for the U.S.A. through 2065. Statistical time series methods are combined with demographic models to produce plausible long run forecasts of vital rates, with probability distributions. The resulting mortality forecasts imply gains in future life expectancy that are roughly twice as large as those forecast by the Office of the Social Security Actuary.... Resulting stochastic forecasts of the elderly population, elderly dependency ratios, and payroll tax rates for health, education and pensions are presented." excerpt

Analysis/forecast experiments with a multivariate statistical analysis scheme using FGGE data

NASA Technical Reports Server (NTRS)

Baker, W. E.; Bloom, S. C.; Nestler, M. S.

1985-01-01

A three-dimensional, multivariate, statistical analysis method, optimal interpolation (OI) is described for modeling meteorological data from widely dispersed sites. The model was developed to analyze FGGE data at the NASA-Goddard Laboratory of Atmospherics. The model features a multivariate surface analysis over the oceans, including maintenance of the Ekman balance and a geographically dependent correlation function. Preliminary comparisons are made between the OI model and similar schemes employed at the European Center for Medium Range Weather Forecasts and the National Meteorological Center. The OI scheme is used to provide input to a GCM, and model error correlations are calculated for forecasts of 500 mb vertical water mixing ratios and the wind profiles. Comparisons are made between the predictions and measured data. The model is shown to be as accurate as a successive corrections model out to 4.5 days.

Seasonal Drought Prediction: Advances, Challenges, and Future Prospects

NASA Astrophysics Data System (ADS)

Hao, Zengchao; Singh, Vijay P.; Xia, Youlong

2018-03-01

Drought prediction is of critical importance to early warning for drought managements. This review provides a synthesis of drought prediction based on statistical, dynamical, and hybrid methods. Statistical drought prediction is achieved by modeling the relationship between drought indices of interest and a suite of potential predictors, including large-scale climate indices, local climate variables, and land initial conditions. Dynamical meteorological drought prediction relies on seasonal climate forecast from general circulation models (GCMs), which can be employed to drive hydrological models for agricultural and hydrological drought prediction with the predictability determined by both climate forcings and initial conditions. Challenges still exist in drought prediction at long lead time and under a changing environment resulting from natural and anthropogenic factors. Future research prospects to improve drought prediction include, but are not limited to, high-quality data assimilation, improved model development with key processes related to drought occurrence, optimal ensemble forecast to select or weight ensembles, and hybrid drought prediction to merge statistical and dynamical forecasts.

Improving medium-range ensemble streamflow forecasts through statistical post-processing

NASA Astrophysics Data System (ADS)

Mendoza, Pablo; Wood, Andy; Clark, Elizabeth; Nijssen, Bart; Clark, Martyn; Ramos, Maria-Helena; Nowak, Kenneth; Arnold, Jeffrey

2017-04-01

Probabilistic hydrologic forecasts are a powerful source of information for decision-making in water resources operations. A common approach is the hydrologic model-based generation of streamflow forecast ensembles, which can be implemented to account for different sources of uncertainties - e.g., from initial hydrologic conditions (IHCs), weather forecasts, and hydrologic model structure and parameters. In practice, hydrologic ensemble forecasts typically have biases and spread errors stemming from errors in the aforementioned elements, resulting in a degradation of probabilistic properties. In this work, we compare several statistical post-processing techniques applied to medium-range ensemble streamflow forecasts obtained with the System for Hydromet Applications, Research and Prediction (SHARP). SHARP is a fully automated prediction system for the assessment and demonstration of short-term to seasonal streamflow forecasting applications, developed by the National Center for Atmospheric Research, University of Washington, U.S. Army Corps of Engineers, and U.S. Bureau of Reclamation. The suite of post-processing techniques includes linear blending, quantile mapping, extended logistic regression, quantile regression, ensemble analogs, and the generalized linear model post-processor (GLMPP). We assess and compare these techniques using multi-year hindcasts in several river basins in the western US. This presentation discusses preliminary findings about the effectiveness of the techniques for improving probabilistic skill, reliability, discrimination, sharpness and resolution.

A Study on Predictive Analytics Application to Ship Machinery Maintenance

DTIC Science & Technology

2013-09-01

Looking at the nature of the time series forecasting method , it would be better applied to offline analysis . The application for real- time online...other system attributes in future. Two techniques of statistical analysis , mainly time series models and cumulative sum control charts, are discussed in...statistical tool employed for the two techniques of statistical analysis . Both time series forecasting as well as CUSUM control charts are shown to be

Use of High-resolution WRF Simulations to Forecast Lightning Threat

NASA Technical Reports Server (NTRS)

McCaul, William E.; LaCasse, K.; Goodman, S. J.

2006-01-01

Recent observational studies have confirmed the existence of a robust statistical relationship between lightning flash rates and the amount of large precipitating ice hydrometeors in storms. This relationship is exploited, in conjunction with the capabilities of recent forecast models such as WRF, to forecast the threat of lightning from convective storms using the output fields from the model forecasts. The simulated vertical flux of graupel at -15C is used in this study as a proxy for charge separation processes and their associated lightning risk. Six-h simulations are conducted for a number of case studies for which three-dimensional lightning validation data from the North Alabama Lightning Mapping Array are available. Experiments indicate that initialization of the WRF model on a 2 km grid using Eta boundary conditions, Doppler radar radial velocity and reflectivity fields, and METAR and ACARS data yield the most realistic simulations. An array of subjective and objective statistical metrics are employed to document the utility of the WRF forecasts. The simulation results are also compared to other more traditional means of forecasting convective storms, such as those based on inspection of the convective available potential energy field.

Spatiotemporal drought forecasting using nonlinear models

NASA Astrophysics Data System (ADS)

Vasiliades, Lampros; Loukas, Athanasios

2010-05-01

Spatiotemporal data mining is the extraction of unknown and implicit knowledge, structures, spatiotemporal relationships, or patterns not explicitly stored in spatiotemporal databases. As one of data mining techniques, forecasting is widely used to predict the unknown future based upon the patterns hidden in the current and past data. In order to achieve spatiotemporal forecasting, some mature analysis tools, e.g., time series and spatial statistics are extended to the spatial dimension and the temporal dimension, respectively. Drought forecasting plays an important role in the planning and management of natural resources and water resource systems in a river basin. Early and timelines forecasting of a drought event can help to take proactive measures and set out drought mitigation strategies to alleviate the impacts of drought. Despite the widespread application of nonlinear mathematical models, comparative studies on spatiotemporal drought forecasting using different models are still a huge task for modellers. This study uses a promising approach, the Gamma Test (GT), to select the input variables and the training data length, so that the trial and error workload could be greatly reduced. The GT enables to quickly evaluate and estimate the best mean squared error that can be achieved by a smooth model on any unseen data for a given selection of inputs, prior to model construction. The GT is applied to forecast droughts using monthly Standardized Precipitation Index (SPI) timeseries at multiple timescales in several precipitation stations at Pinios river basin in Thessaly region, Greece. Several nonlinear models have been developed efficiently, with the aid of the GT, for 1-month up to 12-month ahead forecasting. Several temporal and spatial statistical indices were considered for the performance evaluation of the models. The predicted results show reasonably good agreement with the actual data for short lead times, whereas the forecasting accuracy decreases with increase in lead time. Finally, the developed nonlinear models could be used in an early warning system for risk and decision analyses at the study area.

Multiple regression and Artificial Neural Network for long-term rainfall forecasting using large scale climate modes

NASA Astrophysics Data System (ADS)

Mekanik, F.; Imteaz, M. A.; Gato-Trinidad, S.; Elmahdi, A.

2013-10-01

In this study, the application of Artificial Neural Networks (ANN) and Multiple regression analysis (MR) to forecast long-term seasonal spring rainfall in Victoria, Australia was investigated using lagged El Nino Southern Oscillation (ENSO) and Indian Ocean Dipole (IOD) as potential predictors. The use of dual (combined lagged ENSO-IOD) input sets for calibrating and validating ANN and MR Models is proposed to investigate the simultaneous effect of past values of these two major climate modes on long-term spring rainfall prediction. The MR models that did not violate the limits of statistical significance and multicollinearity were selected for future spring rainfall forecast. The ANN was developed in the form of multilayer perceptron using Levenberg-Marquardt algorithm. Both MR and ANN modelling were assessed statistically using mean square error (MSE), mean absolute error (MAE), Pearson correlation (r) and Willmott index of agreement (d). The developed MR and ANN models were tested on out-of-sample test sets; the MR models showed very poor generalisation ability for east Victoria with correlation coefficients of -0.99 to -0.90 compared to ANN with correlation coefficients of 0.42-0.93; ANN models also showed better generalisation ability for central and west Victoria with correlation coefficients of 0.68-0.85 and 0.58-0.97 respectively. The ability of multiple regression models to forecast out-of-sample sets is compatible with ANN for Daylesford in central Victoria and Kaniva in west Victoria (r = 0.92 and 0.67 respectively). The errors of the testing sets for ANN models are generally lower compared to multiple regression models. The statistical analysis suggest the potential of ANN over MR models for rainfall forecasting using large scale climate modes.

Discrete post-processing of total cloud cover ensemble forecasts

NASA Astrophysics Data System (ADS)

Hemri, Stephan; Haiden, Thomas; Pappenberger, Florian

2017-04-01

This contribution presents an approach to post-process ensemble forecasts for the discrete and bounded weather variable of total cloud cover. Two methods for discrete statistical post-processing of ensemble predictions are tested. The first approach is based on multinomial logistic regression, the second involves a proportional odds logistic regression model. Applying them to total cloud cover raw ensemble forecasts from the European Centre for Medium-Range Weather Forecasts improves forecast skill significantly. Based on station-wise post-processing of raw ensemble total cloud cover forecasts for a global set of 3330 stations over the period from 2007 to early 2014, the more parsimonious proportional odds logistic regression model proved to slightly outperform the multinomial logistic regression model. Reference Hemri, S., Haiden, T., & Pappenberger, F. (2016). Discrete post-processing of total cloud cover ensemble forecasts. Monthly Weather Review 144, 2565-2577.

Regional Earthquake Likelihood Models: A realm on shaky grounds?

NASA Astrophysics Data System (ADS)

Kossobokov, V.

2005-12-01

Seismology is juvenile and its appropriate statistical tools to-date may have a "medievil flavor" for those who hurry up to apply a fuzzy language of a highly developed probability theory. To become "quantitatively probabilistic" earthquake forecasts/predictions must be defined with a scientific accuracy. Following the most popular objectivists' viewpoint on probability, we cannot claim "probabilities" adequate without a long series of "yes/no" forecast/prediction outcomes. Without "antiquated binary language" of "yes/no" certainty we cannot judge an outcome ("success/failure"), and, therefore, quantify objectively a forecast/prediction method performance. Likelihood scoring is one of the delicate tools of Statistics, which could be worthless or even misleading when inappropriate probability models are used. This is a basic loophole for a misuse of likelihood as well as other statistical methods on practice. The flaw could be avoided by an accurate verification of generic probability models on the empirical data. It is not an easy task in the frames of the Regional Earthquake Likelihood Models (RELM) methodology, which neither defines the forecast precision nor allows a means to judge the ultimate success or failure in specific cases. Hopefully, the RELM group realizes the problem and its members do their best to close the hole with an adequate, data supported choice. Regretfully, this is not the case with the erroneous choice of Gerstenberger et al., who started the public web site with forecasts of expected ground shaking for `tomorrow' (Nature 435, 19 May 2005). Gerstenberger et al. have inverted the critical evidence of their study, i.e., the 15 years of recent seismic record accumulated just in one figure, which suggests rejecting with confidence above 97% "the generic California clustering model" used in automatic calculations. As a result, since the date of publication in Nature the United States Geological Survey website delivers to the public, emergency planners and the media, a forecast product, which is based on wrong assumptions that violate the best-documented earthquake statistics in California, which accuracy was not investigated, and which forecasts were not tested in a rigorous way.

Hydrodaynamic - Statistical Forecast Method To 36-48h Ahead Of Storm Wind And Tornadoes Over The Territory Of Europe And Siberia

NASA Astrophysics Data System (ADS)

Perekhodtseva, Elvira V.

2010-05-01

Development of successful method of forecast of storm winds, including squalls and tornadoes, that often result in human and material losses, could allow one to take proper measures against destruction of buildings and to protect people. Well-in-advance successful forecast (from 12 hours to 48 hour) makes possible to reduce the losses. Prediction of the phenomena involved is a very difficult problem for synoptic till recently. The existing graphic and calculation methods still depend on subjective decision of an operator. Nowadays in Russia there is no hydrodynamic model for forecast of the maximal wind velocity V> 25m/c, hence the main tools of objective forecast are statistical methods using the dependence of the phenomena involved on a number of atmospheric parameters (predictors). . Statistical decisive rule of the alternative and probability forecast of these events was obtained in accordance with the concept of "perfect prognosis" using the data of objective analysis. For this purpose the different teaching samples of present and absent of this storm wind and rainfalls were automatically arranged that include the values of forty physically substantiated potential predictors. Then the empirical statistical method was used that involved diagonalization of the mean correlation matrix R of the predictors and extraction of diagonal blocks of strongly correlated predictors. Thus for these phenomena the most informative predictors were selected without loosing information. The statistical decisive rules for diagnosis and prognosis of the phenomena involved U(X) were calculated for choosing informative vector-predictor. We used the criterion of distance of Mahalanobis and criterion of minimum of entropy by Vapnik-Chervonenkis for the selection predictors. Successful development of hydrodynamic models for short-term forecast and improvement of 36-48h forecasts of pressure, temperature and others parameters allowed us to use the prognostic fields of those models for calculations of the discriminant functions in the nodes of the grid 75x75km and the values of probabilities P of dangerous wind and thus to get fully automated forecasts. . In order to apply the alternative forecast to European part of Russia and Europe the author proposes the empirical threshold values specified for this phenomenon and advance period 36 hours. According to the Pirsey-Obukhov criterion (T), the success of this hydrometeorological-statistical method of forecast of storm wind and tornadoes to 36 -48 hours ahead in the warm season for the territory of Europe part of Russia and Siberia is T = 1-a-b=0,54-0,78 after independent and author experiments during the period 2004-2009 years. A lot of examples of very successful forecasts are submitted at this report for the territory of Europe and Russia. The same decisive rules were applied to the forecast of these phenomena during cold period in 2009-2010 years too. On the first month of 2010 a lot of cases of storm wind with heavy snowfall were observed and were forecasting over the territory of France, Italy and Germany.

Simple statistical bias correction techniques greatly improve moderate resolution air quality forecast at station level

NASA Astrophysics Data System (ADS)

Curci, Gabriele; Falasca, Serena

2017-04-01

Deterministic air quality forecast is routinely carried out at many local Environmental Agencies in Europe and throughout the world by means of eulerian chemistry-transport models. The skill of these models in predicting the ground-level concentrations of relevant pollutants (ozone, nitrogen dioxide, particulate matter) a few days ahead has greatly improved in recent years, but it is not yet always compliant with the required quality level for decision making (e.g. the European Commission has set a maximum uncertainty of 50% on daily values of relevant pollutants). Post-processing of deterministic model output is thus still regarded as a useful tool to make the forecast more reliable. In this work, we test several bias correction techniques applied to a long-term dataset of air quality forecasts over Europe and Italy. We used the WRF-CHIMERE modelling system, which provides operational experimental chemical weather forecast at CETEMPS (http://pumpkin.aquila.infn.it/forechem/), to simulate the years 2008-2012 at low resolution over Europe (0.5° x 0.5°) and moderate resolution over Italy (0.15° x 0.15°). We compared the simulated dataset with available observation from the European Environmental Agency database (AirBase) and characterized model skill and compliance with EU legislation using the Delta tool from FAIRMODE project (http://fairmode.jrc.ec.europa.eu/). The bias correction techniques adopted are, in order of complexity: (1) application of multiplicative factors calculated as the ratio of model-to-observed concentrations averaged over the previous days; (2) correction of the statistical distribution of model forecasts, in order to make it similar to that of the observations; (3) development and application of Model Output Statistics (MOS) regression equations. We illustrate differences and advantages/disadvantages of the three approaches. All the methods are relatively easy to implement for other modelling systems.

Accuracy of short‐term sea ice drift forecasts using a coupled ice‐ocean model

PubMed Central

Zhang, Jinlun

2015-01-01

Abstract Arctic sea ice drift forecasts of 6 h–9 days for the summer of 2014 are generated using the Marginal Ice Zone Modeling and Assimilation System (MIZMAS); the model is driven by 6 h atmospheric forecasts from the Climate Forecast System (CFSv2). Forecast ice drift speed is compared to drifting buoys and other observational platforms. Forecast positions are compared with actual positions 24 h–8 days since forecast. Forecast results are further compared to those from the forecasts generated using an ice velocity climatology driven by multiyear integrations of the same model. The results are presented in the context of scheduling the acquisition of high‐resolution images that need to follow buoys or scientific research platforms. RMS errors for ice speed are on the order of 5 km/d for 24–48 h since forecast using the sea ice model compared with 9 km/d using climatology. Predicted buoy position RMS errors are 6.3 km for 24 h and 14 km for 72 h since forecast. Model biases in ice speed and direction can be reduced by adjusting the air drag coefficient and water turning angle, but the adjustments do not affect verification statistics. This suggests that improved atmospheric forecast forcing may further reduce the forecast errors. The model remains skillful for 8 days. Using the forecast model increases the probability of tracking a target drifting in sea ice with a 10 km × 10 km image from 60 to 95% for a 24 h forecast and from 27 to 73% for a 48 h forecast. PMID:27818852

A GLM Post-processor to Adjust Ensemble Forecast Traces

NASA Astrophysics Data System (ADS)

Thiemann, M.; Day, G. N.; Schaake, J. C.; Draijer, S.; Wang, L.

2011-12-01

The skill of hydrologic ensemble forecasts has improved in the last years through a better understanding of climate variability, better climate forecasts and new data assimilation techniques. Having been extensively utilized for probabilistic water supply forecasting, interest is developing to utilize these forecasts in operational decision making. Hydrologic ensemble forecast members typically have inherent biases in flow timing and volume caused by (1) structural errors in the models used, (2) systematic errors in the data used to calibrate those models, (3) uncertain initial hydrologic conditions, and (4) uncertainties in the forcing datasets. Furthermore, hydrologic models have often not been developed for operational decision points and ensemble forecasts are thus not always available where needed. A statistical post-processor can be used to address these issues. The post-processor should (1) correct for systematic biases in flow timing and volume, (2) preserve the skill of the available raw forecasts, (3) preserve spatial and temporal correlation as well as the uncertainty in the forecasted flow data, (4) produce adjusted forecast ensembles that represent the variability of the observed hydrograph to be predicted, and (5) preserve individual forecast traces as equally likely. The post-processor should also allow for the translation of available ensemble forecasts to hydrologically similar locations where forecasts are not available. This paper introduces an ensemble post-processor (EPP) developed in support of New York City water supply operations. The EPP employs a general linear model (GLM) to (1) adjust available ensemble forecast traces and (2) create new ensembles for (nearby) locations where only historic flow observations are available. The EPP is calibrated by developing daily and aggregated statistical relationships form historical flow observations and model simulations. These are then used in operation to obtain the conditional probability density function (PDF) of the observations to be predicted, thus jointly adjusting individual ensemble members. These steps are executed in a normalized transformed space ('z'-space) to account for the strong non-linearity in the flow observations involved. A data window centered on each calibration date is used to minimize impacts from sampling errors and data noise. Testing on datasets from California and New York suggests that the EPP can successfully minimize biases in ensemble forecasts, while preserving the raw forecast skill in a 'days to weeks' forecast horizon and reproducing the variability of climatology for 'weeks to years' forecast horizons.

Ensemble Statistical Post-Processing of the National Air Quality Forecast Capability: Enhancing Ozone Forecasts in Baltimore, Maryland

NASA Technical Reports Server (NTRS)

Garner, Gregory G.; Thompson, Anne M.

2013-01-01

An ensemble statistical post-processor (ESP) is developed for the National Air Quality Forecast Capability (NAQFC) to address the unique challenges of forecasting surface ozone in Baltimore, MD. Air quality and meteorological data were collected from the eight monitors that constitute the Baltimore forecast region. These data were used to build the ESP using a moving-block bootstrap, regression tree models, and extreme-value theory. The ESP was evaluated using a 10-fold cross-validation to avoid evaluation with the same data used in the development process. Results indicate that the ESP is conditionally biased, likely due to slight overfitting while training the regression tree models. When viewed from the perspective of a decision-maker, the ESP provides a wealth of additional information previously not available through the NAQFC alone. The user is provided the freedom to tailor the forecast to the decision at hand by using decision-specific probability thresholds that define a forecast for an ozone exceedance. Taking advantage of the ESP, the user not only receives an increase in value over the NAQFC, but also receives value for An ensemble statistical post-processor (ESP) is developed for the National Air Quality Forecast Capability (NAQFC) to address the unique challenges of forecasting surface ozone in Baltimore, MD. Air quality and meteorological data were collected from the eight monitors that constitute the Baltimore forecast region. These data were used to build the ESP using a moving-block bootstrap, regression tree models, and extreme-value theory. The ESP was evaluated using a 10-fold cross-validation to avoid evaluation with the same data used in the development process. Results indicate that the ESP is conditionally biased, likely due to slight overfitting while training the regression tree models. When viewed from the perspective of a decision-maker, the ESP provides a wealth of additional information previously not available through the NAQFC alone. The user is provided the freedom to tailor the forecast to the decision at hand by using decision-specific probability thresholds that define a forecast for an ozone exceedance. Taking advantage of the ESP, the user not only receives an increase in value over the NAQFC, but also receives value for

Studies regarding the quality of numerical weather forecasts of the WRF model integrated at high-resolutions for the Romanian territory

DOE PAGES

Iriza, Amalia; Dumitrache, Rodica C.; Lupascu, Aurelia; ...

2016-01-01

Our paper aims to evaluate the quality of high-resolution weather forecasts from the Weather Research and Forecasting (WRF) numerical weather prediction model. The lateral and boundary conditions were obtained from the numerical output of the Consortium for Small-scale Modeling (COSMO) model at 7 km horizontal resolution. Furthermore, the WRF model was run for January and July 2013 at two horizontal resolutions (3 and 1 km). The numerical forecasts of the WRF model were evaluated using different statistical scores for 2 m temperature and 10 m wind speed. Our results showed a tendency of the WRF model to overestimate the valuesmore » of the analyzed parameters in comparison to observations.« less

Studies regarding the quality of numerical weather forecasts of the WRF model integrated at high-resolutions for the Romanian territory

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

Iriza, Amalia; Dumitrache, Rodica C.; Lupascu, Aurelia

Our paper aims to evaluate the quality of high-resolution weather forecasts from the Weather Research and Forecasting (WRF) numerical weather prediction model. The lateral and boundary conditions were obtained from the numerical output of the Consortium for Small-scale Modeling (COSMO) model at 7 km horizontal resolution. Furthermore, the WRF model was run for January and July 2013 at two horizontal resolutions (3 and 1 km). The numerical forecasts of the WRF model were evaluated using different statistical scores for 2 m temperature and 10 m wind speed. Our results showed a tendency of the WRF model to overestimate the valuesmore » of the analyzed parameters in comparison to observations.« less

Reservoir inflow forecasting with a modified coactive neuro-fuzzy inference system: a case study for a semi-arid region

NASA Astrophysics Data System (ADS)

Allawi, Mohammed Falah; Jaafar, Othman; Mohamad Hamzah, Firdaus; Mohd, Nuruol Syuhadaa; Deo, Ravinesh C.; El-Shafie, Ahmed

2017-10-01

Existing forecast models applied for reservoir inflow forecasting encounter several drawbacks, due to the difficulty of the underlying mathematical procedures being to cope with and to mimic the naturalization and stochasticity of the inflow data patterns. In this study, appropriate adjustments to the conventional coactive neuro-fuzzy inference system (CANFIS) method are proposed to improve the mathematical procedure, thus enabling a better detection of the high nonlinearity patterns found in the reservoir inflow training data. This modification includes the updating of the back propagation algorithm, leading to a consequent update of the membership rules and the induction of the centre-weighted set rather than the global weighted set used in feature extraction. The modification also aids in constructing an integrated model that is able to not only detect the nonlinearity in the training data but also the wide range of features within the training data records used to simulate the forecasting model. To demonstrate the model's efficacy, the proposed CANFIS method has been applied to forecast monthly inflow data at Aswan High Dam (AHD), located in southern Egypt. Comparative analyses of the forecasting skill of the modified CANFIS and the conventional ANFIS model are carried out with statistical score indicators to assess the reliability of the developed method. The statistical metrics support the better performance of the developed CANFIS model, which significantly outperforms the ANFIS model to attain a low relative error value (23%), mean absolute error (1.4 BCM month-1), root mean square error (1.14 BCM month-1), and a relative large coefficient of determination (0.94). The present study ascertains the better utility of the modified CANFIS model in respect to the traditional ANFIS model applied in reservoir inflow forecasting for a semi-arid region.

Tracking signal test to monitor an intelligent time series forecasting model

NASA Astrophysics Data System (ADS)

Deng, Yan; Jaraiedi, Majid; Iskander, Wafik H.

2004-03-01

Extensive research has been conducted on the subject of Intelligent Time Series forecasting, including many variations on the use of neural networks. However, investigation of model adequacy over time, after the training processes is completed, remains to be fully explored. In this paper we demonstrate a how a smoothed error tracking signals test can be incorporated into a neuro-fuzzy model to monitor the forecasting process and as a statistical measure for keeping the forecasting model up-to-date. The proposed monitoring procedure is effective in the detection of nonrandom changes, due to model inadequacy or lack of unbiasedness in the estimation of model parameters and deviations from the existing patterns. This powerful detection device will result in improved forecast accuracy in the long run. An example data set has been used to demonstrate the application of the proposed method.

ASSESSMENT OF AN ENSEMBLE OF SEVEN REAL-TIME OZONE FORECASTS OVER EASTERN NORTH AMERICA DURING THE SUMMER OF 2004

EPA Science Inventory

The real-time forecasts of ozone (O₃) from seven air quality forecast models (AQFMs) are statistically evaluated against observations collected during July and August of 2004 (53 days) through the Aerometric Information Retrieval Now (AIRNow) network at roughly 340 mon...

Integrated Wind Power Planning Tool

NASA Astrophysics Data System (ADS)

Rosgaard, Martin; Giebel, Gregor; Skov Nielsen, Torben; Hahmann, Andrea; Sørensen, Poul; Madsen, Henrik

2013-04-01

This poster presents the current state of the public service obligation (PSO) funded project PSO 10464, with the title "Integrated Wind Power Planning Tool". The goal is to integrate a mesoscale numerical weather prediction (NWP) model with purely statistical tools in order to assess wind power fluctuations, with focus on long term power system planning for future wind farms as well as short term forecasting for existing wind farms. Currently, wind power fluctuation models are either purely statistical or integrated with NWP models of limited resolution. Using the state-of-the-art mesoscale NWP model Weather Research & Forecasting model (WRF) the forecast error is sought quantified in dependence of the time scale involved. This task constitutes a preparative study for later implementation of features accounting for NWP forecast errors in the DTU Wind Energy maintained Corwind code - a long term wind power planning tool. Within the framework of PSO 10464 research related to operational short term wind power prediction will be carried out, including a comparison of forecast quality at different mesoscale NWP model resolutions and development of a statistical wind power prediction tool taking input from WRF. The short term prediction part of the project is carried out in collaboration with ENFOR A/S; a Danish company that specialises in forecasting and optimisation for the energy sector. The integrated prediction model will allow for the description of the expected variability in wind power production in the coming hours to days, accounting for its spatio-temporal dependencies, and depending on the prevailing weather conditions defined by the WRF output. The output from the integrated short term prediction tool constitutes scenario forecasts for the coming period, which can then be fed into any type of system model or decision making problem to be solved. The high resolution of the WRF results loaded into the integrated prediction model will ensure a high accuracy data basis is available for use in the decision making process of the Danish transmission system operator. The need for high accuracy predictions will only increase over the next decade as Denmark approaches the goal of 50% wind power based electricity in 2025 from the current 20%.

Evaluating the performance of infectious disease forecasts: A comparison of climate-driven and seasonal dengue forecasts for Mexico.

PubMed

Johansson, Michael A; Reich, Nicholas G; Hota, Aditi; Brownstein, John S; Santillana, Mauricio

2016-09-26

Dengue viruses, which infect millions of people per year worldwide, cause large epidemics that strain healthcare systems. Despite diverse efforts to develop forecasting tools including autoregressive time series, climate-driven statistical, and mechanistic biological models, little work has been done to understand the contribution of different components to improved prediction. We developed a framework to assess and compare dengue forecasts produced from different types of models and evaluated the performance of seasonal autoregressive models with and without climate variables for forecasting dengue incidence in Mexico. Climate data did not significantly improve the predictive power of seasonal autoregressive models. Short-term and seasonal autocorrelation were key to improving short-term and long-term forecasts, respectively. Seasonal autoregressive models captured a substantial amount of dengue variability, but better models are needed to improve dengue forecasting. This framework contributes to the sparse literature of infectious disease prediction model evaluation, using state-of-the-art validation techniques such as out-of-sample testing and comparison to an appropriate reference model.

Evaluating the performance of infectious disease forecasts: A comparison of climate-driven and seasonal dengue forecasts for Mexico

PubMed Central

Johansson, Michael A.; Reich, Nicholas G.; Hota, Aditi; Brownstein, John S.; Santillana, Mauricio

2016-01-01

Dengue viruses, which infect millions of people per year worldwide, cause large epidemics that strain healthcare systems. Despite diverse efforts to develop forecasting tools including autoregressive time series, climate-driven statistical, and mechanistic biological models, little work has been done to understand the contribution of different components to improved prediction. We developed a framework to assess and compare dengue forecasts produced from different types of models and evaluated the performance of seasonal autoregressive models with and without climate variables for forecasting dengue incidence in Mexico. Climate data did not significantly improve the predictive power of seasonal autoregressive models. Short-term and seasonal autocorrelation were key to improving short-term and long-term forecasts, respectively. Seasonal autoregressive models captured a substantial amount of dengue variability, but better models are needed to improve dengue forecasting. This framework contributes to the sparse literature of infectious disease prediction model evaluation, using state-of-the-art validation techniques such as out-of-sample testing and comparison to an appropriate reference model. PMID:27665707

A national framework for flood forecasting model assessment for use in operations and investment planning over England and Wales

NASA Astrophysics Data System (ADS)

Moore, Robert J.; Wells, Steven C.; Cole, Steven J.

2016-04-01

It has been common for flood forecasting systems to be commissioned at a catchment or regional level in response to local priorities and hydrological conditions, leading to variety in system design and model choice. As systems mature and efficiencies of national management are sought, there can be a drive towards system rationalisation, gaining an overview of model performance and consideration of simplification through model-type convergence. Flood forecasting model assessments, whilst overseen at a national level, may be commissioned and managed at a catchment and regional level, take a variety of forms and be large in number. This presents a challenge when an integrated national assessment is required to guide operational use of flood forecasts and plan future investment in flood forecasting models and supporting hydrometric monitoring. This contribution reports on how a nationally consistent framework for flood forecasting model performance has been developed to embrace many past, ongoing and future assessments for local river systems by engineering consultants across England & Wales. The outcome is a Performance Summary for every site model assessed which, on a single page, contains relevant catchment information for context, a selection of overlain forecast and observed hydrographs and a set of performance statistics with associated displays of novel condensed form. One display provides performance comparison with other models that may exist for the site. The performance statistics include skill scores for forecasting events (flow/level threshold crossings) of differing severity/rarity, indicating their probability and likely timing, which have real value in an operational setting. The local models assessed can be of any type and span rainfall-runoff (conceptual and transfer function) and flow routing (hydrological and hydrodynamic) forms. Also accommodated by the framework is the national G2G (Grid-to-Grid) distributed hydrological model, providing area-wide coverage across the fluvial rivers of England and Wales, which can be assessed at gauged sites. Thus the performance of the national G2G model forecasts can be directly compared with that from the local models. The Performance Summary for each site model is complemented by a national spatial analysis of model performance stratified by model-type, geographical region and forecast lead-time. The map displays provide an extensive evidence-base that can be interrogated, through a Flood Forecasting Model Performance web portal, to reveal fresh insights into comparative performance across locations, lead-times and models. This work was commissioned by the Environment Agency in partnership with Natural Resources Wales and the Flood Forecasting Centre for England and Wales.

Nowcasting of Low-Visibility Procedure States with Ordered Logistic Regression at Vienna International Airport

NASA Astrophysics Data System (ADS)

Kneringer, Philipp; Dietz, Sebastian; Mayr, Georg J.; Zeileis, Achim

2017-04-01

Low-visibility conditions have a large impact on aviation safety and economic efficiency of airports and airlines. To support decision makers, we develop a statistical probabilistic nowcasting tool for the occurrence of capacity-reducing operations related to low visibility. The probabilities of four different low visibility classes are predicted with an ordered logistic regression model based on time series of meteorological point measurements. Potential predictor variables for the statistical models are visibility, humidity, temperature and wind measurements at several measurement sites. A stepwise variable selection method indicates that visibility and humidity measurements are the most important model inputs. The forecasts are tested with a 30 minute forecast interval up to two hours, which is a sufficient time span for tactical planning at Vienna Airport. The ordered logistic regression models outperform persistence and are competitive with human forecasters.

Evaluation of Satellite and Model Precipitation Products Over Turkey

NASA Astrophysics Data System (ADS)

Yilmaz, M. T.; Amjad, M.

2017-12-01

Satellite-based remote sensing, gauge stations, and models are the three major platforms to acquire precipitation dataset. Among them satellites and models have the advantage of retrieving spatially and temporally continuous and consistent datasets, while the uncertainty estimates of these retrievals are often required for many hydrological studies to understand the source and the magnitude of the uncertainty in hydrological response parameters. In this study, satellite and model precipitation data products are validated over various temporal scales (daily, 3-daily, 7-daily, 10-daily and monthly) using in-situ measured precipitation observations from a network of 733 gauges from all over the Turkey. Tropical Rainfall Measurement Mission (TRMM) Multi-satellite Precipitation Analysis (TMPA) 3B42 version 7 and European Center of Medium-Range Weather Forecast (ECMWF) model estimates (daily, 3-daily, 7-daily and 10-daily accumulated forecast) are used in this study. Retrievals are evaluated for their mean and standard deviation and their accuracies are evaluated via bias, root mean square error, error standard deviation and correlation coefficient statistics. Intensity vs frequency analysis and some contingency table statistics like percent correct, probability of detection, false alarm ratio and critical success index are determined using daily time-series. Both ECMWF forecasts and TRMM observations, on average, overestimate the precipitation compared to gauge estimates; wet biases are 10.26 mm/month and 8.65 mm/month, respectively for ECMWF and TRMM. RMSE values of ECMWF forecasts and TRMM estimates are 39.69 mm/month and 41.55 mm/month, respectively. Monthly correlations between Gauges-ECMWF, Gauges-TRMM and ECMWF-TRMM are 0.76, 0.73 and 0.81, respectively. The model and the satellite error statistics are further compared against the gauges error statistics based on inverse distance weighting (IWD) analysis. Both the model and satellite data have less IWD errors (14.72 mm/month and 10.75 mm/month, respectively) compared to gauges IWD error (21.58 mm/month). These results show that, on average, ECMWF forecast data have higher skill than TRMM observations. Overall, both ECMWF forecast data and TRMM observations show good potential for catchment scale hydrological analysis.

A statistical approach to quasi-extinction forecasting.

PubMed

Holmes, Elizabeth Eli; Sabo, John L; Viscido, Steven Vincent; Fagan, William Fredric

2007-12-01

Forecasting population decline to a certain critical threshold (the quasi-extinction risk) is one of the central objectives of population viability analysis (PVA), and such predictions figure prominently in the decisions of major conservation organizations. In this paper, we argue that accurate forecasting of a population's quasi-extinction risk does not necessarily require knowledge of the underlying biological mechanisms. Because of the stochastic and multiplicative nature of population growth, the ensemble behaviour of population trajectories converges to common statistical forms across a wide variety of stochastic population processes. This paper provides a theoretical basis for this argument. We show that the quasi-extinction surfaces of a variety of complex stochastic population processes (including age-structured, density-dependent and spatially structured populations) can be modelled by a simple stochastic approximation: the stochastic exponential growth process overlaid with Gaussian errors. Using simulated and real data, we show that this model can be estimated with 20-30 years of data and can provide relatively unbiased quasi-extinction risk with confidence intervals considerably smaller than (0,1). This was found to be true even for simulated data derived from some of the noisiest population processes (density-dependent feedback, species interactions and strong age-structure cycling). A key advantage of statistical models is that their parameters and the uncertainty of those parameters can be estimated from time series data using standard statistical methods. In contrast for most species of conservation concern, biologically realistic models must often be specified rather than estimated because of the limited data available for all the various parameters. Biologically realistic models will always have a prominent place in PVA for evaluating specific management options which affect a single segment of a population, a single demographic rate, or different geographic areas. However, for forecasting quasi-extinction risk, statistical models that are based on the convergent statistical properties of population processes offer many advantages over biologically realistic models.

Stationarity test with a direct test for heteroskedasticity in exchange rate forecasting models

NASA Astrophysics Data System (ADS)

Khin, Aye Aye; Chau, Wong Hong; Seong, Lim Chee; Bin, Raymond Ling Leh; Teng, Kevin Low Lock

2017-05-01

Global economic has been decreasing in the recent years, manifested by the greater exchange rates volatility on international commodity market. This study attempts to analyze some prominent exchange rate forecasting models on Malaysian commodity trading: univariate ARIMA, ARCH and GARCH models in conjunction with stationarity test on residual diagnosis direct testing of heteroskedasticity. All forecasting models utilized the monthly data from 1990 to 2015. Given a total of 312 observations, the data used to forecast both short-term and long-term exchange rate. The forecasting power statistics suggested that the forecasting performance of ARIMA (1, 1, 1) model is more efficient than the ARCH (1) and GARCH (1, 1) models. For ex-post forecast, exchange rate was increased from RM 3.50 per USD in January 2015 to RM 4.47 per USD in December 2015 based on the baseline data. For short-term ex-ante forecast, the analysis results indicate a decrease in exchange rate on 2016 June (RM 4.27 per USD) as compared with 2015 December. A more appropriate forecasting method of exchange rate is vital to aid the decision-making process and planning on the sustainable commodities' production in the world economy.

Using statistical and artificial neural network models to forecast potentiometric levels at a deep well in South Texas

NASA Astrophysics Data System (ADS)

Uddameri, V.

2007-01-01

Reliable forecasts of monthly and quarterly fluctuations in groundwater levels are necessary for short- and medium-term planning and management of aquifers to ensure proper service of seasonal demands within a region. Development of physically based transient mathematical models at this time scale poses considerable challenges due to lack of suitable data and other uncertainties. Artificial neural networks (ANN) possess flexible mathematical structures and are capable of mapping highly nonlinear relationships. Feed-forward neural network models were constructed and trained using the back-percolation algorithm to forecast monthly and quarterly time-series water levels at a well that taps into the deeper Evangeline formation of the Gulf Coast aquifer in Victoria, TX. Unlike unconfined formations, no causal relationships exist between water levels and hydro-meteorological variables measured near the vicinity of the well. As such, an endogenous forecasting model using dummy variables to capture short-term seasonal fluctuations and longer-term (decadal) trends was constructed. The root mean square error, mean absolute deviation and correlation coefficient ( R) were noted to be 1.40, 0.33 and 0.77 m, respectively, for an evaluation dataset of quarterly measurements and 1.17, 0.46, and 0.88 m for an evaluative monthly dataset not used to train or test the model. These statistics were better for the ANN model than those developed using statistical regression techniques.

A data-driven multi-model methodology with deep feature selection for short-term wind forecasting

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

Feng, Cong; Cui, Mingjian; Hodge, Bri-Mathias

With the growing wind penetration into the power system worldwide, improving wind power forecasting accuracy is becoming increasingly important to ensure continued economic and reliable power system operations. In this paper, a data-driven multi-model wind forecasting methodology is developed with a two-layer ensemble machine learning technique. The first layer is composed of multiple machine learning models that generate individual forecasts. A deep feature selection framework is developed to determine the most suitable inputs to the first layer machine learning models. Then, a blending algorithm is applied in the second layer to create an ensemble of the forecasts produced by firstmore » layer models and generate both deterministic and probabilistic forecasts. This two-layer model seeks to utilize the statistically different characteristics of each machine learning algorithm. A number of machine learning algorithms are selected and compared in both layers. This developed multi-model wind forecasting methodology is compared to several benchmarks. The effectiveness of the proposed methodology is evaluated to provide 1-hour-ahead wind speed forecasting at seven locations of the Surface Radiation network. Numerical results show that comparing to the single-algorithm models, the developed multi-model framework with deep feature selection procedure has improved the forecasting accuracy by up to 30%.« less

Real-Time Kennedy Space Center and Cape Canaveral Air Force Station High-Resolution Model Implementation and Verification

NASA Technical Reports Server (NTRS)

Shafer, Jaclyn A.; Watson, Leela R.

2015-01-01

Customer: NASA's Launch Services Program (LSP), Ground Systems Development and Operations (GSDO), and Space Launch System (SLS) programs. NASA's LSP, GSDO, SLS and other programs at Kennedy Space Center (KSC) and Cape Canaveral Air Force Station (CCAFS) use the daily and weekly weather forecasts issued by the 45th Weather Squadron (45 WS) as decision tools for their day-to-day and launch operations on the Eastern Range (ER). For example, to determine if they need to limit activities such as vehicle transport to the launch pad, protect people, structures or exposed launch vehicles given a threat of severe weather, or reschedule other critical operations. The 45 WS uses numerical weather prediction models as a guide for these weather forecasts, particularly the Air Force Weather Agency (AFWA) 1.67 kilometer Weather Research and Forecasting (WRF) model. Considering the 45 WS forecasters' and Launch Weather Officers' (LWO) extensive use of the AFWA model, the 45 WS proposed a task at the September 2013 Applied Meteorology Unit (AMU) Tasking Meeting requesting the AMU verify this model. Due to the lack of archived model data available from AFWA, verification is not yet possible. Instead, the AMU proposed to implement and verify the performance of an ER version of the AMU high-resolution WRF Environmental Modeling System (EMS) model (Watson 2013) in real-time. The tasking group agreed to this proposal; therefore the AMU implemented the WRF-EMS model on the second of two NASA AMU modeling clusters. The model was set up with a triple-nested grid configuration over KSC/CCAFS based on previous AMU work (Watson 2013). The outer domain (D01) has 12-kilometer grid spacing, the middle domain (D02) has 4-kilometer grid spacing, and the inner domain (D03) has 1.33-kilometer grid spacing. The model runs a 12-hour forecast every hour, D01 and D02 domain outputs are available once an hour and D03 is every 15 minutes during the forecast period. The AMU assessed the WRF-EMS 1.33-kilometer domain model performance for the 2014 warm season (May-September). Verification statistics were computed using the Model Evaluation Tools, which compared the model forecasts to observations. The mean error values were close to 0 and the root mean square error values were less than 1.8 for mean sea-level pressure (millibars), temperature (degrees Kelvin), dewpoint temperature (degrees Kelvin), and wind speed (per millisecond), all very small differences between the forecast and observations considering the normal magnitudes of the parameters. The precipitation forecast verification results showed consistent under-forecasting of the precipitation object size. This could be an artifact of calculating the statistics for each hour rather than for the entire 12-hour period. The AMU will continue to generate verification statistics for the 1.33-kilometer WRF-EMS domain as data become available in future cool and warm seasons. More data will produce more robust statistics and reveal a more accurate assessment of model performance. Once the formal task was complete, the AMU conducted additional work to better understand the wind direction results. The results were stratified diurnally and by wind speed to determine what effects the stratifications would have on the model wind direction verification statistics. The results are summarized in the addendum at the end of this report. In addition to verifying the model's performance, the AMU also made the output available in the Advanced Weather Interactive Processing System II (AWIPS II). This allows the 45 WS and AMU staff to customize the model output display on the AMU and Range Weather Operations AWIPS II client computers and conduct real-time subjective analyses. In the future, the AMU will implement an updated version of the WRF-EMS model that incorporates local data assimilation. This model will also run in real-time and be made available in AWIPS II.

Statistical models for fever forecasting based on advanced body temperature monitoring.

PubMed

Jordan, Jorge; Miro-Martinez, Pau; Vargas, Borja; Varela-Entrecanales, Manuel; Cuesta-Frau, David

2017-02-01

Body temperature monitoring provides health carers with key clinical information about the physiological status of patients. Temperature readings are taken periodically to detect febrile episodes and consequently implement the appropriate medical countermeasures. However, fever is often difficult to assess at early stages, or remains undetected until the next reading, probably a few hours later. The objective of this article is to develop a statistical model to forecast fever before a temperature threshold is exceeded to improve the therapeutic approach to the subjects involved. To this end, temperature series of 9 patients admitted to a general internal medicine ward were obtained with a continuous monitoring Holter device, collecting measurements of peripheral and core temperature once per minute. These series were used to develop different statistical models that could quantify the probability of having a fever spike in the following 60 minutes. A validation series was collected to assess the accuracy of the models. Finally, the results were compared with the analysis of some series by experienced clinicians. Two different models were developed: a logistic regression model and a linear discrimination analysis model. Both of them exhibited a fever peak forecasting accuracy greater than 84%. When compared with experts' assessment, both models identified 35 (97.2%) of 36 fever spikes. The models proposed are highly accurate in forecasting the appearance of fever spikes within a short period in patients with suspected or confirmed febrile-related illnesses. Copyright © 2016 Elsevier Inc. All rights reserved.

Probabilistic forecasting of extreme weather events based on extreme value theory

NASA Astrophysics Data System (ADS)

Van De Vyver, Hans; Van Schaeybroeck, Bert

2016-04-01

Extreme events in weather and climate such as high wind gusts, heavy precipitation or extreme temperatures are commonly associated with high impacts on both environment and society. Forecasting extreme weather events is difficult, and very high-resolution models are needed to describe explicitly extreme weather phenomena. A prediction system for such events should therefore preferably be probabilistic in nature. Probabilistic forecasts and state estimations are nowadays common in the numerical weather prediction community. In this work, we develop a new probabilistic framework based on extreme value theory that aims to provide early warnings up to several days in advance. We consider the combined events when an observation variable Y (for instance wind speed) exceeds a high threshold y and its corresponding deterministic forecasts X also exceeds a high forecast threshold y. More specifically two problems are addressed:} We consider pairs (X,Y) of extreme events where X represents a deterministic forecast, and Y the observation variable (for instance wind speed). More specifically two problems are addressed: Given a high forecast X=x_0, what is the probability that Y>y? In other words: provide inference on the conditional probability: [ Pr{Y>y|X=x_0}. ] Given a probabilistic model for Problem 1, what is the impact on the verification analysis of extreme events. These problems can be solved with bivariate extremes (Coles, 2001), and the verification analysis in (Ferro, 2007). We apply the Ramos and Ledford (2009) parametric model for bivariate tail estimation of the pair (X,Y). The model accommodates different types of extremal dependence and asymmetry within a parsimonious representation. Results are presented using the ensemble reforecast system of the European Centre of Weather Forecasts (Hagedorn, 2008). Coles, S. (2001) An Introduction to Statistical modelling of Extreme Values. Springer-Verlag.Ferro, C.A.T. (2007) A probability model for verifying deterministic forecasts of extreme events. Wea. Forecasting {22}, 1089-1100.Hagedorn, R. (2008) Using the ECMWF reforecast dataset to calibrate EPS forecasts. ECMWF Newsletter, {117}, 8-13.Ramos, A., Ledford, A. (2009) A new class of models for bivariate joint tails. J.R. Statist. Soc. B {71}, 219-241.

Global Positioning System (GPS) Precipitable Water in Forecasting Lightning at Spaceport Canaveral

NASA Technical Reports Server (NTRS)

Kehrer, Kristen C.; Graf, Brian; Roeder, William

2006-01-01

This paper evaluates the use of precipitable water (PW) from Global Positioning System (GPS) in lightning prediction. Additional independent verification of an earlier model is performed. This earlier model used binary logistic regression with the following four predictor variables optimally selected from a candidate list of 23 candidate predictors: the current precipitable water value for a given time of the day, the change in GPS-PW over the past 9 hours, the KIndex, and the electric field mill value. This earlier model was not optimized for any specific forecast interval, but showed promise for 6 hour and 1.5 hour forecasts. Two new models were developed and verified. These new models were optimized for two operationally significant forecast intervals. The first model was optimized for the 0.5 hour lightning advisories issued by the 45th Weather Squadron. An additional 1.5 hours was allowed for sensor dwell, communication, calculation, analysis, and advisory decision by the forecaster. Therefore the 0.5 hour advisory model became a 2 hour forecast model for lightning within the 45th Weather Squadron advisory areas. The second model was optimized for major ground processing operations supported by the 45th Weather Squadron, which can require lightning forecasts with a lead-time of up to 7.5 hours. Using the same 1.5 lag as in the other new model, this became a 9 hour forecast model for lightning within 37 km (20 NM)) of the 45th Weather Squadron advisory areas. The two new models were built using binary logistic regression from a list of 26 candidate predictor variables: the current GPS-PW value, the change of GPS-PW over 0.5 hour increments from 0.5 to 12 hours, and the K-index. The new 2 hour model found the following for predictors to be statistically significant, listed in decreasing order of contribution to the forecast: the 0.5 hour change in GPS-PW, the 7.5 hour change in GPS-PW, the current GPS-PW value, and the KIndex. The new 9 hour forecast model found the following five independent variables to be statistically significant, listed in decreasing order of contribution to the forecast: the current GPSPW value, the 8.5 hour change in GPS-PW, the 3.5 hour change in GPS-PW, the 12 hour change in GPS-PW, and the K-Index. In both models, the GPS-PW parameters had better correlation to the lightning forecast than the K-Index, a widely used thunderstorm index. Possible future improvements to this study are discussed.

An application of ensemble/multi model approach for wind power production forecast.

NASA Astrophysics Data System (ADS)

Alessandrini, S.; Decimi, G.; Hagedorn, R.; Sperati, S.

2010-09-01

The wind power forecast of the 3 days ahead period are becoming always more useful and important in reducing the problem of grid integration and energy price trading due to the increasing wind power penetration. Therefore it's clear that the accuracy of this forecast is one of the most important requirements for a successful application. The wind power forecast is based on a mesoscale meteorological models that provides the 3 days ahead wind data. A Model Output Statistic correction is then performed to reduce systematic error caused, for instance, by a wrong representation of surface roughness or topography in the meteorological models. The corrected wind data are then used as input in the wind farm power curve to obtain the power forecast. These computations require historical time series of wind measured data (by an anemometer located in the wind farm or on the nacelle) and power data in order to be able to perform the statistical analysis on the past. For this purpose a Neural Network (NN) is trained on the past data and then applied in the forecast task. Considering that the anemometer measurements are not always available in a wind farm a different approach has also been adopted. A training of the NN to link directly the forecasted meteorological data and the power data has also been performed. The normalized RMSE forecast error seems to be lower in most cases by following the second approach. We have examined two wind farms, one located in Denmark on flat terrain and one located in a mountain area in the south of Italy (Sicily). In both cases we compare the performances of a prediction based on meteorological data coming from a single model with those obtained by using two or more models (RAMS, ECMWF deterministic, LAMI, HIRLAM). It is shown that the multi models approach reduces the day-ahead normalized RMSE forecast error of at least 1% compared to the singles models approach. Moreover the use of a deterministic global model, (e.g. ECMWF deterministic model) seems to reach similar level of accuracy of those of the mesocale models (LAMI and RAMS). Finally we have focused on the possibility of using the ensemble model (ECMWF) to estimate the hourly, three days ahead, power forecast accuracy. Contingency diagram between RMSE of the deterministic power forecast and the ensemble members spread of wind forecast have been produced. From this first analysis it seems that ensemble spread could be used as an indicator of the forecast's accuracy at least for the first day ahead period. In fact low spreads often correspond to low forecast error. For longer forecast horizon the correlation between RMSE and ensemble spread decrease becoming too low to be used for this purpose.

Multivariate postprocessing techniques for probabilistic hydrological forecasting

NASA Astrophysics Data System (ADS)

Hemri, Stephan; Lisniak, Dmytro; Klein, Bastian

2016-04-01

Hydrologic ensemble forecasts driven by atmospheric ensemble prediction systems need statistical postprocessing in order to account for systematic errors in terms of both mean and spread. Runoff is an inherently multivariate process with typical events lasting from hours in case of floods to weeks or even months in case of droughts. This calls for multivariate postprocessing techniques that yield well calibrated forecasts in univariate terms and ensure a realistic temporal dependence structure at the same time. To this end, the univariate ensemble model output statistics (EMOS; Gneiting et al., 2005) postprocessing method is combined with two different copula approaches that ensure multivariate calibration throughout the entire forecast horizon. These approaches comprise ensemble copula coupling (ECC; Schefzik et al., 2013), which preserves the dependence structure of the raw ensemble, and a Gaussian copula approach (GCA; Pinson and Girard, 2012), which estimates the temporal correlations from training observations. Both methods are tested in a case study covering three subcatchments of the river Rhine that represent different sizes and hydrological regimes: the Upper Rhine up to the gauge Maxau, the river Moselle up to the gauge Trier, and the river Lahn up to the gauge Kalkofen. The results indicate that both ECC and GCA are suitable for modelling the temporal dependences of probabilistic hydrologic forecasts (Hemri et al., 2015). References Gneiting, T., A. E. Raftery, A. H. Westveld, and T. Goldman (2005), Calibrated probabilistic forecasting using ensemble model output statistics and minimum CRPS estimation, Monthly Weather Review, 133(5), 1098-1118, DOI: 10.1175/MWR2904.1. Hemri, S., D. Lisniak, and B. Klein, Multivariate postprocessing techniques for probabilistic hydrological forecasting, Water Resources Research, 51(9), 7436-7451, DOI: 10.1002/2014WR016473. Pinson, P., and R. Girard (2012), Evaluating the quality of scenarios of short-term wind power generation, Applied Energy, 96, 12-20, DOI: 10.1016/j.apenergy.2011.11.004. Schefzik, R., T. L. Thorarinsdottir, and T. Gneiting (2013), Uncertainty quantification in complex simulation models using ensemble copula coupling, Statistical Science, 28, 616-640, DOI: 10.1214/13-STS443.

Seasonal forecasting of hydrological drought in the Limpopo Basin: a comparison of statistical methods

NASA Astrophysics Data System (ADS)

Seibert, Mathias; Merz, Bruno; Apel, Heiko

2017-03-01

The Limpopo Basin in southern Africa is prone to droughts which affect the livelihood of millions of people in South Africa, Botswana, Zimbabwe and Mozambique. Seasonal drought early warning is thus vital for the whole region. In this study, the predictability of hydrological droughts during the main runoff period from December to May is assessed using statistical approaches. Three methods (multiple linear models, artificial neural networks, random forest regression trees) are compared in terms of their ability to forecast streamflow with up to 12 months of lead time. The following four main findings result from the study. 1. There are stations in the basin at which standardised streamflow is predictable with lead times up to 12 months. The results show high inter-station differences of forecast skill but reach a coefficient of determination as high as 0.73 (cross validated). 2. A large range of potential predictors is considered in this study, comprising well-established climate indices, customised teleconnection indices derived from sea surface temperatures and antecedent streamflow as a proxy of catchment conditions. El Niño and customised indices, representing sea surface temperature in the Atlantic and Indian oceans, prove to be important teleconnection predictors for the region. Antecedent streamflow is a strong predictor in small catchments (with median 42 % explained variance), whereas teleconnections exert a stronger influence in large catchments. 3. Multiple linear models show the best forecast skill in this study and the greatest robustness compared to artificial neural networks and random forest regression trees, despite their capabilities to represent nonlinear relationships. 4. Employed in early warning, the models can be used to forecast a specific drought level. Even if the coefficient of determination is low, the forecast models have a skill better than a climatological forecast, which is shown by analysis of receiver operating characteristics (ROCs). Seasonal statistical forecasts in the Limpopo show promising results, and thus it is recommended to employ them as complementary to existing forecasts in order to strengthen preparedness for droughts.

Probabilistic Forecasting of Surface Ozone with a Novel Statistical Approach

NASA Technical Reports Server (NTRS)

Balashov, Nikolay V.; Thompson, Anne M.; Young, George S.

2017-01-01

The recent change in the Environmental Protection Agency's surface ozone regulation, lowering the surface ozone daily maximum 8-h average (MDA8) exceedance threshold from 75 to 70 ppbv, poses significant challenges to U.S. air quality (AQ) forecasters responsible for ozone MDA8 forecasts. The forecasters, supplied by only a few AQ model products, end up relying heavily on self-developed tools. To help U.S. AQ forecasters, this study explores a surface ozone MDA8 forecasting tool that is based solely on statistical methods and standard meteorological variables from the numerical weather prediction (NWP) models. The model combines the self-organizing map (SOM), which is a clustering technique, with a step wise weighted quadratic regression using meteorological variables as predictors for ozone MDA8. The SOM method identifies different weather regimes, to distinguish between various modes of ozone variability, and groups them according to similarity. In this way, when a regression is developed for a specific regime, data from the other regimes are also used, with weights that are based on their similarity to this specific regime. This approach, regression in SOM (REGiS), yields a distinct model for each regime taking into account both the training cases for that regime and other similar training cases. To produce probabilistic MDA8 ozone forecasts, REGiS weighs and combines all of the developed regression models on the basis of the weather patterns predicted by an NWP model. REGiS is evaluated over the San Joaquin Valley in California and the northeastern plains of Colorado. The results suggest that the model performs best when trained and adjusted separately for an individual AQ station and its corresponding meteorological site.

Neural networks and traditional time series methods: a synergistic combination in state economic forecasts.

PubMed

Hansen, J V; Nelson, R D

1997-01-01

Ever since the initial planning for the 1997 Utah legislative session, neural-network forecasting techniques have provided valuable insights for analysts forecasting tax revenues. These revenue estimates are critically important since agency budgets, support for education, and improvements to infrastructure all depend on their accuracy. Underforecasting generates windfalls that concern taxpayers, whereas overforecasting produces budget shortfalls that cause inadequately funded commitments. The pattern finding ability of neural networks gives insightful and alternative views of the seasonal and cyclical components commonly found in economic time series data. Two applications of neural networks to revenue forecasting clearly demonstrate how these models complement traditional time series techniques. In the first, preoccupation with a potential downturn in the economy distracts analysis based on traditional time series methods so that it overlooks an emerging new phenomenon in the data. In this case, neural networks identify the new pattern that then allows modification of the time series models and finally gives more accurate forecasts. In the second application, data structure found by traditional statistical tools allows analysts to provide neural networks with important information that the networks then use to create more accurate models. In summary, for the Utah revenue outlook, the insights that result from a portfolio of forecasts that includes neural networks exceeds the understanding generated from strictly statistical forecasting techniques. In this case, the synergy clearly results in the whole of the portfolio of forecasts being more accurate than the sum of the individual parts.

Improvements in approaches to forecasting and evaluation techniques

NASA Astrophysics Data System (ADS)

Weatherhead, Elizabeth

2014-05-01

The US is embarking on an experiment to make significant and sustained improvements in weather forecasting. The effort stems from a series of community conversations that recognized the rapid advancements in observations, modeling and computing techniques in the academic, governmental and private sectors. The new directions and initial efforts will be summarized, including information on possibilities for international collaboration. Most new projects are scheduled to start in the last half of 2014. Several advancements include ensemble forecasting with global models, and new sharing of computing resources. Newly developed techniques for evaluating weather forecast models will be presented in detail. The approaches use statistical techniques that incorporate pair-wise comparisons of forecasts with observations and account for daily auto-correlation to assess appropriate uncertainty in forecast changes. Some of the new projects allow for international collaboration, particularly on the research components of the projects.

Post-processing of multi-model ensemble river discharge forecasts using censored EMOS

NASA Astrophysics Data System (ADS)

Hemri, Stephan; Lisniak, Dmytro; Klein, Bastian

2014-05-01

When forecasting water levels and river discharge, ensemble weather forecasts are used as meteorological input to hydrologic process models. As hydrologic models are imperfect and the input ensembles tend to be biased and underdispersed, the output ensemble forecasts for river runoff typically are biased and underdispersed, too. Thus, statistical post-processing is required in order to achieve calibrated and sharp predictions. Standard post-processing methods such as Ensemble Model Output Statistics (EMOS) that have their origins in meteorological forecasting are now increasingly being used in hydrologic applications. Here we consider two sub-catchments of River Rhine, for which the forecasting system of the Federal Institute of Hydrology (BfG) uses runoff data that are censored below predefined thresholds. To address this methodological challenge, we develop a censored EMOS method that is tailored to such data. The censored EMOS forecast distribution can be understood as a mixture of a point mass at the censoring threshold and a continuous part based on a truncated normal distribution. Parameter estimates of the censored EMOS model are obtained by minimizing the Continuous Ranked Probability Score (CRPS) over the training dataset. Model fitting on Box-Cox transformed data allows us to take account of the positive skewness of river discharge distributions. In order to achieve realistic forecast scenarios over an entire range of lead-times, there is a need for multivariate extensions. To this end, we smooth the marginal parameter estimates over lead-times. In order to obtain realistic scenarios of discharge evolution over time, the marginal distributions have to be linked with each other. To this end, the multivariate dependence structure can either be adopted from the raw ensemble like in Ensemble Copula Coupling (ECC), or be estimated from observations in a training period. The censored EMOS model has been applied to multi-model ensemble forecasts issued on a daily basis over a period of three years. For the two catchments considered, this resulted in well calibrated and sharp forecast distributions over all lead-times from 1 to 114 h. Training observations tended to be better indicators for the dependence structure than the raw ensemble.

The Art and Science of Long-Range Space Weather Forecasting

NASA Technical Reports Server (NTRS)

Hathaway, David H.; Wilson, Robert M.

2006-01-01

Long-range space weather forecasts are akin to seasonal forecasts of terrestrial weather. We don t expect to forecast individual events but we do hope to forecast the underlying level of activity important for satellite operations and mission pl&g. Forecasting space weather conditions years or decades into the future has traditionally been based on empirical models of the solar cycle. Models for the shape of the cycle as a function of its amplitude become reliable once the amplitude is well determined - usually two to three years after minimum. Forecasting the amplitude of a cycle well before that time has been more of an art than a science - usually based on cycle statistics and trends. Recent developments in dynamo theory -the theory explaining the generation of the Sun s magnetic field and the solar activity cycle - have now produced models with predictive capabilities. Testing these models with historical sunspot cycle data indicates that these predictions may be highly reliable one, or even two, cycles into the future.

Randomly correcting model errors in the ARPEGE-Climate v6.1 component of CNRM-CM: applications for seasonal forecasts

NASA Astrophysics Data System (ADS)

Batté, Lauriane; Déqué, Michel

2016-06-01

Stochastic methods are increasingly used in global coupled model climate forecasting systems to account for model uncertainties. In this paper, we describe in more detail the stochastic dynamics technique introduced by Batté and Déqué (2012) in the ARPEGE-Climate atmospheric model. We present new results with an updated version of CNRM-CM using ARPEGE-Climate v6.1, and show that the technique can be used both as a means of analyzing model error statistics and accounting for model inadequacies in a seasonal forecasting framework.The perturbations are designed as corrections of model drift errors estimated from a preliminary weakly nudged re-forecast run over an extended reference period of 34 boreal winter seasons. A detailed statistical analysis of these corrections is provided, and shows that they are mainly made of intra-month variance, thereby justifying their use as in-run perturbations of the model in seasonal forecasts. However, the interannual and systematic error correction terms cannot be neglected. Time correlation of the errors is limited, but some consistency is found between the errors of up to 3 consecutive days.These findings encourage us to test several settings of the random draws of perturbations in seasonal forecast mode. Perturbations are drawn randomly but consistently for all three prognostic variables perturbed. We explore the impact of using monthly mean perturbations throughout a given forecast month in a first ensemble re-forecast (SMM, for stochastic monthly means), and test the use of 5-day sequences of perturbations in a second ensemble re-forecast (S5D, for stochastic 5-day sequences). Both experiments are compared in the light of a REF reference ensemble with initial perturbations only. Results in terms of forecast quality are contrasted depending on the region and variable of interest, but very few areas exhibit a clear degradation of forecasting skill with the introduction of stochastic dynamics. We highlight some positive impacts of the method, mainly on Northern Hemisphere extra-tropics. The 500 hPa geopotential height bias is reduced, and improvements project onto the representation of North Atlantic weather regimes. A modest impact on ensemble spread is found over most regions, which suggests that this method could be complemented by other stochastic perturbation techniques in seasonal forecasting mode.

Improving wave forecasting by integrating ensemble modelling and machine learning

NASA Astrophysics Data System (ADS)

O'Donncha, F.; Zhang, Y.; James, S. C.

2017-12-01

Modern smart-grid networks use technologies to instantly relay information on supply and demand to support effective decision making. Integration of renewable-energy resources with these systems demands accurate forecasting of energy production (and demand) capacities. For wave-energy converters, this requires wave-condition forecasting to enable estimates of energy production. Current operational wave forecasting systems exhibit substantial errors with wave-height RMSEs of 40 to 60 cm being typical, which limits the reliability of energy-generation predictions thereby impeding integration with the distribution grid. In this study, we integrate physics-based models with statistical learning aggregation techniques that combine forecasts from multiple, independent models into a single "best-estimate" prediction of the true state. The Simulating Waves Nearshore physics-based model is used to compute wind- and currents-augmented waves in the Monterey Bay area. Ensembles are developed based on multiple simulations perturbing input data (wave characteristics supplied at the model boundaries and winds) to the model. A learning-aggregation technique uses past observations and past model forecasts to calculate a weight for each model. The aggregated forecasts are compared to observation data to quantify the performance of the model ensemble and aggregation techniques. The appropriately weighted ensemble model outperforms an individual ensemble member with regard to forecasting wave conditions.

Simplified methods for real-time prediction of storm surge uncertainty: The city of Venice case study

NASA Astrophysics Data System (ADS)

Mel, Riccardo; Viero, Daniele Pietro; Carniello, Luca; Defina, Andrea; D'Alpaos, Luigi

2014-09-01

Providing reliable and accurate storm surge forecasts is important for a wide range of problems related to coastal environments. In order to adequately support decision-making processes, it also become increasingly important to be able to estimate the uncertainty associated with the storm surge forecast. The procedure commonly adopted to do this uses the results of a hydrodynamic model forced by a set of different meteorological forecasts; however, this approach requires a considerable, if not prohibitive, computational cost for real-time application. In the present paper we present two simplified methods for estimating the uncertainty affecting storm surge prediction with moderate computational effort. In the first approach we use a computationally fast, statistical tidal model instead of a hydrodynamic numerical model to estimate storm surge uncertainty. The second approach is based on the observation that the uncertainty in the sea level forecast mainly stems from the uncertainty affecting the meteorological fields; this has led to the idea to estimate forecast uncertainty via a linear combination of suitable meteorological variances, directly extracted from the meteorological fields. The proposed methods were applied to estimate the uncertainty in the storm surge forecast in the Venice Lagoon. The results clearly show that the uncertainty estimated through a linear combination of suitable meteorological variances nicely matches the one obtained using the deterministic approach and overcomes some intrinsic limitations in the use of a statistical tidal model.

Evaluation of NMME temperature and precipitation bias and forecast skill for South Asia

NASA Astrophysics Data System (ADS)

Cash, Benjamin A.; Manganello, Julia V.; Kinter, James L.

2017-08-01

Systematic error and forecast skill for temperature and precipitation in two regions of Southern Asia are investigated using hindcasts initialized May 1 from the North American Multi-Model Ensemble. We focus on two contiguous but geographically and dynamically diverse regions: the Extended Indian Monsoon Rainfall (70-100E, 10-30 N) and the nearby mountainous area of Pakistan and Afghanistan (60-75E, 23-39 N). Forecast skill is assessed using the Sign test framework, a rigorous statistical method that can be applied to non-Gaussian variables such as precipitation and to different ensemble sizes without introducing bias. We find that models show significant systematic error in both precipitation and temperature for both regions. The multi-model ensemble mean (MMEM) consistently yields the lowest systematic error and the highest forecast skill for both regions and variables. However, we also find that the MMEM consistently provides a statistically significant increase in skill over climatology only in the first month of the forecast. While the MMEM tends to provide higher overall skill than climatology later in the forecast, the differences are not significant at the 95% level. We also find that MMEMs constructed with a relatively small number of ensemble members per model can equal or outperform MMEMs constructed with more members in skill. This suggests some ensemble members either provide no contribution to overall skill or even detract from it.

Maximizing Statistical Power When Verifying Probabilistic Forecasts of Hydrometeorological Events

NASA Astrophysics Data System (ADS)

DeChant, C. M.; Moradkhani, H.

2014-12-01

Hydrometeorological events (i.e. floods, droughts, precipitation) are increasingly being forecasted probabilistically, owing to the uncertainties in the underlying causes of the phenomenon. In these forecasts, the probability of the event, over some lead time, is estimated based on some model simulations or predictive indicators. By issuing probabilistic forecasts, agencies may communicate the uncertainty in the event occurring. Assuming that the assigned probability of the event is correct, which is referred to as a reliable forecast, the end user may perform some risk management based on the potential damages resulting from the event. Alternatively, an unreliable forecast may give false impressions of the actual risk, leading to improper decision making when protecting resources from extreme events. Due to this requisite for reliable forecasts to perform effective risk management, this study takes a renewed look at reliability assessment in event forecasts. Illustrative experiments will be presented, showing deficiencies in the commonly available approaches (Brier Score, Reliability Diagram). Overall, it is shown that the conventional reliability assessment techniques do not maximize the ability to distinguish between a reliable and unreliable forecast. In this regard, a theoretical formulation of the probabilistic event forecast verification framework will be presented. From this analysis, hypothesis testing with the Poisson-Binomial distribution is the most exact model available for the verification framework, and therefore maximizes one's ability to distinguish between a reliable and unreliable forecast. Application of this verification system was also examined within a real forecasting case study, highlighting the additional statistical power provided with the use of the Poisson-Binomial distribution.

Modelling eWork in Europe: Estimates, Models and Forecasts from the EMERGENCE Project. IES Report.

ERIC Educational Resources Information Center

Bates, P.; Huws, U.

A study combined results of a survey of employers in 18 European countries to establish the extent to which they are currently using eWork with European official statistics to develop models, estimates, and forecasts of the numbers of eWorkers in Europe. These four types of "individual" eWork were identified: telehomeworking;…

A real-time evaluation and demonstration of strategies for 'Over-The-Loop' ensemble streamflow forecasting in US watersheds

NASA Astrophysics Data System (ADS)

Wood, Andy; Clark, Elizabeth; Mendoza, Pablo; Nijssen, Bart; Newman, Andy; Clark, Martyn; Nowak, Kenneth; Arnold, Jeffrey

2017-04-01

Many if not most national operational streamflow prediction systems rely on a forecaster-in-the-loop approach that require the hands-on-effort of an experienced human forecaster. This approach evolved from the need to correct for long-standing deficiencies in the models and datasets used in forecasting, and the practice often leads to skillful flow predictions despite the use of relatively simple, conceptual models. Yet the 'in-the-loop' forecast process is not reproducible, which limits opportunities to assess and incorporate new techniques systematically, and the effort required to make forecasts in this way is an obstacle to expanding forecast services - e.g., though adding new forecast locations or more frequent forecast updates, running more complex models, or producing forecast and hindcasts that can support verification. In the last decade, the hydrologic forecasting community has begun develop more centralized, 'over-the-loop' systems. The quality of these new forecast products will depend on their ability to leverage research in areas including earth system modeling, parameter estimation, data assimilation, statistical post-processing, weather and climate prediction, verification, and uncertainty estimation through the use of ensembles. Currently, many national operational streamflow forecasting and water management communities have little experience with the strengths and weaknesses of over-the-loop approaches, even as such systems are beginning to be deployed operationally in centers such as ECMWF. There is thus a need both to evaluate these forecasting advances and to demonstrate their potential in a public arena, raising awareness in forecast user communities and development programs alike. To address this need, the US National Center for Atmospheric Research is collaborating with the University of Washington, the Bureau of Reclamation and the US Army Corps of Engineers, using the NCAR 'System for Hydromet Analysis Research and Prediction Applications' (SHARP) to implement, assess and demonstrate real-time over-the-loop ensemble flow forecasts in a range of US watersheds. The system relies on fully ensemble techniques, including: an 100-member ensemble of meteorological model forcings and an ensemble particle filter data assimilation for initializing watershed states; analog/regression-based downscaling of ensemble weather forecasts from GEFS; and statistical post-processing of ensemble forecast outputs, all of which run in real-time within a workflow managed by ECWMF's ecFlow libraries over large US regional domains. We describe SHARP and present early hindcast and verification results for short to seasonal range streamflow forecasts in a number of US case study watersheds.

Evaluation of regression and neural network models for solar forecasting over different short-term horizons

DOE PAGES

Inanlouganji, Alireza; Reddy, T. Agami; Katipamula, Srinivas

2018-04-13

Forecasting solar irradiation has acquired immense importance in view of the exponential increase in the number of solar photovoltaic (PV) system installations. In this article, analyses results involving statistical and machine-learning techniques to predict solar irradiation for different forecasting horizons are reported. Yearlong typical meteorological year 3 (TMY3) datasets from three cities in the United States with different climatic conditions have been used in this analysis. A simple forecast approach that assumes consecutive days to be identical serves as a baseline model to compare forecasting alternatives. To account for seasonal variability and to capture short-term fluctuations, different variants of themore » lagged moving average (LMX) model with cloud cover as the input variable are evaluated. Finally, the proposed LMX model is evaluated against an artificial neural network (ANN) model. How the one-hour and 24-hour models can be used in conjunction to predict different short-term rolling horizons is discussed, and this joint application is illustrated for a four-hour rolling horizon forecast scheme. Lastly, the effect of using predicted cloud cover values, instead of measured ones, on the accuracy of the models is assessed. Results show that LMX models do not degrade in forecast accuracy if models are trained with the forecast cloud cover data.« less

Skill of Ensemble Seasonal Probability Forecasts

NASA Astrophysics Data System (ADS)

Smith, Leonard A.; Binter, Roman; Du, Hailiang; Niehoerster, Falk

2010-05-01

In operational forecasting, the computational complexity of large simulation models is, ideally, justified by enhanced performance over simpler models. We will consider probability forecasts and contrast the skill of ENSEMBLES-based seasonal probability forecasts of interest to the finance sector (specifically temperature forecasts for Nino 3.4 and the Atlantic Main Development Region (MDR)). The ENSEMBLES model simulations will be contrasted against forecasts from statistical models based on the observations (climatological distributions) and empirical dynamics based on the observations but conditioned on the current state (dynamical climatology). For some start dates, individual ENSEMBLES models yield significant skill even at a lead-time of 14 months. The nature of this skill is discussed, and chances of application are noted. Questions surrounding the interpretation of probability forecasts based on these multi-model ensemble simulations are then considered; the distributions considered are formed by kernel dressing the ensemble and blending with the climatology. The sources of apparent (RMS) skill in distributions based on multi-model simulations is discussed, and it is demonstrated that the inclusion of "zero-skill" models in the long range can improve Root-Mean-Square-Error scores, casting some doubt on the common justification for the claim that all models should be included in forming an operational probability forecast. It is argued that the rational response varies with lead time.

Evaluation of regression and neural network models for solar forecasting over different short-term horizons

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

Inanlouganji, Alireza; Reddy, T. Agami; Katipamula, Srinivas

Forecasting solar irradiation has acquired immense importance in view of the exponential increase in the number of solar photovoltaic (PV) system installations. In this article, analyses results involving statistical and machine-learning techniques to predict solar irradiation for different forecasting horizons are reported. Yearlong typical meteorological year 3 (TMY3) datasets from three cities in the United States with different climatic conditions have been used in this analysis. A simple forecast approach that assumes consecutive days to be identical serves as a baseline model to compare forecasting alternatives. To account for seasonal variability and to capture short-term fluctuations, different variants of themore » lagged moving average (LMX) model with cloud cover as the input variable are evaluated. Finally, the proposed LMX model is evaluated against an artificial neural network (ANN) model. How the one-hour and 24-hour models can be used in conjunction to predict different short-term rolling horizons is discussed, and this joint application is illustrated for a four-hour rolling horizon forecast scheme. Lastly, the effect of using predicted cloud cover values, instead of measured ones, on the accuracy of the models is assessed. Results show that LMX models do not degrade in forecast accuracy if models are trained with the forecast cloud cover data.« less

Scenarios for Evolving Seismic Crises: Possible Communication Strategies

NASA Astrophysics Data System (ADS)

Steacy, S.

2015-12-01

Recent advances in operational earthquake forecasting mean that we are very close to being able to confidently compute changes in earthquake probability as seismic crises develop. For instance, we now have statistical models such as ETAS and STEP which demonstrate considerable skill in forecasting earthquake rates and recent advances in Coulomb based models are also showing much promise. Communicating changes in earthquake probability is likely be very difficult, however, as the absolute probability of a damaging event is likely to remain quite small despite a significant increase in the relative value. Here, we use a hybrid Coulomb/statistical model to compute probability changes for a series of earthquake scenarios in New Zealand. We discuss the strengths and limitations of the forecasts and suggest a number of possible mechanisms that might be used to communicate results in an actual developing seismic crisis.

Using Bayes Model Averaging for Wind Power Forecasts

NASA Astrophysics Data System (ADS)

Preede Revheim, Pål; Beyer, Hans Georg

2014-05-01

For operational purposes predictions of the forecasts of the lumped output of groups of wind farms spread over larger geographic areas will often be of interest. A naive approach is to make forecasts for each individual site and sum them up to get the group forecast. It is however well documented that a better choice is to use a model that also takes advantage of spatial smoothing effects. It might however be the case that some sites tends to more accurately reflect the total output of the region, either in general or for certain wind directions. It will then be of interest giving these a greater influence over the group forecast. Bayesian model averaging (BMA) is a statistical post-processing method for producing probabilistic forecasts from ensembles. Raftery et al. [1] show how BMA can be used for statistical post processing of forecast ensembles, producing PDFs of future weather quantities. The BMA predictive PDF of a future weather quantity is a weighted average of the ensemble members' PDFs, where the weights can be interpreted as posterior probabilities and reflect the ensemble members' contribution to overall forecasting skill over a training period. In Revheim and Beyer [2] the BMA procedure used in Sloughter, Gneiting and Raftery [3] were found to produce fairly accurate PDFs for the future mean wind speed of a group of sites from the single sites wind speeds. However, when the procedure was attempted applied to wind power it resulted in either problems with the estimation of the parameters (mainly caused by longer consecutive periods of no power production) or severe underestimation (mainly caused by problems with reflecting the power curve). In this paper the problems that arose when applying BMA to wind power forecasting is met through two strategies. First, the BMA procedure is run with a combination of single site wind speeds and single site wind power production as input. This solves the problem with longer consecutive periods where the input data does not contain information, but it has the disadvantage of nearly doubling the number of model parameters to be estimated. Second, the BMA procedure is run with group mean wind power as the response variable instead of group mean wind speed. This also solves the problem with longer consecutive periods without information in the input data, but it leaves the power curve to also be estimated from the data. [1] Raftery, A. E., et al. (2005). Using Bayesian Model Averaging to Calibrate Forecast Ensembles. Monthly Weather Review, 133, 1155-1174. [2]Revheim, P. P. and H. G. Beyer (2013). Using Bayesian Model Averaging for wind farm group forecasts. EWEA Wind Power Forecasting Technology Workshop,Rotterdam, 4-5 December 2013. [3]Sloughter, J. M., T. Gneiting and A. E. Raftery (2010). Probabilistic Wind Speed Forecasting Using Ensembles and Bayesian Model Averaging. Journal of the American Statistical Association, Vol. 105, No. 489, 25-35

The epistemological status of general circulation models

NASA Astrophysics Data System (ADS)

Loehle, Craig

2018-03-01

Forecasts of both likely anthropogenic effects on climate and consequent effects on nature and society are based on large, complex software tools called general circulation models (GCMs). Forecasts generated by GCMs have been used extensively in policy decisions related to climate change. However, the relation between underlying physical theories and results produced by GCMs is unclear. In the case of GCMs, many discretizations and approximations are made, and simulating Earth system processes is far from simple and currently leads to some results with unknown energy balance implications. Statistical testing of GCM forecasts for degree of agreement with data would facilitate assessment of fitness for use. If model results need to be put on an anomaly basis due to model bias, then both visual and quantitative measures of model fit depend strongly on the reference period used for normalization, making testing problematic. Epistemology is here applied to problems of statistical inference during testing, the relationship between the underlying physics and the models, the epistemic meaning of ensemble statistics, problems of spatial and temporal scale, the existence or not of an unforced null for climate fluctuations, the meaning of existing uncertainty estimates, and other issues. Rigorous reasoning entails carefully quantifying levels of uncertainty.

Seasonal Atmospheric and Oceanic Predictions

NASA Technical Reports Server (NTRS)

Roads, John; Rienecker, Michele (Technical Monitor)

2003-01-01

Several projects associated with dynamical, statistical, single column, and ocean models are presented. The projects include: 1) Regional Climate Modeling; 2) Statistical Downscaling; 3) Evaluation of SCM and NSIPP AGCM Results at the ARM Program Sites; and 4) Ocean Forecasts.

Modeling and forecasting the distribution of Vibrio vulnificus in Chesapeake Bay.

PubMed

Jacobs, J M; Rhodes, M; Brown, C W; Hood, R R; Leight, A; Long, W; Wood, R

2014-11-01

To construct statistical models to predict the presence, abundance and potential virulence of Vibrio vulnificus in surface waters of Chesapeake Bay for implementation in ecological forecasting systems. We evaluated and applied previously published qPCR assays to water samples (n = 1636) collected from Chesapeake Bay from 2007-2010 in conjunction with State water quality monitoring programmes. A variety of statistical techniques were used in concert to identify water quality parameters associated with V. vulnificus presence, abundance and virulence markers in the interest of developing strong predictive models for use in regional oceanographic modeling systems. A suite of models are provided to represent the best model fit and alternatives using environmental variables that allow them to be put to immediate use in current ecological forecasting efforts. Environmental parameters such as temperature, salinity and turbidity are capable of accurately predicting abundance and distribution of V. vulnificus in Chesapeake Bay. Forcing these empirical models with output from ocean modeling systems allows for spatially explicit forecasts for up to 48 h in the future. This study uses one of the largest data sets compiled to model Vibrio in an estuary, enhances our understanding of environmental correlates with abundance, distribution and presence of potentially virulent strains and offers a method to forecast these pathogens that may be replicated in other regions. This article has been contributed to by US Government employees and their work is in the public domain in the USA.

Evaluation of annual, global seismicity forecasts, including ensemble models

NASA Astrophysics Data System (ADS)

Taroni, Matteo; Zechar, Jeremy; Marzocchi, Warner

2013-04-01

In 2009, the Collaboratory for the Study of the Earthquake Predictability (CSEP) initiated a prototype global earthquake forecast experiment. Three models participated in this experiment for 2009, 2010 and 2011—each model forecast the number of earthquakes above magnitude 6 in 1x1 degree cells that span the globe. Here we use likelihood-based metrics to evaluate the consistency of the forecasts with the observed seismicity. We compare model performance with statistical tests and a new method based on the peer-to-peer gambling score. The results of the comparisons are used to build ensemble models that are a weighted combination of the individual models. Notably, in these experiments the ensemble model always performs significantly better than the single best-performing model. Our results indicate the following: i) time-varying forecasts, if not updated after each major shock, may not provide significant advantages with respect to time-invariant models in 1-year forecast experiments; ii) the spatial distribution seems to be the most important feature to characterize the different forecasting performances of the models; iii) the interpretation of consistency tests may be misleading because some good models may be rejected while trivial models may pass consistency tests; iv) a proper ensemble modeling seems to be a valuable procedure to get the best performing model for practical purposes.

Value versus Accuracy: application of seasonal forecasts to a hydro-economic optimization model for the Sudanese Blue Nile

NASA Astrophysics Data System (ADS)

Satti, S.; Zaitchik, B. F.; Siddiqui, S.; Badr, H. S.; Shukla, S.; Peters-Lidard, C. D.

2015-12-01

The unpredictable nature of precipitation within the East African (EA) region makes it one of the most vulnerable, food insecure regions in the world. There is a vital need for forecasts to inform decision makers, both local and regional, and to help formulate the region's climate change adaptation strategies. Here, we present a suite of different seasonal forecast models, both statistical and dynamical, for the EA region. Objective regionalization is performed for EA on the basis of interannual variability in precipitation in both observations and models. This regionalization is applied as the basis for calculating a number of standard skill scores to evaluate each model's forecast accuracy. A dynamically linked Land Surface Model (LSM) is then applied to determine forecasted flows, which drive the Sudanese Hydroeconomic Optimization Model (SHOM). SHOM combines hydrologic, agronomic and economic inputs to determine the optimal decisions that maximize economic benefits along the Sudanese Blue Nile. This modeling sequence is designed to derive the potential added value of information of each forecasting model to agriculture and hydropower management. A rank of each model's forecasting skill score along with its added value of information is analyzed in order compare the performance of each forecast. This research aims to improve understanding of how characteristics of accuracy, lead time, and uncertainty of seasonal forecasts influence their utility to water resources decision makers who utilize them.

Forecasting coconut production in the Philippines with ARIMA model

NASA Astrophysics Data System (ADS)

Lim, Cristina Teresa

2015-02-01

The study aimed to depict the situation of the coconut industry in the Philippines for the future years applying Autoregressive Integrated Moving Average (ARIMA) method. Data on coconut production, one of the major industrial crops of the country, for the period of 1990 to 2012 were analyzed using time-series methods. Autocorrelation (ACF) and partial autocorrelation functions (PACF) were calculated for the data. Appropriate Box-Jenkins autoregressive moving average model was fitted. Validity of the model was tested using standard statistical techniques. The forecasting power of autoregressive moving average (ARMA) model was used to forecast coconut production for the eight leading years.

Total probabilities of ensemble runoff forecasts

NASA Astrophysics Data System (ADS)

Olav Skøien, Jon; Bogner, Konrad; Salamon, Peter; Smith, Paul; Pappenberger, Florian

2017-04-01

Ensemble forecasting has a long history from meteorological modelling, as an indication of the uncertainty of the forecasts. However, it is necessary to calibrate and post-process the ensembles as the they often exhibit both bias and dispersion errors. Two of the most common methods for this are Bayesian Model Averaging (Raftery et al., 2005) and Ensemble Model Output Statistics (EMOS) (Gneiting et al., 2005). There are also methods for regionalizing these methods (Berrocal et al., 2007) and for incorporating the correlation between lead times (Hemri et al., 2013). Engeland and Steinsland Engeland and Steinsland (2014) developed a framework which can estimate post-processing parameters varying in space and time, while giving a spatially and temporally consistent output. However, their method is computationally complex for our larger number of stations, which makes it unsuitable for our purpose. Our post-processing method of the ensembles is developed in the framework of the European Flood Awareness System (EFAS - http://www.efas.eu), where we are making forecasts for whole Europe, and based on observations from around 700 catchments. As the target is flood forecasting, we are also more interested in improving the forecast skill for high-flows rather than in a good prediction of the entire flow regime. EFAS uses a combination of ensemble forecasts and deterministic forecasts from different meteorological forecasters to force a distributed hydrologic model and to compute runoff ensembles for each river pixel within the model domain. Instead of showing the mean and the variability of each forecast ensemble individually, we will now post-process all model outputs to estimate the total probability, the post-processed mean and uncertainty of all ensembles. The post-processing parameters are first calibrated for each calibration location, but we are adding a spatial penalty in the calibration process to force a spatial correlation of the parameters. The penalty takes distance, stream-connectivity and size of the catchment areas into account. This can in some cases have a slight negative impact on the calibration error, but avoids large differences between parameters of nearby locations, whether stream connected or not. The spatial calibration also makes it easier to interpolate the post-processing parameters to uncalibrated locations. We also look into different methods for handling the non-normal distributions of runoff data and the effect of different data transformations on forecasts skills in general and for floods in particular. Berrocal, V. J., Raftery, A. E. and Gneiting, T.: Combining Spatial Statistical and Ensemble Information in Probabilistic Weather Forecasts, Mon. Weather Rev., 135(4), 1386-1402, doi:10.1175/MWR3341.1, 2007. Engeland, K. and Steinsland, I.: Probabilistic postprocessing models for flow forecasts for a system of catchments and several lead times, Water Resour. Res., 50(1), 182-197, doi:10.1002/2012WR012757, 2014. Gneiting, T., Raftery, A. E., Westveld, A. H. and Goldman, T.: Calibrated Probabilistic Forecasting Using Ensemble Model Output Statistics and Minimum CRPS Estimation, Mon. Weather Rev., 133(5), 1098-1118, doi:10.1175/MWR2904.1, 2005. Hemri, S., Fundel, F. and Zappa, M.: Simultaneous calibration of ensemble river flow predictions over an entire range of lead times, Water Resour. Res., 49(10), 6744-6755, doi:10.1002/wrcr.20542, 2013. Raftery, A. E., Gneiting, T., Balabdaoui, F. and Polakowski, M.: Using Bayesian Model Averaging to Calibrate Forecast Ensembles, Mon. Weather Rev., 133(5), 1155-1174, doi:10.1175/MWR2906.1, 2005.

Total probabilities of ensemble runoff forecasts

NASA Astrophysics Data System (ADS)

Olav Skøien, Jon; Bogner, Konrad; Salamon, Peter; Smith, Paul; Pappenberger, Florian

2016-04-01

Ensemble forecasting has for a long time been used as a method in meteorological modelling to indicate the uncertainty of the forecasts. However, as the ensembles often exhibit both bias and dispersion errors, it is necessary to calibrate and post-process them. Two of the most common methods for this are Bayesian Model Averaging (Raftery et al., 2005) and Ensemble Model Output Statistics (EMOS) (Gneiting et al., 2005). There are also methods for regionalizing these methods (Berrocal et al., 2007) and for incorporating the correlation between lead times (Hemri et al., 2013). Engeland and Steinsland Engeland and Steinsland (2014) developed a framework which can estimate post-processing parameters which are different in space and time, but still can give a spatially and temporally consistent output. However, their method is computationally complex for our larger number of stations, and cannot directly be regionalized in the way we would like, so we suggest a different path below. The target of our work is to create a mean forecast with uncertainty bounds for a large number of locations in the framework of the European Flood Awareness System (EFAS - http://www.efas.eu) We are therefore more interested in improving the forecast skill for high-flows rather than the forecast skill of lower runoff levels. EFAS uses a combination of ensemble forecasts and deterministic forecasts from different forecasters to force a distributed hydrologic model and to compute runoff ensembles for each river pixel within the model domain. Instead of showing the mean and the variability of each forecast ensemble individually, we will now post-process all model outputs to find a total probability, the post-processed mean and uncertainty of all ensembles. The post-processing parameters are first calibrated for each calibration location, but assuring that they have some spatial correlation, by adding a spatial penalty in the calibration process. This can in some cases have a slight negative impact on the calibration error, but makes it easier to interpolate the post-processing parameters to uncalibrated locations. We also look into different methods for handling the non-normal distributions of runoff data and the effect of different data transformations on forecasts skills in general and for floods in particular. Berrocal, V. J., Raftery, A. E. and Gneiting, T.: Combining Spatial Statistical and Ensemble Information in Probabilistic Weather Forecasts, Mon. Weather Rev., 135(4), 1386-1402, doi:10.1175/MWR3341.1, 2007. Engeland, K. and Steinsland, I.: Probabilistic postprocessing models for flow forecasts for a system of catchments and several lead times, Water Resour. Res., 50(1), 182-197, doi:10.1002/2012WR012757, 2014. Gneiting, T., Raftery, A. E., Westveld, A. H. and Goldman, T.: Calibrated Probabilistic Forecasting Using Ensemble Model Output Statistics and Minimum CRPS Estimation, Mon. Weather Rev., 133(5), 1098-1118, doi:10.1175/MWR2904.1, 2005. Hemri, S., Fundel, F. and Zappa, M.: Simultaneous calibration of ensemble river flow predictions over an entire range of lead times, Water Resour. Res., 49(10), 6744-6755, doi:10.1002/wrcr.20542, 2013. Raftery, A. E., Gneiting, T., Balabdaoui, F. and Polakowski, M.: Using Bayesian Model Averaging to Calibrate Forecast Ensembles, Mon. Weather Rev., 133(5), 1155-1174, doi:10.1175/MWR2906.1, 2005.

Forecasting infectious disease emergence subject to seasonal forcing.

PubMed

Miller, Paige B; O'Dea, Eamon B; Rohani, Pejman; Drake, John M

2017-09-06

Despite high vaccination coverage, many childhood infections pose a growing threat to human populations. Accurate disease forecasting would be of tremendous value to public health. Forecasting disease emergence using early warning signals (EWS) is possible in non-seasonal models of infectious diseases. Here, we assessed whether EWS also anticipate disease emergence in seasonal models. We simulated the dynamics of an immunizing infectious pathogen approaching the tipping point to disease endemicity. To explore the effect of seasonality on the reliability of early warning statistics, we varied the amplitude of fluctuations around the average transmission. We proposed and analyzed two new early warning signals based on the wavelet spectrum. We measured the reliability of the early warning signals depending on the strength of their trend preceding the tipping point and then calculated the Area Under the Curve (AUC) statistic. Early warning signals were reliable when disease transmission was subject to seasonal forcing. Wavelet-based early warning signals were as reliable as other conventional early warning signals. We found that removing seasonal trends, prior to analysis, did not improve early warning statistics uniformly. Early warning signals anticipate the onset of critical transitions for infectious diseases which are subject to seasonal forcing. Wavelet-based early warning statistics can also be used to forecast infectious disease.

Evolving forecasting classifications and applications in health forecasting

PubMed Central

Soyiri, Ireneous N; Reidpath, Daniel D

2012-01-01

Health forecasting forewarns the health community about future health situations and disease episodes so that health systems can better allocate resources and manage demand. The tools used for developing and measuring the accuracy and validity of health forecasts commonly are not defined although they are usually adapted forms of statistical procedures. This review identifies previous typologies used in classifying the forecasting methods commonly used in forecasting health conditions or situations. It then discusses the strengths and weaknesses of these methods and presents the choices available for measuring the accuracy of health-forecasting models, including a note on the discrepancies in the modes of validation. PMID:22615533

Skill of Global Raw and Postprocessed Ensemble Predictions of Rainfall over Northern Tropical Africa

NASA Astrophysics Data System (ADS)

Vogel, Peter; Knippertz, Peter; Fink, Andreas H.; Schlueter, Andreas; Gneiting, Tilmann

2018-04-01

Accumulated precipitation forecasts are of high socioeconomic importance for agriculturally dominated societies in northern tropical Africa. In this study, we analyze the performance of nine operational global ensemble prediction systems (EPSs) relative to climatology-based forecasts for 1 to 5-day accumulated precipitation based on the monsoon seasons 2007-2014 for three regions within northern tropical Africa. To assess the full potential of raw ensemble forecasts across spatial scales, we apply state-of-the-art statistical postprocessing methods in form of Bayesian Model Averaging (BMA) and Ensemble Model Output Statistics (EMOS), and verify against station and spatially aggregated, satellite-based gridded observations. Raw ensemble forecasts are uncalibrated, unreliable, and underperform relative to climatology, independently of region, accumulation time, monsoon season, and ensemble. Differences between raw ensemble and climatological forecasts are large, and partly stem from poor prediction for low precipitation amounts. BMA and EMOS postprocessed forecasts are calibrated, reliable, and strongly improve on the raw ensembles, but - somewhat disappointingly - typically do not outperform climatology. Most EPSs exhibit slight improvements over the period 2007-2014, but overall have little added value compared to climatology. We suspect that the parametrization of convection is a potential cause for the sobering lack of ensemble forecast skill in a region dominated by mesoscale convective systems.

Surface wave effect on the upper ocean in marine forecast

NASA Astrophysics Data System (ADS)

Wang, Guansuo; Qiao, Fangli; Xia, Changshui; Zhao, Chang

2015-04-01

An Operational Coupled Forecast System for the seas off China and adjacent (OCFS-C) is constructed based on the paralleled wave-circulation coupled model, which is tested with comprehensive experiments and operational since November 1st, 2007. The main feature of the system is that the wave-induced mixing is considered in circulation model. Daily analyses and three day forecasts of three-dimensional temperature, salinity, currents and wave height are produced. Coverage is global at 1/2 degreed resolution with nested models up to 1/24 degree resolution in China Sea. Daily remote sensing sea surface temperatures (SST) are taken to relax to an analytical product as hot restarting fields for OCFS-C by the Nudging techniques. Forecasting-data inter-comparisons are performed to measure the effectiveness of OCFS-C in predicting upper-ocean quantities including SST, mixed layer depth (MLD) and subsurface temperature. The variety of performance with lead time and real-time is discussed as well using the daily statistic results for SST between forecast and satellite data. Several buoy observations and many Argo profiles are used for this validation. Except the conventional statistical metrics, non-dimension skill scores (SS) is taken to estimate forecast skill. Model SST comparisons with more one year-long SST time series from 2 buoys given a large SS value (more than 0.90). And skill in predicting the seasonal variability of SST is confirmed. Model subsurface temperature comparisons with that from a lot of Argo profiles indicated that OCFS-C has low skill in predicting subsurface temperatures between 80m and 120m. Inter-comparisons of MLD reveal that MLD from model is shallower than that from Argo profiles by about 12m. QCFS-C is successful and steady in predicting MLD. The daily statistic results for SST between 1-d, 2-d and 3-d forecast and data is adopted to describe variability of Skill in predicting SST with lead time or real time. In a word QCFS-C shows reasonable accuracy over a series of studies designed to test ability to predict upper ocean conditions.

Forecasting influenza in Hong Kong with Google search queries and statistical model fusion.

PubMed

Xu, Qinneng; Gel, Yulia R; Ramirez Ramirez, L Leticia; Nezafati, Kusha; Zhang, Qingpeng; Tsui, Kwok-Leung

2017-01-01

The objective of this study is to investigate predictive utility of online social media and web search queries, particularly, Google search data, to forecast new cases of influenza-like-illness (ILI) in general outpatient clinics (GOPC) in Hong Kong. To mitigate the impact of sensitivity to self-excitement (i.e., fickle media interest) and other artifacts of online social media data, in our approach we fuse multiple offline and online data sources. Four individual models: generalized linear model (GLM), least absolute shrinkage and selection operator (LASSO), autoregressive integrated moving average (ARIMA), and deep learning (DL) with Feedforward Neural Networks (FNN) are employed to forecast ILI-GOPC both one week and two weeks in advance. The covariates include Google search queries, meteorological data, and previously recorded offline ILI. To our knowledge, this is the first study that introduces deep learning methodology into surveillance of infectious diseases and investigates its predictive utility. Furthermore, to exploit the strength from each individual forecasting models, we use statistical model fusion, using Bayesian model averaging (BMA), which allows a systematic integration of multiple forecast scenarios. For each model, an adaptive approach is used to capture the recent relationship between ILI and covariates. DL with FNN appears to deliver the most competitive predictive performance among the four considered individual models. Combing all four models in a comprehensive BMA framework allows to further improve such predictive evaluation metrics as root mean squared error (RMSE) and mean absolute predictive error (MAPE). Nevertheless, DL with FNN remains the preferred method for predicting locations of influenza peaks. The proposed approach can be viewed a feasible alternative to forecast ILI in Hong Kong or other countries where ILI has no constant seasonal trend and influenza data resources are limited. The proposed methodology is easily tractable and computationally efficient.

Monthly mean forecast experiments with the GISS model

NASA Technical Reports Server (NTRS)

Spar, J.; Atlas, R. M.; Kuo, E.

1976-01-01

The GISS general circulation model was used to compute global monthly mean forecasts for January 1973, 1974, and 1975 from initial conditions on the first day of each month and constant sea surface temperatures. Forecasts were evaluated in terms of global and hemispheric energetics, zonally averaged meridional and vertical profiles, forecast error statistics, and monthly mean synoptic fields. Although it generated a realistic mean meridional structure, the model did not adequately reproduce the observed interannual variations in the large scale monthly mean energetics and zonally averaged circulation. The monthly mean sea level pressure field was not predicted satisfactorily, but annual changes in the Icelandic low were simulated. The impact of temporal sea surface temperature variations on the forecasts was investigated by comparing two parallel forecasts for January 1974, one using climatological ocean temperatures and the other observed daily ocean temperatures. The use of daily updated sea surface temperatures produced no discernible beneficial effect.

Investigating market efficiency through a forecasting model based on differential equations

NASA Astrophysics Data System (ADS)

de Resende, Charlene C.; Pereira, Adriano C. M.; Cardoso, Rodrigo T. N.; de Magalhães, A. R. Bosco

2017-05-01

A new differential equation based model for stock price trend forecast is proposed as a tool to investigate efficiency in an emerging market. Its predictive power showed statistically to be higher than the one of a completely random model, signaling towards the presence of arbitrage opportunities. Conditions for accuracy to be enhanced are investigated, and application of the model as part of a trading strategy is discussed.

Forecasting United States heartworm Dirofilaria immitis prevalence in dogs.

PubMed

Bowman, Dwight D; Liu, Yan; McMahan, Christopher S; Nordone, Shila K; Yabsley, Michael J; Lund, Robert B

2016-10-10

This paper forecasts next year's canine heartworm prevalence in the United States from 16 climate, geographic and societal factors. The forecast's construction and an assessment of its performance are described. The forecast is based on a spatial-temporal conditional autoregressive model fitted to over 31 million antigen heartworm tests conducted in the 48 contiguous United States during 2011-2015. The forecast uses county-level data on 16 predictive factors, including temperature, precipitation, median household income, local forest and surface water coverage, and presence/absence of eight mosquito species. Non-static factors are extrapolated into the forthcoming year with various statistical methods. The fitted model and factor extrapolations are used to estimate next year's regional prevalence. The correlation between the observed and model-estimated county-by-county heartworm prevalence for the 5-year period 2011-2015 is 0.727, demonstrating reasonable model accuracy. The correlation between 2015 observed and forecasted county-by-county heartworm prevalence is 0.940, demonstrating significant skill and showing that heartworm prevalence can be forecasted reasonably accurately. The forecast presented herein can a priori alert veterinarians to areas expected to see higher than normal heartworm activity. The proposed methods may prove useful for forecasting other diseases.

Forecasting Performance of Grey Prediction for Education Expenditure and School Enrollment

ERIC Educational Resources Information Center

Tang, Hui-Wen Vivian; Yin, Mu-Shang

2012-01-01

GM(1,1) and GM(1,1) rolling models derived from grey system theory were estimated using time-series data from projection studies by National Center for Education Statistics (NCES). An out-of-sample forecasting competition between the two grey prediction models and exponential smoothing used by NCES was conducted for education expenditure and…

Wind-Farm Forecasting Using the HARMONIE Weather Forecast Model and Bayes Model Averaging for Bias Removal.

NASA Astrophysics Data System (ADS)

O'Brien, Enda; McKinstry, Alastair; Ralph, Adam

2015-04-01

Building on previous work presented at EGU 2013 (http://www.sciencedirect.com/science/article/pii/S1876610213016068 ), more results are available now from a different wind-farm in complex terrain in southwest Ireland. The basic approach is to interpolate wind-speed forecasts from an operational weather forecast model (i.e., HARMONIE in the case of Ireland) to the precise location of each wind-turbine, and then use Bayes Model Averaging (BMA; with statistical information collected from a prior training-period of e.g., 25 days) to remove systematic biases. Bias-corrected wind-speed forecasts (and associated power-generation forecasts) are then provided twice daily (at 5am and 5pm) out to 30 hours, with each forecast validation fed back to BMA for future learning. 30-hr forecasts from the operational Met Éireann HARMONIE model at 2.5km resolution have been validated against turbine SCADA observations since Jan. 2014. An extra high-resolution (0.5km grid-spacing) HARMONIE configuration has been run since Nov. 2014 as an extra member of the forecast "ensemble". A new version of HARMONIE with extra filters designed to stabilize high-resolution configurations has been run since Jan. 2015. Measures of forecast skill and forecast errors will be provided, and the contributions made by the various physical and computational enhancements to HARMONIE will be quantified.

Wind power application research on the fusion of the determination and ensemble prediction

NASA Astrophysics Data System (ADS)

Lan, Shi; Lina, Xu; Yuzhu, Hao

2017-07-01

The fused product of wind speed for the wind farm is designed through the use of wind speed products of ensemble prediction from the European Centre for Medium-Range Weather Forecasts (ECMWF) and professional numerical model products on wind power based on Mesoscale Model5 (MM5) and Beijing Rapid Update Cycle (BJ-RUC), which are suitable for short-term wind power forecasting and electric dispatch. The single-valued forecast is formed by calculating the different ensemble statistics of the Bayesian probabilistic forecasting representing the uncertainty of ECMWF ensemble prediction. Using autoregressive integrated moving average (ARIMA) model to improve the time resolution of the single-valued forecast, and based on the Bayesian model averaging (BMA) and the deterministic numerical model prediction, the optimal wind speed forecasting curve and the confidence interval are provided. The result shows that the fusion forecast has made obvious improvement to the accuracy relative to the existing numerical forecasting products. Compared with the 0-24 h existing deterministic forecast in the validation period, the mean absolute error (MAE) is decreased by 24.3 % and the correlation coefficient (R) is increased by 12.5 %. In comparison with the ECMWF ensemble forecast, the MAE is reduced by 11.7 %, and R is increased 14.5 %. Additionally, MAE did not increase with the prolongation of the forecast ahead.

Statistical Post-Processing of Wind Speed Forecasts to Estimate Relative Economic Value

NASA Astrophysics Data System (ADS)

Courtney, Jennifer; Lynch, Peter; Sweeney, Conor

2013-04-01

The objective of this research is to get the best possible wind speed forecasts for the wind energy industry by using an optimal combination of well-established forecasting and post-processing methods. We start with the ECMWF 51 member ensemble prediction system (EPS) which is underdispersive and hence uncalibrated. We aim to produce wind speed forecasts that are more accurate and calibrated than the EPS. The 51 members of the EPS are clustered to 8 weighted representative members (RMs), chosen to minimize the within-cluster spread, while maximizing the inter-cluster spread. The forecasts are then downscaled using two limited area models, WRF and COSMO, at two resolutions, 14km and 3km. This process creates four distinguishable ensembles which are used as input to statistical post-processes requiring multi-model forecasts. Two such processes are presented here. The first, Bayesian Model Averaging, has been proven to provide more calibrated and accurate wind speed forecasts than the ECMWF EPS using this multi-model input data. The second, heteroscedastic censored regression is indicating positive results also. We compare the two post-processing methods, applied to a year of hindcast wind speed data around Ireland, using an array of deterministic and probabilistic verification techniques, such as MAE, CRPS, probability transform integrals and verification rank histograms, to show which method provides the most accurate and calibrated forecasts. However, the value of a forecast to an end-user cannot be fully quantified by just the accuracy and calibration measurements mentioned, as the relationship between skill and value is complex. Capturing the full potential of the forecast benefits also requires detailed knowledge of the end-users' weather sensitive decision-making processes and most importantly the economic impact it will have on their income. Finally, we present the continuous relative economic value of both post-processing methods to identify which is more beneficial to the wind energy industry of Ireland.

Forecasting volatility with neural regression: a contribution to model adequacy.

PubMed

Refenes, A N; Holt, W T

2001-01-01

Neural nets' usefulness for forecasting is limited by problems of overfitting and the lack of rigorous procedures for model identification, selection and adequacy testing. This paper describes a methodology for neural model misspecification testing. We introduce a generalization of the Durbin-Watson statistic for neural regression and discuss the general issues of misspecification testing using residual analysis. We derive a generalized influence matrix for neural estimators which enables us to evaluate the distribution of the statistic. We deploy Monte Carlo simulation to compare the power of the test for neural and linear regressors. While residual testing is not a sufficient condition for model adequacy, it is nevertheless a necessary condition to demonstrate that the model is a good approximation to the data generating process, particularly as neural-network estimation procedures are susceptible to partial convergence. The work is also an important step toward developing rigorous procedures for neural model identification, selection and adequacy testing which have started to appear in the literature. We demonstrate its applicability in the nontrivial problem of forecasting implied volatility innovations using high-frequency stock index options. Each step of the model building process is validated using statistical tests to verify variable significance and model adequacy with the results confirming the presence of nonlinear relationships in implied volatility innovations.

Determining relevant parameters for a statistical tropical cyclone genesis tool based upon global model output

NASA Astrophysics Data System (ADS)

Halperin, D.; Hart, R. E.; Fuelberg, H. E.; Cossuth, J.

2013-12-01

Predicting tropical cyclone (TC) genesis has been a vexing problem for forecasters. While the literature describes environmental conditions which are necessary for TC genesis, predicting if and when a specific disturbance will organize and become a TC remains a challenge. As recently as 5-10 years ago, global models possessed little if any skill in forecasting TC genesis. However, due to increased resolution and more advanced model parameterizations, we have reached the point where global models can provide useful TC genesis guidance to operational forecasters. A recent study evaluated five global models' ability to predict TC genesis out to four days over the North Atlantic basin (Halperin et al. 2013). The results indicate that the models are indeed able to capture the genesis time and location correctly a fair percentage of the time. The study also uncovered model biases. For example, probability of detection and false alarm rate varies spatially within the basin. Also, as expected, the models' performance decreases with increasing lead time. In order to explain these and other biases, it is useful to analyze the model-indicated genesis events further to determine whether or not there are systematic differences between successful forecasts (hits), false alarms, and miss events. This study will examine composites of a number of physically-relevant environmental parameters (e.g., magnitude of vertical wind shear, aerially averaged mid-level relative humidity) and disturbance-based parameters (e.g., 925 hPa maximum wind speed, vertical alignment of relative vorticity) among each TC genesis event classification (i.e., hit, false alarm, miss). We will use standard statistical tests (e.g., Student's t test, Mann-Whitney-U Test) to calculate whether or not any differences are statistically significant. We also plan to discuss how these composite results apply to a few illustrative case studies. The results may help determine which aspects of the forecast are (in)correct and whether the incorrect aspects can be bias-corrected. This, in turn, may allow us to further enhance probabilistic forecasts of TC genesis.

Evaluating NMME Seasonal Forecast Skill for use in NASA SERVIR Hub Regions

NASA Technical Reports Server (NTRS)

Roberts, J. Brent; Roberts, Franklin R.

2013-01-01

The U.S. National Multi-Model Ensemble seasonal forecasting system is providing hindcast and real-time data streams to be used in assessing and improving seasonal predictive capacity. The coupled forecasts have numerous potential applications, both national and international in scope. The NASA / USAID SERVIR project, which leverages satellite and modeling-based resources for environmental decision making in developing nations, is focusing on the evaluation of NMME forecasts specifically for use in driving applications models in hub regions including East Africa, the Hindu Kush- Himalayan (HKH) region and Mesoamerica. A prerequisite for seasonal forecast use in application modeling (e.g. hydrology, agriculture) is bias correction and skill assessment. Efforts to address systematic biases and multi-model combination in support of NASA SERVIR impact modeling requirements will be highlighted. Specifically, quantilequantile mapping for bias correction has been implemented for all archived NMME hindcasts. Both deterministic and probabilistic skill estimates for raw, bias-corrected, and multi-model ensemble forecasts as a function of forecast lead will be presented for temperature and precipitation. Complementing this statistical assessment will be case studies of significant events, for example, the ability of the NMME forecasts suite to anticipate the 2010/2011 drought in the Horn of Africa and its relationship to evolving SST patterns.

Northern Hemisphere climate trends in reanalysis and forecast model predictions: The 500 hPa annual means

NASA Astrophysics Data System (ADS)

Bordi, I.; Fraedrich, K.; Sutera, A.

2010-06-01

The lead time dependent climates of the ECMWF weather prediction model, initialized with ERA-40 reanalysis, are analysed using 44 years of day-1 to day-10 forecasts of the northern hemispheric 500-hPa geopotential height fields. The study addresses the question whether short-term tendencies have an impact on long-term trends. Comparing climate trends of ERA-40 with those of the forecasts, it seems that the forecast model rapidly loses the memory of initial conditions creating its own climate. All forecast trends show a high degree of consistency. Comparison results suggest that: (i) Only centers characterized by an upward trend are statistical significant when increasing the lead time. (ii) In midilatitudes an upward trend larger than the one observed in the reanalysis characterizes the forecasts, while in the tropics there is a good agreement. (iii) The downward trend in reanalysis at high latitudes characterizes also the day-1 forecast which, however, increasing lead time approaches zero.

National Centers for Environmental Prediction

Science.gov Websites

Statistics Observational Data Processing Data Assimilation Monsoon Desk Model Transition Seminars Seminar (CFS) HURRICANE WEATHER RESEARCH and FORECASTING (HWRF) GLOBAL ENSEMBLE FORECAST SYSTEM (GEFS) NATIONAL Climate Prediction (NCWCP) 5830 University Research Court College Park, MD 20740 Page Author: EMC

Deep Learning Based Solar Flare Forecasting Model. I. Results for Line-of-sight Magnetograms

NASA Astrophysics Data System (ADS)

Huang, Xin; Wang, Huaning; Xu, Long; Liu, Jinfu; Li, Rong; Dai, Xinghua

2018-03-01

Solar flares originate from the release of the energy stored in the magnetic field of solar active regions, the triggering mechanism for these flares, however, remains unknown. For this reason, the conventional solar flare forecast is essentially based on the statistic relationship between solar flares and measures extracted from observational data. In the current work, the deep learning method is applied to set up the solar flare forecasting model, in which forecasting patterns can be learned from line-of-sight magnetograms of solar active regions. In order to obtain a large amount of observational data to train the forecasting model and test its performance, a data set is created from line-of-sight magnetogarms of active regions observed by SOHO/MDI and SDO/HMI from 1996 April to 2015 October and corresponding soft X-ray solar flares observed by GOES. The testing results of the forecasting model indicate that (1) the forecasting patterns can be automatically reached with the MDI data and they can also be applied to the HMI data; furthermore, these forecasting patterns are robust to the noise in the observational data; (2) the performance of the deep learning forecasting model is not sensitive to the given forecasting periods (6, 12, 24, or 48 hr); (3) the performance of the proposed forecasting model is comparable to that of the state-of-the-art flare forecasting models, even if the duration of the total magnetograms continuously spans 19.5 years. Case analyses demonstrate that the deep learning based solar flare forecasting model pays attention to areas with the magnetic polarity-inversion line or the strong magnetic field in magnetograms of active regions.

CEREF: A hybrid data-driven model for forecasting annual streamflow from a socio-hydrological system

NASA Astrophysics Data System (ADS)

Zhang, Hongbo; Singh, Vijay P.; Wang, Bin; Yu, Yinghao

2016-09-01

Hydrological forecasting is complicated by flow regime alterations in a coupled socio-hydrologic system, encountering increasingly non-stationary, nonlinear and irregular changes, which make decision support difficult for future water resources management. Currently, many hybrid data-driven models, based on the decomposition-prediction-reconstruction principle, have been developed to improve the ability to make predictions of annual streamflow. However, there exist many problems that require further investigation, the chief among which is the direction of trend components decomposed from annual streamflow series and is always difficult to ascertain. In this paper, a hybrid data-driven model was proposed to capture this issue, which combined empirical mode decomposition (EMD), radial basis function neural networks (RBFNN), and external forces (EF) variable, also called the CEREF model. The hybrid model employed EMD for decomposition and RBFNN for intrinsic mode function (IMF) forecasting, and determined future trend component directions by regression with EF as basin water demand representing the social component in the socio-hydrologic system. The Wuding River basin was considered for the case study, and two standard statistical measures, root mean squared error (RMSE) and mean absolute error (MAE), were used to evaluate the performance of CEREF model and compare with other models: the autoregressive (AR), RBFNN and EMD-RBFNN. Results indicated that the CEREF model had lower RMSE and MAE statistics, 42.8% and 7.6%, respectively, than did other models, and provided a superior alternative for forecasting annual runoff in the Wuding River basin. Moreover, the CEREF model can enlarge the effective intervals of streamflow forecasting compared to the EMD-RBFNN model by introducing the water demand planned by the government department to improve long-term prediction accuracy. In addition, we considered the high-frequency component, a frequent subject of concern in EMD-based forecasting, and results showed that removing high-frequency component is an effective measure to improve forecasting precision and is suggested for use with the CEREF model for better performance. Finally, the study concluded that the CEREF model can be used to forecast non-stationary annual streamflow change as a co-evolution of hydrologic and social systems with better accuracy. Also, the modification about removing high-frequency can further improve the performance of the CEREF model. It should be noted that the CEREF model is beneficial for data-driven hydrologic forecasting in complex socio-hydrologic systems, and as a simple data-driven socio-hydrologic forecasting model, deserves more attention.

High-order fuzzy time-series based on multi-period adaptation model for forecasting stock markets

NASA Astrophysics Data System (ADS)

Chen, Tai-Liang; Cheng, Ching-Hsue; Teoh, Hia-Jong

2008-02-01

Stock investors usually make their short-term investment decisions according to recent stock information such as the late market news, technical analysis reports, and price fluctuations. To reflect these short-term factors which impact stock price, this paper proposes a comprehensive fuzzy time-series, which factors linear relationships between recent periods of stock prices and fuzzy logical relationships (nonlinear relationships) mined from time-series into forecasting processes. In empirical analysis, the TAIEX (Taiwan Stock Exchange Capitalization Weighted Stock Index) and HSI (Heng Seng Index) are employed as experimental datasets, and four recent fuzzy time-series models, Chen’s (1996), Yu’s (2005), Cheng’s (2006) and Chen’s (2007), are used as comparison models. Besides, to compare with conventional statistic method, the method of least squares is utilized to estimate the auto-regressive models of the testing periods within the databases. From analysis results, the performance comparisons indicate that the multi-period adaptation model, proposed in this paper, can effectively improve the forecasting performance of conventional fuzzy time-series models which only factor fuzzy logical relationships in forecasting processes. From the empirical study, the traditional statistic method and the proposed model both reveal that stock price patterns in the Taiwan stock and Hong Kong stock markets are short-term.

Satellite Sounder Data Assimilation for Improving Alaska Region Weather Forecast

NASA Technical Reports Server (NTRS)

Zhu, Jiang; Stevens, E.; Zavodsky, B. T.; Zhang, X.; Heinrichs, T.; Broderson, D.

2014-01-01

Data assimilation has been demonstrated very useful in improving both global and regional numerical weather prediction. Alaska has very coarser surface observation sites. On the other hand, it gets much more satellite overpass than lower 48 states. How to utilize satellite data to improve numerical prediction is one of hot topics among weather forecast community in Alaska. The Geographic Information Network of Alaska (GINA) at University of Alaska is conducting study on satellite data assimilation for WRF model. AIRS/CRIS sounder profile data are used to assimilate the initial condition for the customized regional WRF model (GINA-WRF model). Normalized standard deviation, RMSE, and correlation statistic analysis methods are applied to analyze one case of 48 hours forecasts and one month of 24-hour forecasts in order to evaluate the improvement of regional numerical model from Data assimilation. The final goal of the research is to provide improved real-time short-time forecast for Alaska regions.

Forecasting monthly inflow discharge of the Iffezheim reservoir using data-driven models

NASA Astrophysics Data System (ADS)

Zhang, Qing; Aljoumani, Basem; Hillebrand, Gudrun; Hoffmann, Thomas; Hinkelmann, Reinhard

2017-04-01

River stream flow is an essential element in hydrology study fields, especially for reservoir management, since it defines input into reservoirs. Forecasting this stream flow plays an important role in short or long-term planning and management in the reservoir, e.g. optimized reservoir and hydroelectric operation or agricultural irrigation. Highly accurate flow forecasting can significantly reduce economic losses and is always pursued by reservoir operators. Therefore, hydrologic time series forecasting has received tremendous attention of researchers. Many models have been proposed to improve the hydrological forecasting. Due to the fact that most natural phenomena occurring in environmental systems appear to behave in random or probabilistic ways, different cases may need a different methods to forecast the inflow and even a unique treatment to improve the forecast accuracy. The purpose of this study is to determine an appropriate model for forecasting monthly inflow to the Iffezheim reservoir in Germany, which is the last of the barrages in the Upper Rhine. Monthly time series of discharges, measured from 1946 to 2001 at the Plittersdorf station, which is located 6 km downstream of the Iffezheim reservoir, were applied. The accuracies of the used stochastic models - Fiering model and Auto-Regressive Integrated Moving Average models (ARIMA) are compared with Artificial Intelligence (AI) models - single Artificial Neural Network (ANN) and Wavelet ANN models (WANN). The Fiering model is a linear stochastic model and used for generating synthetic monthly data. The basic idea in modeling time series using ARIMA is to identify a simple model with as few model parameters as possible in order to provide a good statistical fit to the data. To identify and fit the ARIMA models, four phase approaches were used: identification, parameter estimation, diagnostic checking, and forecasting. An automatic selection criterion, such as the Akaike information criterion, is utilized to enhance this flexible approach to set up the model. As distinct from both stochastic models, the ANN and its related conjunction methods Wavelet-ANN (WANN) models are effective to handle non-linear systems and have been developed with antecedent flows as inputs to forecast up to 12-months lead-time for the Iffezheim reservoir. In the ANN and WANN models, the Feed Forward Back Propagation method (FFBP) is applied. The sigmoid activity and linear functions were used with several different neurons for the hidden layers and for the output layer, respectively. To compare the accuracy of the different models and identify the most suitable model for reliable forecasting, four quantitative standard statistical performance evaluation measures, the root mean square error (RMSE), the mean bias error (MAE) and the determination correlation coefficient (DC), are employed. The results reveal that the ARIMA (2, 1, 2) performs better than Fiering, ANN and WANN models. Further, the WANN model is found to be slightly better than the ANN model for forecasting monthly inflow of the Iffezheim reservoir. As a result, by using the ARIMA model, the predicted and observed values agree reasonably well.

Predicting and downscaling ENSO impacts on intraseasonal precipitation statistics in California: The 1997/98 event

USGS Publications Warehouse

Gershunov, A.; Barnett, T.P.; Cayan, D.R.; Tubbs, T.; Goddard, L.

2000-01-01

Three long-range forecasting methods have been evaluated for prediction and downscaling of seasonal and intraseasonal precipitation statistics in California. Full-statistical, hybrid-dynamical - statistical and full-dynamical approaches have been used to forecast El Nin??o - Southern Oscillation (ENSO) - related total precipitation, daily precipitation frequency, and average intensity anomalies during the January - March season. For El Nin??o winters, the hybrid approach emerges as the best performer, while La Nin??a forecasting skill is poor. The full-statistical forecasting method features reasonable forecasting skill for both La Nin??a and El Nin??o winters. The performance of the full-dynamical approach could not be evaluated as rigorously as that of the other two forecasting schemes. Although the full-dynamical forecasting approach is expected to outperform simpler forecasting schemes in the long run, evidence is presented to conclude that, at present, the full-dynamical forecasting approach is the least viable of the three, at least in California. The authors suggest that operational forecasting of any intraseasonal temperature, precipitation, or streamflow statistic derivable from the available records is possible now for ENSO-extreme years.

Urban pavement surface temperature. Comparison of numerical and statistical approach

NASA Astrophysics Data System (ADS)

Marchetti, Mario; Khalifa, Abderrahmen; Bues, Michel; Bouilloud, Ludovic; Martin, Eric; Chancibaut, Katia

2015-04-01

The forecast of pavement surface temperature is very specific in the context of urban winter maintenance. to manage snow plowing and salting of roads. Such forecast mainly relies on numerical models based on a description of the energy balance between the atmosphere, the buildings and the pavement, with a canyon configuration. Nevertheless, there is a specific need in the physical description and the numerical implementation of the traffic in the energy flux balance. This traffic was originally considered as a constant. Many changes were performed in a numerical model to describe as accurately as possible the traffic effects on this urban energy balance, such as tires friction, pavement-air exchange coefficient, and infrared flux neat balance. Some experiments based on infrared thermography and radiometry were then conducted to quantify the effect fo traffic on urban pavement surface. Based on meteorological data, corresponding pavement temperature forecast were calculated and were compared with fiels measurements. Results indicated a good agreement between the forecast from the numerical model based on this energy balance approach. A complementary forecast approach based on principal component analysis (PCA) and partial least-square regression (PLS) was also developed, with data from thermal mapping usng infrared radiometry. The forecast of pavement surface temperature with air temperature was obtained in the specific case of urban configurtation, and considering traffic into measurements used for the statistical analysis. A comparison between results from the numerical model based on energy balance, and PCA/PLS was then conducted, indicating the advantages and limits of each approach.

Forecasting the discomfort levels within the greater Athens area, Greece using artificial neural networks and multiple criteria analysis

NASA Astrophysics Data System (ADS)

Vouterakos, P. A.; Moustris, K. P.; Bartzokas, A.; Ziomas, I. C.; Nastos, P. T.; Paliatsos, A. G.

2012-12-01

In this work, artificial neural networks (ANNs) were developed and applied in order to forecast the discomfort levels due to the combination of high temperature and air humidity, during the hot season of the year, in eight different regions within the Greater Athens area (GAA), Greece. For the selection of the best type and architecture of ANNs-forecasting models, the multiple criteria analysis (MCA) technique was applied. Three different types of ANNs were developed and tested with the MCA method. Concretely, the multilayer perceptron, the generalized feed forward networks (GFFN), and the time-lag recurrent networks were developed and tested. Results showed that the best ANNs type performance was achieved by using the GFFN model for the prediction of discomfort levels due to high temperature and air humidity within GAA. For the evaluation of the constructed ANNs, appropriate statistical indices were used. The analysis proved that the forecasting ability of the developed ANNs models is very satisfactory at a significant statistical level of p < 0.01.

An application of seasonal ARIMA models on group commodities to forecast Philippine merchandise exports performance

NASA Astrophysics Data System (ADS)

Natividad, Gina May R.; Cawiding, Olive R.; Addawe, Rizavel C.

2017-11-01

The increase in the merchandise exports of the country offers information about the Philippines' trading role within the global economy. Merchandise exports statistics are used to monitor the country's overall production that is consumed overseas. This paper investigates the comparison between two models obtained by a) clustering the commodity groups into two based on its proportional contribution to the total exports, and b) treating only the total exports. Different seasonal autoregressive integrated moving average (SARIMA) models were then developed for the clustered commodities and for the total exports based on the monthly merchandise exports of the Philippines from 2011 to 2016. The data set used in this study was retrieved from the Philippine Statistics Authority (PSA) which is the central statistical authority in the country responsible for primary data collection. A test for significance of the difference between means at 0.05 level of significance was then performed on the forecasts produced. The result indicates that there is a significant difference between the mean of the forecasts of the two models. Moreover, upon a comparison of the root mean square error (RMSE) and mean absolute error (MAE) of the models, it was found that the models used for the clustered groups outperform the model for the total exports.

An Evaluation of Real-time Air Quality Forecasts and their Urban Emissions over Eastern Texas During the Summer of 2006 Second Texas Air Quality Study Field Study

EPA Science Inventory

Forecasts of ozone (O₃) and particulate matter (diameter less than 2.5 µm, PM_2.5) from seven air quality forecast models (AQFMs) are statistically evaluated against observations collected during August and September of 2006 (49 days) through the AIRNow netwo...

Forecasting distributions of large federal-lands fires utilizing satellite and gridded weather information

Treesearch

H.K. Preisler; R.E. Burgan; J.C. Eidenshink; J.M. Klaver; R.W. Klaver

2009-01-01

The current study presents a statistical model for assessing the skill of fire danger indices and for forecasting the distribution of the expected numbers of large fires over a given region and for the upcoming week. The procedure permits development of daily maps that forecast, for the forthcoming week and within federal lands, percentiles of the distributions of (i)...

Status of the NASA GMAO Observing System Simulation Experiment

NASA Technical Reports Server (NTRS)

Prive, Nikki C.; Errico, Ronald M.

2014-01-01

An Observing System Simulation Experiment (OSSE) is a pure modeling study used when actual observations are too expensive or difficult to obtain. OSSEs are valuable tools for determining the potential impact of new observing systems on numerical weather forecasts and for evaluation of data assimilation systems (DAS). An OSSE has been developed at the NASA Global Modeling and Assimilation Office (GMAO, Errico et al 2013). The GMAO OSSE uses a 13-month integration of the European Centre for Medium- Range Weather Forecasts 2005 operational model at T511/L91 resolution for the Nature Run (NR). Synthetic observations have been updated so that they are based on real observations during the summer of 2013. The emulated observation types include AMSU-A, MHS, IASI, AIRS, and HIRS4 radiance data, GPS-RO, and conventional types including aircraft, rawinsonde, profiler, surface, and satellite winds. The synthetic satellite wind observations are colocated with the NR cloud fields, and the rawinsondes are advected during ascent using the NR wind fields. Data counts for the synthetic observations are matched as closely as possible to real data counts, as shown in Figure 2. Errors are added to the synthetic observations to emulate representativeness and instrument errors. The synthetic errors are calibrated so that the statistics of observation innovation and analysis increments in the OSSE are similar to the same statistics for assimilation of real observations, in an iterative method described by Errico et al (2013). The standard deviations of observation minus forecast (xo-H(xb)) are compared for the OSSE and real data in Figure 3. The synthetic errors include both random, uncorrelated errors, and an additional correlated error component for some observational types. Vertically correlated errors are included for conventional sounding data and GPS-RO, and channel correlated errors are introduced to AIRS and IASI (Figure 4). HIRS, AMSU-A, and MHS have a component of horizontally correlated error. The forecast model used by the GMAO OSSE is the Goddard Earth Observing System Model, Version 5 (GEOS-5) with Gridpoint Statistical Interpolation (GSI) DAS. The model version has been updated to v. 5.13.3, corresponding to the current operational model. Forecasts are run on a cube-sphere grid with 180 points along each edge of the cube (approximately 0.5 degree horizontal resolution) with 72 vertical levels. The DAS is cycled at 6-hour intervals, with 240 hour forecasts launched daily at 0000 UTC. Evaluation of the forecasting skill for July and August is currently underway. Prior versions of the GMAO OSSE have been found to have greater forecasting skill than real world forecasts. It is anticipated that similar forecast skill will be found in the updated OSSE.

Bias correction of satellite precipitation products for flood forecasting application at the Upper Mahanadi River Basin in Eastern India

NASA Astrophysics Data System (ADS)

Beria, H.; Nanda, T., Sr.; Chatterjee, C.

2015-12-01

High resolution satellite precipitation products such as Tropical Rainfall Measuring Mission (TRMM), Climate Forecast System Reanalysis (CFSR), European Centre for Medium-Range Weather Forecasts (ECMWF), etc., offer a promising alternative to flood forecasting in data scarce regions. At the current state-of-art, these products cannot be used in the raw form for flood forecasting, even at smaller lead times. In the current study, these precipitation products are bias corrected using statistical techniques, such as additive and multiplicative bias corrections, and wavelet multi-resolution analysis (MRA) with India Meteorological Department (IMD) gridded precipitation product,obtained from gauge-based rainfall estimates. Neural network based rainfall-runoff modeling using these bias corrected products provide encouraging results for flood forecasting upto 48 hours lead time. We will present various statistical and graphical interpretations of catchment response to high rainfall events using both the raw and bias corrected precipitation products at different lead times.

The statistical extended-range (10-30-day) forecast of summer rainfall anomalies over the entire China

NASA Astrophysics Data System (ADS)

Zhu, Zhiwei; Li, Tim

2017-01-01

The extended-range (10-30-day) rainfall forecast over the entire China was carried out using spatial-temporal projection models (STPMs). Using a rotated empirical orthogonal function analysis of intraseasonal (10-80-day) rainfall anomalies, China is divided into ten sub-regions. Different predictability sources were selected for each of the ten regions. The forecast skills are ranked for each region. Based on temporal correlation coefficient (TCC) and Gerrity skill score, useful skills are found for most parts of China at a 20-25-day lead. The southern China and the mid-lower reaches of Yangtze River Valley show the highest predictive skills, whereas southwestern China and Huang-Huai region have the lowest predictive skills. By combining forecast results from ten regional STPMs, the TCC distribution of 8-year (2003-2010) independent forecast for the entire China is investigated. The combined forecast results from ten STPMs show significantly higher skills than the forecast with just one single STPM for the entire China. Independent forecast examples of summer rainfall anomalies around the period of Beijing Olympic Games in 2008 and Shanghai World Expo in 2010 are presented. The result shows that the current model is able to reproduce the gross pattern of the summer intraseasonal rainfall over China at a 20-day lead. The present study provides, for the first time, a guide on the statistical extended-range forecast of summer rainfall anomalies for the entire China. It is anticipated that the ideas and methods proposed here will facilitate the extended-range forecast in China.

September Arctic Sea Ice minimum prediction - a new skillful statistical approach

NASA Astrophysics Data System (ADS)

Ionita-Scholz, Monica; Grosfeld, Klaus; Scholz, Patrick; Treffeisen, Renate; Lohmann, Gerrit

2017-04-01

Sea ice in both Polar Regions is an important indicator for the expression of global climate change and its polar amplification. Consequently, a broad interest exists on sea ice, its coverage, variability and long term change. Knowledge on sea ice requires high quality data on ice extent, thickness and its dynamics. However, its predictability is complex and it depends on various climate and oceanic parameters and conditions. In order to provide insights into the potential development of a monthly/seasonal signal of sea ice evolution, we developed a robust statistical model based on ocean heat content, sea surface temperature and different atmospheric variables to calculate an estimate of the September Sea ice extent (SSIE) on monthly time scale. Although previous statistical attempts at monthly/seasonal forecasts of SSIE show a relatively reduced skill, we show here that more than 92% (r = 0.96) of the September sea ice extent can be predicted at the end of May by using previous months' climate and oceanic conditions. The skill of the model increases with a decrease in the time lag used for the forecast. At the end of August, our predictions are even able to explain 99% of the SSIE. Our statistical model captures both the general trend as well as the interannual variability of the SSIE. Moreover, it is able to properly forecast the years with extreme high/low SSIE (e.g. 1996/ 2007, 2012, 2013). Besides its forecast skill for SSIE, the model could provide a valuable tool for identifying relevant regions and climate parameters that are important for the sea ice development in the Arctic and for detecting sensitive and critical regions in global coupled climate models with focus on sea ice formation.

A stochastic HMM-based forecasting model for fuzzy time series.

PubMed

Li, Sheng-Tun; Cheng, Yi-Chung

2010-10-01

Recently, fuzzy time series have attracted more academic attention than traditional time series due to their capability of dealing with the uncertainty and vagueness inherent in the data collected. The formulation of fuzzy relations is one of the key issues affecting forecasting results. Most of the present works adopt IF-THEN rules for relationship representation, which leads to higher computational overhead and rule redundancy. Sullivan and Woodall proposed a Markov-based formulation and a forecasting model to reduce computational overhead; however, its applicability is limited to handling one-factor problems. In this paper, we propose a novel forecasting model based on the hidden Markov model by enhancing Sullivan and Woodall's work to allow handling of two-factor forecasting problems. Moreover, in order to make the nature of conjecture and randomness of forecasting more realistic, the Monte Carlo method is adopted to estimate the outcome. To test the effectiveness of the resulting stochastic model, we conduct two experiments and compare the results with those from other models. The first experiment consists of forecasting the daily average temperature and cloud density in Taipei, Taiwan, and the second experiment is based on the Taiwan Weighted Stock Index by forecasting the exchange rate of the New Taiwan dollar against the U.S. dollar. In addition to improving forecasting accuracy, the proposed model adheres to the central limit theorem, and thus, the result statistically approximates to the real mean of the target value being forecast.

Extended Kalman Filter framework for forecasting shoreline evolution

USGS Publications Warehouse

Long, Joseph; Plant, Nathaniel G.

2012-01-01

A shoreline change model incorporating both long- and short-term evolution is integrated into a data assimilation framework that uses sparse observations to generate an updated forecast of shoreline position and to estimate unobserved geophysical variables and model parameters. Application of the assimilation algorithm provides quantitative statistical estimates of combined model-data forecast uncertainty which is crucial for developing hazard vulnerability assessments, evaluation of prediction skill, and identifying future data collection needs. Significant attention is given to the estimation of four non-observable parameter values and separating two scales of shoreline evolution using only one observable morphological quantity (i.e. shoreline position).

European Wintertime Windstorms and its Links to Large-Scale Variability Modes

NASA Astrophysics Data System (ADS)

Befort, D. J.; Wild, S.; Walz, M. A.; Knight, J. R.; Lockwood, J. F.; Thornton, H. E.; Hermanson, L.; Bett, P.; Weisheimer, A.; Leckebusch, G. C.

2017-12-01

Winter storms associated with extreme wind speeds and heavy precipitation are the most costly natural hazard in several European countries. Improved understanding and seasonal forecast skill of winter storms will thus help society, policy-makers and (re-) insurance industry to be better prepared for such events. We firstly assess the ability to represent extra-tropical windstorms over the Northern Hemisphere of three seasonal forecast ensemble suites: ECMWF System3, ECMWF System4 and GloSea5. Our results show significant skill for inter-annual variability of windstorm frequency over parts of Europe in two of these forecast suites (ECMWF-S4 and GloSea5) indicating the potential use of current seasonal forecast systems. In a regression model we further derive windstorm variability using the forecasted NAO from the seasonal model suites thus estimating the suitability of the NAO as the only predictor. We find that the NAO as the main large-scale mode over Europe can explain some of the achieved skill and is therefore an important source of variability in the seasonal models. However, our results show that the regression model fails to reproduce the skill level of the directly forecast windstorm frequency over large areas of central Europe. This suggests that the seasonal models also capture other sources of variability/predictability of windstorms than the NAO. In order to investigate which other large-scale variability modes steer the interannual variability of windstorms we develop a statistical model using a Poisson GLM. We find that the Scandinavian Pattern (SCA) in fact explains a larger amount of variability for Central Europe during the 20th century than the NAO. This statistical model is able to skilfully reproduce the interannual variability of windstorm frequency especially for the British Isles and Central Europe with correlations up to 0.8.

On the effect of model parameters on forecast objects

NASA Astrophysics Data System (ADS)

Marzban, Caren; Jones, Corinne; Li, Ning; Sandgathe, Scott

2018-04-01

Many physics-based numerical models produce a gridded, spatial field of forecasts, e.g., a temperature map. The field for some quantities generally consists of spatially coherent and disconnected objects. Such objects arise in many problems, including precipitation forecasts in atmospheric models, eddy currents in ocean models, and models of forest fires. Certain features of these objects (e.g., location, size, intensity, and shape) are generally of interest. Here, a methodology is developed for assessing the impact of model parameters on the features of forecast objects. The main ingredients of the methodology include the use of (1) Latin hypercube sampling for varying the values of the model parameters, (2) statistical clustering algorithms for identifying objects, (3) multivariate multiple regression for assessing the impact of multiple model parameters on the distribution (across the forecast domain) of object features, and (4) methods for reducing the number of hypothesis tests and controlling the resulting errors. The final output of the methodology is a series of box plots and confidence intervals that visually display the sensitivities. The methodology is demonstrated on precipitation forecasts from a mesoscale numerical weather prediction model.

Technical note: Combining quantile forecasts and predictive distributions of streamflows

NASA Astrophysics Data System (ADS)

Bogner, Konrad; Liechti, Katharina; Zappa, Massimiliano

2017-11-01

The enhanced availability of many different hydro-meteorological modelling and forecasting systems raises the issue of how to optimally combine this great deal of information. Especially the usage of deterministic and probabilistic forecasts with sometimes widely divergent predicted future streamflow values makes it even more complicated for decision makers to sift out the relevant information. In this study multiple streamflow forecast information will be aggregated based on several different predictive distributions, and quantile forecasts. For this combination the Bayesian model averaging (BMA) approach, the non-homogeneous Gaussian regression (NGR), also known as the ensemble model output statistic (EMOS) techniques, and a novel method called Beta-transformed linear pooling (BLP) will be applied. By the help of the quantile score (QS) and the continuous ranked probability score (CRPS), the combination results for the Sihl River in Switzerland with about 5 years of forecast data will be compared and the differences between the raw and optimally combined forecasts will be highlighted. The results demonstrate the importance of applying proper forecast combination methods for decision makers in the field of flood and water resource management.

Visualizing Uncertainty for Probabilistic Weather Forecasting based on Reforecast Analogs

NASA Astrophysics Data System (ADS)

Pelorosso, Leandro; Diehl, Alexandra; Matković, Krešimir; Delrieux, Claudio; Ruiz, Juan; Gröeller, M. Eduard; Bruckner, Stefan

2016-04-01

Numerical weather forecasts are prone to uncertainty coming from inaccuracies in the initial and boundary conditions and lack of precision in numerical models. Ensemble of forecasts partially addresses these problems by considering several runs of the numerical model. Each forecast is generated with different initial and boundary conditions and different model configurations [GR05]. The ensembles can be expressed as probabilistic forecasts, which have proven to be very effective in the decision-making processes [DE06]. The ensemble of forecasts represents only some of the possible future atmospheric states, usually underestimating the degree of uncertainty in the predictions [KAL03, PH06]. Hamill and Whitaker [HW06] introduced the "Reforecast Analog Regression" (RAR) technique to overcome the limitations of ensemble forecasting. This technique produces probabilistic predictions based on the analysis of historical forecasts and observations. Visual analytics provides tools for processing, visualizing, and exploring data to get new insights and discover hidden information patterns in an interactive exchange between the user and the application [KMS08]. In this work, we introduce Albero, a visual analytics solution for probabilistic weather forecasting based on the RAR technique. Albero targets at least two different type of users: "forecasters", who are meteorologists working in operational weather forecasting and "researchers", who work in the construction of numerical prediction models. Albero is an efficient tool for analyzing precipitation forecasts, allowing forecasters to make and communicate quick decisions. Our solution facilitates the analysis of a set of probabilistic forecasts, associated statistical data, observations and uncertainty. A dashboard with small-multiples of probabilistic forecasts allows the forecasters to analyze at a glance the distribution of probabilities as a function of time, space, and magnitude. It provides the user with a more accurate measure of forecast uncertainty that could result in better decision-making. It offers different level of abstractions to help with the recalibration of the RAR method. It also has an inspection tool that displays the selected analogs, their observations and statistical data. It gives the users access to inner parts of the method, unveiling hidden information. References [GR05] GNEITING T., RAFTERY A. E.: Weather forecasting with ensemble methods. Science 310, 5746, 248-249, 2005. [KAL03] KALNAY E.: Atmospheric modeling, data assimilation and predictability. Cambridge University Press, 2003. [PH06] PALMER T., HAGEDORN R.: Predictability of weather and climate. Cambridge University Press, 2006. [HW06] HAMILL T. M., WHITAKER J. S.: Probabilistic quantitative precipitation forecasts based on reforecast analogs: Theory and application. Monthly Weather Review 134, 11, 3209-3229, 2006. [DE06] DEITRICK S., EDSALL R.: The influence of uncertainty visualization on decision making: An empirical evaluation. Springer, 2006. [KMS08] KEIM D. A., MANSMANN F., SCHNEIDEWIND J., THOMAS J., ZIEGLER H.: Visual analytics: Scope and challenges. Springer, 2008.

A scoping review of malaria forecasting: past work and future directions

PubMed Central

Zinszer, Kate; Verma, Aman D; Charland, Katia; Brewer, Timothy F; Brownstein, John S; Sun, Zhuoyu; Buckeridge, David L

2012-01-01

Objectives There is a growing body of literature on malaria forecasting methods and the objective of our review is to identify and assess methods, including predictors, used to forecast malaria. Design Scoping review. Two independent reviewers searched information sources, assessed studies for inclusion and extracted data from each study. Information sources Search strategies were developed and the following databases were searched: CAB Abstracts, EMBASE, Global Health, MEDLINE, ProQuest Dissertations & Theses and Web of Science. Key journals and websites were also manually searched. Eligibility criteria for included studies We included studies that forecasted incidence, prevalence or epidemics of malaria over time. A description of the forecasting model and an assessment of the forecast accuracy of the model were requirements for inclusion. Studies were restricted to human populations and to autochthonous transmission settings. Results We identified 29 different studies that met our inclusion criteria for this review. The forecasting approaches included statistical modelling, mathematical modelling and machine learning methods. Climate-related predictors were used consistently in forecasting models, with the most common predictors being rainfall, relative humidity, temperature and the normalised difference vegetation index. Model evaluation was typically based on a reserved portion of data and accuracy was measured in a variety of ways including mean-squared error and correlation coefficients. We could not compare the forecast accuracy of models from the different studies as the evaluation measures differed across the studies. Conclusions Applying different forecasting methods to the same data, exploring the predictive ability of non-environmental variables, including transmission reducing interventions and using common forecast accuracy measures will allow malaria researchers to compare and improve models and methods, which should improve the quality of malaria forecasting. PMID:23180505

Do quantitative decadal forecasts from GCMs provide decision relevant skill?

NASA Astrophysics Data System (ADS)

Suckling, E. B.; Smith, L. A.

2012-04-01

It is widely held that only physics-based simulation models can capture the dynamics required to provide decision-relevant probabilistic climate predictions. This fact in itself provides no evidence that predictions from today's GCMs are fit for purpose. Empirical (data-based) models are employed to make probability forecasts on decadal timescales, where it is argued that these 'physics free' forecasts provide a quantitative 'zero skill' target for the evaluation of forecasts based on more complicated models. It is demonstrated that these zero skill models are competitive with GCMs on decadal scales for probability forecasts evaluated over the last 50 years. Complications of statistical interpretation due to the 'hindcast' nature of this experiment, and the likely relevance of arguments that the lack of hindcast skill is irrelevant as the signal will soon 'come out of the noise' are discussed. A lack of decision relevant quantiative skill does not bring the science-based insights of anthropogenic warming into doubt, but it does call for a clear quantification of limits, as a function of lead time, for spatial and temporal scales on which decisions based on such model output are expected to prove maladaptive. Failing to do so may risk the credibility of science in support of policy in the long term. The performance amongst a collection of simulation models is evaluated, having transformed ensembles of point forecasts into probability distributions through the kernel dressing procedure [1], according to a selection of proper skill scores [2] and contrasted with purely data-based empirical models. Data-based models are unlikely to yield realistic forecasts for future climate change if the Earth system moves away from the conditions observed in the past, upon which the models are constructed; in this sense the empirical model defines zero skill. When should a decision relevant simulation model be expected to significantly outperform such empirical models? Probability forecasts up to ten years ahead (decadal forecasts) are considered, both on global and regional spatial scales for surface air temperature. Such decadal forecasts are not only important in terms of providing information on the impacts of near-term climate change, but also from the perspective of climate model validation, as hindcast experiments and a sufficient database of historical observations allow standard forecast verification methods to be used. Simulation models from the ENSEMBLES hindcast experiment [3] are evaluated and contrasted with static forecasts of the observed climatology, persistence forecasts and against simple statistical models, called dynamic climatology (DC). It is argued that DC is a more apropriate benchmark in the case of a non-stationary climate. It is found that the ENSEMBLES models do not demonstrate a significant increase in skill relative to the empirical models even at global scales over any lead time up to a decade ahead. It is suggested that the contsruction and co-evaluation with the data-based models become a regular component of the reporting of large simulation model forecasts. The methodology presented may easily be adapted to other forecasting experiments and is expected to influence the design of future experiments. The inclusion of comparisons with dynamic climatology and other data-based approaches provide important information to both scientists and decision makers on which aspects of state-of-the-art simulation forecasts are likely to be fit for purpose. [1] J. Bröcker and L. A. Smith. From ensemble forecasts to predictive distributions, Tellus A, 60(4), 663-678 (2007). [2] J. Bröcker and L. A. Smith. Scoring probabilistic forecasts: The importance of being proper, Weather and Forecasting, 22, 382-388 (2006). [3] F. J. Doblas-Reyes, A. Weisheimer, T. N. Palmer, J. M. Murphy and D. Smith. Forecast quality asessment of the ENSEMBLES seasonal-to-decadal stream 2 hindcasts, ECMWF Technical Memorandum, 621 (2010).

Air Pollution Forecasts: An Overview

PubMed Central

Bai, Lu; Wang, Jianzhou; Lu, Haiyan

2018-01-01

Air pollution is defined as a phenomenon harmful to the ecological system and the normal conditions of human existence and development when some substances in the atmosphere exceed a certain concentration. In the face of increasingly serious environmental pollution problems, scholars have conducted a significant quantity of related research, and in those studies, the forecasting of air pollution has been of paramount importance. As a precaution, the air pollution forecast is the basis for taking effective pollution control measures, and accurate forecasting of air pollution has become an important task. Extensive research indicates that the methods of air pollution forecasting can be broadly divided into three classical categories: statistical forecasting methods, artificial intelligence methods, and numerical forecasting methods. More recently, some hybrid models have been proposed, which can improve the forecast accuracy. To provide a clear perspective on air pollution forecasting, this study reviews the theory and application of those forecasting models. In addition, based on a comparison of different forecasting methods, the advantages and disadvantages of some methods of forecasting are also provided. This study aims to provide an overview of air pollution forecasting methods for easy access and reference by researchers, which will be helpful in further studies. PMID:29673227

Air Pollution Forecasts: An Overview.

PubMed

Bai, Lu; Wang, Jianzhou; Ma, Xuejiao; Lu, Haiyan

2018-04-17

Air pollution is defined as a phenomenon harmful to the ecological system and the normal conditions of human existence and development when some substances in the atmosphere exceed a certain concentration. In the face of increasingly serious environmental pollution problems, scholars have conducted a significant quantity of related research, and in those studies, the forecasting of air pollution has been of paramount importance. As a precaution, the air pollution forecast is the basis for taking effective pollution control measures, and accurate forecasting of air pollution has become an important task. Extensive research indicates that the methods of air pollution forecasting can be broadly divided into three classical categories: statistical forecasting methods, artificial intelligence methods, and numerical forecasting methods. More recently, some hybrid models have been proposed, which can improve the forecast accuracy. To provide a clear perspective on air pollution forecasting, this study reviews the theory and application of those forecasting models. In addition, based on a comparison of different forecasting methods, the advantages and disadvantages of some methods of forecasting are also provided. This study aims to provide an overview of air pollution forecasting methods for easy access and reference by researchers, which will be helpful in further studies.

A seasonal hydrologic ensemble prediction system for water resource management

NASA Astrophysics Data System (ADS)

Luo, L.; Wood, E. F.

2006-12-01

A seasonal hydrologic ensemble prediction system, developed for the Ohio River basin, has been improved and expanded to several other regions including the Eastern U.S., Africa and East Asia. The prediction system adopts the traditional Extended Streamflow Prediction (ESP) approach, utilizing the VIC (Variable Infiltration Capacity) hydrological model as the central tool for producing ensemble prediction of soil moisture, snow and streamflow with lead times up to 6-month. VIC is forced by observed meteorology to estimate the hydrological initial condition prior to the forecast, but during the forecast period the atmospheric forcing comes from statistically downscaled, seasonal forecast from dynamic climate models. The seasonal hydrologic ensemble prediction system is currently producing realtime seasonal hydrologic forecast for these regions on a monthly basis. Using hindcasts from a 19-year period (1981-1999), during which seasonal hindcasts from NCEP Climate Forecast System (CFS) and European Union DEMETER project are available, we evaluate the performance of the forecast system over our forecast regions. The evaluation shows that the prediction system using the current forecast approach is able to produce reliable and accurate precipitation, soil moisture and streamflow predictions. The overall skill is much higher then the traditional ESP. In particular, forecasts based on multiple climate model forecast are more skillful than single model-based forecast. This emphasizes the significant need for producing seasonal climate forecast with multiple climate models for hydrologic applications. Forecast from this system is expected to provide very valuable information about future hydrologic states and associated risks for end users, including water resource management and financial sectors.

Satellite temperature monitoring and prediction system

NASA Technical Reports Server (NTRS)

Barnett, U. R.; Martsolf, J. D.; Crosby, F. L.

1980-01-01

The paper describes the Florida Satellite Freeze Forecast System (SFFS) in its current state. All data collection options have been demonstrated, and data collected over a three year period have been stored for future analysis. Presently, specific minimum temperature forecasts are issued routinely from November through March. The procedures for issuing these forecast are discussed. The automated data acquisition and processing system is described, and the physical and statistical models employed are examined.

Uncertainty quantification and reliability assessment in operational oil spill forecast modeling system.

PubMed

Hou, Xianlong; Hodges, Ben R; Feng, Dongyu; Liu, Qixiao

2017-03-15

As oil transport increasing in the Texas bays, greater risks of ship collisions will become a challenge, yielding oil spill accidents as a consequence. To minimize the ecological damage and optimize rapid response, emergency managers need to be informed with how fast and where oil will spread as soon as possible after a spill. The state-of-the-art operational oil spill forecast modeling system improves the oil spill response into a new stage. However uncertainty due to predicted data inputs often elicits compromise on the reliability of the forecast result, leading to misdirection in contingency planning. Thus understanding the forecast uncertainty and reliability become significant. In this paper, Monte Carlo simulation is implemented to provide parameters to generate forecast probability maps. The oil spill forecast uncertainty is thus quantified by comparing the forecast probability map and the associated hindcast simulation. A HyosPy-based simple statistic model is developed to assess the reliability of an oil spill forecast in term of belief degree. The technologies developed in this study create a prototype for uncertainty and reliability analysis in numerical oil spill forecast modeling system, providing emergency managers to improve the capability of real time operational oil spill response and impact assessment. Copyright © 2017 Elsevier Ltd. All rights reserved.

The GISS sounding temperature impact test

NASA Technical Reports Server (NTRS)

Halem, M.; Ghil, M.; Atlas, R.; Susskind, J.; Quirk, W. J.

1978-01-01

The impact of DST 5 and DST 6 satellite sounding data on mid-range forecasting was studied. The GISS temperature sounding technique, the GISS time-continuous four-dimensional assimilation procedure based on optimal statistical analysis, the GISS forecast model, and the verification techniques developed, including impact on local precipitation forecasts are described. It is found that the impact of sounding data was substantial and beneficial for the winter test period, Jan. 29 - Feb. 21. 1976. Forecasts started from initial state obtained with the aid of satellite data showed a mean improvement of about 4 points in the 48 and 772 hours Sub 1 scores as verified over North America and Europe. This corresponds to an 8 to 12 hour forecast improvement in the forecast range at 48 hours. An automated local precipitation forecast model applied to 128 cities in the United States showed on an average 15% improvement when satellite data was used for numerical forecasts. The improvement was 75% in the midwest.

Atmospheric and oceanographic research review, 1978. [global weather, ocean/air interactions, and climate

NASA Technical Reports Server (NTRS)

1978-01-01

Research activities related to global weather, ocean/air interactions, and climate are reported. The global weather research is aimed at improving the assimilation of satellite-derived data in weather forecast models, developing analysis/forecast models that can more fully utilize satellite data, and developing new measures of forecast skill to properly assess the impact of satellite data on weather forecasting. The oceanographic research goal is to understand and model the processes that determine the general circulation of the oceans, focusing on those processes that affect sea surface temperature and oceanic heat storage, which are the oceanographic variables with the greatest influence on climate. The climate research objective is to support the development and effective utilization of space-acquired data systems in climate forecast models and to conduct sensitivity studies to determine the affect of lower boundary conditions on climate and predictability studies to determine which global climate features can be modeled either deterministically or statistically.

A New Multivariate Approach in Generating Ensemble Meteorological Forcings for Hydrological Forecasting

NASA Astrophysics Data System (ADS)

Khajehei, Sepideh; Moradkhani, Hamid

2015-04-01

Producing reliable and accurate hydrologic ensemble forecasts are subject to various sources of uncertainty, including meteorological forcing, initial conditions, model structure, and model parameters. Producing reliable and skillful precipitation ensemble forecasts is one approach to reduce the total uncertainty in hydrological applications. Currently, National Weather Prediction (NWP) models are developing ensemble forecasts for various temporal ranges. It is proven that raw products from NWP models are biased in mean and spread. Given the above state, there is a need for methods that are able to generate reliable ensemble forecasts for hydrological applications. One of the common techniques is to apply statistical procedures in order to generate ensemble forecast from NWP-generated single-value forecasts. The procedure is based on the bivariate probability distribution between the observation and single-value precipitation forecast. However, one of the assumptions of the current method is fitting Gaussian distribution to the marginal distributions of observed and modeled climate variable. Here, we have described and evaluated a Bayesian approach based on Copula functions to develop an ensemble precipitation forecast from the conditional distribution of single-value precipitation forecasts. Copula functions are known as the multivariate joint distribution of univariate marginal distributions, which are presented as an alternative procedure in capturing the uncertainties related to meteorological forcing. Copulas are capable of modeling the joint distribution of two variables with any level of correlation and dependency. This study is conducted over a sub-basin in the Columbia River Basin in USA using the monthly precipitation forecasts from Climate Forecast System (CFS) with 0.5x0.5 Deg. spatial resolution to reproduce the observations. The verification is conducted on a different period and the superiority of the procedure is compared with Ensemble Pre-Processor approach currently used by National Weather Service River Forecast Centers in USA.

Monthly monsoon rainfall forecasting using artificial neural networks

NASA Astrophysics Data System (ADS)

Ganti, Ravikumar

2014-10-01

Indian agriculture sector heavily depends on monsoon rainfall for successful harvesting. In the past, prediction of rainfall was mainly performed using regression models, which provide reasonable accuracy in the modelling and forecasting of complex physical systems. Recently, Artificial Neural Networks (ANNs) have been proposed as efficient tools for modelling and forecasting. A feed-forward multi-layer perceptron type of ANN architecture trained using the popular back-propagation algorithm was employed in this study. Other techniques investigated for modeling monthly monsoon rainfall include linear and non-linear regression models for comparison purposes. The data employed in this study include monthly rainfall and monthly average of the daily maximum temperature in the North Central region in India. Specifically, four regression models and two ANN model's were developed. The performance of various models was evaluated using a wide variety of standard statistical parameters and scatter plots. The results obtained in this study for forecasting monsoon rainfalls using ANNs have been encouraging. India's economy and agricultural activities can be effectively managed with the help of the availability of the accurate monsoon rainfall forecasts.

An online tool for Operational Probabilistic Drought Forecasting System (OPDFS): a Statistical-Dynamical Framework

NASA Astrophysics Data System (ADS)

Zarekarizi, M.; Moradkhani, H.; Yan, H.

2017-12-01

The Operational Probabilistic Drought Forecasting System (OPDFS) is an online tool recently developed at Portland State University for operational agricultural drought forecasting. This is an integrated statistical-dynamical framework issuing probabilistic drought forecasts monthly for the lead times of 1, 2, and 3 months. The statistical drought forecasting method utilizes copula functions in order to condition the future soil moisture values on the antecedent states. Due to stochastic nature of land surface properties, the antecedent soil moisture states are uncertain; therefore, data assimilation system based on Particle Filtering (PF) is employed to quantify the uncertainties associated with the initial condition of the land state, i.e. soil moisture. PF assimilates the satellite soil moisture data to Variable Infiltration Capacity (VIC) land surface model and ultimately updates the simulated soil moisture. The OPDFS builds on the NOAA's seasonal drought outlook by offering drought probabilities instead of qualitative ordinal categories and provides the user with the probability maps associated with a particular drought category. A retrospective assessment of the OPDFS showed that the forecasting of the 2012 Great Plains and 2014 California droughts were possible at least one month in advance. The OPDFS offers a timely assistance to water managers, stakeholders and decision-makers to develop resilience against uncertain upcoming droughts.

A Bayesian spatio-temporal model for forecasting Anaplasma species seroprevalence in domestic dogs within the contiguous United States.

PubMed

Liu, Yan; Watson, Stella C; Gettings, Jenna R; Lund, Robert B; Nordone, Shila K; Yabsley, Michael J; McMahan, Christopher S

2017-01-01

This paper forecasts the 2016 canine Anaplasma spp. seroprevalence in the United States from eight climate, geographic and societal factors. The forecast's construction and an assessment of its performance are described. The forecast is based on a spatial-temporal conditional autoregressive model fitted to over 11 million Anaplasma spp. seroprevalence test results for dogs conducted in the 48 contiguous United States during 2011-2015. The forecast uses county-level data on eight predictive factors, including annual temperature, precipitation, relative humidity, county elevation, forestation coverage, surface water coverage, population density and median household income. Non-static factors are extrapolated into the forthcoming year with various statistical methods. The fitted model and factor extrapolations are used to estimate next year's regional prevalence. The correlation between the observed and model-estimated county-by-county Anaplasma spp. seroprevalence for the five-year period 2011-2015 is 0.902, demonstrating reasonable model accuracy. The weighted correlation (accounting for different sample sizes) between 2015 observed and forecasted county-by-county Anaplasma spp. seroprevalence is 0.987, exhibiting that the proposed approach can be used to accurately forecast Anaplasma spp. seroprevalence. The forecast presented herein can a priori alert veterinarians to areas expected to see Anaplasma spp. seroprevalence beyond the accepted endemic range. The proposed methods may prove useful for forecasting other diseases.

Implementation and Research on the Operational Use of the Mesoscale Prediction Model COAMPS in Poland

DTIC Science & Technology

2007-09-30

COAMPS model. Bogumil Jakubiak, University of Warsaw – participated in EGU General Assembly , Vienna Austria 15-20 April 2007 giving one oral and two...conditional forecast (background) error probability density function using an ensemble of the model forecast to generate background error statistics...COAMPS system on ICM machines at Warsaw University for the purpose of providing operational support to the general public using the ICM meteorological

Evaluation of precipitation forecasts from 3D-Var and hybrid GSI-based system during Indian summer monsoon 2015

NASA Astrophysics Data System (ADS)

Singh, Sanjeev Kumar; Prasad, V. S.

2018-02-01

This paper presents a systematic investigation of medium-range rainfall forecasts from two versions of the National Centre for Medium Range Weather Forecasting (NCMRWF)-Global Forecast System based on three-dimensional variational (3D-Var) and hybrid analysis system namely, NGFS and HNGFS, respectively, during Indian summer monsoon (June-September) 2015. The NGFS uses gridpoint statistical interpolation (GSI) 3D-Var data assimilation system, whereas HNGFS uses hybrid 3D ensemble-variational scheme. The analysis includes the evaluation of rainfall fields and comparisons of rainfall using statistical score such as mean precipitation, bias, correlation coefficient, root mean square error and forecast improvement factor. In addition to these, categorical scores like Peirce skill score and bias score are also computed to describe particular aspects of forecasts performance. The comparison results of mean precipitation reveal that both the versions of model produced similar large-scale feature of Indian summer monsoon rainfall for day-1 through day-5 forecasts. The inclusion of fully flow-dependent background error covariance significantly improved the wet biases in HNGFS over the Indian Ocean. The forecast improvement factor and Peirce skill score in the HNGFS have also found better than NGFS for day-1 through day-5 forecasts.

An application of a multi model approach for solar energy prediction in Southern Italy

NASA Astrophysics Data System (ADS)

Avolio, Elenio; Lo Feudo, Teresa; Calidonna, Claudia Roberta; Contini, Daniele; Torcasio, Rosa Claudia; Tiriolo, Luca; Montesanti, Stefania; Transerici, Claudio; Federico, Stefano

2015-04-01

The accuracy of the short and medium range forecast of solar irradiance is very important for solar energy integration into the grid. This issue is particularly important for Southern Italy where a significant availability of solar energy is associated with a poor development of the grid. In this work we analyse the performance of two deterministic models for the prediction of surface temperature and short-wavelength radiance for two sites in southern Italy. Both parameters are needed to forecast the power production from solar power plants, so the performance of the forecast for these meteorological parameters is of paramount importance. The models considered in this work are the RAMS (Regional Atmospheric Modeling System) and the WRF (Weather Research and Forecasting Model) and they were run for the summer 2013 at 4 km horizontal resolution over Italy. The forecast lasts three days. Initial and dynamic boundary conditions are given by the 12 UTC deterministic forecast of the ECMWF-IFS (European Centre for Medium Weather Range Forecast - Integrated Forecasting System) model, and were available every 6 hours. Verification is given against two surface stations located in Southern Italy, Lamezia Terme and Lecce, and are based on hourly output of models forecast. Results for the whole period for temperature show a positive bias for the RAMS model and a negative bias for the WRF model. RMSE is between 1 and 2 °C for both models. Results for the whole period for the short-wavelength radiance show a positive bias for both models (about 30 W/m2 for both models) and a RMSE of 100 W/m2. To reduce the model errors, a statistical post-processing technique, i.e the multi-model, is adopted. In this approach the two model's outputs are weighted with an adequate set of weights computed for a training period. In general, the performance is improved by the application of the technique, and the RMSE is reduced by a sizeable fraction (i.e. larger than 10% of the initial RMSE) depending on the forecasting time and parameter. The performance of the multi model is discussed as a function of the length of the training period and is compared with the performance of the MOS (Model Output Statistics) approach. ACKNOWLEDGMENTS This work is partially supported by projects PON04a2E Sinergreen-ResNovae - "Smart Energy Master for the energetic government of the territory" and PONa3_00363 "High Technology Infrastructure for Climate and Environment Monitoring" (I-AMICA) founded by Italian Ministry of University and Research (MIUR) PON 2007-2013. The ECMWF and CNMCA (Centro Nazionale di Meteorologia e Climatologia Aeronautica) are acknowledged for the use of the MARS (Meteorological Archive and Retrieval System).

Estimating predictive hydrological uncertainty by dressing deterministic and ensemble forecasts; a comparison, with application to Meuse and Rhine

NASA Astrophysics Data System (ADS)

Verkade, J. S.; Brown, J. D.; Davids, F.; Reggiani, P.; Weerts, A. H.

2017-12-01

Two statistical post-processing approaches for estimation of predictive hydrological uncertainty are compared: (i) 'dressing' of a deterministic forecast by adding a single, combined estimate of both hydrological and meteorological uncertainty and (ii) 'dressing' of an ensemble streamflow forecast by adding an estimate of hydrological uncertainty to each individual streamflow ensemble member. Both approaches aim to produce an estimate of the 'total uncertainty' that captures both the meteorological and hydrological uncertainties. They differ in the degree to which they make use of statistical post-processing techniques. In the 'lumped' approach, both sources of uncertainty are lumped by post-processing deterministic forecasts using their verifying observations. In the 'source-specific' approach, the meteorological uncertainties are estimated by an ensemble of weather forecasts. These ensemble members are routed through a hydrological model and a realization of the probability distribution of hydrological uncertainties (only) is then added to each ensemble member to arrive at an estimate of the total uncertainty. The techniques are applied to one location in the Meuse basin and three locations in the Rhine basin. Resulting forecasts are assessed for their reliability and sharpness, as well as compared in terms of multiple verification scores including the relative mean error, Brier Skill Score, Mean Continuous Ranked Probability Skill Score, Relative Operating Characteristic Score and Relative Economic Value. The dressed deterministic forecasts are generally more reliable than the dressed ensemble forecasts, but the latter are sharper. On balance, however, they show similar quality across a range of verification metrics, with the dressed ensembles coming out slightly better. Some additional analyses are suggested. Notably, these include statistical post-processing of the meteorological forecasts in order to increase their reliability, thus increasing the reliability of the streamflow forecasts produced with ensemble meteorological forcings.

Adaptive correction of ensemble forecasts

NASA Astrophysics Data System (ADS)

Pelosi, Anna; Battista Chirico, Giovanni; Van den Bergh, Joris; Vannitsem, Stephane

2017-04-01

Forecasts from numerical weather prediction (NWP) models often suffer from both systematic and non-systematic errors. These are present in both deterministic and ensemble forecasts, and originate from various sources such as model error and subgrid variability. Statistical post-processing techniques can partly remove such errors, which is particularly important when NWP outputs concerning surface weather variables are employed for site specific applications. Many different post-processing techniques have been developed. For deterministic forecasts, adaptive methods such as the Kalman filter are often used, which sequentially post-process the forecasts by continuously updating the correction parameters as new ground observations become available. These methods are especially valuable when long training data sets do not exist. For ensemble forecasts, well-known techniques are ensemble model output statistics (EMOS), and so-called "member-by-member" approaches (MBM). Here, we introduce a new adaptive post-processing technique for ensemble predictions. The proposed method is a sequential Kalman filtering technique that fully exploits the information content of the ensemble. One correction equation is retrieved and applied to all members, however the parameters of the regression equations are retrieved by exploiting the second order statistics of the forecast ensemble. We compare our new method with two other techniques: a simple method that makes use of a running bias correction of the ensemble mean, and an MBM post-processing approach that rescales the ensemble mean and spread, based on minimization of the Continuous Ranked Probability Score (CRPS). We perform a verification study for the region of Campania in southern Italy. We use two years (2014-2015) of daily meteorological observations of 2-meter temperature and 10-meter wind speed from 18 ground-based automatic weather stations distributed across the region, comparing them with the corresponding COSMO-LEPS ensemble forecasts. Deterministic verification scores (e.g., mean absolute error, bias) and probabilistic scores (e.g., CRPS) are used to evaluate the post-processing techniques. We conclude that the new adaptive method outperforms the simpler running bias-correction. The proposed adaptive method often outperforms the MBM method in removing bias. The MBM method has the advantage of correcting the ensemble spread, although it needs more training data.

Application of time series analysis in modelling and forecasting emergency department visits in a medical centre in Southern Taiwan

PubMed Central

Juang, Wang-Chuan; Huang, Sin-Jhih; Huang, Fong-Dee; Cheng, Pei-Wen; Wann, Shue-Ren

2017-01-01

Objective Emergency department (ED) overcrowding is acknowledged as an increasingly important issue worldwide. Hospital managers are increasingly paying attention to ED crowding in order to provide higher quality medical services to patients. One of the crucial elements for a good management strategy is demand forecasting. Our study sought to construct an adequate model and to forecast monthly ED visits. Methods We retrospectively gathered monthly ED visits from January 2009 to December 2016 to carry out a time series autoregressive integrated moving average (ARIMA) analysis. Initial development of the model was based on past ED visits from 2009 to 2016. A best-fit model was further employed to forecast the monthly data of ED visits for the next year (2016). Finally, we evaluated the predicted accuracy of the identified model with the mean absolute percentage error (MAPE). The software packages SAS/ETS V.9.4 and Office Excel 2016 were used for all statistical analyses. Results A series of statistical tests showed that six models, including ARIMA (0, 0, 1), ARIMA (1, 0, 0), ARIMA (1, 0, 1), ARIMA (2, 0, 1), ARIMA (3, 0, 1) and ARIMA (5, 0, 1), were candidate models. The model that gave the minimum Akaike information criterion and Schwartz Bayesian criterion and followed the assumptions of residual independence was selected as the adequate model. Finally, a suitable ARIMA (0, 0, 1) structure, yielding a MAPE of 8.91%, was identified and obtained as Visitt=7111.161+(at+0.37462 at−1). Conclusion The ARIMA (0, 0, 1) model can be considered adequate for predicting future ED visits, and its forecast results can be used to aid decision-making processes. PMID:29196487

CSEP-Japan: The Japanese node of the collaboratory for the study of earthquake predictability

NASA Astrophysics Data System (ADS)

Yokoi, S.; Tsuruoka, H.; Nanjo, K.; Hirata, N.

2011-12-01

Collaboratory for the Study of Earthquake Predictability (CSEP) is a global project of earthquake predictability research. The final goal of this project is to have a look for the intrinsic predictability of the earthquake rupture process through forecast testing experiments. The Earthquake Research Institute, the University of Tokyo joined the CSEP and started the Japanese testing center called as CSEP-Japan. This testing center constitutes an open access to researchers contributing earthquake forecast models for applied to Japan. A total of 91 earthquake forecast models were submitted on the prospective experiment starting from 1 November 2009. The models are separated into 4 testing classes (1 day, 3 months, 1 year and 3 years) and 3 testing regions covering an area of Japan including sea area, Japanese mainland and Kanto district. We evaluate the performance of the models in the official suite of tests defined by the CSEP. The experiments of 1-day, 3-month, 1-year and 3-year forecasting classes were implemented for 92 rounds, 4 rounds, 1round and 0 round (now in progress), respectively. The results of the 3-month class gave us new knowledge concerning statistical forecasting models. All models showed a good performance for magnitude forecasting. On the other hand, observation is hardly consistent in space-distribution with most models in some cases where many earthquakes occurred at the same spot. Throughout the experiment, it has been clarified that some properties of the CSEP's evaluation tests such as the L-test show strong correlation with the N-test. We are now processing to own (cyber-) infrastructure to support the forecast experiment as follows. (1) Japanese seismicity has changed since the 2011 Tohoku earthquake. The 3rd call for forecasting models was announced in order to promote model improvement for forecasting earthquakes after this earthquake. So, we provide Japanese seismicity catalog maintained by JMA for modelers to study how seismicity changes in Japan. (2) Now we prepare the 3-D forecasting experiment with a depth range of 0 to 100 km in Kanto region. (3) The testing center improved an evaluation system for 1-day class experiment because this testing class required fast calculation ability to finish forecasting and testing results within one day. This development will make a real-time forecasting system come true. (4) The special issue of 1st part titled Earthquake Forecast Testing Experiment in Japan was published on the Earth, Planets and Space Vol. 63, No.3, 2011 on March, 2011. This issue includes papers of algorithm of statistical models participating our experiment and outline of the experiment in Japan. The 2nd part of this issue, which is now on line, will be published soon. In this presentation, we will overview CSEP-Japan and results of the experiments, and discuss direction of our activity. An outline of the experiment and activities of the Japanese Testing Center are published on our WEB site;

Fuzzy logic-based analogue forecasting and hybrid modelling of horizontal visibility

NASA Astrophysics Data System (ADS)

Tuba, Zoltán; Bottyán, Zsolt

2018-04-01

Forecasting visibility is one of the greatest challenges in aviation meteorology. At the same time, high accuracy visibility forecasts can significantly reduce or make avoidable weather-related risk in aviation as well. To improve forecasting visibility, this research links fuzzy logic-based analogue forecasting and post-processed numerical weather prediction model outputs in hybrid forecast. Performance of analogue forecasting model was improved by the application of Analytic Hierarchy Process. Then, linear combination of the mentioned outputs was applied to create ultra-short term hybrid visibility prediction which gradually shifts the focus from statistical to numerical products taking their advantages during the forecast period. It gives the opportunity to bring closer the numerical visibility forecast to the observations even it is wrong initially. Complete verification of categorical forecasts was carried out; results are available for persistence and terminal aerodrome forecasts (TAF) as well in order to compare. The average value of Heidke Skill Score (HSS) of examined airports of analogue and hybrid forecasts shows very similar results even at the end of forecast period where the rate of analogue prediction in the final hybrid output is 0.1-0.2 only. However, in case of poor visibility (1000-2500 m), hybrid (0.65) and analogue forecasts (0.64) have similar average of HSS in the first 6 h of forecast period, and have better performance than persistence (0.60) or TAF (0.56). Important achievement that hybrid model takes into consideration physics and dynamics of the atmosphere due to the increasing part of the numerical weather prediction. In spite of this, its performance is similar to the most effective visibility forecasting methods and does not follow the poor verification results of clearly numerical outputs.

Improving medium-range and seasonal hydroclimate forecasts in the southeast USA

NASA Astrophysics Data System (ADS)

Tian, Di

Accurate hydro-climate forecasts are important for decision making by water managers, agricultural producers, and other stake holders. Numerical weather prediction models and general circulation models may have potential for improving hydro-climate forecasts at different scales. In this study, forecast analogs of the Global Forecast System (GFS) and Global Ensemble Forecast System (GEFS) based on different approaches were evaluated for medium-range reference evapotranspiration (ETo), irrigation scheduling, and urban water demand forecasts in the southeast United States; the Climate Forecast System version 2 (CFSv2) and the North American national multi-model ensemble (NMME) were statistically downscaled for seasonal forecasts of ETo, precipitation (P) and 2-m temperature (T2M) at the regional level. The GFS mean temperature (Tmean), relative humidity, and wind speed (Wind) reforecasts combined with the climatology of Reanalysis 2 solar radiation (Rs) produced higher skill than using the direct GFS output only. Constructed analogs showed slightly higher skill than natural analogs for deterministic forecasts. Both irrigation scheduling driven by the GEFS-based ETo forecasts and GEFS-based ETo forecast skill were generally positive up to one week throughout the year. The GEFS improved ETo forecast skill compared to the GFS. The GEFS-based analog forecasts for the input variables of an operational urban water demand model were skillful when applied in the Tampa Bay area. The modified operational models driven by GEFS analog forecasts showed higher forecast skill than the operational model based on persistence. The results for CFSv2 seasonal forecasts showed maximum temperature (Tmax) and Rs had the greatest influence on ETo. The downscaled Tmax showed the highest predictability, followed by Tmean, Tmin, Rs, and Wind. The CFSv2 model could better predict ETo in cold seasons during El Nino Southern Oscillation (ENSO) events only when the forecast initial condition was in ENSO. Downscaled P and T2M forecasts were produced by directly downscaling the NMME P and T2M output or indirectly using the NMME forecasts of Nino3.4 sea surface temperatures to predict local-scale P and T2M. The indirect method generally showed the highest forecast skill which occurs in cold seasons. The bias-corrected NMME ensemble forecast skill did not outperform the best single model.

A Bayesian joint probability modeling approach for seasonal forecasting of streamflows at multiple sites

NASA Astrophysics Data System (ADS)

Wang, Q. J.; Robertson, D. E.; Chiew, F. H. S.

2009-05-01

Seasonal forecasting of streamflows can be highly valuable for water resources management. In this paper, a Bayesian joint probability (BJP) modeling approach for seasonal forecasting of streamflows at multiple sites is presented. A Box-Cox transformed multivariate normal distribution is proposed to model the joint distribution of future streamflows and their predictors such as antecedent streamflows and El Niño-Southern Oscillation indices and other climate indicators. Bayesian inference of model parameters and uncertainties is implemented using Markov chain Monte Carlo sampling, leading to joint probabilistic forecasts of streamflows at multiple sites. The model provides a parametric structure for quantifying relationships between variables, including intersite correlations. The Box-Cox transformed multivariate normal distribution has considerable flexibility for modeling a wide range of predictors and predictands. The Bayesian inference formulated allows the use of data that contain nonconcurrent and missing records. The model flexibility and data-handling ability means that the BJP modeling approach is potentially of wide practical application. The paper also presents a number of statistical measures and graphical methods for verification of probabilistic forecasts of continuous variables. Results for streamflows at three river gauges in the Murrumbidgee River catchment in southeast Australia show that the BJP modeling approach has good forecast quality and that the fitted model is consistent with observed data.

Benchmark analysis of forecasted seasonal temperature over different climatic areas

NASA Astrophysics Data System (ADS)

Giunta, G.; Salerno, R.; Ceppi, A.; Ercolani, G.; Mancini, M.

2015-12-01

From a long-term perspective, an improvement of seasonal forecasting, which is often exclusively based on climatology, could provide a new capability for the management of energy resources in a time scale of just a few months. This paper regards a benchmark analysis in relation to long-term temperature forecasts over Italy in the year 2010, comparing the eni-kassandra meteo forecast (e-kmf®) model, the Climate Forecast System-National Centers for Environmental Prediction (CFS-NCEP) model, and the climatological reference (based on 25-year data) with observations. Statistical indexes are used to understand the reliability of the prediction of 2-m monthly air temperatures with a perspective of 12 weeks ahead. The results show how the best performance is achieved by the e-kmf® system which improves the reliability for long-term forecasts compared to climatology and the CFS-NCEP model. By using the reliable high-performance forecast system, it is possible to optimize the natural gas portfolio and management operations, thereby obtaining a competitive advantage in the European energy market.

A hybrid least squares support vector machines and GMDH approach for river flow forecasting

NASA Astrophysics Data System (ADS)

Samsudin, R.; Saad, P.; Shabri, A.

2010-06-01

This paper proposes a novel hybrid forecasting model, which combines the group method of data handling (GMDH) and the least squares support vector machine (LSSVM), known as GLSSVM. The GMDH is used to determine the useful input variables for LSSVM model and the LSSVM model which works as time series forecasting. In this study the application of GLSSVM for monthly river flow forecasting of Selangor and Bernam River are investigated. The results of the proposed GLSSVM approach are compared with the conventional artificial neural network (ANN) models, Autoregressive Integrated Moving Average (ARIMA) model, GMDH and LSSVM models using the long term observations of monthly river flow discharge. The standard statistical, the root mean square error (RMSE) and coefficient of correlation (R) are employed to evaluate the performance of various models developed. Experiment result indicates that the hybrid model was powerful tools to model discharge time series and can be applied successfully in complex hydrological modeling.

Daily air quality index forecasting with hybrid models: A case in China.

PubMed

Zhu, Suling; Lian, Xiuyuan; Liu, Haixia; Hu, Jianming; Wang, Yuanyuan; Che, Jinxing

2017-12-01

Air quality is closely related to quality of life. Air pollution forecasting plays a vital role in air pollution warnings and controlling. However, it is difficult to attain accurate forecasts for air pollution indexes because the original data are non-stationary and chaotic. The existing forecasting methods, such as multiple linear models, autoregressive integrated moving average (ARIMA) and support vector regression (SVR), cannot fully capture the information from series of pollution indexes. Therefore, new effective techniques need to be proposed to forecast air pollution indexes. The main purpose of this research is to develop effective forecasting models for regional air quality indexes (AQI) to address the problems above and enhance forecasting accuracy. Therefore, two hybrid models (EMD-SVR-Hybrid and EMD-IMFs-Hybrid) are proposed to forecast AQI data. The main steps of the EMD-SVR-Hybrid model are as follows: the data preprocessing technique EMD (empirical mode decomposition) is utilized to sift the original AQI data to obtain one group of smoother IMFs (intrinsic mode functions) and a noise series, where the IMFs contain the important information (level, fluctuations and others) from the original AQI series. LS-SVR is applied to forecast the sum of the IMFs, and then, S-ARIMA (seasonal ARIMA) is employed to forecast the residual sequence of LS-SVR. In addition, EMD-IMFs-Hybrid first separately forecasts the IMFs via statistical models and sums the forecasting results of the IMFs as EMD-IMFs. Then, S-ARIMA is employed to forecast the residuals of EMD-IMFs. To certify the proposed hybrid model, AQI data from June 2014 to August 2015 collected from Xingtai in China are utilized as a test case to investigate the empirical research. In terms of some of the forecasting assessment measures, the AQI forecasting results of Xingtai show that the two proposed hybrid models are superior to ARIMA, SVR, GRNN, EMD-GRNN, Wavelet-GRNN and Wavelet-SVR. Therefore, the proposed hybrid models can be used as effective and simple tools for air pollution forecasting and warning as well as for management. Copyright © 2017 Elsevier Ltd. All rights reserved.

Cloud Impacts on Pavement Temperature in Energy Balance Models

NASA Astrophysics Data System (ADS)

Walker, C. L.

2013-12-01

Forecast systems provide decision support for end-users ranging from the solar energy industry to municipalities concerned with road safety. Pavement temperature is an important variable when considering vehicle response to various weather conditions. A complex, yet direct relationship exists between tire and pavement temperatures. Literature has shown that as tire temperature increases, friction decreases which affects vehicle performance. Many forecast systems suffer from inaccurate radiation forecasts resulting in part from the inability to model different types of clouds and their influence on radiation. This research focused on forecast improvement by determining how cloud type impacts the amount of shortwave radiation reaching the surface and subsequent pavement temperatures. The study region was the Great Plains where surface solar radiation data were obtained from the High Plains Regional Climate Center's Automated Weather Data Network stations. Road pavement temperature data were obtained from the Meteorological Assimilation Data Ingest System. Cloud properties and radiative transfer quantities were obtained from the Clouds and Earth's Radiant Energy System mission via Aqua and Terra Moderate Resolution Imaging Spectroradiometer satellite products. An additional cloud data set was incorporated from the Naval Research Laboratory Cloud Classification algorithm. Statistical analyses using a modified nearest neighbor approach were first performed relating shortwave radiation variability with road pavement temperature fluctuations. Then statistical associations were determined between the shortwave radiation and cloud property data sets. Preliminary results suggest that substantial pavement forecasting improvement is possible with the inclusion of cloud-specific information. Future model sensitivity testing seeks to quantify the magnitude of forecast improvement.

Can bias correction and statistical downscaling methods improve the skill of seasonal precipitation forecasts?

NASA Astrophysics Data System (ADS)

Manzanas, R.; Lucero, A.; Weisheimer, A.; Gutiérrez, J. M.

2018-02-01

Statistical downscaling methods are popular post-processing tools which are widely used in many sectors to adapt the coarse-resolution biased outputs from global climate simulations to the regional-to-local scale typically required by users. They range from simple and pragmatic Bias Correction (BC) methods, which directly adjust the model outputs of interest (e.g. precipitation) according to the available local observations, to more complex Perfect Prognosis (PP) ones, which indirectly derive local predictions (e.g. precipitation) from appropriate upper-air large-scale model variables (predictors). Statistical downscaling methods have been extensively used and critically assessed in climate change applications; however, their advantages and limitations in seasonal forecasting are not well understood yet. In particular, a key problem in this context is whether they serve to improve the forecast quality/skill of raw model outputs beyond the adjustment of their systematic biases. In this paper we analyze this issue by applying two state-of-the-art BC and two PP methods to downscale precipitation from a multimodel seasonal hindcast in a challenging tropical region, the Philippines. To properly assess the potential added value beyond the reduction of model biases, we consider two validation scores which are not sensitive to changes in the mean (correlation and reliability categories). Our results show that, whereas BC methods maintain or worsen the skill of the raw model forecasts, PP methods can yield significant skill improvement (worsening) in cases for which the large-scale predictor variables considered are better (worse) predicted by the model than precipitation. For instance, PP methods are found to increase (decrease) model reliability in nearly 40% of the stations considered in boreal summer (autumn). Therefore, the choice of a convenient downscaling approach (either BC or PP) depends on the region and the season.

scoringRules - A software package for probabilistic model evaluation

NASA Astrophysics Data System (ADS)

Lerch, Sebastian; Jordan, Alexander; Krüger, Fabian

2016-04-01

Models in the geosciences are generally surrounded by uncertainty, and being able to quantify this uncertainty is key to good decision making. Accordingly, probabilistic forecasts in the form of predictive distributions have become popular over the last decades. With the proliferation of probabilistic models arises the need for decision theoretically principled tools to evaluate the appropriateness of models and forecasts in a generalized way. Various scoring rules have been developed over the past decades to address this demand. Proper scoring rules are functions S(F,y) which evaluate the accuracy of a forecast distribution F , given that an outcome y was observed. As such, they allow to compare alternative models, a crucial ability given the variety of theories, data sources and statistical specifications that is available in many situations. This poster presents the software package scoringRules for the statistical programming language R, which contains functions to compute popular scoring rules such as the continuous ranked probability score for a variety of distributions F that come up in applied work. Two main classes are parametric distributions like normal, t, or gamma distributions, and distributions that are not known analytically, but are indirectly described through a sample of simulation draws. For example, Bayesian forecasts produced via Markov Chain Monte Carlo take this form. Thereby, the scoringRules package provides a framework for generalized model evaluation that both includes Bayesian as well as classical parametric models. The scoringRules package aims to be a convenient dictionary-like reference for computing scoring rules. We offer state of the art implementations of several known (but not routinely applied) formulas, and implement closed-form expressions that were previously unavailable. Whenever more than one implementation variant exists, we offer statistically principled default choices.

Advanced inflow forecasting for a hydropower plant in an Alpine hydropower regulated catchment - coupling of operational and hydrological forecasts

NASA Astrophysics Data System (ADS)

Tilg, Anna-Maria; Schöber, Johannes; Huttenlau, Matthias; Messner, Jakob; Achleitner, Stefan

2017-04-01

Hydropower is a renewable energy source which can help to stabilize fluctuations in the volatile energy market. Especially pumped-storage infrastructures in the European Alps play an important role within the European energy grid system. Today, the runoff of rivers in the Alps is often influenced by cascades of hydropower infrastructures where the operational procedures are triggered by energy market demands, water deliveries and flood control aspects rather than by hydro-meteorological variables. An example for such a highly hydropower regulated river is the catchment of the river Inn in the Eastern European Alps, originating in the Engadin (Switzerland). A new hydropower plant is going to be built as transboundary project at the boarder of Switzerland and Austria using the water of the Inn River. For the operation, a runoff forecast to the plant is required. The challenge in this case is that a high proportion of runoff is turbine water from an upstream situated hydropower cascade. The newly developed physically based hydrological forecasting system is mainly capable to cover natural hydrological runoff processes caused by storms and snow melt but can model only a small degree of human impact. These discontinuous parts of the runoff downstream of the pumped storage are described by means of an additional statistical model which has been developed. The main goal of the statistical model is to forecast the turbine water up to five days in advance. The lead time of the data driven model exceeds the lead time of the used energy production forecast. Additionally, the amount of turbine water is linked to the need of electricity production and the electricity price. It has been shown that especially the parameters day-ahead prognosis of the energy production and turbine inflow of the previous week are good predictors and are therefore used as input parameters for the model. As the data is restricted due to technical conditions, so-called Tobit models have been used to develop a linear regression for the runoff forecast. Although the day-ahead prognosis cannot always be kept, the regression model delivers, especially during office hours, very reasonable results. In the remaining hours the error between measurement and the forecast increases. Overall, the inflow forecast can be substantially improved by the implementation of the developed regression in the hydrological modelling system.

An application of ensemble/multi model approach for wind power production forecasting

NASA Astrophysics Data System (ADS)

Alessandrini, S.; Pinson, P.; Hagedorn, R.; Decimi, G.; Sperati, S.

2011-02-01

The wind power forecasts of the 3 days ahead period are becoming always more useful and important in reducing the problem of grid integration and energy price trading due to the increasing wind power penetration. Therefore it's clear that the accuracy of this forecast is one of the most important requirements for a successful application. The wind power forecast applied in this study is based on meteorological models that provide the 3 days ahead wind data. A Model Output Statistic correction is then performed to reduce systematic error caused, for instance, by a wrong representation of surface roughness or topography in the meteorological models. For this purpose a training of a Neural Network (NN) to link directly the forecasted meteorological data and the power data has been performed. One wind farm has been examined located in a mountain area in the south of Italy (Sicily). First we compare the performances of a prediction based on meteorological data coming from a single model with those obtained by the combination of models (RAMS, ECMWF deterministic, LAMI). It is shown that the multi models approach reduces the day-ahead normalized RMSE forecast error (normalized by nominal power) of at least 1% compared to the singles models approach. Finally we have focused on the possibility of using the ensemble model system (EPS by ECMWF) to estimate the hourly, three days ahead, power forecast accuracy. Contingency diagram between RMSE of the deterministic power forecast and the ensemble members spread of wind forecast have been produced. From this first analysis it seems that ensemble spread could be used as an indicator of the forecast's accuracy at least for the first three days ahead period.

Analysis and Synthesis of Load Forecasting Data for Renewable Integration Studies: Preprint

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

Steckler, N.; Florita, A.; Zhang, J.

2013-11-01

As renewable energy constitutes greater portions of the generation fleet, the importance of modeling uncertainty as part of integration studies also increases. In pursuit of optimal system operations, it is important to capture not only the definitive behavior of power plants, but also the risks associated with systemwide interactions. This research examines the dependence of load forecast errors on external predictor variables such as temperature, day type, and time of day. The analysis was utilized to create statistically relevant instances of sequential load forecasts with only a time series of historic, measured load available. The creation of such load forecastsmore » relies on Bayesian techniques for informing and updating the model, thus providing a basis for networked and adaptive load forecast models in future operational applications.« less

Combining empirical approaches and error modelling to enhance predictive uncertainty estimation in extrapolation for operational flood forecasting. Tests on flood events on the Loire basin, France.

NASA Astrophysics Data System (ADS)

Berthet, Lionel; Marty, Renaud; Bourgin, François; Viatgé, Julie; Piotte, Olivier; Perrin, Charles

2017-04-01

An increasing number of operational flood forecasting centres assess the predictive uncertainty associated with their forecasts and communicate it to the end users. This information can match the end-users needs (i.e. prove to be useful for an efficient crisis management) only if it is reliable: reliability is therefore a key quality for operational flood forecasts. In 2015, the French flood forecasting national and regional services (Vigicrues network; www.vigicrues.gouv.fr) implemented a framework to compute quantitative discharge and water level forecasts and to assess the predictive uncertainty. Among the possible technical options to achieve this goal, a statistical analysis of past forecasting errors of deterministic models has been selected (QUOIQUE method, Bourgin, 2014). It is a data-based and non-parametric approach based on as few assumptions as possible about the forecasting error mathematical structure. In particular, a very simple assumption is made regarding the predictive uncertainty distributions for large events outside the range of the calibration data: the multiplicative error distribution is assumed to be constant, whatever the magnitude of the flood. Indeed, the predictive distributions may not be reliable in extrapolation. However, estimating the predictive uncertainty for these rare events is crucial when major floods are of concern. In order to improve the forecasts reliability for major floods, an attempt at combining the operational strength of the empirical statistical analysis and a simple error modelling is done. Since the heteroscedasticity of forecast errors can considerably weaken the predictive reliability for large floods, this error modelling is based on the log-sinh transformation which proved to reduce significantly the heteroscedasticity of the transformed error in a simulation context, even for flood peaks (Wang et al., 2012). Exploratory tests on some operational forecasts issued during the recent floods experienced in France (major spring floods in June 2016 on the Loire river tributaries and flash floods in fall 2016) will be shown and discussed. References Bourgin, F. (2014). How to assess the predictive uncertainty in hydrological modelling? An exploratory work on a large sample of watersheds, AgroParisTech Wang, Q. J., Shrestha, D. L., Robertson, D. E. and Pokhrel, P (2012). A log-sinh transformation for data normalization and variance stabilization. Water Resources Research, , W05514, doi:10.1029/2011WR010973

Comparative assessment of several post-processing methods for correcting evapotranspiration forecasts derived from TIGGE datasets.

NASA Astrophysics Data System (ADS)

Tian, D.; Medina, H.

2017-12-01

Post-processing of medium range reference evapotranspiration (ETo) forecasts based on numerical weather prediction (NWP) models has the potential of improving the quality and utility of these forecasts. This work compares the performance of several post-processing methods for correcting ETo forecasts over the continental U.S. generated from The Observing System Research and Predictability Experiment (THORPEX) Interactive Grand Global Ensemble (TIGGE) database using data from Europe (EC), the United Kingdom (MO), and the United States (NCEP). The pondered post-processing techniques are: simple bias correction, the use of multimodels, the Ensemble Model Output Statistics (EMOS, Gneitting et al., 2005) and the Bayesian Model Averaging (BMA, Raftery et al., 2005). ETo estimates based on quality-controlled U.S. Regional Climate Reference Network measurements, and computed with the FAO 56 Penman Monteith equation, are adopted as baseline. EMOS and BMA are generally the most efficient post-processing techniques of the ETo forecasts. Nevertheless, the simple bias correction of the best model is commonly much more rewarding than using multimodel raw forecasts. Our results demonstrate the potential of different forecasting and post-processing frameworks in operational evapotranspiration and irrigation advisory systems at national scale.

Bayesian model averaging using particle filtering and Gaussian mixture modeling: Theory, concepts, and simulation experiments

NASA Astrophysics Data System (ADS)

Rings, Joerg; Vrugt, Jasper A.; Schoups, Gerrit; Huisman, Johan A.; Vereecken, Harry

2012-05-01

Bayesian model averaging (BMA) is a standard method for combining predictive distributions from different models. In recent years, this method has enjoyed widespread application and use in many fields of study to improve the spread-skill relationship of forecast ensembles. The BMA predictive probability density function (pdf) of any quantity of interest is a weighted average of pdfs centered around the individual (possibly bias-corrected) forecasts, where the weights are equal to posterior probabilities of the models generating the forecasts, and reflect the individual models skill over a training (calibration) period. The original BMA approach presented by Raftery et al. (2005) assumes that the conditional pdf of each individual model is adequately described with a rather standard Gaussian or Gamma statistical distribution, possibly with a heteroscedastic variance. Here we analyze the advantages of using BMA with a flexible representation of the conditional pdf. A joint particle filtering and Gaussian mixture modeling framework is presented to derive analytically, as closely and consistently as possible, the evolving forecast density (conditional pdf) of each constituent ensemble member. The median forecasts and evolving conditional pdfs of the constituent models are subsequently combined using BMA to derive one overall predictive distribution. This paper introduces the theory and concepts of this new ensemble postprocessing method, and demonstrates its usefulness and applicability by numerical simulation of the rainfall-runoff transformation using discharge data from three different catchments in the contiguous United States. The revised BMA method receives significantly lower-prediction errors than the original default BMA method (due to filtering) with predictive uncertainty intervals that are substantially smaller but still statistically coherent (due to the use of a time-variant conditional pdf).

Forecasting influenza in Hong Kong with Google search queries and statistical model fusion

PubMed Central

Ramirez Ramirez, L. Leticia; Nezafati, Kusha; Zhang, Qingpeng; Tsui, Kwok-Leung

2017-01-01

Background The objective of this study is to investigate predictive utility of online social media and web search queries, particularly, Google search data, to forecast new cases of influenza-like-illness (ILI) in general outpatient clinics (GOPC) in Hong Kong. To mitigate the impact of sensitivity to self-excitement (i.e., fickle media interest) and other artifacts of online social media data, in our approach we fuse multiple offline and online data sources. Methods Four individual models: generalized linear model (GLM), least absolute shrinkage and selection operator (LASSO), autoregressive integrated moving average (ARIMA), and deep learning (DL) with Feedforward Neural Networks (FNN) are employed to forecast ILI-GOPC both one week and two weeks in advance. The covariates include Google search queries, meteorological data, and previously recorded offline ILI. To our knowledge, this is the first study that introduces deep learning methodology into surveillance of infectious diseases and investigates its predictive utility. Furthermore, to exploit the strength from each individual forecasting models, we use statistical model fusion, using Bayesian model averaging (BMA), which allows a systematic integration of multiple forecast scenarios. For each model, an adaptive approach is used to capture the recent relationship between ILI and covariates. Results DL with FNN appears to deliver the most competitive predictive performance among the four considered individual models. Combing all four models in a comprehensive BMA framework allows to further improve such predictive evaluation metrics as root mean squared error (RMSE) and mean absolute predictive error (MAPE). Nevertheless, DL with FNN remains the preferred method for predicting locations of influenza peaks. Conclusions The proposed approach can be viewed a feasible alternative to forecast ILI in Hong Kong or other countries where ILI has no constant seasonal trend and influenza data resources are limited. The proposed methodology is easily tractable and computationally efficient. PMID:28464015

MesoNAM Verification Phase II

NASA Technical Reports Server (NTRS)

Watson, Leela R.

2011-01-01

The 45th Weather Squadron Launch Weather Officers use the 12-km resolution North American Mesoscale model (MesoNAM) forecasts to support launch weather operations. In Phase I, the performance of the model at KSC/CCAFS was measured objectively by conducting a detailed statistical analysis of model output compared to observed values. The objective analysis compared the MesoNAM forecast winds, temperature, and dew point to the observed values from the sensors in the KSC/CCAFS wind tower network. In Phase II, the AMU modified the current tool by adding an additional 15 months of model output to the database and recalculating the verification statistics. The bias, standard deviation of bias, Root Mean Square Error, and Hypothesis test for bias were calculated to verify the performance of the model. The results indicated that the accuracy decreased as the forecast progressed, there was a diurnal signal in temperature with a cool bias during the late night and a warm bias during the afternoon, and there was a diurnal signal in dewpoint temperature with a low bias during the afternoon and a high bias during the late night.

Real-time Mainshock Forecast by Statistical Discrimination of Foreshock Clusters

NASA Astrophysics Data System (ADS)

Nomura, S.; Ogata, Y.

2016-12-01

Foreshock discremination is one of the most effective ways for short-time forecast of large main shocks. Though many large earthquakes accompany their foreshocks, discreminating them from enormous small earthquakes is difficult and only probabilistic evaluation from their spatio-temporal features and magnitude evolution may be available. Logistic regression is the statistical learning method best suited to such binary pattern recognition problems where estimates of a-posteriori probability of class membership are required. Statistical learning methods can keep learning discreminating features from updating catalog and give probabilistic recognition of forecast in real time. We estimated a non-linear function of foreshock proportion by smooth spline bases and evaluate the possibility of foreshocks by the logit function. In this study, we classified foreshocks from earthquake catalog by the Japan Meteorological Agency by single-link clustering methods and learned spatial and temporal features of foreshocks by the probability density ratio estimation. We use the epicentral locations, time spans and difference in magnitudes for learning and forecasting. Magnitudes of main shocks are also predicted our method by incorporating b-values into our method. We discuss the spatial pattern of foreshocks from the classifier composed by our model. We also implement a back test to validate predictive performance of the model by this catalog.

The development and evaluation of a hydrological seasonal forecast system prototype for predicting spring flood volumes in Swedish rivers

NASA Astrophysics Data System (ADS)

Foster, Kean; Bertacchi Uvo, Cintia; Olsson, Jonas

2018-05-01

Hydropower makes up nearly half of Sweden's electrical energy production. However, the distribution of the water resources is not aligned with demand, as most of the inflows to the reservoirs occur during the spring flood period. This means that carefully planned reservoir management is required to help redistribute water resources to ensure optimal production and accurate forecasts of the spring flood volume (SFV) is essential for this. The current operational SFV forecasts use a historical ensemble approach where the HBV model is forced with historical observations of precipitation and temperature. In this work we develop and test a multi-model prototype, building on previous work, and evaluate its ability to forecast the SFV in 84 sub-basins in northern Sweden. The hypothesis explored in this work is that a multi-model seasonal forecast system incorporating different modelling approaches is generally more skilful at forecasting the SFV in snow dominated regions than a forecast system that utilises only one approach. The testing is done using cross-validated hindcasts for the period 1981-2015 and the results are evaluated against both climatology and the current system to determine skill. Both the multi-model methods considered showed skill over the reference forecasts. The version that combined the historical modelling chain, dynamical modelling chain, and statistical modelling chain performed better than the other and was chosen for the prototype. The prototype was able to outperform the current operational system 57 % of the time on average and reduce the error in the SFV by ˜ 6 % across all sub-basins and forecast dates.

Moment-Based Physical Models of Broadband Clutter due to Aggregations of Fish

DTIC Science & Technology

2013-09-30

statistical models for signal-processing algorithm development. These in turn will help to develop a capability to statistically forecast the impact of...aggregations of fish based on higher-order statistical measures describable in terms of physical and system parameters. Environmentally , these models...processing. In this experiment, we had good ground truth on (1) and (2), and had control over (3) and (4) except for environmentally -imposed restrictions

Long-range forecast of all India summer monsoon rainfall using adaptive neuro-fuzzy inference system: skill comparison with CFSv2 model simulation and real-time forecast for the year 2015

NASA Astrophysics Data System (ADS)

Chaudhuri, S.; Das, D.; Goswami, S.; Das, S. K.

2016-11-01

All India summer monsoon rainfall (AISMR) characteristics play a vital role for the policy planning and national economy of the country. In view of the significant impact of monsoon system on regional as well as global climate systems, accurate prediction of summer monsoon rainfall has become a challenge. The objective of this study is to develop an adaptive neuro-fuzzy inference system (ANFIS) for long range forecast of AISMR. The NCEP/NCAR reanalysis data of temperature, zonal and meridional wind at different pressure levels have been taken to construct the input matrix of ANFIS. The membership of the input parameters for AISMR as high, medium or low is estimated with trapezoidal membership function. The fuzzified standardized input parameters and the de-fuzzified target output are trained with artificial neural network models. The forecast of AISMR with ANFIS is compared with non-hybrid multi-layer perceptron model (MLP), radial basis functions network (RBFN) and multiple linear regression (MLR) models. The forecast error analyses of the models reveal that ANFIS provides the best forecast of AISMR with minimum prediction error of 0.076, whereas the errors with MLP, RBFN and MLR models are 0.22, 0.18 and 0.73 respectively. During validation with observations, ANFIS shows its potency over the said comparative models. Performance of the ANFIS model is verified through different statistical skill scores, which also confirms the aptitude of ANFIS in forecasting AISMR. The forecast skill of ANFIS is also observed to be better than Climate Forecast System version 2. The real-time forecast with ANFIS shows possibility of deficit (65-75 cm) AISMR in the year 2015.

Machine Learning Based Multi-Physical-Model Blending for Enhancing Renewable Energy Forecast -- Improvement via Situation Dependent Error Correction

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

Lu, Siyuan; Hwang, Youngdeok; Khabibrakhmanov, Ildar

With increasing penetration of solar and wind energy to the total energy supply mix, the pressing need for accurate energy forecasting has become well-recognized. Here we report the development of a machine-learning based model blending approach for statistically combining multiple meteorological models for improving the accuracy of solar/wind power forecast. Importantly, we demonstrate that in addition to parameters to be predicted (such as solar irradiance and power), including additional atmospheric state parameters which collectively define weather situations as machine learning input provides further enhanced accuracy for the blended result. Functional analysis of variance shows that the error of individual modelmore » has substantial dependence on the weather situation. The machine-learning approach effectively reduces such situation dependent error thus produces more accurate results compared to conventional multi-model ensemble approaches based on simplistic equally or unequally weighted model averaging. Validation over an extended period of time results show over 30% improvement in solar irradiance/power forecast accuracy compared to forecasts based on the best individual model.« less

Dynamic Statistical Models for Pyroclastic Density Current Generation at Soufrière Hills Volcano

NASA Astrophysics Data System (ADS)

Wolpert, Robert L.; Spiller, Elaine T.; Calder, Eliza S.

2018-05-01

To mitigate volcanic hazards from pyroclastic density currents, volcanologists generate hazard maps that provide long-term forecasts of areas of potential impact. Several recent efforts in the field develop new statistical methods for application of flow models to generate fully probabilistic hazard maps that both account for, and quantify, uncertainty. However a limitation to the use of most statistical hazard models, and a key source of uncertainty within them, is the time-averaged nature of the datasets by which the volcanic activity is statistically characterized. Where the level, or directionality, of volcanic activity frequently changes, e.g. during protracted eruptive episodes, or at volcanoes that are classified as persistently active, it is not appropriate to make short term forecasts based on longer time-averaged metrics of the activity. Thus, here we build, fit and explore dynamic statistical models for the generation of pyroclastic density current from Soufrière Hills Volcano (SHV) on Montserrat including their respective collapse direction and flow volumes based on 1996-2008 flow datasets. The development of this approach allows for short-term behavioral changes to be taken into account in probabilistic volcanic hazard assessments. We show that collapses from the SHV lava dome follow a clear pattern, and that a series of smaller flows in a given direction often culminate in a larger collapse and thereafter directionality of the flows change. Such models enable short term forecasting (weeks to months) that can reflect evolving conditions such as dome and crater morphology changes and non-stationary eruptive behavior such as extrusion rate variations. For example, the probability of inundation of the Belham Valley in the first 180 days of a forecast period is about twice as high for lava domes facing Northwest toward that valley as it is for domes pointing East toward the Tar River Valley. As rich multi-parametric volcano monitoring dataset become increasingly available, eruption forecasting is becoming an increasingly viable and important research field. We demonstrate an approach to utilize such data in order to appropriately 'tune' probabilistic hazard assessments for pyroclastic flows. Our broader objective with development of this method is to help advance time-dependent volcanic hazard assessment, by bridging the

Objective Lightning Forecasting at Kennedy Space Center and Cape Canaveral Air Force Station using Cloud-to-Ground Lightning Surveillance System Data

NASA Technical Reports Server (NTRS)

Lambert, Winfred; Wheeler, Mark; Roeder, William

2005-01-01

The 45th Weather Squadron (45 WS) at Cape Canaveral Air-Force Station (CCAFS)ln Florida issues a probability of lightning occurrence in their daily 24-hour and weekly planning forecasts. This information is used for general planning of operations at CCAFS and Kennedy Space Center (KSC). These facilities are located in east-central Florida at the east end of a corridor known as 'Lightning Alley', an indication that lightning has a large impact on space-lift operations. Much of the current lightning probability forecast is based on a subjective analysis of model and observational data and an objective forecast tool developed over 30 years ago. The 45 WS requested that a new lightning probability forecast tool based on statistical analysis of more recent historical warm season (May-September) data be developed in order to increase the objectivity of the daily thunderstorm probability forecast. The resulting tool is a set of statistical lightning forecast equations, one for each month of the warm season, that provide a lightning occurrence probability for the day by 1100 UTC (0700 EDT) during the warm season.

Comparative Analysis of River Flow Modelling by Using Supervised Learning Technique

NASA Astrophysics Data System (ADS)

Ismail, Shuhaida; Mohamad Pandiahi, Siraj; Shabri, Ani; Mustapha, Aida

2018-04-01

The goal of this research is to investigate the efficiency of three supervised learning algorithms for forecasting monthly river flow of the Indus River in Pakistan, spread over 550 square miles or 1800 square kilometres. The algorithms include the Least Square Support Vector Machine (LSSVM), Artificial Neural Network (ANN) and Wavelet Regression (WR). The forecasting models predict the monthly river flow obtained from the three models individually for river flow data and the accuracy of the all models were then compared against each other. The monthly river flow of the said river has been forecasted using these three models. The obtained results were compared and statistically analysed. Then, the results of this analytical comparison showed that LSSVM model is more precise in the monthly river flow forecasting. It was found that LSSVM has he higher r with the value of 0.934 compared to other models. This indicate that LSSVM is more accurate and efficient as compared to the ANN and WR model.

Statistical forecasting of repetitious dome failures during the waning eruption of Redoubt Volcano, Alaska, February-April 1990

USGS Publications Warehouse

Page, R.A.; Lahr, J.C.; Chouet, B.A.; Power, J.A.; Stephens, C.D.

1994-01-01

The waning phase of the 1989-1990 eruption of Redoubt Volcano in the Cook Inlet region of south-central Alaska comprised a quasi-regular pattern of repetitious dome growth and destruction that lasted from February 15 to late April 1990. The dome failures produced ash plumes hazardous to airline traffic. In response to this hazard, the Alaska Volcano Observatory sought to forecast these ash-producing events using two approaches. One approach built on early successes in issuing warnings before major eruptions on December 14, 1989 and January 2, 1990. These warnings were based largely on changes in seismic activity related to the occurrence of precursory swarms of long-period seismic events. The search for precursory swarms of long-period seismicity was continued through the waning phase of the eruption and led to warnings before tephra eruptions on March 23 and April 6. The observed regularity of dome failures after February 15 suggested that a statistical forecasting method based on a constant-rate failure model might also be successful. The first statistical forecast was issued on March 16 after seven events had occurred, at an average interval of 4.5 days. At this time, the interval between dome failures abruptly lengthened. Accordingly, the forecast was unsuccessful and further forecasting was suspended until the regularity of subsequent failures could be confirmed. Statistical forecasting resumed on April 12, after four dome failure episodes separated by an average of 7.8 days. One dome failure (April 15) was successfully forecast using a 70% confidence window, and a second event (April 21) was narrowly missed before the end of the activity. The cessation of dome failures after April 21 resulted in a concluding false alarm. Although forecasting success during the eruption was limited, retrospective analysis shows that early and consistent application of the statistical method using a constant-rate failure model and a 90% confidence window could have yielded five successful forecasts and two false alarms; no events would have been missed. On closer examination, the intervals between successive dome failures are not uniform but tend to increase with time. This increase attests to the continuous, slowly decreasing supply of magma to the surface vent during the waning phase of the eruption. The domes formed in a precarious position in a breach in the summit crater rim where they were susceptible to gravitational collapse. The instability of the February 15-April 21 domes relative to the earlier domes is attributed to reaming the lip of the vent by a laterally directed explosion during the major dome-destroying eruption of February 15, a process which would leave a less secure foundation for subsequent domes. ?? 1994.

Forecasting the duration of volcanic eruptions: an empirical probabilistic model

NASA Astrophysics Data System (ADS)

Gunn, L. S.; Blake, S.; Jones, M. C.; Rymer, H.

2014-01-01

The ability to forecast future volcanic eruption durations would greatly benefit emergency response planning prior to and during a volcanic crises. This paper introduces a probabilistic model to forecast the duration of future and on-going eruptions. The model fits theoretical distributions to observed duration data and relies on past eruptions being a good indicator of future activity. A dataset of historical Mt. Etna flank eruptions is presented and used to demonstrate the model. The data have been compiled through critical examination of existing literature along with careful consideration of uncertainties on reported eruption start and end dates between the years 1300 AD and 2010. Data following 1600 is considered to be reliable and free of reporting biases. The distribution of eruption duration between the years 1600 and 1669 is found to be statistically different from that following it and the forecasting model is run on two datasets of Mt. Etna flank eruption durations: 1600-2010 and 1670-2010. Each dataset is modelled using a log-logistic distribution with parameter values found by maximum likelihood estimation. Survivor function statistics are applied to the model distributions to forecast (a) the probability of an eruption exceeding a given duration, (b) the probability of an eruption that has already lasted a particular number of days exceeding a given total duration and (c) the duration with a given probability of being exceeded. Results show that excluding the 1600-1670 data has little effect on the forecasting model result, especially where short durations are involved. By assigning the terms `likely' and `unlikely' to probabilities of 66 % or more and 33 % or less, respectively, the forecasting model based on the 1600-2010 dataset indicates that a future flank eruption on Mt. Etna would be likely to exceed 20 days (± 7 days) but unlikely to exceed 86 days (± 29 days). This approach can easily be adapted for use on other highly active, well-documented volcanoes or for different duration data such as the duration of explosive episodes or the duration of repose periods between eruptions.

Statistical Properties of Global Precipitation in the NCEP GFS Model and TMPA Observations for Data Assimilation

NASA Technical Reports Server (NTRS)

Lien, Guo-Yuan; Kalnay, Eugenia; Miyoshi, Takemasa; Huffman, George J.

2016-01-01

Assimilation of satellite precipitation data into numerical models presents several difficulties, with two of the most important being the non-Gaussian error distributions associated with precipitation, and large model and observation errors. As a result, improving the model forecast beyond a few hours by assimilating precipitation has been found to be difficult. To identify the challenges and propose practical solutions to assimilation of precipitation, statistics are calculated for global precipitation in a low-resolution NCEP Global Forecast System (GFS) model and the TRMM Multisatellite Precipitation Analysis (TMPA). The samples are constructed using the same model with the same forecast period, observation variables, and resolution as in the follow-on GFSTMPA precipitation assimilation experiments presented in the companion paper.The statistical results indicate that the T62 and T126 GFS models generally have positive bias in precipitation compared to the TMPA observations, and that the simulation of the marine stratocumulus precipitation is not realistic in the T62 GFS model. It is necessary to apply to precipitation either the commonly used logarithm transformation or the newly proposed Gaussian transformation to obtain a better relationship between the model and observational precipitation. When the Gaussian transformations are separately applied to the model and observational precipitation, they serve as a bias correction that corrects the amplitude-dependent biases. In addition, using a spatially andor temporally averaged precipitation variable, such as the 6-h accumulated precipitation, should be advantageous for precipitation assimilation.

Dynamical Downscaling of Seasonal Climate Prediction over Nordeste Brazil with ECHAM3 and NCEP's Regional Spectral Models at IRI.

NASA Astrophysics Data System (ADS)

Nobre, Paulo; Moura, Antonio D.; Sun, Liqiang

2001-12-01

This study presents an evaluation of a seasonal climate forecast done with the International Research Institute for Climate Prediction (IRI) dynamical forecast system (regional model nested into a general circulation model) over northern South America for January-April 1999, encompassing the rainy season over Brazil's Nordeste. The one-way nesting is one in two tiers: first the NCEP's Regional Spectral Model (RSM) runs with an 80-km grid mesh forced by the ECHAM3 atmospheric general circulation model (AGCM) outputs; then the RSM runs with a finer grid mesh (20 km) forced by the forecasts generated by the RSM-80. An ensemble of three realizations is done. Lower boundary conditions over the oceans for both ECHAM and RSM model runs are sea surface temperature forecasts over the tropical oceans. Soil moisture is initialized by ECHAM's inputs. The rainfall forecasts generated by the regional model are compared with those of the AGCM and observations. It is shown that the regional model at 80-km resolution improves upon the AGCM rainfall forecast, reducing both seasonal bias and root-mean-square error. On the other hand, the RSM-20 forecasts presented larger errors, with spatial patterns that resemble those of local topography. The better forecast of the position and width of the intertropical convergence zone (ITCZ) over the tropical Atlantic by the RSM-80 model is one of the principal reasons for better-forecast scores of the RSM-80 relative to the AGCM. The regional model improved the spatial as well as the temporal details of rainfall distribution, and also presenting the minimum spread among the ensemble members. The statistics of synoptic-scale weather variability on seasonal timescales were best forecast with the regional 80-km model over the Nordeste. The possibility of forecasting the frequency distribution of dry and wet spells within the rainy season is encouraging.

Automated turbulence forecasts for aviation hazards

NASA Astrophysics Data System (ADS)

Sharman, R.; Frehlich, R.; Vandenberghe, F.

2010-09-01

An operational turbulence forecast system for commercial and aviation use is described that is based on an ensemble of turbulence diagnostics derived from standard NWP model outputs. In the U. S. this forecast product is named GTG (Graphical Turbulence Guidance) and has been described in detail in Sharman et al., WAF 2006. Since turbulence has many sources in the atmosphere, the ensemble approach of combining diagnostics has been shown to provide greater statistical accuracy than the use of a single diagnostic, or of a subgrid tke parameterization. GTG is sponsored by the FAA, and has undergone rigorous accuracy, safety, and usability evaluations. The GTG product is now hosted on NOAA's Aviation Data Service (ADDS), web site (http://aviationweather.gov/), for access by pilots, air traffic controllers, and dispatchers. During this talk the various turbulence diagnostics, their statistical properties, and their relative performance (based on comparisons to observations) will be presented. Importantly, the model output is ɛ1/3 (where ɛ is the eddy dissipation rate), so is aircraft independent. The diagnostics are individually and collectively calibrated so that their PDFs satisfy the expected log normal distribution of ɛ^1/3. Some of the diagnostics try to take into account the role of gravity waves and inertia-gravity waves in the turbulence generation process. Although the current GTG product is based on the RUC forecast model running over the CONUS, it is transitioning to a WRF based product, and in fact WRF-based versions are currently running operationally over Taiwan and has also been implemented for use by the French Navy in climatological studies. Yet another version has been developed which uses GFS model output to provide global turbulence forecasts. Thus the forecast product is available as a postprocessing program for WRF or other model output and provides 3D maps of turbulence likelihood of any region where NWP model data is available. Although the current GTG has been used mainly for large commercial aircraft, since the output is aircraft independent it could readily be scaled to smaller aircraft such as UAVs. Further, the ensemble technique allows the diagnostics to be used to form probabilistic forecasts, in a manner similar to ensemble NWP forecasts.

Remote and Local Influences in Forecasting Pacific SST: a Linear Inverse Model and a Multimodel Ensemble Study

NASA Astrophysics Data System (ADS)

Faggiani Dias, D.; Subramanian, A. C.; Zanna, L.; Miller, A. J.

2017-12-01

Sea surface temperature (SST) in the Pacific sector is well known to vary on time scales from seasonal to decadal, and the ability to predict these SST fluctuations has many societal and economical benefits. Therefore, we use a suite of statistical linear inverse models (LIMs) to understand the remote and local SST variability that influences SST predictions over the North Pacific region and further improve our understanding on how the long-observed SST record can help better guide multi-model ensemble forecasts. Observed monthly SST anomalies in the Pacific sector (between 15oS and 60oN) are used to construct different regional LIMs for seasonal to decadal prediction. The forecast skills of the LIMs are compared to that from two operational forecast systems in the North American Multi-Model Ensemble (NMME) revealing that the LIM has better skill in the Northeastern Pacific than NMME models. The LIM is also found to have comparable forecast skill for SST in the Tropical Pacific with NMME models. This skill, however, is highly dependent on the initialization month, with forecasts initialized during the summer having better skill than those initialized during the winter. The forecast skill with LIM is also influenced by the verification period utilized to make the predictions, likely due to the changing character of El Niño in the 20th century. The North Pacific seems to be a source of predictability for the Tropics on seasonal to interannual time scales, while the Tropics act to worsen the skill for the forecast in the North Pacific. The data were also bandpassed into seasonal, interannual and decadal time scales to identify the relationships between time scales using the structure of the propagator matrix. For the decadal component, this coupling occurs the other way around: Tropics seem to be a source of predictability for the Extratropics, but the Extratropics don't improve the predictability for the Tropics. These results indicate the importance of temporal scale interactions in improving predictability on decadal timescales. Hence, we show that LIMs are not only useful as benchmarks for estimates of statistical skill, but also to isolate contributions to the forecast skills from different timescales, spatial scales or even model components.

Implementation and Research on the Operational Use of the Mesoscale Prediction Model COAMPS in Poland

DTIC Science & Technology

2008-09-30

participated in EGU General Assembly , Vienna Austria 13-18 April 2008, giving a poster presentation. Bogumil Jakubiak, University of Warsaw...participated in EGU General Assembly , Vienna Austria 13-18 April 2008, giving two posters presentation. Mikolaj Sierzega, University of Warwick – participated...model forecast to generate background error statistics. This helps us to identify and understand the uncertainties in high-resolution NWP forecasts

The quality and value of seasonal precipitation forecasts for an early warning of large-scale droughts and floods in West Africa

NASA Astrophysics Data System (ADS)

Bliefernicht, Jan; Seidel, Jochen; Salack, Seyni; Waongo, Moussa; Laux, Patrick; Kunstmann, Harald

2017-04-01

Seasonal precipitation forecasts are a crucial source of information for an early warning of hydro-meteorological extremes in West Africa. However, the current seasonal forecasting system used by the West African weather services in the framework of the West African Climate Outlook forum (PRESAO) is limited to probabilistic precipitation forecasts of 1-month lead time. To improve this provision, we use an ensemble-based quantile-quantile transformation for bias correction of precipitation forecasts provided by a global seasonal ensemble prediction system, the Climate Forecast System Version 2 (CFS2). The statistical technique eliminates systematic differences between global forecasts and observations with the potential to preserve the signal from the model. The technique has also the advantage that it can be easily implemented at national weather services with low capacities. The statistical technique is used to generate probabilistic forecasts of monthly and seasonal precipitation amount and other precipitation indices useful for an early warning of large-scale drought and floods in West Africa. The evaluation of the statistical technique is done using CFS hindcasts (1982 to 2009) in a cross-validation mode to determine the performance of the precipitation forecasts for several lead times focusing on drought and flood events depicted over the Volta and Niger basins. In addition, operational forecasts provided by PRESAO are analyzed from 1998 to 2015. The precipitation forecasts are compared to low-skill reference forecasts generated from gridded observations (i.e. GPCC, CHIRPS) and a novel in-situ gauge database from national observation networks (see Poster EGU2017-10271). The forecasts are evaluated using state-of-the-art verification techniques to determine specific quality attributes of probabilistic forecasts such as reliability, accuracy and skill. In addition, cost-loss approaches are used to determine the value of probabilistic forecasts for multiple users in warning situations. The outcomes of the hindcasts experiment for the Volta basin illustrate that the statistical technique can clearly improve the CFS precipitation forecasts with the potential to provide skillful and valuable early precipitation warnings for large-scale drought and flood situations several months in ahead. In this presentation we give a detailed overview about the ensemble-based quantile-quantile-transformation, its validation and verification and the possibilities of this technique to complement PRESAO. We also highlight the performance of this technique for extremes such as the Sahel drought in the 80ties and in comparison to the various reference data sets (e.g. CFS2, PRESAO, observational data sets) used in this study.

Ensemble flare forecasting: using numerical weather prediction techniques to improve space weather operations

NASA Astrophysics Data System (ADS)

Murray, S.; Guerra, J. A.

2017-12-01

One essential component of operational space weather forecasting is the prediction of solar flares. Early flare forecasting work focused on statistical methods based on historical flaring rates, but more complex machine learning methods have been developed in recent years. A multitude of flare forecasting methods are now available, however it is still unclear which of these methods performs best, and none are substantially better than climatological forecasts. Current operational space weather centres cannot rely on automated methods, and generally use statistical forecasts with a little human intervention. Space weather researchers are increasingly looking towards methods used in terrestrial weather to improve current forecasting techniques. Ensemble forecasting has been used in numerical weather prediction for many years as a way to combine different predictions in order to obtain a more accurate result. It has proved useful in areas such as magnetospheric modelling and coronal mass ejection arrival analysis, however has not yet been implemented in operational flare forecasting. Here we construct ensemble forecasts for major solar flares by linearly combining the full-disk probabilistic forecasts from a group of operational forecasting methods (ASSA, ASAP, MAG4, MOSWOC, NOAA, and Solar Monitor). Forecasts from each method are weighted by a factor that accounts for the method's ability to predict previous events, and several performance metrics (both probabilistic and categorical) are considered. The results provide space weather forecasters with a set of parameters (combination weights, thresholds) that allow them to select the most appropriate values for constructing the 'best' ensemble forecast probability value, according to the performance metric of their choice. In this way different forecasts can be made to fit different end-user needs.

EMC Global Climate And Weather Modeling Branch Personnel

Science.gov Websites

Comparison Statistics which includes: NCEP Raw and Bias-Corrected Ensemble Domain Averaged Bias NCEP Raw and Bias-Corrected Ensemble Domain Averaged Bias Reduction (Percents) CMC Raw and Bias-Corrected Control Forecast Domain Averaged Bias CMC Raw and Bias-Corrected Control Forecast Domain Averaged Bias Reduction

A Bayesian spatio-temporal model for forecasting Anaplasma species seroprevalence in domestic dogs within the contiguous United States

PubMed Central

Liu, Yan; Watson, Stella C.; Gettings, Jenna R.; Lund, Robert B.; Nordone, Shila K.; McMahan, Christopher S.

2017-01-01

This paper forecasts the 2016 canine Anaplasma spp. seroprevalence in the United States from eight climate, geographic and societal factors. The forecast’s construction and an assessment of its performance are described. The forecast is based on a spatial-temporal conditional autoregressive model fitted to over 11 million Anaplasma spp. seroprevalence test results for dogs conducted in the 48 contiguous United States during 2011–2015. The forecast uses county-level data on eight predictive factors, including annual temperature, precipitation, relative humidity, county elevation, forestation coverage, surface water coverage, population density and median household income. Non-static factors are extrapolated into the forthcoming year with various statistical methods. The fitted model and factor extrapolations are used to estimate next year’s regional prevalence. The correlation between the observed and model-estimated county-by-county Anaplasma spp. seroprevalence for the five-year period 2011–2015 is 0.902, demonstrating reasonable model accuracy. The weighted correlation (accounting for different sample sizes) between 2015 observed and forecasted county-by-county Anaplasma spp. seroprevalence is 0.987, exhibiting that the proposed approach can be used to accurately forecast Anaplasma spp. seroprevalence. The forecast presented herein can a priori alert veterinarians to areas expected to see Anaplasma spp. seroprevalence beyond the accepted endemic range. The proposed methods may prove useful for forecasting other diseases. PMID:28738085

Targeted observations to improve tropical cyclone track forecasts in the Atlantic and eastern Pacific basins

NASA Astrophysics Data System (ADS)

Aberson, Sim David

In 1997, the National Hurricane Center and the Hurricane Research Division began conducting operational synoptic surveillance missions with the Gulfstream IV-SP jet aircraft to improve operational forecast models. During the first two years, twenty-four missions were conducted around tropical cyclones threatening the continental United States, Puerto Rico, and the Virgin Islands. Global Positioning System dropwindsondes were released from the aircraft at 150--200 km intervals along the flight track in the tropical cyclone environment to obtain wind, temperature, and humidity profiles from flight level (around 150 hPa) to the surface. The observations were processed and formatted aboard the aircraft and transmitted to the National Centers for Environmental Prediction (NCEP). There, they were ingested into the Global Data Assimilation System that subsequently provides initial and time-dependent boundary conditions for numerical models that forecast tropical cyclone track and intensity. Three dynamical models were employed in testing the targeting and sampling strategies. With the assimilation into the numerical guidance of all the observations gathered during the surveillance missions, only the 12-h Geophysical Fluid Dynamics Laboratory Hurricane Model forecast showed statistically significant improvement. Neither the forecasts from the Aviation run of the Global Spectral Model nor the shallow-water VICBAR model were improved with the assimilation of the dropwindsonde data. This mediocre result is found to be due mainly to the difficulty in operationally quantifying the storm-motion vector used to create accurate synthetic data to represent the tropical cyclone vortex in the models. A secondary limit on forecast improvements from the surveillance missions is the limited amount of data provided by the one surveillance aircraft in regular missions. The inability of some surveillance missions to surround the tropical cyclone with dropwindsonde observations is a possible third limit, though the results are inconclusive. Due to limited aircraft resources, optimal observing strategies for these missions must be developed. Since observations in areas of decaying error modes are unlikely to have large impact on subsequent forecasts, such strategies should be based on taking observations in those geographic locations corresponding to the most rapidly growing error modes in the numerical models and on known deficiencies in current data assimilation systems. Here, the most rapidly growing modes are represented by areas of large forecast spread in the NCEP bred-mode global ensemble forecasting system. The sampling strategy requires sampling the entire target region at approximately the same resolution as the North American rawinsonde network to limit the possibly spurious spread of information from dropwindsonde observations into data-sparse regions where errors are likely to grow. When only the subset of data in these fully-sampled target regions is assimilated into the numerical models, statistically significant reduction of the track forecast errors of up to 25% within the critical first two days of the forecast are seen. These model improvements are comparable with the cumulative business-as-usual track forecast model improvements expected over eighteen years.

Statistical forecast of seasonal discharge in Central Asia using observational records: development of a generic linear modelling tool for operational water resource management

NASA Astrophysics Data System (ADS)

Apel, Heiko; Abdykerimova, Zharkinay; Agalhanova, Marina; Baimaganbetov, Azamat; Gavrilenko, Nadejda; Gerlitz, Lars; Kalashnikova, Olga; Unger-Shayesteh, Katy; Vorogushyn, Sergiy; Gafurov, Abror

2018-04-01

The semi-arid regions of Central Asia crucially depend on the water resources supplied by the mountainous areas of the Tien Shan and Pamir and Altai mountains. During the summer months the snow-melt- and glacier-melt-dominated river discharge originating in the mountains provides the main water resource available for agricultural production, but also for storage in reservoirs for energy generation during the winter months. Thus a reliable seasonal forecast of the water resources is crucial for sustainable management and planning of water resources. In fact, seasonal forecasts are mandatory tasks of all national hydro-meteorological services in the region. In order to support the operational seasonal forecast procedures of hydro-meteorological services, this study aims to develop a generic tool for deriving statistical forecast models of seasonal river discharge based solely on observational records. The generic model structure is kept as simple as possible in order to be driven by meteorological and hydrological data readily available at the hydro-meteorological services, and to be applicable for all catchments in the region. As snow melt dominates summer runoff, the main meteorological predictors for the forecast models are monthly values of winter precipitation and temperature, satellite-based snow cover data, and antecedent discharge. This basic predictor set was further extended by multi-monthly means of the individual predictors, as well as composites of the predictors. Forecast models are derived based on these predictors as linear combinations of up to four predictors. A user-selectable number of the best models is extracted automatically by the developed model fitting algorithm, which includes a test for robustness by a leave-one-out cross-validation. Based on the cross-validation the predictive uncertainty was quantified for every prediction model. Forecasts of the mean seasonal discharge of the period April to September are derived every month from January until June. The application of the model for several catchments in Central Asia - ranging from small to the largest rivers (240 to 290 000 km2 catchment area) - for the period 2000-2015 provided skilful forecasts for most catchments already in January, with adjusted R2 values of the best model in the range of 0.6-0.8 for most of the catchments. The skill of the prediction increased every following month, i.e. with reduced lead time, with adjusted R2 values usually in the range 0.8-0.9 for the best and 0.7-0.8 on average for the set of models in April just before the prediction period. The later forecasts in May and June improve further due to the high predictive power of the discharge in the first 2 months of the snow melt period. The improved skill of the set of forecast models with decreasing lead time resulted in narrow predictive uncertainty bands at the beginning of the snow melt period. In summary, the proposed generic automatic forecast model development tool provides robust predictions for seasonal water availability in Central Asia, which will be tested against the official forecasts in the upcoming years, with the vision of operational implementation.

Ensemble forecast of human West Nile virus cases and mosquito infection rates

NASA Astrophysics Data System (ADS)

Defelice, Nicholas B.; Little, Eliza; Campbell, Scott R.; Shaman, Jeffrey

2017-02-01

West Nile virus (WNV) is now endemic in the continental United States; however, our ability to predict spillover transmission risk and human WNV cases remains limited. Here we develop a model depicting WNV transmission dynamics, which we optimize using a data assimilation method and two observed data streams, mosquito infection rates and reported human WNV cases. The coupled model-inference framework is then used to generate retrospective ensemble forecasts of historical WNV outbreaks in Long Island, New York for 2001-2014. Accurate forecasts of mosquito infection rates are generated before peak infection, and >65% of forecasts accurately predict seasonal total human WNV cases up to 9 weeks before the past reported case. This work provides the foundation for implementation of a statistically rigorous system for real-time forecast of seasonal outbreaks of WNV.

Ensemble forecast of human West Nile virus cases and mosquito infection rates.

PubMed

DeFelice, Nicholas B; Little, Eliza; Campbell, Scott R; Shaman, Jeffrey

2017-02-24

West Nile virus (WNV) is now endemic in the continental United States; however, our ability to predict spillover transmission risk and human WNV cases remains limited. Here we develop a model depicting WNV transmission dynamics, which we optimize using a data assimilation method and two observed data streams, mosquito infection rates and reported human WNV cases. The coupled model-inference framework is then used to generate retrospective ensemble forecasts of historical WNV outbreaks in Long Island, New York for 2001-2014. Accurate forecasts of mosquito infection rates are generated before peak infection, and >65% of forecasts accurately predict seasonal total human WNV cases up to 9 weeks before the past reported case. This work provides the foundation for implementation of a statistically rigorous system for real-time forecast of seasonal outbreaks of WNV.

Ensemble-based methods for forecasting census in hospital units

PubMed Central

2013-01-01

Background The ability to accurately forecast census counts in hospital departments has considerable implications for hospital resource allocation. In recent years several different methods have been proposed forecasting census counts, however many of these approaches do not use available patient-specific information. Methods In this paper we present an ensemble-based methodology for forecasting the census under a framework that simultaneously incorporates both (i) arrival trends over time and (ii) patient-specific baseline and time-varying information. The proposed model for predicting census has three components, namely: current census count, number of daily arrivals and number of daily departures. To model the number of daily arrivals, we use a seasonality adjusted Poisson Autoregressive (PAR) model where the parameter estimates are obtained via conditional maximum likelihood. The number of daily departures is predicted by modeling the probability of departure from the census using logistic regression models that are adjusted for the amount of time spent in the census and incorporate both patient-specific baseline and time varying patient-specific covariate information. We illustrate our approach using neonatal intensive care unit (NICU) data collected at Women & Infants Hospital, Providence RI, which consists of 1001 consecutive NICU admissions between April 1st 2008 and March 31st 2009. Results Our results demonstrate statistically significant improved prediction accuracy for 3, 5, and 7 day census forecasts and increased precision of our forecasting model compared to a forecasting approach that ignores patient-specific information. Conclusions Forecasting models that utilize patient-specific baseline and time-varying information make the most of data typically available and have the capacity to substantially improve census forecasts. PMID:23721123

Ensemble-based methods for forecasting census in hospital units.

PubMed

Koestler, Devin C; Ombao, Hernando; Bender, Jesse

2013-05-30

The ability to accurately forecast census counts in hospital departments has considerable implications for hospital resource allocation. In recent years several different methods have been proposed forecasting census counts, however many of these approaches do not use available patient-specific information. In this paper we present an ensemble-based methodology for forecasting the census under a framework that simultaneously incorporates both (i) arrival trends over time and (ii) patient-specific baseline and time-varying information. The proposed model for predicting census has three components, namely: current census count, number of daily arrivals and number of daily departures. To model the number of daily arrivals, we use a seasonality adjusted Poisson Autoregressive (PAR) model where the parameter estimates are obtained via conditional maximum likelihood. The number of daily departures is predicted by modeling the probability of departure from the census using logistic regression models that are adjusted for the amount of time spent in the census and incorporate both patient-specific baseline and time varying patient-specific covariate information. We illustrate our approach using neonatal intensive care unit (NICU) data collected at Women & Infants Hospital, Providence RI, which consists of 1001 consecutive NICU admissions between April 1st 2008 and March 31st 2009. Our results demonstrate statistically significant improved prediction accuracy for 3, 5, and 7 day census forecasts and increased precision of our forecasting model compared to a forecasting approach that ignores patient-specific information. Forecasting models that utilize patient-specific baseline and time-varying information make the most of data typically available and have the capacity to substantially improve census forecasts.

Development of predictive weather scenarios for early prediction of rice yield in South Korea

NASA Astrophysics Data System (ADS)

Shin, Y.; Cho, J.; Jung, I.

2017-12-01

International grain prices are becoming unstable due to frequent occurrence of abnormal weather phenomena caused by climate change. Early prediction of grain yield using weather forecast data is important for stabilization of international grain prices. The APEC Climate Center (APCC) is providing seasonal forecast data based on monthly climate prediction models for global seasonal forecasting services. The 3-month and 6-month seasonal forecast data using the multi-model ensemble (MME) technique are provided in their own website, ADSS (APCC Data Service System, http://adss.apcc21.org/). The spatial resolution of seasonal forecast data for each individual model is 2.5°×2.5°(about 250km) and the time scale is created as monthly. In this study, we developed customized weather forecast scenarios that are combined seasonal forecast data and observational data apply to early rice yield prediction model. Statistical downscale method was applied to produce meteorological input data of crop model because field scale crop model (ORYZA2000) requires daily weather data. In order to determine whether the forecasting data is suitable for the crop model, we produced spatio-temporal downscaled weather scenarios and evaluated the predictability by comparison with observed weather data at 57 ASOS stations in South Korea. The customized weather forecast scenarios can be applied to various application fields not only early rice yield prediction. Acknowledgement This work was carried out with the support of "Cooperative Research Program for Agriculture Science and Technology Development (Project No: PJ012855022017)" Rural Development Administration, Republic of Korea.

Verification of Advances in a Coupled Snow-runoff Modeling Framework for Operational Streamflow Forecasts

NASA Astrophysics Data System (ADS)

Barik, M. G.; Hogue, T. S.; Franz, K. J.; He, M.

2011-12-01

The National Oceanic and Atmospheric Administration's (NOAA's) River Forecast Centers (RFCs) issue hydrologic forecasts related to flood events, reservoir operations for water supply, streamflow regulation, and recreation on the nation's streams and rivers. The RFCs use the National Weather Service River Forecast System (NWSRFS) for streamflow forecasting which relies on a coupled snow model (i.e. SNOW17) and rainfall-runoff model (i.e. SAC-SMA) in snow-dominated regions of the US. Errors arise in various steps of the forecasting system from input data, model structure, model parameters, and initial states. The goal of the current study is to undertake verification of potential improvements in the SNOW17-SAC-SMA modeling framework developed for operational streamflow forecasts. We undertake verification for a range of parameters sets (i.e. RFC, DREAM (Differential Evolution Adaptive Metropolis)) as well as a data assimilation (DA) framework developed for the coupled models. Verification is also undertaken for various initial conditions to observe the influence of variability in initial conditions on the forecast. The study basin is the North Fork America River Basin (NFARB) located on the western side of the Sierra Nevada Mountains in northern California. Hindcasts are verified using both deterministic (i.e. Nash Sutcliffe efficiency, root mean square error, and joint distribution) and probabilistic (i.e. reliability diagram, discrimination diagram, containing ratio, and Quantile plots) statistics. Our presentation includes comparison of the performance of different optimized parameters and the DA framework as well as assessment of the impact associated with the initial conditions used for streamflow forecasts for the NFARB.

Comparison of hourly surface downwelling solar radiation estimated from MSG-SEVIRI and forecast by the RAMS model with pyranometers over Italy

NASA Astrophysics Data System (ADS)

Federico, Stefano; Torcasio, Rosa Claudia; Sanò, Paolo; Casella, Daniele; Campanelli, Monica; Fokke Meirink, Jan; Wang, Ping; Vergari, Stefania; Diémoz, Henri; Dietrich, Stefano

2017-06-01

In this paper, we evaluate the performance of two global horizontal solar irradiance (GHI) estimates, one derived from Meteosat Second Generation (MSG) and another from the 1-day forecast of the Regional Atmospheric Modeling System (RAMS) mesoscale model. The horizontal resolution of the MSG-GHI is 3 × 5 km2 over Italy, which is the focus area of this study. For this paper, RAMS has the horizontal resolution of 4 km.The performances of the MSG-GHI estimate and RAMS-GHI 1-day forecast are evaluated for 1 year (1 June 2013-31 May 2014) against data of 12 ground-based pyranometers over Italy spanning a range of climatic conditions, i.e. from maritime Mediterranean to Alpine climate.Statistics for hourly GHI and daily integrated GHI are presented for the four seasons and the whole year for all the measurement sites. Different sky conditions are considered in the analysisResults for hourly data show an evident dependence on the sky conditions, with the root mean square error (RMSE) increasing from clear to cloudy conditions. The RMSE is substantially higher for Alpine stations in all the seasons, mainly because of the increase of the cloud coverage for these stations, which is not well represented at the satellite and model resolutions. Considering the yearly statistics computed from hourly data for the RAMS model, the RMSE ranges from 152 W m-2 (31 %) obtained for Cozzo Spadaro, a maritime station, to 287 W m-2 (82 %) for Aosta, an Alpine site. Considering the yearly statistics computed from hourly data for MSG-GHI, the minimum RMSE is for Cozzo Spadaro (71 W m-2, 14 %), while the maximum is for Aosta (181 W m-2, 51 %). The mean bias error (MBE) shows the tendency of RAMS to over-forecast the GHI, while no specific behaviour is found for MSG-GHI.Results for daily integrated GHI show a lower RMSE compared to hourly GHI evaluation for both RAMS-GHI 1-day forecast and MSG-GHI estimate. Considering the yearly evaluation, the RMSE of daily integrated GHI is at least 9 % lower (in percentage units, from 31 to 22 % for RAMS in Cozzo Spadaro) than the RMSE computed for hourly data for each station. A partial compensation of underestimation and overestimation of the GHI contributes to the RMSE reduction. Furthermore, a post-processing technique, namely model output statistics (MOS), is applied to improve the GHI forecast at hourly and daily temporal scales. The application of MOS shows an improvement of RAMS-GHI forecast, which depends on the site considered, while the impact of MOS on MSG-GHI RMSE is small.

Long-term ensemble forecast of snowmelt inflow into the Cheboksary Reservoir under two different weather scenarios

NASA Astrophysics Data System (ADS)

Gelfan, Alexander; Moreydo, Vsevolod; Motovilov, Yury; Solomatine, Dimitri P.

2018-04-01

A long-term forecasting ensemble methodology, applied to water inflows into the Cheboksary Reservoir (Russia), is presented. The methodology is based on a version of the semi-distributed hydrological model ECOMAG (ECOlogical Model for Applied Geophysics) that allows for the calculation of an ensemble of inflow hydrographs using two different sets of weather ensembles for the lead time period: observed weather data, constructed on the basis of the Ensemble Streamflow Prediction methodology (ESP-based forecast), and synthetic weather data, simulated by a multi-site weather generator (WG-based forecast). We have studied the following: (1) whether there is any advantage of the developed ensemble forecasts in comparison with the currently issued operational forecasts of water inflow into the Cheboksary Reservoir, and (2) whether there is any noticeable improvement in probabilistic forecasts when using the WG-simulated ensemble compared to the ESP-based ensemble. We have found that for a 35-year period beginning from the reservoir filling in 1982, both continuous and binary model-based ensemble forecasts (issued in the deterministic form) outperform the operational forecasts of the April-June inflow volume actually used and, additionally, provide acceptable forecasts of additional water regime characteristics besides the inflow volume. We have also demonstrated that the model performance measures (in the verification period) obtained from the WG-based probabilistic forecasts, which are based on a large number of possible weather scenarios, appeared to be more statistically reliable than the corresponding measures calculated from the ESP-based forecasts based on the observed weather scenarios.

Supporting inland waterway transport on German waterways by operational forecasting services - water-levels, discharges, river ice

NASA Astrophysics Data System (ADS)

Meißner, Dennis; Klein, Bastian; Ionita, Monica; Hemri, Stephan; Rademacher, Silke

2017-04-01

Inland waterway transport (IWT) is an important commercial sector significantly vulnerable to hydrological impacts. River ice and floods limit the availability of the waterway network and may cause considerable damages to waterway infrastructure. Low flows significantly affect IWT's operation efficiency usually several months a year due to the close correlation of (low) water levels / water depths and (high) transport costs. Therefore "navigation-related" hydrological forecasts focussing on the specific requirements of water-bound transport (relevant forecast locations, target parameters, skill characteristics etc.) play a major role in order to mitigate IWT's vulnerability to hydro-meteorological impacts. In light of continuing transport growth within the European Union, hydrological forecasts for the waterways are essential to stimulate the use of the free capacity IWT still offers more consequently. An overview of the current operational and pre-operational forecasting systems for the German waterways predicting water levels, discharges and river ice thickness on various time-scales will be presented. While short-term (deterministic) forecasts have a long tradition in navigation-related forecasting, (probabilistic) forecasting services offering extended lead-times are not yet well-established and are still subject to current research and development activities (e.g. within the EU-projects EUPORIAS and IMPREX). The focus is on improving technical aspects as well as on exploring adequate ways of disseminating and communicating probabilistic forecast information. For the German stretch of the River Rhine, one of the most frequented inland waterways worldwide, the existing deterministic forecast scheme has been extended by ensemble forecasts combined with statistical post-processing modules applying EMOS (Ensemble Model Output Statistics) and ECC (Ensemble Copula Coupling) in order to generate water level predictions up to 10 days and to estimate its predictive uncertainty properly. Additionally for the key locations at the international waterways Rhine, Elbe and Danube three competing forecast approaches are currently tested in a pre-operational set-up in order to generate monthly to seasonal (up to 3 months) forecasts: (1) the well-known Ensemble Streamflow Prediction approach (ensemble based on historical meteorology), (2) coupling hydrological models with post-processed outputs from ECMWF's general circulation model (System 4), and (3) a purely statistical approach based on the stable relationship (teleconnection) of global or regional oceanic, climate and hydrological data with river flows. The current results, still pre-operational, reveal the existence of a valuable predictability of water levels and streamflow also at monthly up to seasonal time-scales along the larger rivers used as waterways in Germany. Last but not least insight into the technical set-up of the aforementioned forecasting systems operated at the Federal Institute of Hydrology, which are based on a Delft-FEWS application, will be given focussing on the step-wise extension of the former system by integrating new components in order to meet the growing needs of the customers and to improve and extend the forecast portfolio for waterway users.

Using Meteorological Analogues for Reordering Postprocessed Precipitation Ensembles in Hydrological Forecasting

NASA Astrophysics Data System (ADS)

Bellier, Joseph; Bontron, Guillaume; Zin, Isabella

2017-12-01

Meteorological ensemble forecasts are nowadays widely used as input of hydrological models for probabilistic streamflow forecasting. These forcings are frequently biased and have to be statistically postprocessed, using most of the time univariate techniques that apply independently to individual locations, lead times and weather variables. Postprocessed ensemble forecasts therefore need to be reordered so as to reconstruct suitable multivariate dependence structures. The Schaake shuffle and ensemble copula coupling are the two most popular methods for this purpose. This paper proposes two adaptations of them that make use of meteorological analogues for reconstructing spatiotemporal dependence structures of precipitation forecasts. Performances of the original and adapted techniques are compared through a multistep verification experiment using real forecasts from the European Centre for Medium-Range Weather Forecasts. This experiment evaluates not only multivariate precipitation forecasts but also the corresponding streamflow forecasts that derive from hydrological modeling. Results show that the relative performances of the different reordering methods vary depending on the verification step. In particular, the standard Schaake shuffle is found to perform poorly when evaluated on streamflow. This emphasizes the crucial role of the precipitation spatiotemporal dependence structure in hydrological ensemble forecasting.

Earthquake likelihood model testing

USGS Publications Warehouse

Schorlemmer, D.; Gerstenberger, M.C.; Wiemer, S.; Jackson, D.D.; Rhoades, D.A.

2007-01-01

INTRODUCTIONThe Regional Earthquake Likelihood Models (RELM) project aims to produce and evaluate alternate models of earthquake potential (probability per unit volume, magnitude, and time) for California. Based on differing assumptions, these models are produced to test the validity of their assumptions and to explore which models should be incorporated in seismic hazard and risk evaluation. Tests based on physical and geological criteria are useful but we focus on statistical methods using future earthquake catalog data only. We envision two evaluations: a test of consistency with observed data and a comparison of all pairs of models for relative consistency. Both tests are based on the likelihood method, and both are fully prospective (i.e., the models are not adjusted to fit the test data). To be tested, each model must assign a probability to any possible event within a specified region of space, time, and magnitude. For our tests the models must use a common format: earthquake rates in specified “bins” with location, magnitude, time, and focal mechanism limits.Seismology cannot yet deterministically predict individual earthquakes; however, it should seek the best possible models for forecasting earthquake occurrence. This paper describes the statistical rules of an experiment to examine and test earthquake forecasts. The primary purposes of the tests described below are to evaluate physical models for earthquakes, assure that source models used in seismic hazard and risk studies are consistent with earthquake data, and provide quantitative measures by which models can be assigned weights in a consensus model or be judged as suitable for particular regions.In this paper we develop a statistical method for testing earthquake likelihood models. A companion paper (Schorlemmer and Gerstenberger 2007, this issue) discusses the actual implementation of these tests in the framework of the RELM initiative.Statistical testing of hypotheses is a common task and a wide range of possible testing procedures exist. Jolliffe and Stephenson (2003) present different forecast verifications from atmospheric science, among them likelihood testing of probability forecasts and testing the occurrence of binary events. Testing binary events requires that for each forecasted event, the spatial, temporal and magnitude limits be given. Although major earthquakes can be considered binary events, the models within the RELM project express their forecasts on a spatial grid and in 0.1 magnitude units; thus the results are a distribution of rates over space and magnitude. These forecasts can be tested with likelihood tests.In general, likelihood tests assume a valid null hypothesis against which a given hypothesis is tested. The outcome is either a rejection of the null hypothesis in favor of the test hypothesis or a nonrejection, meaning the test hypothesis cannot outperform the null hypothesis at a given significance level. Within RELM, there is no accepted null hypothesis and thus the likelihood test needs to be expanded to allow comparable testing of equipollent hypotheses.To test models against one another, we require that forecasts are expressed in a standard format: the average rate of earthquake occurrence within pre-specified limits of hypocentral latitude, longitude, depth, magnitude, time period, and focal mechanisms. Focal mechanisms should either be described as the inclination of P-axis, declination of P-axis, and inclination of the T-axis, or as strike, dip, and rake angles. Schorlemmer and Gerstenberger (2007, this issue) designed classes of these parameters such that similar models will be tested against each other. These classes make the forecasts comparable between models. Additionally, we are limited to testing only what is precisely defined and consistently reported in earthquake catalogs. Therefore it is currently not possible to test such information as fault rupture length or area, asperity location, etc. Also, to account for data quality issues, we allow for location and magnitude uncertainties as well as the probability that an event is dependent on another event.As we mentioned above, only models with comparable forecasts can be tested against each other. Our current tests are designed to examine grid-based models. This requires that any fault-based model be adapted to a grid before testing is possible. While this is a limitation of the testing, it is an inherent difficulty in any such comparative testing. Please refer to appendix B for a statistical evaluation of the application of the Poisson hypothesis to fault-based models.The testing suite we present consists of three different tests: L-Test, N-Test, and R-Test. These tests are defined similarily to Kagan and Jackson (1995). The first two tests examine the consistency of the hypotheses with the observations while the last test compares the spatial performances of the models.

The MSFC Solar Activity Future Estimation (MSAFE) Model

NASA Technical Reports Server (NTRS)

Suggs, Ron

2017-01-01

The MSAFE model provides forecasts for the solar indices SSN, F10.7, and Ap. These solar indices are used as inputs to space environment models used in orbital spacecraft operations and space mission analysis. Forecasts from the MSAFE model are provided on the MSFC Natural Environments Branch's solar web page and are updated as new monthly observations become available. The MSAFE prediction routine employs a statistical technique that calculates deviations of past solar cycles from the mean cycle and performs a regression analysis to calculate the deviation from the mean cycle of the solar index at the next future time interval. The forecasts are initiated for a given cycle after about 8 to 9 monthly observations from the start of the cycle are collected. A forecast made at the beginning of cycle 24 using the MSAFE program captured the cycle fairly well with some difficulty in discerning the double peak that occurred at solar cycle maximum.

A statistical inference for concentrations of benzo[a]pyrene partially measured in the ambient air of an industrial city in Korea

NASA Astrophysics Data System (ADS)

Kim, Yongku; Seo, Young-Kyo; Baek, Sung-Ok

2013-12-01

Although large quantities of air pollutants are released into the atmosphere, they are partially monitored and routinely assessed for their health implications. This paper proposes a statistical model describing the temporal behavior of hazardous air pollutants (HAPs), which can have negative effects on human health. Benzo[a]pyrene (BaP) is selected for statistical modeling. The proposed model incorporates the linkage between BaP and meteorology and is specifically formulated to identify meteorological effects and allow for seasonal trends. The model is used to estimate and forecast temporal fields of BaP conditional on observed (or forecasted) meteorological conditions, including temperature, precipitation, wind speed, and air quality. The effects of BaP on human health are examined by characterizing health indicators, namely the cancer risk and the hazard quotient. The model provides useful information for the optimal monitoring period and projection of future BaP concentrations for both industrial and residential areas in Korea.

Influenza forecasting in human populations: a scoping review.

PubMed

Chretien, Jean-Paul; George, Dylan; Shaman, Jeffrey; Chitale, Rohit A; McKenzie, F Ellis

2014-01-01

Forecasts of influenza activity in human populations could help guide key preparedness tasks. We conducted a scoping review to characterize these methodological approaches and identify research gaps. Adapting the PRISMA methodology for systematic reviews, we searched PubMed, CINAHL, Project Euclid, and Cochrane Database of Systematic Reviews for publications in English since January 1, 2000 using the terms "influenza AND (forecast* OR predict*)", excluding studies that did not validate forecasts against independent data or incorporate influenza-related surveillance data from the season or pandemic for which the forecasts were applied. We included 35 publications describing population-based (N = 27), medical facility-based (N = 4), and regional or global pandemic spread (N = 4) forecasts. They included areas of North America (N = 15), Europe (N = 14), and/or Asia-Pacific region (N = 4), or had global scope (N = 3). Forecasting models were statistical (N = 18) or epidemiological (N = 17). Five studies used data assimilation methods to update forecasts with new surveillance data. Models used virological (N = 14), syndromic (N = 13), meteorological (N = 6), internet search query (N = 4), and/or other surveillance data as inputs. Forecasting outcomes and validation metrics varied widely. Two studies compared distinct modeling approaches using common data, 2 assessed model calibration, and 1 systematically incorporated expert input. Of the 17 studies using epidemiological models, 8 included sensitivity analysis. This review suggests need for use of good practices in influenza forecasting (e.g., sensitivity analysis); direct comparisons of diverse approaches; assessment of model calibration; integration of subjective expert input; operational research in pilot, real-world applications; and improved mutual understanding among modelers and public health officials.

Influenza Forecasting in Human Populations: A Scoping Review

PubMed Central

Chretien, Jean-Paul; George, Dylan; Shaman, Jeffrey; Chitale, Rohit A.; McKenzie, F. Ellis

2014-01-01

Forecasts of influenza activity in human populations could help guide key preparedness tasks. We conducted a scoping review to characterize these methodological approaches and identify research gaps. Adapting the PRISMA methodology for systematic reviews, we searched PubMed, CINAHL, Project Euclid, and Cochrane Database of Systematic Reviews for publications in English since January 1, 2000 using the terms “influenza AND (forecast* OR predict*)”, excluding studies that did not validate forecasts against independent data or incorporate influenza-related surveillance data from the season or pandemic for which the forecasts were applied. We included 35 publications describing population-based (N = 27), medical facility-based (N = 4), and regional or global pandemic spread (N = 4) forecasts. They included areas of North America (N = 15), Europe (N = 14), and/or Asia-Pacific region (N = 4), or had global scope (N = 3). Forecasting models were statistical (N = 18) or epidemiological (N = 17). Five studies used data assimilation methods to update forecasts with new surveillance data. Models used virological (N = 14), syndromic (N = 13), meteorological (N = 6), internet search query (N = 4), and/or other surveillance data as inputs. Forecasting outcomes and validation metrics varied widely. Two studies compared distinct modeling approaches using common data, 2 assessed model calibration, and 1 systematically incorporated expert input. Of the 17 studies using epidemiological models, 8 included sensitivity analysis. This review suggests need for use of good practices in influenza forecasting (e.g., sensitivity analysis); direct comparisons of diverse approaches; assessment of model calibration; integration of subjective expert input; operational research in pilot, real-world applications; and improved mutual understanding among modelers and public health officials. PMID:24714027

A Simulation Optimization Approach to Epidemic Forecasting

PubMed Central

Nsoesie, Elaine O.; Beckman, Richard J.; Shashaani, Sara; Nagaraj, Kalyani S.; Marathe, Madhav V.

2013-01-01

Reliable forecasts of influenza can aid in the control of both seasonal and pandemic outbreaks. We introduce a simulation optimization (SIMOP) approach for forecasting the influenza epidemic curve. This study represents the final step of a project aimed at using a combination of simulation, classification, statistical and optimization techniques to forecast the epidemic curve and infer underlying model parameters during an influenza outbreak. The SIMOP procedure combines an individual-based model and the Nelder-Mead simplex optimization method. The method is used to forecast epidemics simulated over synthetic social networks representing Montgomery County in Virginia, Miami, Seattle and surrounding metropolitan regions. The results are presented for the first four weeks. Depending on the synthetic network, the peak time could be predicted within a 95% CI as early as seven weeks before the actual peak. The peak infected and total infected were also accurately forecasted for Montgomery County in Virginia within the forecasting period. Forecasting of the epidemic curve for both seasonal and pandemic influenza outbreaks is a complex problem, however this is a preliminary step and the results suggest that more can be achieved in this area. PMID:23826222

A Simulation Optimization Approach to Epidemic Forecasting.

PubMed

Nsoesie, Elaine O; Beckman, Richard J; Shashaani, Sara; Nagaraj, Kalyani S; Marathe, Madhav V

2013-01-01

Reliable forecasts of influenza can aid in the control of both seasonal and pandemic outbreaks. We introduce a simulation optimization (SIMOP) approach for forecasting the influenza epidemic curve. This study represents the final step of a project aimed at using a combination of simulation, classification, statistical and optimization techniques to forecast the epidemic curve and infer underlying model parameters during an influenza outbreak. The SIMOP procedure combines an individual-based model and the Nelder-Mead simplex optimization method. The method is used to forecast epidemics simulated over synthetic social networks representing Montgomery County in Virginia, Miami, Seattle and surrounding metropolitan regions. The results are presented for the first four weeks. Depending on the synthetic network, the peak time could be predicted within a 95% CI as early as seven weeks before the actual peak. The peak infected and total infected were also accurately forecasted for Montgomery County in Virginia within the forecasting period. Forecasting of the epidemic curve for both seasonal and pandemic influenza outbreaks is a complex problem, however this is a preliminary step and the results suggest that more can be achieved in this area.

Short-term ensemble streamflow forecasting using operationally-produced single-valued streamflow forecasts - A Hydrologic Model Output Statistics (HMOS) approach

NASA Astrophysics Data System (ADS)

Regonda, Satish Kumar; Seo, Dong-Jun; Lawrence, Bill; Brown, James D.; Demargne, Julie

2013-08-01

We present a statistical procedure for generating short-term ensemble streamflow forecasts from single-valued, or deterministic, streamflow forecasts produced operationally by the U.S. National Weather Service (NWS) River Forecast Centers (RFCs). The resulting ensemble streamflow forecast provides an estimate of the predictive uncertainty associated with the single-valued forecast to support risk-based decision making by the forecasters and by the users of the forecast products, such as emergency managers. Forced by single-valued quantitative precipitation and temperature forecasts (QPF, QTF), the single-valued streamflow forecasts are produced at a 6-h time step nominally out to 5 days into the future. The single-valued streamflow forecasts reflect various run-time modifications, or "manual data assimilation", applied by the human forecasters in an attempt to reduce error from various sources in the end-to-end forecast process. The proposed procedure generates ensemble traces of streamflow from a parsimonious approximation of the conditional multivariate probability distribution of future streamflow given the single-valued streamflow forecast, QPF, and the most recent streamflow observation. For parameter estimation and evaluation, we used a multiyear archive of the single-valued river stage forecast produced operationally by the NWS Arkansas-Red River Basin River Forecast Center (ABRFC) in Tulsa, Oklahoma. As a by-product of parameter estimation, the procedure provides a categorical assessment of the effective lead time of the operational hydrologic forecasts for different QPF and forecast flow conditions. To evaluate the procedure, we carried out hindcasting experiments in dependent and cross-validation modes. The results indicate that the short-term streamflow ensemble hindcasts generated from the procedure are generally reliable within the effective lead time of the single-valued forecasts and well capture the skill of the single-valued forecasts. For smaller basins, however, the effective lead time is significantly reduced by short basin memory and reduced skill in the single-valued QPF.

Long-term flow forecasts based on climate and hydrologic modeling: Uruguay River basin

NASA Astrophysics Data System (ADS)

Tucci, Carlos Eduardo Morelli; Clarke, Robin Thomas; Collischonn, Walter; da Silva Dias, Pedro Leite; de Oliveira, Gilvan Sampaio

2003-07-01

This paper describes a procedure for predicting seasonal flow in the Rio Uruguay drainage basin (area 75,000 km2, lying in Brazilian territory), using sequences of future daily rainfall given by the global climate model (GCM) of the Brazilian agency for climate prediction (Centro de Previsão de Tempo e Clima, or CPTEC). Sequences of future daily rainfall given by this model were used as input to a rainfall-runoff model appropriate for large drainage basins. Forecasts of flow in the Rio Uruguay were made for the period 1995-2001 of the full record, which began in 1940. Analysis showed that GCM forecasts underestimated rainfall over almost all the basin, particularly in winter, although interannual variability in regional rainfall was reproduced relatively well. A statistical procedure was used to correct for the underestimation of rainfall. When the corrected rainfall sequences were transformed to flow by the hydrologic model, forecasts of flow in the Rio Uruguay basin were better than forecasts based on historic mean or median flows by 37% for monthly flows and by 54% for 3-monthly flows.

Multicomponent ensemble models to forecast induced seismicity

NASA Astrophysics Data System (ADS)

Király-Proag, E.; Gischig, V.; Zechar, J. D.; Wiemer, S.

2018-01-01

In recent years, human-induced seismicity has become a more and more relevant topic due to its economic and social implications. Several models and approaches have been developed to explain underlying physical processes or forecast induced seismicity. They range from simple statistical models to coupled numerical models incorporating complex physics. We advocate the need for forecast testing as currently the best method for ascertaining if models are capable to reasonably accounting for key physical governing processes—or not. Moreover, operational forecast models are of great interest to help on-site decision-making in projects entailing induced earthquakes. We previously introduced a standardized framework following the guidelines of the Collaboratory for the Study of Earthquake Predictability, the Induced Seismicity Test Bench, to test, validate, and rank induced seismicity models. In this study, we describe how to construct multicomponent ensemble models based on Bayesian weightings that deliver more accurate forecasts than individual models in the case of Basel 2006 and Soultz-sous-Forêts 2004 enhanced geothermal stimulation projects. For this, we examine five calibrated variants of two significantly different model groups: (1) Shapiro and Smoothed Seismicity based on the seismogenic index, simple modified Omori-law-type seismicity decay, and temporally weighted smoothed seismicity; (2) Hydraulics and Seismicity based on numerically modelled pore pressure evolution that triggers seismicity using the Mohr-Coulomb failure criterion. We also demonstrate how the individual and ensemble models would perform as part of an operational Adaptive Traffic Light System. Investigating seismicity forecasts based on a range of potential injection scenarios, we use forecast periods of different durations to compute the occurrence probabilities of seismic events M ≥ 3. We show that in the case of the Basel 2006 geothermal stimulation the models forecast hazardous levels of seismicity days before the occurrence of felt events.

COP21 climate negotiators' responses to climate model forecasts

NASA Astrophysics Data System (ADS)

Bosetti, Valentina; Weber, Elke; Berger, Loïc; Budescu, David V.; Liu, Ning; Tavoni, Massimo

2017-02-01

Policymakers involved in climate change negotiations are key users of climate science. It is therefore vital to understand how to communicate scientific information most effectively to this group. We tested how a unique sample of policymakers and negotiators at the Paris COP21 conference update their beliefs on year 2100 global mean temperature increases in response to a statistical summary of climate models' forecasts. We randomized the way information was provided across participants using three different formats similar to those used in Intergovernmental Panel on Climate Change reports. In spite of having received all available relevant scientific information, policymakers adopted such information very conservatively, assigning it less weight than their own prior beliefs. However, providing individual model estimates in addition to the statistical range was more effective in mitigating such inertia. The experiment was repeated with a population of European MBA students who, despite starting from similar priors, reported conditional probabilities closer to the provided models' forecasts than policymakers. There was also no effect of presentation format in the MBA sample. These results highlight the importance of testing visualization tools directly on the population of interest.

Verify MesoNAM Performance

NASA Technical Reports Server (NTRS)

Bauman, William H., III

2010-01-01

The AMU conducted an objective analysis of the MesoNAM forecasts compared to observed values from sensors at specified KSC/CCAFS wind towers by calculating the following statistics to verify the performance of the model: 1) Bias (mean difference), 2) Standard deviation of Bias, 3) Root Mean Square Error (RMSE), and 4) Hypothesis test for Bias = O. The 45 WS LWOs use the MesoNAM to support launch weather operations. However, the actual performance of the model at KSC and CCAFS had not been measured objectively. The analysis compared the MesoNAM forecast winds, temperature and dew point to the observed values from the sensors on wind towers. The data were stratified by tower sensor, month and onshore/offshore wind direction based on the orientation of the coastline to each tower's location. The model's performance statistics were then calculated for each wind tower based on sensor height and model initialization time. The period of record for the data used in this task was based on the operational start of the current MesoNAM in mid-August 2006 and so the task began with the first full month of data, September 2006, through May 2010. The analysis of model performance indicated: a) The accuracy decreased as the forecast valid time from the model initialization increased, b) There was a diurnal signal in T with a cool bias during the late night and a warm bias during the afternoon, c) There was a diurnal signal in Td with a low bias during the afternoon and a high bias during the late night, and d) The model parameters at each vertical level most closely matched the observed parameters at heights closest to those vertical levels. The AMU developed a GUI that consists of a multi-level drop-down menu written in JavaScript embedded within the HTML code. This tool allows the LWO to easily and efficiently navigate among the charts and spreadsheet files containing the model performance statistics. The objective statistics give the LWOs knowledge of the model's strengths and weaknesses and the GUI allows quick access to the data which will result in improved forecasts for operations.

The role of multiple-scale modelling of epilepsy in seizure forecasting

PubMed Central

Kuhlmann, Levin; Grayden, David B.; Wendling, Fabrice; Schiff, Steven J.

2014-01-01

Over the past three decades, a number of seizure prediction, or forecasting, methods have been developed. Although major achievements were accomplished regarding the statistical evaluation of proposed algorithms, it is recognized that further progress is still necessary for clinical application in patients. The lack of physiological motivation can partly explain this limitation. Therefore, a natural question is raised: can computational models of epilepsy be used to improve these methods? Here we review the literature on the multiple-scale neural modelling of epilepsy and the use of such models to infer physiological changes underlying epilepsy and epileptic seizures. We argue how these methods can be applied to advance the state-of-the-art in seizure forecasting. PMID:26035674

Application of a medium-range global hydrologic probabilistic forecast scheme to the Ohio River Basin

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

Voisin, Nathalie; Pappenberger, Florian; Lettenmaier, D. P.

2011-08-15

A 10-day globally applicable flood prediction scheme was evaluated using the Ohio River basin as a test site for the period 2003-2007. The Variable Infiltration Capacity (VIC) hydrology model was initialized with the European Centre for Medium Range Weather Forecasts (ECMWF) analysis temperatures and wind, and Tropical Rainfall Monitoring Mission Multi Satellite Precipitation Analysis (TMPA) precipitation up to the day of forecast. In forecast mode, the VIC model was then forced with a calibrated and statistically downscaled ECMWF ensemble prediction system (EPS) 10-day ensemble forecast. A parallel set up was used where ECMWF EPS forecasts were interpolated to the spatialmore » scale of the hydrology model. Each set of forecasts was extended by 5 days using monthly mean climatological variables and zero precipitation in order to account for the effect of initial conditions. The 15-day spatially distributed ensemble runoff forecasts were then routed to four locations in the basin, each with different drainage areas. Surrogates for observed daily runoff and flow were provided by the reference run, specifically VIC simulation forced with ECMWF analysis fields and TMPA precipitation fields. The flood prediction scheme using the calibrated and downscaled ECMWF EPS forecasts was shown to be more accurate and reliable than interpolated forecasts for both daily distributed runoff forecasts and daily flow forecasts. Initial and antecedent conditions dominated the flow forecasts for lead times shorter than the time of concentration depending on the flow forecast amounts and the drainage area sizes. The flood prediction scheme had useful skill for the 10 following days at all sites.« less

Seasonal Drought Prediction in East Africa: Can National Multi-Model Ensemble Forecasts Help?

NASA Technical Reports Server (NTRS)

Shukla, Shraddhanand; Roberts, J. B.; Funk, Christopher; Robertson, F. R.; Hoell, Andrew

2015-01-01

The increasing food and water demands of East Africa's growing population are stressing the region's inconsistent water resources and rain-fed agriculture. As recently as in 2011 part of this region underwent one of the worst famine events in its history. Timely and skillful drought forecasts at seasonal scale for this region can inform better water and agro-pastoral management decisions, support optimal allocation of the region's water resources, and mitigate socio-economic losses incurred by droughts. However seasonal drought prediction in this region faces several challenges. Lack of skillful seasonal rainfall forecasts; the focus of this presentation, is one of those major challenges. In the past few decades, major strides have been taken towards improvement of seasonal scale dynamical climate forecasts. The National Centers for Environmental Prediction's (NCEP) National Multi-model Ensemble (NMME) is one such state-of-the-art dynamical climate forecast system. The NMME incorporates climate forecasts from 6+ fully coupled dynamical models resulting in 100+ ensemble member forecasts. Recent studies have indicated that in general NMME offers improvement over forecasts from any single model. However thus far the skill of NMME for forecasting rainfall in a vulnerable region like the East Africa has been unexplored. In this presentation we report findings of a comprehensive analysis that examines the strength and weakness of NMME in forecasting rainfall at seasonal scale in East Africa for all three of the prominent seasons for the region. (i.e. March-April-May, July-August-September and October-November- December). Simultaneously we also describe hybrid approaches; that combine statistical approaches with NMME forecasts; to improve rainfall forecast skill in the region when raw NMME forecasts lack in skill.

Seasonal Drought Prediction in East Africa: Can National Multi-Model Ensemble Forecasts Help?

NASA Technical Reports Server (NTRS)

Shukla, Shraddhanand; Roberts, J. B.; Funk, Christopher; Robertson, F. R.; Hoell, Andrew

2014-01-01

The increasing food and water demands of East Africa's growing population are stressing the region's inconsistent water resources and rain-fed agriculture. As recently as in 2011 part of this region underwent one of the worst famine events in its history. Timely and skillful drought forecasts at seasonal scale for this region can inform better water and agro-pastoral management decisions, support optimal allocation of the region's water resources, and mitigate socio-economic losses incurred by droughts. However seasonal drought prediction in this region faces several challenges. Lack of skillful seasonal rainfall forecasts; the focus of this presentation, is one of those major challenges. In the past few decades, major strides have been taken towards improvement of seasonal scale dynamical climate forecasts. The National Centers for Environmental Prediction's (NCEP) National Multi-model Ensemble (NMME) is one such state-of-the-art dynamical climate forecast system. The NMME incorporates climate forecasts from 6+ fully coupled dynamical models resulting in 100+ ensemble member forecasts. Recent studies have indicated that in general NMME offers improvement over forecasts from any single model. However thus far the skill of NMME for forecasting rainfall in a vulnerable region like the East Africa has been unexplored. In this presentation we report findings of a comprehensive analysis that examines the strength and weakness of NMME in forecasting rainfall at seasonal scale in East Africa for all three of the prominent seasons for the region. (i.e. March-April-May, July-August-September and October-November- December). Simultaneously we also describe hybrid approaches; that combine statistical approaches with NMME forecasts; to improve rainfall forecast skill in the region when raw NMME forecasts lack in skill.

Regional Model Nesting Within GFS Daily Forecasts Over West Africa

NASA Technical Reports Server (NTRS)

Druyan, Leonard M.; Fulakeza, Matthew; Lonergan, Patrick; Worrell, Ruben

2010-01-01

The study uses the RM3, the regional climate model at the Center for Climate Systems Research of Columbia University and the NASA/Goddard Institute for Space Studies (CCSR/GISS). The paper evaluates 30 48-hour RM3 weather forecasts over West Africa during September 2006 made on a 0.5 grid nested within 1 Global Forecast System (GFS) global forecasts. September 2006 was the Special Observing Period #3 of the African Monsoon Multidisciplinary Analysis (AMMA). Archived GFS initial conditions and lateral boundary conditions for the simulations from the US National Weather Service, National Oceanographic and Atmospheric Administration were interpolated four times daily. Results for precipitation forecasts are validated against Tropical Rainfall Measurement Mission (TRMM) satellite estimates and data from the Famine Early Warning System (FEWS), which includes rain gauge measurements, and forecasts of circulation are compared to reanalysis 2. Performance statistics for the precipitation forecasts include bias, root-mean-square errors and spatial correlation coefficients. The nested regional model forecasts are compared to GFS forecasts to gauge whether nesting provides additional realistic information. They are also compared to RM3 simulations driven by reanalysis 2, representing high potential skill forecasts, to gauge the sensitivity of results to lateral boundary conditions. Nested RM3/GFS forecasts generate excessive moisture advection toward West Africa, which in turn causes prodigious amounts of model precipitation. This problem is corrected by empirical adjustments in the preparation of lateral boundary conditions and initial conditions. The resulting modified simulations improve on the GFS precipitation forecasts, achieving time-space correlations with TRMM of 0.77 on the first day and 0.63 on the second day. One realtime RM3/GFS precipitation forecast made at and posted by the African Centre of Meteorological Application for Development (ACMAD) in Niamey, Niger is shown.

Developing International Guidelines on Volcanic Hazard Assessments for Nuclear Facilities

NASA Astrophysics Data System (ADS)

Connor, Charles

2014-05-01

Worldwide, tremendous progress has been made in recent decades in forecasting volcanic events, such as episodes of volcanic unrest, eruptions, and the potential impacts of eruptions. Generally these forecasts are divided into two categories. Short-term forecasts are prepared in response to unrest at volcanoes, rely on geophysical monitoring and related observations, and have the goal of forecasting events on timescales of hours to weeks to provide time for evacuation of people, shutdown of facilities, and implementation of related safety measures. Long-term forecasts are prepared to better understand the potential impacts of volcanism in the future and to plan for potential volcanic activity. Long-term forecasts are particularly useful to better understand and communicate the potential consequences of volcanic events for populated areas around volcanoes and for siting critical infrastructure, such as nuclear facilities. Recent work by an international team, through the auspices of the International Atomic Energy Agency, has focused on developing guidelines for long-term volcanic hazard assessments. These guidelines have now been implemented for hazard assessment for nuclear facilities in nations including Indonesia, the Philippines, Armenia, Chile, and the United States. One any time scale, all volcanic hazard assessments rely on a geologically reasonable conceptual model of volcanism. Such conceptual models are usually built upon years or decades of geological studies of specific volcanic systems, analogous systems, and development of a process-level understanding of volcanic activity. Conceptual models are used to bound potential rates of volcanic activity, potential magnitudes of eruptions, and to understand temporal and spatial trends in volcanic activity. It is these conceptual models that provide essential justification for assumptions made in statistical model development and the application of numerical models to generate quantitative forecasts. It is a tremendous challenge in quantitative volcanic hazard assessments to encompass alternative conceptual models, and to create models that are robust to evolving understanding of specific volcanic systems by the scientific community. A central question in volcanic hazards forecasts is quantifying rates of volcanic activity. Especially for long-dormant volcanic systems, data from the geologic record may be sparse, individual events may be missing or unrecognized in the geologic record, patterns of activity may be episodic or otherwise nonstationary. This leads to uncertainty in forecasting long-term rates of activity. Hazard assessments strive to quantify such uncertainty, for example by comparing observed rates of activity with alternative parametric and nonparametric models. Numerical models are presented that characterize the spatial distribution of potential volcanic events. These spatial density models serve as the basis for application of numerical models of specific phenomena such as development of lava flow, tephra fallout, and a host of other volcanic phenomena. Monte Carlo techniques (random sampling, stratified sampling, importance sampling) are methods used to sample vent location and other key eruption parameters, such as eruption volume, magma rheology, and eruption column height for probabilistic models. The development of coupled scenarios (e.g., the probability of tephra accumulation on a slope resulting in subsequent debris flows) is also assessed through these methods, usually with the aid of event trees. The primary products of long-term forecasts are a statistical model of the conditional probability of the potential effects of volcanism, should an eruption occur, and the probability of such activity occurring. It is emphasized that hazard forecasting is an iterative process, and board consideration must be given to alternative conceptual models of volcanism, weighting of volcanological data in the analyses, and alternative statistical and numerical models. This structure is amenable to expert elicitation in order to weight alternative models and to explore alternative scenarios.

Using a Software Tool in Forecasting: a Case Study of Sales Forecasting Taking into Account Data Uncertainty

NASA Astrophysics Data System (ADS)

Fabianová, Jana; Kačmáry, Peter; Molnár, Vieroslav; Michalik, Peter

2016-10-01

Forecasting is one of the logistics activities and a sales forecast is the starting point for the elaboration of business plans. Forecast accuracy affects the business outcomes and ultimately may significantly affect the economic stability of the company. The accuracy of the prediction depends on the suitability of the use of forecasting methods, experience, quality of input data, time period and other factors. The input data are usually not deterministic but they are often of random nature. They are affected by uncertainties of the market environment, and many other factors. Taking into account the input data uncertainty, the forecast error can by reduced. This article deals with the use of the software tool for incorporating data uncertainty into forecasting. Proposals are presented of a forecasting approach and simulation of the impact of uncertain input parameters to the target forecasted value by this case study model. The statistical analysis and risk analysis of the forecast results is carried out including sensitivity analysis and variables impact analysis.

Optimising seasonal streamflow forecast lead time for operational decision making in Australia

NASA Astrophysics Data System (ADS)

Schepen, Andrew; Zhao, Tongtiegang; Wang, Q. J.; Zhou, Senlin; Feikema, Paul

2016-10-01

Statistical seasonal forecasts of 3-month streamflow totals are released in Australia by the Bureau of Meteorology and updated on a monthly basis. The forecasts are often released in the second week of the forecast period, due to the onerous forecast production process. The current service relies on models built using data for complete calendar months, meaning the forecast production process cannot begin until the first day of the forecast period. Somehow, the bureau needs to transition to a service that provides forecasts before the beginning of the forecast period; timelier forecast release will become critical as sub-seasonal (monthly) forecasts are developed. Increasing the forecast lead time to one month ahead is not considered a viable option for Australian catchments that typically lack any predictability associated with snowmelt. The bureau's forecasts are built around Bayesian joint probability models that have antecedent streamflow, rainfall and climate indices as predictors. In this study, we adapt the modelling approach so that forecasts have any number of days of lead time. Daily streamflow and sea surface temperatures are used to develop predictors based on 28-day sliding windows. Forecasts are produced for 23 forecast locations with 0-14- and 21-day lead time. The forecasts are assessed in terms of continuous ranked probability score (CRPS) skill score and reliability metrics. CRPS skill scores, on average, reduce monotonically with increase in days of lead time, although both positive and negative differences are observed. Considering only skilful forecast locations, CRPS skill scores at 7-day lead time are reduced on average by 4 percentage points, with differences largely contained within +5 to -15 percentage points. A flexible forecasting system that allows for any number of days of lead time could benefit Australian seasonal streamflow forecast users by allowing more time for forecasts to be disseminated, comprehended and made use of prior to the commencement of a forecast season. The system would allow for forecasts to be updated if necessary.

Environmental forecasting and turbulence modeling

NASA Astrophysics Data System (ADS)

Hunt, J. C. R.

This review describes the fundamental assumptions and current methodologies of the two main kinds of environmental forecast; the first is valid for a limited period of time into the future and over a limited space-time ‘target’, and is largely determined by the initial and preceding state of the environment, such as the weather or pollution levels, up to the time when the forecast is issued and by its state at the edges of the region being considered; the second kind provides statistical information over long periods of time and/or over large space-time targets, so that they only depend on the statistical averages of the initial and ‘edge’ conditions. Environmental forecasts depend on the various ways that models are constructed. These range from those based on the ‘reductionist’ methodology (i.e., the combination of separate, scientifically based, models for the relevant processes) to those based on statistical methodologies, using a mixture of data and scientifically based empirical modeling. These are, as a rule, focused on specific quantities required for the forecast. The persistence and predictability of events associated with environmental and turbulent flows and the reasons for variation in the accuracy of their forecasts (of the first and second kinds) are now better understood and better modeled. This has partly resulted from using analogous results of disordered chaotic systems, and using the techniques of calculating ensembles of realizations, ideally involving several different models, so as to incorporate in the probabilistic forecasts a wider range of possible events. The rationale for such an approach needs to be developed. However, other insights have resulted from the recognition of the ordered, though randomly occurring, nature of the persistent motions in these flows, whose scales range from those of synoptic weather patterns (whether storms or ‘blocked’ anticyclones) to small scale vortices. These eigen states can be predicted from the reductionist models or may be modeled specifically, for example, in terms of ‘self-organized’ critical phenomena. It is noted how in certain applications of turbulent modeling its methods are beginning to resemble those of environmental simulations, because of the trend to introduce ‘on-line’ controls of the turbulent flows in advanced flows in advanced engineering fluid systems. In real time simulations, for both local environmental processes and these engineering systems, maximum information is needed about the likely flow patterns in order to optimize both the assimilation of limited real-time data and the use of limited real-time computing capacity. It is concluded that philosophical studies of how scientific models develop and of the concept of determinism in science are helpful in considering these complex issues.

Forecasting daily patient volumes in the emergency department.

PubMed

Jones, Spencer S; Thomas, Alun; Evans, R Scott; Welch, Shari J; Haug, Peter J; Snow, Gregory L

2008-02-01

Shifts in the supply of and demand for emergency department (ED) resources make the efficient allocation of ED resources increasingly important. Forecasting is a vital activity that guides decision-making in many areas of economic, industrial, and scientific planning, but has gained little traction in the health care industry. There are few studies that explore the use of forecasting methods to predict patient volumes in the ED. The goals of this study are to explore and evaluate the use of several statistical forecasting methods to predict daily ED patient volumes at three diverse hospital EDs and to compare the accuracy of these methods to the accuracy of a previously proposed forecasting method. Daily patient arrivals at three hospital EDs were collected for the period January 1, 2005, through March 31, 2007. The authors evaluated the use of seasonal autoregressive integrated moving average, time series regression, exponential smoothing, and artificial neural network models to forecast daily patient volumes at each facility. Forecasts were made for horizons ranging from 1 to 30 days in advance. The forecast accuracy achieved by the various forecasting methods was compared to the forecast accuracy achieved when using a benchmark forecasting method already available in the emergency medicine literature. All time series methods considered in this analysis provided improved in-sample model goodness of fit. However, post-sample analysis revealed that time series regression models that augment linear regression models by accounting for serial autocorrelation offered only small improvements in terms of post-sample forecast accuracy, relative to multiple linear regression models, while seasonal autoregressive integrated moving average, exponential smoothing, and artificial neural network forecasting models did not provide consistently accurate forecasts of daily ED volumes. This study confirms the widely held belief that daily demand for ED services is characterized by seasonal and weekly patterns. The authors compared several time series forecasting methods to a benchmark multiple linear regression model. The results suggest that the existing methodology proposed in the literature, multiple linear regression based on calendar variables, is a reasonable approach to forecasting daily patient volumes in the ED. However, the authors conclude that regression-based models that incorporate calendar variables, account for site-specific special-day effects, and allow for residual autocorrelation provide a more appropriate, informative, and consistently accurate approach to forecasting daily ED patient volumes.

Time series modeling and forecasting using memetic algorithms for regime-switching models.

PubMed

Bergmeir, Christoph; Triguero, Isaac; Molina, Daniel; Aznarte, José Luis; Benitez, José Manuel

2012-11-01

In this brief, we present a novel model fitting procedure for the neuro-coefficient smooth transition autoregressive model (NCSTAR), as presented by Medeiros and Veiga. The model is endowed with a statistically founded iterative building procedure and can be interpreted in terms of fuzzy rule-based systems. The interpretability of the generated models and a mathematically sound building procedure are two very important properties of forecasting models. The model fitting procedure employed by the original NCSTAR is a combination of initial parameter estimation by a grid search procedure with a traditional local search algorithm. We propose a different fitting procedure, using a memetic algorithm, in order to obtain more accurate models. An empirical evaluation of the method is performed, applying it to various real-world time series originating from three forecasting competitions. The results indicate that we can significantly enhance the accuracy of the models, making them competitive to models commonly used in the field.

Open-source Software for Demand Forecasting of Clinical Laboratory Test Volumes Using Time-series Analysis.

PubMed

Mohammed, Emad A; Naugler, Christopher

2017-01-01

Demand forecasting is the area of predictive analytics devoted to predicting future volumes of services or consumables. Fair understanding and estimation of how demand will vary facilitates the optimal utilization of resources. In a medical laboratory, accurate forecasting of future demand, that is, test volumes, can increase efficiency and facilitate long-term laboratory planning. Importantly, in an era of utilization management initiatives, accurately predicted volumes compared to the realized test volumes can form a precise way to evaluate utilization management initiatives. Laboratory test volumes are often highly amenable to forecasting by time-series models; however, the statistical software needed to do this is generally either expensive or highly technical. In this paper, we describe an open-source web-based software tool for time-series forecasting and explain how to use it as a demand forecasting tool in clinical laboratories to estimate test volumes. This tool has three different models, that is, Holt-Winters multiplicative, Holt-Winters additive, and simple linear regression. Moreover, these models are ranked and the best one is highlighted. This tool will allow anyone with historic test volume data to model future demand.

Open-source Software for Demand Forecasting of Clinical Laboratory Test Volumes Using Time-series Analysis

PubMed Central

Mohammed, Emad A.; Naugler, Christopher

2017-01-01

Background: Demand forecasting is the area of predictive analytics devoted to predicting future volumes of services or consumables. Fair understanding and estimation of how demand will vary facilitates the optimal utilization of resources. In a medical laboratory, accurate forecasting of future demand, that is, test volumes, can increase efficiency and facilitate long-term laboratory planning. Importantly, in an era of utilization management initiatives, accurately predicted volumes compared to the realized test volumes can form a precise way to evaluate utilization management initiatives. Laboratory test volumes are often highly amenable to forecasting by time-series models; however, the statistical software needed to do this is generally either expensive or highly technical. Method: In this paper, we describe an open-source web-based software tool for time-series forecasting and explain how to use it as a demand forecasting tool in clinical laboratories to estimate test volumes. Results: This tool has three different models, that is, Holt-Winters multiplicative, Holt-Winters additive, and simple linear regression. Moreover, these models are ranked and the best one is highlighted. Conclusion: This tool will allow anyone with historic test volume data to model future demand. PMID:28400996

A Wind Forecasting System for Energy Application

NASA Astrophysics Data System (ADS)

Courtney, Jennifer; Lynch, Peter; Sweeney, Conor

2010-05-01

Accurate forecasting of available energy is crucial for the efficient management and use of wind power in the national power grid. With energy output critically dependent upon wind strength there is a need to reduce the errors associated wind forecasting. The objective of this research is to get the best possible wind forecasts for the wind energy industry. To achieve this goal, three methods are being applied. First, a mesoscale numerical weather prediction (NWP) model called WRF (Weather Research and Forecasting) is being used to predict wind values over Ireland. Currently, a gird resolution of 10km is used and higher model resolutions are being evaluated to establish whether they are economically viable given the forecast skill improvement they produce. Second, the WRF model is being used in conjunction with ECMWF (European Centre for Medium-Range Weather Forecasts) ensemble forecasts to produce a probabilistic weather forecasting product. Due to the chaotic nature of the atmosphere, a single, deterministic weather forecast can only have limited skill. The ECMWF ensemble methods produce an ensemble of 51 global forecasts, twice a day, by perturbing initial conditions of a 'control' forecast which is the best estimate of the initial state of the atmosphere. This method provides an indication of the reliability of the forecast and a quantitative basis for probabilistic forecasting. The limitation of ensemble forecasting lies in the fact that the perturbed model runs behave differently under different weather patterns and each model run is equally likely to be closest to the observed weather situation. Models have biases, and involve assumptions about physical processes and forcing factors such as underlying topography. Third, Bayesian Model Averaging (BMA) is being applied to the output from the ensemble forecasts in order to statistically post-process the results and achieve a better wind forecasting system. BMA is a promising technique that will offer calibrated probabilistic wind forecasts which will be invaluable in wind energy management. In brief, this method turns the ensemble forecasts into a calibrated predictive probability distribution. Each ensemble member is provided with a 'weight' determined by its relative predictive skill over a training period of around 30 days. Verification of data is carried out using observed wind data from operational wind farms. These are then compared to existing forecasts produced by ECMWF and Met Eireann in relation to skill scores. We are developing decision-making models to show the benefits achieved using the data produced by our wind energy forecasting system. An energy trading model will be developed, based on the rules currently used by the Single Electricity Market Operator for energy trading in Ireland. This trading model will illustrate the potential for financial savings by using the forecast data generated by this research.

Effect of data quality on a hybrid Coulomb/STEP model for earthquake forecasting

NASA Astrophysics Data System (ADS)

Steacy, Sandy; Jimenez, Abigail; Gerstenberger, Matt; Christophersen, Annemarie

2014-05-01

Operational earthquake forecasting is rapidly becoming a 'hot topic' as civil protection authorities seek quantitative information on likely near future earthquake distributions during seismic crises. At present, most of the models in public domain are statistical and use information about past and present seismicity as well as b-value and Omori's law to forecast future rates. A limited number of researchers, however, are developing hybrid models which add spatial constraints from Coulomb stress modeling to existing statistical approaches. Steacy et al. (2013), for instance, recently tested a model that combines Coulomb stress patterns with the STEP (short-term earthquake probability) approach against seismicity observed during the 2010-2012 Canterbury earthquake sequence. They found that the new model performed at least as well as, and often better than, STEP when tested against retrospective data but that STEP was generally better in pseudo-prospective tests that involved data actually available within the first 10 days of each event of interest. They suggested that the major reason for this discrepancy was uncertainty in the slip models and, in particular, in the geometries of the faults involved in each complex major event. Here we test this hypothesis by developing a number of retrospective forecasts for the Landers earthquake using hypothetical slip distributions developed by Steacy et al. (2004) to investigate the sensitivity of Coulomb stress models to fault geometry and earthquake slip. Specifically, we consider slip models based on the NEIC location, the CMT solution, surface rupture, and published inversions and find significant variation in the relative performance of the models depending upon the input data.

Use of temperature to improve West Nile virus forecasts

PubMed Central

Schneider, Zachary D.; Caillouet, Kevin A.; Campbell, Scott R.; Damian, Dan; Irwin, Patrick; Jones, Herff M. P.; Townsend, John

2018-01-01

Ecological and laboratory studies have demonstrated that temperature modulates West Nile virus (WNV) transmission dynamics and spillover infection to humans. Here we explore whether inclusion of temperature forcing in a model depicting WNV transmission improves WNV forecast accuracy relative to a baseline model depicting WNV transmission without temperature forcing. Both models are optimized using a data assimilation method and two observed data streams: mosquito infection rates and reported human WNV cases. Each coupled model-inference framework is then used to generate retrospective ensemble forecasts of WNV for 110 outbreak years from among 12 geographically diverse United States counties. The temperature-forced model improves forecast accuracy for much of the outbreak season. From the end of July until the beginning of October, a timespan during which 70% of human cases are reported, the temperature-forced model generated forecasts of the total number of human cases over the next 3 weeks, total number of human cases over the season, the week with the highest percentage of infectious mosquitoes, and the peak percentage of infectious mosquitoes that on average increased absolute forecast accuracy 5%, 10%, 12%, and 6%, respectively, over the non-temperature forced baseline model. These results indicate that use of temperature forcing improves WNV forecast accuracy and provide further evidence that temperature influences rates of WNV transmission. The findings provide a foundation for implementation of a statistically rigorous system for real-time forecast of seasonal WNV outbreaks and their use as a quantitative decision support tool for public health officials and mosquito control programs. PMID:29522514

Combination of synoptical-analogous and dynamical methods to increase skill score of monthly air temperature forecasts over Northern Eurasia

NASA Astrophysics Data System (ADS)

Khan, Valentina; Tscepelev, Valery; Vilfand, Roman; Kulikova, Irina; Kruglova, Ekaterina; Tischenko, Vladimir

2016-04-01

Long-range forecasts at monthly-seasonal time scale are in great demand of socio-economic sectors for exploiting climate-related risks and opportunities. At the same time, the quality of long-range forecasts is not fully responding to user application necessities. Different approaches, including combination of different prognostic models, are used in forecast centers to increase the prediction skill for specific regions and globally. In the present study, two forecasting methods are considered which are exploited in operational practice of Hydrometeorological Center of Russia. One of them is synoptical-analogous method of forecasting of surface air temperature at monthly scale. Another one is dynamical system based on the global semi-Lagrangian model SL-AV, developed in collaboration of Institute of Numerical Mathematics and Hydrometeorological Centre of Russia. The seasonal version of this model has been used to issue global and regional forecasts at monthly-seasonal time scales. This study presents results of the evaluation of surface air temperature forecasts generated with using above mentioned synoptical-statistical and dynamical models, and their combination to potentially increase skill score over Northern Eurasia. The test sample of operational forecasts is encompassing period from 2010 through 2015. The seasonal and interannual variability of skill scores of these methods has been discussed. It was noticed that the quality of all forecasts is highly dependent on the inertia of macro-circulation processes. The skill scores of forecasts are decreasing during significant alterations of synoptical fields for both dynamical and empirical schemes. Procedure of combination of forecasts from different methods, in some cases, has demonstrated its effectiveness. For this study the support has been provided by Grant of Russian Science Foundation (№14-37-00053).

Crude oil price forecasting based on hybridizing wavelet multiple linear regression model, particle swarm optimization techniques, and principal component analysis.

PubMed

Shabri, Ani; Samsudin, Ruhaidah

2014-01-01

Crude oil prices do play significant role in the global economy and are a key input into option pricing formulas, portfolio allocation, and risk measurement. In this paper, a hybrid model integrating wavelet and multiple linear regressions (MLR) is proposed for crude oil price forecasting. In this model, Mallat wavelet transform is first selected to decompose an original time series into several subseries with different scale. Then, the principal component analysis (PCA) is used in processing subseries data in MLR for crude oil price forecasting. The particle swarm optimization (PSO) is used to adopt the optimal parameters of the MLR model. To assess the effectiveness of this model, daily crude oil market, West Texas Intermediate (WTI), has been used as the case study. Time series prediction capability performance of the WMLR model is compared with the MLR, ARIMA, and GARCH models using various statistics measures. The experimental results show that the proposed model outperforms the individual models in forecasting of the crude oil prices series.

Crude Oil Price Forecasting Based on Hybridizing Wavelet Multiple Linear Regression Model, Particle Swarm Optimization Techniques, and Principal Component Analysis

PubMed Central

Shabri, Ani; Samsudin, Ruhaidah

2014-01-01

Crude oil prices do play significant role in the global economy and are a key input into option pricing formulas, portfolio allocation, and risk measurement. In this paper, a hybrid model integrating wavelet and multiple linear regressions (MLR) is proposed for crude oil price forecasting. In this model, Mallat wavelet transform is first selected to decompose an original time series into several subseries with different scale. Then, the principal component analysis (PCA) is used in processing subseries data in MLR for crude oil price forecasting. The particle swarm optimization (PSO) is used to adopt the optimal parameters of the MLR model. To assess the effectiveness of this model, daily crude oil market, West Texas Intermediate (WTI), has been used as the case study. Time series prediction capability performance of the WMLR model is compared with the MLR, ARIMA, and GARCH models using various statistics measures. The experimental results show that the proposed model outperforms the individual models in forecasting of the crude oil prices series. PMID:24895666

Impact of Land Surface Initialization Approach on Subseasonal Forecast Skill: a Regional Analysis in the Southern Hemisphere

NASA Technical Reports Server (NTRS)

Hirsch, Annette L.; Kala, Jatin; Pitman, Andy J.; Carouge, Claire; Evans, Jason P.; Haverd, Vanessa; Mocko, David

2014-01-01

The authors use a sophisticated coupled land-atmosphere modeling system for a Southern Hemisphere subdomain centered over southeastern Australia to evaluate differences in simulation skill from two different land surface initialization approaches. The first approach uses equilibrated land surface states obtained from offline simulations of the land surface model, and the second uses land surface states obtained from reanalyses. The authors find that land surface initialization using prior offline simulations contribute to relative gains in subseasonal forecast skill. In particular, relative gains in forecast skill for temperature of 10%-20% within the first 30 days of the forecast can be attributed to the land surface initialization method using offline states. For precipitation there is no distinct preference for the land surface initialization method, with limited gains in forecast skill irrespective of the lead time. The authors evaluated the asymmetry between maximum and minimum temperatures and found that maximum temperatures had the largest gains in relative forecast skill, exceeding 20% in some regions. These results were statistically significant at the 98% confidence level at up to 60 days into the forecast period. For minimum temperature, using reanalyses to initialize the land surface contributed to relative gains in forecast skill, reaching 40% in parts of the domain that were statistically significant at the 98% confidence level. The contrasting impact of the land surface initialization method between maximum and minimum temperature was associated with different soil moisture coupling mechanisms. Therefore, land surface initialization from prior offline simulations does improve predictability for temperature, particularly maximum temperature, but with less obvious improvements for precipitation and minimum temperature over southeastern Australia.

A genetic-algorithm-based remnant grey prediction model for energy demand forecasting.

PubMed

Hu, Yi-Chung

2017-01-01

Energy demand is an important economic index, and demand forecasting has played a significant role in drawing up energy development plans for cities or countries. As the use of large datasets and statistical assumptions is often impractical to forecast energy demand, the GM(1,1) model is commonly used because of its simplicity and ability to characterize an unknown system by using a limited number of data points to construct a time series model. This paper proposes a genetic-algorithm-based remnant GM(1,1) (GARGM(1,1)) with sign estimation to further improve the forecasting accuracy of the original GM(1,1) model. The distinctive feature of GARGM(1,1) is that it simultaneously optimizes the parameter specifications of the original and its residual models by using the GA. The results of experiments pertaining to a real case of energy demand in China showed that the proposed GARGM(1,1) outperforms other remnant GM(1,1) variants.

A genetic-algorithm-based remnant grey prediction model for energy demand forecasting

PubMed Central

2017-01-01

Energy demand is an important economic index, and demand forecasting has played a significant role in drawing up energy development plans for cities or countries. As the use of large datasets and statistical assumptions is often impractical to forecast energy demand, the GM(1,1) model is commonly used because of its simplicity and ability to characterize an unknown system by using a limited number of data points to construct a time series model. This paper proposes a genetic-algorithm-based remnant GM(1,1) (GARGM(1,1)) with sign estimation to further improve the forecasting accuracy of the original GM(1,1) model. The distinctive feature of GARGM(1,1) is that it simultaneously optimizes the parameter specifications of the original and its residual models by using the GA. The results of experiments pertaining to a real case of energy demand in China showed that the proposed GARGM(1,1) outperforms other remnant GM(1,1) variants. PMID:28981548

Verification of space weather forecasts at the UK Met Office

NASA Astrophysics Data System (ADS)

Bingham, S.; Sharpe, M.; Jackson, D.; Murray, S.

2017-12-01

The UK Met Office Space Weather Operations Centre (MOSWOC) has produced space weather guidance twice a day since its official opening in 2014. Guidance includes 4-day probabilistic forecasts of X-ray flares, geomagnetic storms, high-energy electron events and high-energy proton events. Evaluation of such forecasts is important to forecasters, stakeholders, model developers and users to understand the performance of these forecasts and also strengths and weaknesses to enable further development. Met Office terrestrial near real-time verification systems have been adapted to provide verification of X-ray flare and geomagnetic storm forecasts. Verification is updated daily to produce Relative Operating Characteristic (ROC) curves and Reliability diagrams, and rolling Ranked Probability Skill Scores (RPSSs) thus providing understanding of forecast performance and skill. Results suggest that the MOSWOC issued X-ray flare forecasts are usually not statistically significantly better than a benchmark climatological forecast (where the climatology is based on observations from the previous few months). By contrast, the issued geomagnetic storm activity forecast typically performs better against this climatological benchmark.

ANEMOS: Development of a next generation wind power forecasting system for the large-scale integration of onshore and offshore wind farms.

NASA Astrophysics Data System (ADS)

Kariniotakis, G.; Anemos Team

2003-04-01

Objectives: Accurate forecasting of the wind energy production up to two days ahead is recognized as a major contribution for reliable large-scale wind power integration. Especially, in a liberalized electricity market, prediction tools enhance the position of wind energy compared to other forms of dispatchable generation. ANEMOS, is a new 3.5 years R&D project supported by the European Commission, that resembles research organizations and end-users with an important experience on the domain. The project aims to develop advanced forecasting models that will substantially outperform current methods. Emphasis is given to situations like complex terrain, extreme weather conditions, as well as to offshore prediction for which no specific tools currently exist. The prediction models will be implemented in a software platform and installed for online operation at onshore and offshore wind farms by the end-users participating in the project. Approach: The paper presents the methodology of the project. Initially, the prediction requirements are identified according to the profiles of the end-users. The project develops prediction models based on both a physical and an alternative statistical approach. Research on physical models gives emphasis to techniques for use in complex terrain and the development of prediction tools based on CFD techniques, advanced model output statistics or high-resolution meteorological information. Statistical models (i.e. based on artificial intelligence) are developed for downscaling, power curve representation, upscaling for prediction at regional or national level, etc. A benchmarking process is set-up to evaluate the performance of the developed models and to compare them with existing ones using a number of case studies. The synergy between statistical and physical approaches is examined to identify promising areas for further improvement of forecasting accuracy. Appropriate physical and statistical prediction models are also developed for offshore wind farms taking into account advances in marine meteorology (interaction between wind and waves, coastal effects). The benefits from the use of satellite radar images for modeling local weather patterns are investigated. A next generation forecasting software, ANEMOS, will be developed to integrate the various models. The tool is enhanced by advanced Information Communication Technology (ICT) functionality and can operate both in stand alone, or remote mode, or be interfaced with standard Energy or Distribution Management Systems (EMS/DMS) systems. Contribution: The project provides an advanced technology for wind resource forecasting applicable in a large scale: at a single wind farm, regional or national level and for both interconnected and island systems. A major milestone is the on-line operation of the developed software by the participating utilities for onshore and offshore wind farms and the demonstration of the economic benefits. The outcome of the ANEMOS project will help consistently the increase of wind integration in two levels; in an operational level due to better management of wind farms, but also, it will contribute to increasing the installed capacity of wind farms. This is because accurate prediction of the resource reduces the risk of wind farm developers, who are then more willing to undertake new wind farm installations especially in a liberalized electricity market environment.

Forecasting space weather: Can new econometric methods improve accuracy?

NASA Astrophysics Data System (ADS)

Reikard, Gordon

2011-06-01

Space weather forecasts are currently used in areas ranging from navigation and communication to electric power system operations. The relevant forecast horizons can range from as little as 24 h to several days. This paper analyzes the predictability of two major space weather measures using new time series methods, many of them derived from econometrics. The data sets are the A p geomagnetic index and the solar radio flux at 10.7 cm. The methods tested include nonlinear regressions, neural networks, frequency domain algorithms, GARCH models (which utilize the residual variance), state transition models, and models that combine elements of several techniques. While combined models are complex, they can be programmed using modern statistical software. The data frequency is daily, and forecasting experiments are run over horizons ranging from 1 to 7 days. Two major conclusions stand out. First, the frequency domain method forecasts the A p index more accurately than any time domain model, including both regressions and neural networks. This finding is very robust, and holds for all forecast horizons. Combining the frequency domain method with other techniques yields a further small improvement in accuracy. Second, the neural network forecasts the solar flux more accurately than any other method, although at short horizons (2 days or less) the regression and net yield similar results. The neural net does best when it includes measures of the long-term component in the data.

Use of Temperature to Improve West Nile Virus Forecasts

NASA Astrophysics Data System (ADS)

Shaman, J. L.; DeFelice, N.; Schneider, Z.; Little, E.; Barker, C.; Caillouet, K.; Campbell, S.; Damian, D.; Irwin, P.; Jones, H.; Townsend, J.

2017-12-01

Ecological and laboratory studies have demonstrated that temperature modulates West Nile virus (WNV) transmission dynamics and spillover infection to humans. Here we explore whether the inclusion of temperature forcing in a model depicting WNV transmission improves WNV forecast accuracy relative to a baseline model depicting WNV transmission without temperature forcing. Both models are optimized using a data assimilation method and two observed data streams: mosquito infection rates and reported human WNV cases. Each coupled model-inference framework is then used to generate retrospective ensemble forecasts of WNV for 110 outbreak years from among 12 geographically diverse United States counties. The temperature-forced model improves forecast accuracy for much of the outbreak season. From the end of July until the beginning of October, a timespan during which 70% of human cases are reported, the temperature-forced model generated forecasts of the total number of human cases over the next 3 weeks, total number of human cases over the season, the week with the highest percentage of infectious mosquitoes, and the peak percentage of infectious mosquitoes that were on average 5%, 10%, 12%, and 6% more accurate, respectively, than the baseline model. These results indicate that use of temperature forcing improves WNV forecast accuracy and provide further evidence that temperatures influence rates of WNV transmission. The findings help build a foundation for implementation of a statistically rigorous system for real-time forecast of seasonal WNV outbreaks and their use as a quantitative decision support tool for public health officials and mosquito control programs.

Municipal water consumption forecast accuracy

NASA Astrophysics Data System (ADS)

Fullerton, Thomas M.; Molina, Angel L.

2010-06-01

Municipal water consumption planning is an active area of research because of infrastructure construction and maintenance costs, supply constraints, and water quality assurance. In spite of that, relatively few water forecast accuracy assessments have been completed to date, although some internal documentation may exist as part of the proprietary "grey literature." This study utilizes a data set of previously published municipal consumption forecasts to partially fill that gap in the empirical water economics literature. Previously published municipal water econometric forecasts for three public utilities are examined for predictive accuracy against two random walk benchmarks commonly used in regional analyses. Descriptive metrics used to quantify forecast accuracy include root-mean-square error and Theil inequality statistics. Formal statistical assessments are completed using four-pronged error differential regression F tests. Similar to studies for other metropolitan econometric forecasts in areas with similar demographic and labor market characteristics, model predictive performances for the municipal water aggregates in this effort are mixed for each of the municipalities included in the sample. Given the competitiveness of the benchmarks, analysts should employ care when utilizing econometric forecasts of municipal water consumption for planning purposes, comparing them to recent historical observations and trends to insure reliability. Comparative results using data from other markets, including regions facing differing labor and demographic conditions, would also be helpful.

Rate/state Coulomb stress transfer model for the CSEP Japan seismicity forecast

NASA Astrophysics Data System (ADS)

Toda, Shinji; Enescu, Bogdan

2011-03-01

Numerous studies retrospectively found that seismicity rate jumps (drops) by coseismic Coulomb stress increase (decrease). The Collaboratory for the Study of Earthquake Prediction (CSEP) instead provides us an opportunity for prospective testing of the Coulomb hypothesis. Here we adapt our stress transfer model incorporating rate and state dependent friction law to the CSEP Japan seismicity forecast. We demonstrate how to compute the forecast rates of large shocks in 2009 using the large earthquakes during the past 120 years. The time dependent impact of the coseismic stress perturbations explains qualitatively well the occurrence of the recent moderate size shocks. Such ability is partly similar to that of statistical earthquake clustering models. However, our model differs from them as follows: the off-fault aftershock zones can be simulated using finite fault sources; the regional areal patterns of triggered seismicity are modified by the dominant mechanisms of the potential sources; the imparted stresses due to large earthquakes produce stress shadows that lead to a reduction of the forecasted number of earthquakes. Although the model relies on several unknown parameters, it is the first physics based model submitted to the CSEP Japan test center and has the potential to be tuned for short-term earthquake forecasts.

Operational foreshock forecasting: Fifteen years after

NASA Astrophysics Data System (ADS)

Ogata, Y.

2010-12-01

We are concerned with operational forecasting of the probability that events are foreshocks of a forthcoming earthquake that is significantly larger (mainshock). Specifically, we define foreshocks as the preshocks substantially smaller than the mainshock by a magnitude gap of 0.5 or larger. The probability gain of foreshock forecast is extremely high compare to long-term forecast by renewal processes or various alarm-based intermediate-term forecasts because of a large event’s low occurrence rate in a short period and a narrow target region. Thus, it is desired to establish operational foreshock probability forecasting as seismologists have done for aftershocks. When a series of earthquakes occurs in a region, we attempt to discriminate foreshocks from a swarm or mainshock-aftershock sequence. Namely, after real time identification of an earthquake cluster using methods such as the single-link algorithm, the probability is calculated by applying statistical features that discriminate foreshocks from other types of clusters, by considering the events' stronger proximity in time and space and tendency towards chronologically increasing magnitudes. These features were modeled for probability forecasting and the coefficients of the model were estimated in Ogata et al. (1996) for the JMA hypocenter data (M≧4, 1926-1993). Currently, fifteen years has passed since the publication of the above-stated work so that we are able to present the performance and validation of the forecasts (1994-2009) by using the same model. Taking isolated events into consideration, the probability of the first events in a potential cluster being a foreshock vary in a range between 0+% and 10+% depending on their locations. This conditional forecasting performs significantly better than the unconditional (average) foreshock probability of 3.7% throughout Japan region. Furthermore, when we have the additional events in a cluster, the forecast probabilities range more widely from nearly 0% to about 40% depending on the discrimination features among the events in the cluster. This conditional forecasting further performs significantly better than the unconditional foreshock probability of 7.3%, which is the average probability of the plural events in the earthquake clusters. Indeed, the frequency ratios of the actual foreshocks are consistent with the forecasted probabilities. Reference: Ogata, Y., Utsu, T. and Katsura, K. (1996). Statistical discrimination of foreshocks from other earthquake clusters, Geophys. J. Int. 127, 17-30.

Spectral Analysis of Forecast Error Investigated with an Observing System Simulation Experiment

NASA Technical Reports Server (NTRS)

Prive, N. C.; Errico, Ronald M.

2015-01-01

The spectra of analysis and forecast error are examined using the observing system simulation experiment (OSSE) framework developed at the National Aeronautics and Space Administration Global Modeling and Assimilation Office (NASAGMAO). A global numerical weather prediction model, the Global Earth Observing System version 5 (GEOS-5) with Gridpoint Statistical Interpolation (GSI) data assimilation, is cycled for two months with once-daily forecasts to 336 hours to generate a control case. Verification of forecast errors using the Nature Run as truth is compared with verification of forecast errors using self-analysis; significant underestimation of forecast errors is seen using self-analysis verification for up to 48 hours. Likewise, self analysis verification significantly overestimates the error growth rates of the early forecast, as well as mischaracterizing the spatial scales at which the strongest growth occurs. The Nature Run-verified error variances exhibit a complicated progression of growth, particularly for low wave number errors. In a second experiment, cycling of the model and data assimilation over the same period is repeated, but using synthetic observations with different explicitly added observation errors having the same error variances as the control experiment, thus creating a different realization of the control. The forecast errors of the two experiments become more correlated during the early forecast period, with correlations increasing for up to 72 hours before beginning to decrease.

Experiments with a three-dimensional statistical objective analysis scheme using FGGE data

NASA Technical Reports Server (NTRS)

Baker, Wayman E.; Bloom, Stephen C.; Woollen, John S.; Nestler, Mark S.; Brin, Eugenia

1987-01-01

A three-dimensional (3D), multivariate, statistical objective analysis scheme (referred to as optimum interpolation or OI) has been developed for use in numerical weather prediction studies with the FGGE data. Some novel aspects of the present scheme include: (1) a multivariate surface analysis over the oceans, which employs an Ekman balance instead of the usual geostrophic relationship, to model the pressure-wind error cross correlations, and (2) the capability to use an error correlation function which is geographically dependent. A series of 4-day data assimilation experiments are conducted to examine the importance of some of the key features of the OI in terms of their effects on forecast skill, as well as to compare the forecast skill using the OI with that utilizing a successive correction method (SCM) of analysis developed earlier. For the three cases examined, the forecast skill is found to be rather insensitive to varying the error correlation function geographically. However, significant differences are noted between forecasts from a two-dimensional (2D) version of the OI and those from the 3D OI, with the 3D OI forecasts exhibiting better forecast skill. The 3D OI forecasts are also more accurate than those from the SCM initial conditions. The 3D OI with the multivariate oceanic surface analysis was found to produce forecasts which were slightly more accurate, on the average, than a univariate version.

A prospective earthquake forecast experiment in the western Pacific

NASA Astrophysics Data System (ADS)

Eberhard, David A. J.; Zechar, J. Douglas; Wiemer, Stefan

2012-09-01

Since the beginning of 2009, the Collaboratory for the Study of Earthquake Predictability (CSEP) has been conducting an earthquake forecast experiment in the western Pacific. This experiment is an extension of the Kagan-Jackson experiments begun 15 years earlier and is a prototype for future global earthquake predictability experiments. At the beginning of each year, seismicity models make a spatially gridded forecast of the number of Mw≥ 5.8 earthquakes expected in the next year. For the three participating statistical models, we analyse the first two years of this experiment. We use likelihood-based metrics to evaluate the consistency of the forecasts with the observed target earthquakes and we apply measures based on Student's t-test and the Wilcoxon signed-rank test to compare the forecasts. Overall, a simple smoothed seismicity model (TripleS) performs the best, but there are some exceptions that indicate continued experiments are vital to fully understand the stability of these models, the robustness of model selection and, more generally, earthquake predictability in this region. We also estimate uncertainties in our results that are caused by uncertainties in earthquake location and seismic moment. Our uncertainty estimates are relatively small and suggest that the evaluation metrics are relatively robust. Finally, we consider the implications of our results for a global earthquake forecast experiment.

Model selection as a science driver for dark energy surveys

NASA Astrophysics Data System (ADS)

Mukherjee, Pia; Parkinson, David; Corasaniti, Pier Stefano; Liddle, Andrew R.; Kunz, Martin

2006-07-01

A key science goal of upcoming dark energy surveys is to seek time-evolution of the dark energy. This problem is one of model selection, where the aim is to differentiate between cosmological models with different numbers of parameters. However, the power of these surveys is traditionally assessed by estimating their ability to constrain parameters, which is a different statistical problem. In this paper, we use Bayesian model selection techniques, specifically forecasting of the Bayes factors, to compare the abilities of different proposed surveys in discovering dark energy evolution. We consider six experiments - supernova luminosity measurements by the Supernova Legacy Survey, SNAP, JEDI and ALPACA, and baryon acoustic oscillation measurements by WFMOS and JEDI - and use Bayes factor plots to compare their statistical constraining power. The concept of Bayes factor forecasting has much broader applicability than dark energy surveys.

Assimilating NOAA SST data into BSH operational circulation model for North and Baltic Seas

NASA Astrophysics Data System (ADS)

Losa, Svetlana; Schroeter, Jens; Nerger, Lars; Janjic, Tijana; Danilov, Sergey; Janssen, Frank

A data assimilation (DA) system is developed for BSH operational circulation model in order to improve forecast of current velocities, sea surface height, temperature and salinity in the North and Baltic Seas. Assimilated data are NOAA sea surface temperature (SST) data for the following period: 01.10.07 -30.09.08. All data assimilation experiments are based on im-plementation of one of the so-called statistical DA methods -Singular Evolutive Interpolated Kalman (SEIK) filter, -with different ways of prescribing assumed model and data errors statis-tics. Results of the experiments will be shown and compared against each other. Hydrographic data from MARNET stations and sea level at series of tide gauges are used as independent information to validate the data assimilation system. Keywords: Operational Oceanography and forecasting

Statistical Analysis of Model Data for Operational Space Launch Weather Support at Kennedy Space Center and Cape Canaveral Air Force Station

NASA Technical Reports Server (NTRS)

Bauman, William H., III

2010-01-01

The 12-km resolution North American Mesoscale (NAM) model (MesoNAM) is used by the 45th Weather Squadron (45 WS) Launch Weather Officers at Kennedy Space Center (KSC) and Cape Canaveral Air Force Station (CCAFS) to support space launch weather operations. The 45 WS tasked the Applied Meteorology Unit to conduct an objective statistics-based analysis of MesoNAM output compared to wind tower mesonet observations and then develop a an operational tool to display the results. The National Centers for Environmental Prediction began running the current version of the MesoNAM in mid-August 2006. The period of record for the dataset was 1 September 2006 - 31 January 2010. The AMU evaluated MesoNAM hourly forecasts from 0 to 84 hours based on model initialization times of 00, 06, 12 and 18 UTC. The MesoNAM forecast winds, temperature and dew point were compared to the observed values of these parameters from the sensors in the KSC/CCAFS wind tower network. The data sets were stratified by model initialization time, month and onshore/offshore flow for each wind tower. Statistics computed included bias (mean difference), standard deviation of the bias, root mean square error (RMSE) and a hypothesis test for bias = O. Twelve wind towers located in close proximity to key launch complexes were used for the statistical analysis with the sensors on the towers positioned at varying heights to include 6 ft, 30 ft, 54 ft, 60 ft, 90 ft, 162 ft, 204 ft and 230 ft depending on the launch vehicle and associated weather launch commit criteria being evaluated. These twelve wind towers support activities for the Space Shuttle (launch and landing), Delta IV, Atlas V and Falcon 9 launch vehicles. For all twelve towers, the results indicate a diurnal signal in the bias of temperature (T) and weaker but discernable diurnal signal in the bias of dewpoint temperature (T(sub d)) in the MesoNAM forecasts. Also, the standard deviation of the bias and RMSE of T, T(sub d), wind speed and wind direction indicated the model error increased with the forecast period all four parameters. The hypothesis testing uses statistics to determine the probability that a given hypothesis is true. The goal of using the hypothesis test was to determine if the model bias of any of the parameters assessed throughout the model forecast period was statistically zero. For th is dataset, if this test produced a value >= -1 .96 or <= 1.96 for a data point, then the bias at that point was effectively zero and the model forecast for that point was considered to have no error. A graphical user interface (GUI) was developed so the 45 WS would have an operational tool at their disposal that would be easy to navigate among the multiple stratifications of information to include tower locations, month, model initialization times, sensor heights and onshore/offshore flow. The AMU developed the GUI using HyperText Markup Language (HTML) so the tool could be used in most popular web browsers with computers running different operating systems such as Microsoft Windows and Linux.

Probabilistic postprocessing models for flow forecasts for a system of catchments and several lead times

NASA Astrophysics Data System (ADS)

Engeland, Kolbjorn; Steinsland, Ingelin

2014-05-01

This study introduces a methodology for the construction of probabilistic inflow forecasts for multiple catchments and lead times, and investigates criterions for evaluation of multi-variate forecasts. A post-processing approach is used, and a Gaussian model is applied for transformed variables. The post processing model has two main components, the mean model and the dependency model. The mean model is used to estimate the marginal distributions for forecasted inflow for each catchment and lead time, whereas the dependency models was used to estimate the full multivariate distribution of forecasts, i.e. co-variances between catchments and lead times. In operational situations, it is a straightforward task to use the models to sample inflow ensembles which inherit the dependencies between catchments and lead times. The methodology was tested and demonstrated in the river systems linked to the Ulla-Førre hydropower complex in southern Norway, where simultaneous probabilistic forecasts for five catchments and ten lead times were constructed. The methodology exhibits sufficient flexibility to utilize deterministic flow forecasts from a numerical hydrological model as well as statistical forecasts such as persistent forecasts and sliding window climatology forecasts. It also deals with variation in the relative weights of these forecasts with both catchment and lead time. When evaluating predictive performance in original space using cross validation, the case study found that it is important to include the persistent forecast for the initial lead times and the hydrological forecast for medium-term lead times. Sliding window climatology forecasts become more important for the latest lead times. Furthermore, operationally important features in this case study such as heteroscedasticity, lead time varying between lead time dependency and lead time varying between catchment dependency are captured. Two criterions were used for evaluating the added value of the dependency model. The first one was the Energy score (ES) that is a multi-dimensional generalization of continuous rank probability score (CRPS). ES was calculated for all lead-times and catchments together, for each catchment across all lead times and for each lead time across all catchments. The second criterion was to use CRPS for forecasted inflows accumulated over several lead times and catchments. The results showed that ES was not very sensitive to correct covariance structure, whereas CRPS for accumulated flows where more suitable for evaluating the dependency model. This indicates that it is more appropriate to evaluate relevant univariate variables that depends on the dependency structure then to evaluate the multivariate forecast directly.

Analysis Monthly Import of Palm Oil Products Using Box-Jenkins Model

NASA Astrophysics Data System (ADS)

Ahmad, Nurul F. Y.; Khalid, Kamil; Saifullah Rusiman, Mohd; Ghazali Kamardan, M.; Roslan, Rozaini; Che-Him, Norziha

2018-04-01

The palm oil industry has been an important component of the national economy especially the agriculture sector. The aim of this study is to identify the pattern of import of palm oil products, to model the time series using Box-Jenkins model and to forecast the monthly import of palm oil products. The method approach is included in the statistical test for verifying the equivalence model and statistical measurement of three models, namely Autoregressive (AR) model, Moving Average (MA) model and Autoregressive Moving Average (ARMA) model. The model identification of all product import palm oil is different in which the AR(1) was found to be the best model for product import palm oil while MA(3) was found to be the best model for products import palm kernel oil. For the palm kernel, MA(4) was found to be the best model. The results forecast for the next four months for products import palm oil, palm kernel oil and palm kernel showed the most significant decrease compared to the actual data.

Error Estimation of An Ensemble Statistical Seasonal Precipitation Prediction Model

NASA Technical Reports Server (NTRS)

Shen, Samuel S. P.; Lau, William K. M.; Kim, Kyu-Myong; Li, Gui-Long

2001-01-01

This NASA Technical Memorandum describes an optimal ensemble canonical correlation forecasting model for seasonal precipitation. Each individual forecast is based on the canonical correlation analysis (CCA) in the spectral spaces whose bases are empirical orthogonal functions (EOF). The optimal weights in the ensemble forecasting crucially depend on the mean square error of each individual forecast. An estimate of the mean square error of a CCA prediction is made also using the spectral method. The error is decomposed onto EOFs of the predictand and decreases linearly according to the correlation between the predictor and predictand. Since new CCA scheme is derived for continuous fields of predictor and predictand, an area-factor is automatically included. Thus our model is an improvement of the spectral CCA scheme of Barnett and Preisendorfer. The improvements include (1) the use of area-factor, (2) the estimation of prediction error, and (3) the optimal ensemble of multiple forecasts. The new CCA model is applied to the seasonal forecasting of the United States (US) precipitation field. The predictor is the sea surface temperature (SST). The US Climate Prediction Center's reconstructed SST is used as the predictor's historical data. The US National Center for Environmental Prediction's optimally interpolated precipitation (1951-2000) is used as the predictand's historical data. Our forecast experiments show that the new ensemble canonical correlation scheme renders a reasonable forecasting skill. For example, when using September-October-November SST to predict the next season December-January-February precipitation, the spatial pattern correlation between the observed and predicted are positive in 46 years among the 50 years of experiments. The positive correlations are close to or greater than 0.4 in 29 years, which indicates excellent performance of the forecasting model. The forecasting skill can be further enhanced when several predictors are used.

Evaluation of the Impact of AIRS Radiance and Profile Data Assimilation in Partly Cloudy Regions

NASA Technical Reports Server (NTRS)

Zavodsky, Bradley; Srikishen, Jayanthi; Jedlovec, Gary

2013-01-01

Improvements to global and regional numerical weather prediction have been demonstrated through assimilation of data from NASA s Atmospheric Infrared Sounder (AIRS). Current operational data assimilation systems use AIRS radiances, but impact on regional forecasts has been much smaller than for global forecasts. Retrieved profiles from AIRS contain much of the information that is contained in the radiances and may be able to reveal reasons for this reduced impact. Assimilating AIRS retrieved profiles in an identical analysis configuration to the radiances, tracking the quantity and quality of the assimilated data in each technique, and examining analysis increments and forecast impact from each data type can yield clues as to the reasons for the reduced impact. By doing this with regional scale models individual synoptic features (and the impact of AIRS on these features) can be more easily tracked. This project examines the assimilation of hyperspectral sounder data used in operational numerical weather prediction by comparing operational techniques used for AIRS radiances and research techniques used for AIRS retrieved profiles. Parallel versions of a configuration of the Weather Research and Forecasting (WRF) model with Gridpoint Statistical Interpolation (GSI) are run to examine the impact AIRS radiances and retrieved profiles. Statistical evaluation of a long-term series of forecast runs will be compared along with preliminary results of in-depth investigations for select case comparing the analysis increments in partly cloudy regions and short-term forecast impacts.

Evaluation of the Impact of Atmospheric Infrared Sounder (AIRS) Radiance and Profile Data Assimilation in Partly Cloudy Regions

NASA Technical Reports Server (NTRS)

Zavodsky, Bradley; Srikishen, Jayanthi; Jedlovec, Gary

2013-01-01

Improvements to global and regional numerical weather prediction have been demonstrated through assimilation of data from NASA s Atmospheric Infrared Sounder (AIRS). Current operational data assimilation systems use AIRS radiances, but impact on regional forecasts has been much smaller than for global forecasts. Retrieved profiles from AIRS contain much of the information that is contained in the radiances and may be able to reveal reasons for this reduced impact. Assimilating AIRS retrieved profiles in an identical analysis configuration to the radiances, tracking the quantity and quality of the assimilated data in each technique, and examining analysis increments and forecast impact from each data type can yield clues as to the reasons for the reduced impact. By doing this with regional scale models individual synoptic features (and the impact of AIRS on these features) can be more easily tracked. This project examines the assimilation of hyperspectral sounder data used in operational numerical weather prediction by comparing operational techniques used for AIRS radiances and research techniques used for AIRS retrieved profiles. Parallel versions of a configuration of the Weather Research and Forecasting (WRF) model with Gridpoint Statistical Interpolation (GSI) are run to examine the impact AIRS radiances and retrieved profiles. Statistical evaluation of 6 weeks of forecast runs will be compared along with preliminary results of in-depth investigations for select case comparing the analysis increments in partly cloudy regions and short-term forecast impacts.

Application of a deep-learning method to the forecast of daily solar flare occurrence using Convolution Neural Network

NASA Astrophysics Data System (ADS)

Shin, Seulki; Moon, Yong-Jae; Chu, Hyoungseok

2017-08-01

As the application of deep-learning methods has been succeeded in various fields, they have a high potential to be applied to space weather forecasting. Convolutional neural network, one of deep learning methods, is specialized in image recognition. In this study, we apply the AlexNet architecture, which is a winner of Imagenet Large Scale Virtual Recognition Challenge (ILSVRC) 2012, to the forecast of daily solar flare occurrence using the MatConvNet software of MATLAB. Our input images are SOHO/MDI, EIT 195Å, and 304Å from January 1996 to December 2010, and output ones are yes or no of flare occurrence. We select training dataset from Jan 1996 to Dec 2000 and from Jan 2003 to Dec 2008. Testing dataset is chosen from Jan 2001 to Dec 2002 and from Jan 2009 to Dec 2010 in order to consider the solar cycle effect. In training dataset, we randomly select one fifth of training data for validation dataset to avoid the overfitting problem. Our model successfully forecasts the flare occurrence with about 0.90 probability of detection (POD) for common flares (C-, M-, and X-class). While POD of major flares (M- and X-class) forecasting is 0.96, false alarm rate (FAR) also scores relatively high(0.60). We also present several statistical parameters such as critical success index (CSI) and true skill statistics (TSS). Our model can immediately be applied to automatic forecasting service when image data are available.

An operational wave forecasting system for the east coast of India

NASA Astrophysics Data System (ADS)

Sandhya, K. G.; Murty, P. L. N.; Deshmukh, Aditya N.; Balakrishnan Nair, T. M.; Shenoi, S. S. C.

2018-03-01

Demand for operational ocean state forecasting is increasing, owing to the ever-increasing marine activities in the context of blue economy. In the present study, an operational wave forecasting system for the east coast of India is proposed using unstructured Simulating WAves Nearshore model (UNSWAN). This modelling system uses very high resolution mesh near the Indian east coast and coarse resolution offshore, and thus avoids the necessity of nesting with a global wave model. The model is forced with European Centre for Medium-Range Weather Forecasts (ECMWF) winds and simulates wave parameters and wave spectra for the next 3 days. The spatial pictures of satellite data overlaid on simulated wave height show that the model is capable of simulating the significant wave heights and their gradients realistically. Spectral validation has been done using the available data to prove the reliability of the model. To further evaluate the model performance, the wave forecast for the entire year 2014 is evaluated against buoy measurements over the region at 4 waverider buoy locations. Seasonal analysis of significant wave height (Hs) at the four locations showed that the correlation between the modelled and observed was the highest (in the range 0.78-0.96) during the post-monsoon season. The variability of Hs was also the highest during this season at all locations. The error statistics showed clear seasonal and geographical location dependence. The root mean square error at Visakhapatnam was the same (0.25) for all seasons, but it was the smallest for pre-monsoon season (0.12 m and 0.17 m) for Puducherry and Gopalpur. The wind sea component showed higher variability compared to the corresponding swell component in all locations and for all seasons. The variability was picked by the model to a reasonable level in most of the cases. The results of statistical analysis show that the modelling system is suitable for use in the operational scenario.

Application of time series analysis in modelling and forecasting emergency department visits in a medical centre in Southern Taiwan.

PubMed

Juang, Wang-Chuan; Huang, Sin-Jhih; Huang, Fong-Dee; Cheng, Pei-Wen; Wann, Shue-Ren

2017-12-01

Emergency department (ED) overcrowding is acknowledged as an increasingly important issue worldwide. Hospital managers are increasingly paying attention to ED crowding in order to provide higher quality medical services to patients. One of the crucial elements for a good management strategy is demand forecasting. Our study sought to construct an adequate model and to forecast monthly ED visits. We retrospectively gathered monthly ED visits from January 2009 to December 2016 to carry out a time series autoregressive integrated moving average (ARIMA) analysis. Initial development of the model was based on past ED visits from 2009 to 2016. A best-fit model was further employed to forecast the monthly data of ED visits for the next year (2016). Finally, we evaluated the predicted accuracy of the identified model with the mean absolute percentage error (MAPE). The software packages SAS/ETS V.9.4 and Office Excel 2016 were used for all statistical analyses. A series of statistical tests showed that six models, including ARIMA (0, 0, 1), ARIMA (1, 0, 0), ARIMA (1, 0, 1), ARIMA (2, 0, 1), ARIMA (3, 0, 1) and ARIMA (5, 0, 1), were candidate models. The model that gave the minimum Akaike information criterion and Schwartz Bayesian criterion and followed the assumptions of residual independence was selected as the adequate model. Finally, a suitable ARIMA (0, 0, 1) structure, yielding a MAPE of 8.91%, was identified and obtained as Visit t =7111.161+(a t +0.37462 a t -1). The ARIMA (0, 0, 1) model can be considered adequate for predicting future ED visits, and its forecast results can be used to aid decision-making processes. © Article author(s) (or their employer(s) unless otherwise stated in the text of the article) 2017. All rights reserved. No commercial use is permitted unless otherwise expressly granted.

Is it growing exponentially fast? -- Impact of assuming exponential growth for characterizing and forecasting epidemics with initial near-exponential growth dynamics.

PubMed

Chowell, Gerardo; Viboud, Cécile

2016-10-01

The increasing use of mathematical models for epidemic forecasting has highlighted the importance of designing models that capture the baseline transmission characteristics in order to generate reliable epidemic forecasts. Improved models for epidemic forecasting could be achieved by identifying signature features of epidemic growth, which could inform the design of models of disease spread and reveal important characteristics of the transmission process. In particular, it is often taken for granted that the early growth phase of different growth processes in nature follow early exponential growth dynamics. In the context of infectious disease spread, this assumption is often convenient to describe a transmission process with mass action kinetics using differential equations and generate analytic expressions and estimates of the reproduction number. In this article, we carry out a simulation study to illustrate the impact of incorrectly assuming an exponential-growth model to characterize the early phase (e.g., 3-5 disease generation intervals) of an infectious disease outbreak that follows near-exponential growth dynamics. Specifically, we assess the impact on: 1) goodness of fit, 2) bias on the growth parameter, and 3) the impact on short-term epidemic forecasts. Designing transmission models and statistical approaches that more flexibly capture the profile of epidemic growth could lead to enhanced model fit, improved estimates of key transmission parameters, and more realistic epidemic forecasts.

Mental Models of Software Forecasting

NASA Technical Reports Server (NTRS)

Hihn, J.; Griesel, A.; Bruno, K.; Fouser, T.; Tausworthe, R.

1993-01-01

The majority of software engineers resist the use of the currently available cost models. One problem is that the mathematical and statistical models that are currently available do not correspond with the mental models of the software engineers. In an earlier JPL funded study (Hihn and Habib-agahi, 1991) it was found that software engineers prefer to use analogical or analogy-like techniques to derive size and cost estimates, whereas curren CER's hide any analogy in the regression equations. In addition, the currently available models depend upon information which is not available during early planning when the most important forecasts must be made.

Landslide early warning based on failure forecast models: the example of Mt. de La Saxe rockslide, northern Italy

NASA Astrophysics Data System (ADS)

Manconi, A.; Giordan, D.

2015-02-01

We investigate the use of landslide failure forecast models by exploiting near-real-time monitoring data. Starting from the inverse velocity theory, we analyze landslide surface displacements on different temporal windows, and apply straightforward statistical methods to obtain confidence intervals on the estimated time of failure. Here we describe the main concepts of our method, and show an example of application to a real emergency scenario, the La Saxe rockslide, Aosta Valley region, northern Italy. Based on the herein presented case study, we identify operational thresholds based on the reliability of the forecast models, in order to support the management of early warning systems in the most critical phases of the landslide emergency.

An analysis, sensitivity and prediction of winter fog events using FASP model over Indo-Gangetic plains, India

NASA Astrophysics Data System (ADS)

Srivastava, S. K., Sr.; Sharma, D. A.; Sachdeva, K.

2017-12-01

Indo-Gangetic plains of India experience severe fog conditions during the peak winter months of December and January every year. In this paper an attempt has been to analyze the spatial and temporal variability of winter fog over Indo-Gangetic plains. Further, an attempt has also been made to configure an efficient meso-scale numerical weather prediction model using different parameterization schemes and develop a forecasting tool for prediction of fog during winter months over Indo-Gangetic plains. The study revealed that an alarming increasing positive trend of fog frequency prevails over many locations of IGP. Hot spot and cluster analysis were conducted to identify the high fog prone zones using GIS and inferential statistical tools respectively. Hot spots on an average experiences fog on 68.27% days, it is followed by moderate and cold spots with 48.03% and 21.79% respectively. The study proposes a new FASP (Fog Analysis, sensitivity and prediction) Model for overall analysis and prediction of fog at a particular location and period over IGP. In the first phase of this model long term climatological fog data of a location is analyzed to determine its characteristics and prevailing trend using various advanced statistical techniques. During a second phase a sensitivity test is conducted with different combination of parameterization schemes to determine the most suitable combination for fog simulation over a particular location and period and in the third and final phase, first ARIMA model is used to predict the number of fog days in future . Thereafter, Numerical model is used to predict the various meteorological parameters favourable for fog forecast. Finally, Hybrid model is used for fog forecast over the study location. The results of the FASP model are validated with actual ground based fog data using statistical tools. Forecast Fog-gram generated using hybrid model during Jan 2017 shows highly encouraging results for fog occurrence/Non occurrence between 25 hrs to 72 hours forecast. The model predicted the fog occurrences/Non occurrence with more than 85 % accuracy over most of the locations across the study area. The minimum visibility departure is within 500 m on 90% occasions over the central IGP and within 1000m on more than 80 % occasions over most of the locations across Indo-Gangetic plains.

Season-ahead water quality forecasts for the Schuylkill River, Pennsylvania

NASA Astrophysics Data System (ADS)

Block, P. J.; Leung, K.

2013-12-01

Anticipating and preparing for elevated water quality parameter levels in critical water sources, using weather forecasts, is not uncommon. In this study, we explore the feasibility of extending this prediction scale to a season-ahead for the Schuylkill River in Philadelphia, utilizing both statistical and dynamical prediction models, to characterize the season. This advance information has relevance for recreational activities, ecosystem health, and water treatment, as the Schuylkill provides 40% of Philadelphia's water supply. The statistical model associates large-scale climate drivers with streamflow and water quality parameter levels; numerous variables from NOAA's CFSv2 model are evaluated for the dynamical approach. A multi-model combination is also assessed. Results indicate moderately skillful prediction of average summertime total coliform and wintertime turbidity, using season-ahead oceanic and atmospheric variables, predominantly from the North Atlantic Ocean. Models predicting the number of elevated turbidity events across the wintertime season are also explored.

NMME Monthly / Seasonal Forecasts for NASA SERVIR Applications Science

NASA Astrophysics Data System (ADS)

Robertson, F. R.; Roberts, J. B.

2014-12-01

This work details use of the North American Multi-Model Ensemble (NMME) experimental forecasts as drivers for Decision Support Systems (DSSs) in the NASA / USAID initiative, SERVIR (a Spanish acronym meaning "to serve"). SERVIR integrates satellite observations, ground-based data and forecast models to monitor and forecast environmental changes and to improve response to natural disasters. Through the use of DSSs whose "front ends" are physically based models, the SERVIR activity provides a natural testbed to determine the extent to which NMME monthly to seasonal projections enable scientists, educators, project managers and policy implementers in developing countries to better use probabilistic outlooks of seasonal hydrologic anomalies in assessing agricultural / food security impacts, water availability, and risk to societal infrastructure. The multi-model NMME framework provides a "best practices" approach to probabilistic forecasting. The NMME forecasts are generated at resolution more coarse than that required to support DSS models; downscaling in both space and time is necessary. The methodology adopted here applied model output statistics where we use NMME ensemble monthly projections of sea-surface temperature (SST) and precipitation from 30 years of hindcasts with observations of precipitation and temperature for target regions. Since raw model forecasts are well-known to have structural biases, a cross-validated multivariate regression methodology (CCA) is used to link the model projected states as predictors to the predictands of the target region. The target regions include a number of basins in East and South Africa as well as the Ganges / Baramaputra / Meghna basin complex. The MOS approach used address spatial downscaling. Temporal disaggregation of monthly seasonal forecasts is achieved through use of a tercile bootstrapping approach. We interpret the results of these studies, the levels of skill by several metrics, and key uncertainties.

NMME Monthly / Seasonal Forecasts for NASA SERVIR Applications Science

NASA Technical Reports Server (NTRS)

Robertson, Franklin R.; Roberts, Jason B.

2014-01-01

This work details use of the North American Multi-Model Ensemble (NMME) experimental forecasts as drivers for Decision Support Systems (DSSs) in the NASA / USAID initiative, SERVIR (a Spanish acronym meaning "to serve"). SERVIR integrates satellite observations, ground-based data and forecast models to monitor and forecast environmental changes and to improve response to natural disasters. Through the use of DSSs whose "front ends" are physically based models, the SERVIR activity provides a natural testbed to determine the extent to which NMME monthly to seasonal projections enable scientists, educators, project managers and policy implementers in developing countries to better use probabilistic outlooks of seasonal hydrologic anomalies in assessing agricultural / food security impacts, water availability, and risk to societal infrastructure. The multi-model NMME framework provides a "best practices" approach to probabilistic forecasting. The NMME forecasts are generated at resolution more coarse than that required to support DSS models; downscaling in both space and time is necessary. The methodology adopted here applied model output statistics where we use NMME ensemble monthly projections of sea-surface temperature (SST) and precipitation from 30 years of hindcasts with observations of precipitation and temperature for target regions. Since raw model forecasts are well-known to have structural biases, a cross-validated multivariate regression methodology (CCA) is used to link the model projected states as predictors to the predictands of the target region. The target regions include a number of basins in East and South Africa as well as the Ganges / Baramaputra / Meghna basin complex. The MOS approach used address spatial downscaling. Temporal disaggregation of monthly seasonal forecasts is achieved through use of a tercile bootstrapping approach. We interpret the results of these studies, the levels of skill by several metrics, and key uncertainties.

Improvement of short-term numerical wind predictions

NASA Astrophysics Data System (ADS)

Bedard, Joel

Geophysic Model Output Statistics (GMOS) are developed to optimize the use of NWP for complex sites. GMOS differs from other MOS that are widely used by meteorological centers in the following aspects: it takes into account the surrounding geophysical parameters such as surface roughness, terrain height, etc., along with wind direction; it can be directly applied without any training, although training will further improve the results. The GMOS was applied to improve the Environment Canada GEM-LAM 2.5km forecasts at North Cape (PEI, Canada): It improves the predictions RMSE by 25-30% for all time horizons and almost all meteorological conditions; the topographic signature of the forecast error due to insufficient grid refinement is eliminated and the NWP combined with GMOS outperform the persistence from a 2h horizon, instead of 4h without GMOS. Finally, GMOS was applied at another site (Bouctouche, NB, Canada): similar improvements were observed, thus showing its general applicability. Keywords: wind energy, wind power forecast, numerical weather prediction, complex sites, model output statistics

An ensemble-ANFIS based uncertainty assessment model for forecasting multi-scalar standardized precipitation index

NASA Astrophysics Data System (ADS)

Ali, Mumtaz; Deo, Ravinesh C.; Downs, Nathan J.; Maraseni, Tek

2018-07-01

Forecasting drought by means of the World Meteorological Organization-approved Standardized Precipitation Index (SPI) is considered to be a fundamental task to support socio-economic initiatives and effectively mitigating the climate-risk. This study aims to develop a robust drought modelling strategy to forecast multi-scalar SPI in drought-rich regions of Pakistan where statistically significant lagged combinations of antecedent SPI are used to forecast future SPI. With ensemble-Adaptive Neuro Fuzzy Inference System ('ensemble-ANFIS') executed via a 10-fold cross-validation procedure, a model is constructed by randomly partitioned input-target data. Resulting in 10-member ensemble-ANFIS outputs, judged by mean square error and correlation coefficient in the training period, the optimal forecasts are attained by the averaged simulations, and the model is benchmarked with M5 Model Tree and Minimax Probability Machine Regression (MPMR). The results show the proposed ensemble-ANFIS model's preciseness was notably better (in terms of the root mean square and mean absolute error including the Willmott's, Nash-Sutcliffe and Legates McCabe's index) for the 6- and 12- month compared to the 3-month forecasts as verified by the largest error proportions that registered in smallest error band. Applying 10-member simulations, ensemble-ANFIS model was validated for its ability to forecast severity (S), duration (D) and intensity (I) of drought (including the error bound). This enabled uncertainty between multi-models to be rationalized more efficiently, leading to a reduction in forecast error caused by stochasticity in drought behaviours. Through cross-validations at diverse sites, a geographic signature in modelled uncertainties was also calculated. Considering the superiority of ensemble-ANFIS approach and its ability to generate uncertainty-based information, the study advocates the versatility of a multi-model approach for drought-risk forecasting and its prime importance for estimating drought properties over confidence intervals to generate better information for strategic decision-making.

Study on the medical meteorological forecast of the number of hypertension inpatient based on SVR

NASA Astrophysics Data System (ADS)

Zhai, Guangyu; Chai, Guorong; Zhang, Haifeng

2017-06-01

The purpose of this study is to build a hypertension prediction model by discussing the meteorological factors for hypertension incidence. The research method is selecting the standard data of relative humidity, air temperature, visibility, wind speed and air pressure of Lanzhou from 2010 to 2012(calculating the maximum, minimum and average value with 5 days as a unit ) as the input variables of Support Vector Regression(SVR) and the standard data of hypertension incidence of the same period as the output dependent variables to obtain the optimal prediction parameters by cross validation algorithm, then by SVR algorithm learning and training, a SVR forecast model for hypertension incidence is built. The result shows that the hypertension prediction model is composed of 15 input independent variables, the training accuracy is 0.005, the final error is 0.0026389. The forecast accuracy based on SVR model is 97.1429%, which is higher than statistical forecast equation and neural network prediction method. It is concluded that SVR model provides a new method for hypertension prediction with its simple calculation, small error as well as higher historical sample fitting and Independent sample forecast capability.

Verifying Operational and Developmental Air Force Weather Cloud Analysis and Forecast Products Using Lidar Data from Department of Energy Atmospheric Radiation Measurement (ARM) Sites

NASA Astrophysics Data System (ADS)

Hildebrand, E. P.

2017-12-01

Air Force Weather has developed various cloud analysis and forecast products designed to support global Department of Defense (DoD) missions. A World-Wide Merged Cloud Analysis (WWMCA) and short term Advected Cloud (ADVCLD) forecast is generated hourly using data from 16 geostationary and polar-orbiting satellites. Additionally, WWMCA and Numerical Weather Prediction (NWP) data are used in a statistical long-term (out to five days) cloud forecast model known as the Diagnostic Cloud Forecast (DCF). The WWMCA and ADVCLD are generated on the same polar stereographic 24 km grid for each hemisphere, whereas the DCF is generated on the same grid as its parent NWP model. When verifying the cloud forecast models, the goal is to understand not only the ability to detect cloud, but also the ability to assign it to the correct vertical layer. ADVCLD and DCF forecasts traditionally have been verified using WWMCA data as truth, but this might over-inflate the performance of those models because WWMCA also is a primary input dataset for those models. Because of this, in recent years, a WWMCA Reanalysis product has been developed, but this too is not a fully independent dataset. This year, work has been done to incorporate data from external, independent sources to verify not only the cloud forecast products, but the WWMCA data itself. One such dataset that has been useful for examining the 3-D performance of the cloud analysis and forecast models is Atmospheric Radiation Measurement (ARM) data from various sites around the globe. This presentation will focus on the use of the Department of Energy (DoE) ARM data to verify Air Force Weather cloud analysis and forecast products. Results will be presented to show relative strengths and weaknesses of the analyses and forecasts.

Moving beyond the cost-loss ratio: economic assessment of streamflow forecasts for a risk-averse decision maker

NASA Astrophysics Data System (ADS)

Matte, Simon; Boucher, Marie-Amélie; Boucher, Vincent; Fortier Filion, Thomas-Charles

2017-06-01

A large effort has been made over the past 10 years to promote the operational use of probabilistic or ensemble streamflow forecasts. Numerous studies have shown that ensemble forecasts are of higher quality than deterministic ones. Many studies also conclude that decisions based on ensemble rather than deterministic forecasts lead to better decisions in the context of flood mitigation. Hence, it is believed that ensemble forecasts possess a greater economic and social value for both decision makers and the general population. However, the vast majority of, if not all, existing hydro-economic studies rely on a cost-loss ratio framework that assumes a risk-neutral decision maker. To overcome this important flaw, this study borrows from economics and evaluates the economic value of early warning flood systems using the well-known Constant Absolute Risk Aversion (CARA) utility function, which explicitly accounts for the level of risk aversion of the decision maker. This new framework allows for the full exploitation of the information related to a forecasts' uncertainty, making it especially suited for the economic assessment of ensemble or probabilistic forecasts. Rather than comparing deterministic and ensemble forecasts, this study focuses on comparing different types of ensemble forecasts. There are multiple ways of assessing and representing forecast uncertainty. Consequently, there exist many different means of building an ensemble forecasting system for future streamflow. One such possibility is to dress deterministic forecasts using the statistics of past error forecasts. Such dressing methods are popular among operational agencies because of their simplicity and intuitiveness. Another approach is the use of ensemble meteorological forecasts for precipitation and temperature, which are then provided as inputs to one or many hydrological model(s). In this study, three concurrent ensemble streamflow forecasting systems are compared: simple statistically dressed deterministic forecasts, forecasts based on meteorological ensembles, and a variant of the latter that also includes an estimation of state variable uncertainty. This comparison takes place for the Montmorency River, a small flood-prone watershed in southern central Quebec, Canada. The assessment of forecasts is performed for lead times of 1 to 5 days, both in terms of forecasts' quality (relative to the corresponding record of observations) and in terms of economic value, using the new proposed framework based on the CARA utility function. It is found that the economic value of a forecast for a risk-averse decision maker is closely linked to the forecast reliability in predicting the upper tail of the streamflow distribution. Hence, post-processing forecasts to avoid over-forecasting could help improve both the quality and the value of forecasts.

A comparison of ensemble post-processing approaches that preserve correlation structures

NASA Astrophysics Data System (ADS)

Schefzik, Roman; Van Schaeybroeck, Bert; Vannitsem, Stéphane

2016-04-01

Despite the fact that ensemble forecasts address the major sources of uncertainty, they exhibit biases and dispersion errors and therefore are known to improve by calibration or statistical post-processing. For instance the ensemble model output statistics (EMOS) method, also known as non-homogeneous regression approach (Gneiting et al., 2005) is known to strongly improve forecast skill. EMOS is based on fitting and adjusting a parametric probability density function (PDF). However, EMOS and other common post-processing approaches apply to a single weather quantity at a single location for a single look-ahead time. They are therefore unable of taking into account spatial, inter-variable and temporal dependence structures. Recently many research efforts have been invested in designing post-processing methods that resolve this drawback but also in verification methods that enable the detection of dependence structures. New verification methods are applied on two classes of post-processing methods, both generating physically coherent ensembles. A first class uses the ensemble copula coupling (ECC) that starts from EMOS but adjusts the rank structure (Schefzik et al., 2013). The second class is a member-by-member post-processing (MBM) approach that maps each raw ensemble member to a corrected one (Van Schaeybroeck and Vannitsem, 2015). We compare variants of the EMOS-ECC and MBM classes and highlight a specific theoretical connection between them. All post-processing variants are applied in the context of the ensemble system of the European Centre of Weather Forecasts (ECMWF) and compared using multivariate verification tools including the energy score, the variogram score (Scheuerer and Hamill, 2015) and the band depth rank histogram (Thorarinsdottir et al., 2015). Gneiting, Raftery, Westveld, and Goldman, 2005: Calibrated probabilistic forecasting using ensemble model output statistics and minimum CRPS estimation. Mon. Wea. Rev., {133}, 1098-1118. Scheuerer and Hamill, 2015. Variogram-based proper scoring rules for probabilistic forecasts of multivariate quantities. Mon. Wea. Rev. {143},1321-1334. Schefzik, Thorarinsdottir, Gneiting. Uncertainty quantification in complex simulation models using ensemble copula coupling. Statistical Science {28},616-640, 2013. Thorarinsdottir, M. Scheuerer, and C. Heinz, 2015. Assessing the calibration of high-dimensional ensemble forecasts using rank histograms, arXiv:1310.0236. Van Schaeybroeck and Vannitsem, 2015: Ensemble post-processing using member-by-member approaches: theoretical aspects. Q.J.R. Meteorol. Soc., 141: 807-818.

Validating the WRF-Chem model for wind energy applications using High Resolution Doppler Lidar data from a Utah 2012 field campaign

NASA Astrophysics Data System (ADS)

Mitchell, M. J.; Pichugina, Y. L.; Banta, R. M.

2015-12-01

Models are important tools for assessing potential of wind energy sites, but the accuracy of these projections has not been properly validated. In this study, High Resolution Doppler Lidar (HRDL) data obtained with high temporal and spatial resolution at heights of modern turbine rotors were compared to output from the WRF-chem model in order to help improve the performance of the model in producing accurate wind forecasts for the industry. HRDL data were collected from January 23-March 1, 2012 during the Uintah Basin Winter Ozone Study (UBWOS) field campaign. A model validation method was based on the qualitative comparison of the wind field images, time-series analysis and statistical analysis of the observed and modeled wind speed and direction, both for case studies and for the whole experiment. To compare the WRF-chem model output to the HRDL observations, the model heights and forecast times were interpolated to match the observed times and heights. Then, time-height cross-sections of the HRDL and WRF-Chem wind speed and directions were plotted to select case studies. Cross-sections of the differences between the observed and forecasted wind speed and directions were also plotted to visually analyze the model performance in different wind flow conditions. A statistical analysis includes the calculation of vertical profiles and time series of bias, correlation coefficient, root mean squared error, and coefficient of determination between two datasets. The results from this analysis reveals where and when the model typically struggles in forecasting winds at heights of modern turbine rotors so that in the future the model can be improved for the industry.

The MSFC Solar Activity Future Estimation (MSAFE) Model

NASA Technical Reports Server (NTRS)

Suggs, Ronnie J.

2017-01-01

The MSAFE model provides forecasts for the solar indices SSN, F10.7, and Ap. These solar indices are used as inputs to many space environment models used in orbital spacecraft operations and space mission analysis. Forecasts from the MSAFE model are provided on the MSFC Natural Environments Branch's solar webpage and are updated as new monthly observations come available. The MSAFE prediction routine employs a statistical technique that calculates deviations of past solar cycles from the mean cycle and performs a regression analysis to predict the deviation from the mean cycle of the solar index at the next future time interval. The prediction algorithm is applied recursively to produce monthly smoothed solar index values for the remaining of the cycle. The forecasts are initiated for a given cycle after about 8 to 12 months of observations are collected. A forecast made at the beginning of cycle 24 using the MSAFE program captured the cycle fairly well with some difficulty in discerning the double peak that occurred at solar cycle maximum.

Statistical bias correction modelling for seasonal rainfall forecast for the case of Bali island

NASA Astrophysics Data System (ADS)

Lealdi, D.; Nurdiati, S.; Sopaheluwakan, A.

2018-04-01

Rainfall is an element of climate which is highly influential to the agricultural sector. Rain pattern and distribution highly determines the sustainability of agricultural activities. Therefore, information on rainfall is very useful for agriculture sector and farmers in anticipating the possibility of extreme events which often cause failures of agricultural production. This research aims to identify the biases from seasonal forecast products from ECMWF (European Centre for Medium-Range Weather Forecasts) rainfall forecast and to build a transfer function in order to correct the distribution biases as a new prediction model using quantile mapping approach. We apply this approach to the case of Bali Island, and as a result, the use of bias correction methods in correcting systematic biases from the model gives better results. The new prediction model obtained with this approach is better than ever. We found generally that during rainy season, the bias correction approach performs better than in dry season.

Assessing the viability of `over-the-loop' real-time short-to-medium range ensemble streamflow forecasts

NASA Astrophysics Data System (ADS)

Wood, A. W.; Clark, E.; Mendoza, P. A.; Nijssen, B.; Newman, A. J.; Clark, M. P.; Arnold, J.; Nowak, K. C.

2016-12-01

Many if not most national operational short-to-medium range streamflow prediction systems rely on a forecaster-in-the-loop approach in which some parts of the forecast workflow are automated, but others require the hands-on-effort of an experienced human forecaster. This approach evolved out of the need to correct for deficiencies in the models and datasets that were available for forecasting, and often leads to skillful predictions despite the use of relatively simple, conceptual models. On the other hand, the process is not reproducible, which limits opportunities to assess and incorporate process variations, and the effort required to make forecasts in this way is an obstacle to expanding forecast services - e.g., though adding new forecast locations or more frequent forecast updates, running more complex models, or producing forecast ensembles and hindcasts that can support verification. In the last decade, the hydrologic forecasting community has begun to develop more centralized, `over-the-loop' systems. The quality of these new forecast products will depend on their ability to leverage research in areas including earth system modeling, parameter estimation, data assimilation, statistical post-processing, weather and climate prediction, verification, and uncertainty estimation through the use of ensembles. Currently, the operational streamflow forecasting and water management communities have little experience with the strengths and weaknesses of over-the-loop approaches, even as the systems are being rolled out in major operational forecasting centers. There is thus a need both to evaluate these forecasting advances and to demonstrate their potential in a public arena, raising awareness in forecast user communities and development programs alike. To address this need, the National Center for Atmospheric Research is collaborating with the University of Washington, the Bureau of Reclamation and the US Army Corps of Engineers, using the NCAR 'System for Hydromet Analysis, Research, and Prediction' (SHARP) to implement, assess and demonstrate real-time over-the-loop forecasts. We present early hindcast and verification results from SHARP for short to medium range streamflow forecasts in a number of US case study watersheds.

Visibility Modeling and Forecasting for Abu Dhabi using Time Series Analysis Method

NASA Astrophysics Data System (ADS)

Eibedingil, I. G.; Abula, B.; Afshari, A.; Temimi, M.

2015-12-01

Land-Atmosphere interactions-their strength, directionality and evolution-are one of the main sources of uncertainty in contemporary climate modeling. A particularly crucial role in sustaining and modulating land-atmosphere interaction is the one of aerosols and dusts. Aerosols are tiny particles suspended in the air ranging from a few nanometers to a few hundred micrometers in diameter. Furthermore, the amount of dust and fog in the atmosphere is an important measure of visibility, which is another dimension of land-atmosphere interactions. Visibility affects all form of traffic, aviation, land and sailing. Being able to predict the change of visibility in the air in advance enables relevant authorities to take necessary actions before the disaster falls. Time Series Analysis (TAS) method is an emerging technique for modeling and forecasting the behavior of land-atmosphere interactions, including visibility. This research assess the dynamics and evolution of visibility around Abu Dhabi International Airport (+24.4320 latitude, +54.6510 longitude, and 27m elevation) using mean daily visibility and mean daily wind speed. TAS has been first used to model and forecast the visibility, and then the Transfer Function Model has been applied, considering the wind speed as an exogenous variable. By considering the Akaike Information Criterion (AIC) and Mean Absolute Percentage Error (MAPE) as a statistical criteria, two forecasting models namely univarite time series model and transfer function model, were developed to forecast the visibility around Abu Dhabi International Airport for three weeks. Transfer function model improved the MAPE of the forecast significantly.

Quantitative precipitation forecasts in the Alps - an assessment from the Forecast Demonstration Project MAP D-PHASE

NASA Astrophysics Data System (ADS)

Ament, F.; Weusthoff, T.; Arpagaus, M.; Rotach, M.

2009-04-01

The main aim of the WWRP Forecast Demonstration Project MAP D-PHASE is to demonstrate the performance of today's models to forecast heavy precipitation and flood events in the Alpine region. Therefore an end-to-end, real-time forecasting system was installed and operated during the D PHASE Operations Period from June to November 2007. Part of this system are 30 numerical weather prediction models (deterministic as well as ensemble systems) operated by weather services and research institutes, which issue alerts if predicted precipitation accumulations exceed critical thresholds. Additionally to the real-time alerts, all relevant model fields of these simulations are stored in a central data archive. This comprehensive data set allows a detailed assessment of today's quantitative precipitation forecast (QPF) performance in the Alpine region. We will present results of QPF verifications against Swiss radar and rain gauge data both from a qualitative point of view, in terms of alerts, as well as from a quantitative perspective, in terms of precipitation rate. Various influencing factors like lead time, accumulation time, selection of warning thresholds, or bias corrections will be discussed. Additional to traditional verifications of area average precipitation amounts, the performance of the models to predict the correct precipitation statistics without requiring a point-to-point match will be described by using modern Fuzzy verification techniques. Both analyses reveal significant advantages of deep convection resolving models compared to coarser models with parameterized convection. An intercomparison of the model forecasts themselves reveals a remarkably high variability between different models, and makes it worthwhile to evaluate the potential of a multi-model ensemble. Various multi-model ensemble strategies will be tested by combining D-PHASE models to virtual ensemble systems.

Time Series Analysis for Forecasting Hospital Census: Application to the Neonatal Intensive Care Unit

PubMed Central

Hoover, Stephen; Jackson, Eric V.; Paul, David; Locke, Robert

2016-01-01

Summary Background Accurate prediction of future patient census in hospital units is essential for patient safety, health outcomes, and resource planning. Forecasting census in the Neonatal Intensive Care Unit (NICU) is particularly challenging due to limited ability to control the census and clinical trajectories. The fixed average census approach, using average census from previous year, is a forecasting alternative used in clinical practice, but has limitations due to census variations. Objective Our objectives are to: (i) analyze the daily NICU census at a single health care facility and develop census forecasting models, (ii) explore models with and without patient data characteristics obtained at the time of admission, and (iii) evaluate accuracy of the models compared with the fixed average census approach. Methods We used five years of retrospective daily NICU census data for model development (January 2008 – December 2012, N=1827 observations) and one year of data for validation (January – December 2013, N=365 observations). Best-fitting models of ARIMA and linear regression were applied to various 7-day prediction periods and compared using error statistics. Results The census showed a slightly increasing linear trend. Best fitting models included a non-seasonal model, ARIMA(1,0,0), seasonal ARIMA models, ARIMA(1,0,0)x(1,1,2)7 and ARIMA(2,1,4)x(1,1,2)14, as well as a seasonal linear regression model. Proposed forecasting models resulted on average in 36.49% improvement in forecasting accuracy compared with the fixed average census approach. Conclusions Time series models provide higher prediction accuracy under different census conditions compared with the fixed average census approach. Presented methodology is easily applicable in clinical practice, can be generalized to other care settings, support short- and long-term census forecasting, and inform staff resource planning. PMID:27437040

Time Series Analysis for Forecasting Hospital Census: Application to the Neonatal Intensive Care Unit.

PubMed

Capan, Muge; Hoover, Stephen; Jackson, Eric V; Paul, David; Locke, Robert

2016-01-01

Accurate prediction of future patient census in hospital units is essential for patient safety, health outcomes, and resource planning. Forecasting census in the Neonatal Intensive Care Unit (NICU) is particularly challenging due to limited ability to control the census and clinical trajectories. The fixed average census approach, using average census from previous year, is a forecasting alternative used in clinical practice, but has limitations due to census variations. Our objectives are to: (i) analyze the daily NICU census at a single health care facility and develop census forecasting models, (ii) explore models with and without patient data characteristics obtained at the time of admission, and (iii) evaluate accuracy of the models compared with the fixed average census approach. We used five years of retrospective daily NICU census data for model development (January 2008 - December 2012, N=1827 observations) and one year of data for validation (January - December 2013, N=365 observations). Best-fitting models of ARIMA and linear regression were applied to various 7-day prediction periods and compared using error statistics. The census showed a slightly increasing linear trend. Best fitting models included a non-seasonal model, ARIMA(1,0,0), seasonal ARIMA models, ARIMA(1,0,0)x(1,1,2)7 and ARIMA(2,1,4)x(1,1,2)14, as well as a seasonal linear regression model. Proposed forecasting models resulted on average in 36.49% improvement in forecasting accuracy compared with the fixed average census approach. Time series models provide higher prediction accuracy under different census conditions compared with the fixed average census approach. Presented methodology is easily applicable in clinical practice, can be generalized to other care settings, support short- and long-term census forecasting, and inform staff resource planning.

Short-range quantitative precipitation forecasting using Deep Learning approaches

NASA Astrophysics Data System (ADS)

Akbari Asanjan, A.; Yang, T.; Gao, X.; Hsu, K. L.; Sorooshian, S.

2017-12-01

Predicting short-range quantitative precipitation is very important for flood forecasting, early flood warning and other hydrometeorological purposes. This study aims to improve the precipitation forecasting skills using a recently developed and advanced machine learning technique named Long Short-Term Memory (LSTM). The proposed LSTM learns the changing patterns of clouds from Cloud-Top Brightness Temperature (CTBT) images, retrieved from the infrared channel of Geostationary Operational Environmental Satellite (GOES), using a sophisticated and effective learning method. After learning the dynamics of clouds, the LSTM model predicts the upcoming rainy CTBT events. The proposed model is then merged with a precipitation estimation algorithm termed Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN) to provide precipitation forecasts. The results of merged LSTM with PERSIANN are compared to the results of an Elman-type Recurrent Neural Network (RNN) merged with PERSIANN and Final Analysis of Global Forecast System model over the states of Oklahoma, Florida and Oregon. The performance of each model is investigated during 3 storm events each located over one of the study regions. The results indicate the outperformance of merged LSTM forecasts comparing to the numerical and statistical baselines in terms of Probability of Detection (POD), False Alarm Ratio (FAR), Critical Success Index (CSI), RMSE and correlation coefficient especially in convective systems. The proposed method shows superior capabilities in short-term forecasting over compared methods.

Forecasting the mortality rates of Malaysian population using Heligman-Pollard model

NASA Astrophysics Data System (ADS)

Ibrahim, Rose Irnawaty; Mohd, Razak; Ngataman, Nuraini; Abrisam, Wan Nur Azifah Wan Mohd

2017-08-01

Actuaries, demographers and other professionals have always been aware of the critical importance of mortality forecasting due to declining trend of mortality and continuous increases in life expectancy. Heligman-Pollard model was introduced in 1980 and has been widely used by researchers in modelling and forecasting future mortality. This paper aims to estimate an eight-parameter model based on Heligman and Pollard's law of mortality. Since the model involves nonlinear equations that are explicitly difficult to solve, the Matrix Laboratory Version 7.0 (MATLAB 7.0) software will be used in order to estimate the parameters. Statistical Package for the Social Sciences (SPSS) will be applied to forecast all the parameters according to Autoregressive Integrated Moving Average (ARIMA). The empirical data sets of Malaysian population for period of 1981 to 2015 for both genders will be considered, which the period of 1981 to 2010 will be used as "training set" and the period of 2011 to 2015 as "testing set". In order to investigate the accuracy of the estimation, the forecast results will be compared against actual data of mortality rates. The result shows that Heligman-Pollard model fit well for male population at all ages while the model seems to underestimate the mortality rates for female population at the older ages.

Evaluation of the 29-km Eta Model for Weather Support to the United States Space Program

NASA Technical Reports Server (NTRS)

Manobianco, John; Nutter, Paul

1997-01-01

The Applied Meteorology Unit (AMU) conducted a year-long evaluation of NCEP's 29-km mesoscale Eta (meso-eta) weather prediction model in order to identify added value to forecast operations in support of the United States space program. The evaluation was stratified over warm and cool seasons and considered both objective and subjective verification methodologies. Objective verification results generally indicate that meso-eta model point forecasts at selected stations exhibit minimal error growth in terms of RMS errors and are reasonably unbiased. Conversely, results from the subjective verification demonstrate that model forecasts of developing weather events such as thunderstorms, sea breezes, and cold fronts, are not always as accurate as implied by the seasonal error statistics. Sea-breeze case studies reveal that the model generates a dynamically-consistent thermally direct circulation over the Florida peninsula, although at a larger scale than observed. Thunderstorm verification reveals that the meso-eta model is capable of predicting areas of organized convection, particularly during the late afternoon hours but is not capable of forecasting individual thunderstorms. Verification of cold fronts during the cool season reveals that the model is capable of forecasting a majority of cold frontal passages through east central Florida to within +1-h of observed frontal passage.

Forecasting Japanese encephalitis incidence from historical morbidity patterns: Statistical analysis with 27 years of observation in Assam, India.

PubMed

Handique, Bijoy K; Khan, Siraj A; Mahanta, J; Sudhakar, S

2014-09-01

Japanese encephalitis (JE) is one of the dreaded mosquito-borne viral diseases mostly prevalent in south Asian countries including India. Early warning of the disease in terms of disease intensity is crucial for taking adequate and appropriate intervention measures. The present study was carried out in Dibrugarh district in the state of Assam located in the northeastern region of India to assess the accuracy of selected forecasting methods based on historical morbidity patterns of JE incidence during the past 22 years (1985-2006). Four selected forecasting methods, viz. seasonal average (SA), seasonal adjustment with last three observations (SAT), modified method adjusting long-term and cyclic trend (MSAT), and autoregressive integrated moving average (ARIMA) have been employed to assess the accuracy of each of the forecasting methods. The forecasting methods were validated for five consecutive years from 2007-2012 and accuracy of each method has been assessed. The forecasting method utilising seasonal adjustment with long-term and cyclic trend emerged as best forecasting method among the four selected forecasting methods and outperformed the even statistically more advanced ARIMA method. Peak of the disease incidence could effectively be predicted with all the methods, but there are significant variations in magnitude of forecast errors among the selected methods. As expected, variation in forecasts at primary health centre (PHC) level is wide as compared to that of district level forecasts. The study showed that adopted forecasting techniques could reasonably forecast the intensity of JE cases at PHC level without considering the external variables. The results indicate that the understanding of long-term and cyclic trend of the disease intensity will improve the accuracy of the forecasts, but there is a need for making the forecast models more robust to explain sudden variation in the disease intensity with detail analysis of parasite and host population dynamics.

Does Specification Matter? Experiments with Simple Multiregional Probabilistic Population Projections

PubMed Central

Raymer, James; Abel, Guy J.; Rogers, Andrei

2012-01-01

Population projection models that introduce uncertainty are a growing subset of projection models in general. In this paper, we focus on the importance of decisions made with regard to the model specifications adopted. We compare the forecasts and prediction intervals associated with four simple regional population projection models: an overall growth rate model, a component model with net migration, a component model with in-migration and out-migration rates, and a multiregional model with destination-specific out-migration rates. Vector autoregressive models are used to forecast future rates of growth, birth, death, net migration, in-migration and out-migration, and destination-specific out-migration for the North, Midlands and South regions in England. They are also used to forecast different international migration measures. The base data represent a time series of annual data provided by the Office for National Statistics from 1976 to 2008. The results illustrate how both the forecasted subpopulation totals and the corresponding prediction intervals differ for the multiregional model in comparison to other simpler models, as well as for different assumptions about international migration. The paper ends end with a discussion of our results and possible directions for future research. PMID:23236221

Forecasting VaR and ES of stock index portfolio: A Vine copula method

NASA Astrophysics Data System (ADS)

Zhang, Bangzheng; Wei, Yu; Yu, Jiang; Lai, Xiaodong; Peng, Zhenfeng

2014-12-01

Risk measurement has both theoretical and practical significance in risk management. Using daily sample of 10 international stock indices, firstly this paper models the internal structures among different stock markets with C-Vine, D-Vine and R-Vine copula models. Secondly, the Value-at-Risk (VaR) and Expected Shortfall (ES) of the international stock markets portfolio are forecasted using Monte Carlo method based on the estimated dependence of different Vine copulas. Finally, the accuracy of VaR and ES measurements obtained from different statistical models are evaluated by UC, IND, CC and Posterior analysis. The empirical results show that the VaR forecasts at the quantile levels of 0.9, 0.95, 0.975 and 0.99 with three kinds of Vine copula models are sufficiently accurate. Several traditional methods, such as historical simulation, mean-variance and DCC-GARCH models, fail to pass the CC backtesting. The Vine copula methods can accurately forecast the ES of the portfolio on the base of VaR measurement, and D-Vine copula model is superior to other Vine copulas.

Simplification of the Kalman filter for meteorological data assimilation

NASA Technical Reports Server (NTRS)

Dee, Dick P.

1991-01-01

The paper proposes a new statistical method of data assimilation that is based on a simplification of the Kalman filter equations. The forecast error covariance evolution is approximated simply by advecting the mass-error covariance field, deriving the remaining covariances geostrophically, and accounting for external model-error forcing only at the end of each forecast cycle. This greatly reduces the cost of computation of the forecast error covariance. In simulations with a linear, one-dimensional shallow-water model and data generated artificially, the performance of the simplified filter is compared with that of the Kalman filter and the optimal interpolation (OI) method. The simplified filter produces analyses that are nearly optimal, and represents a significant improvement over OI.

Cod Collapse and the Climate in the North Atlantic

NASA Astrophysics Data System (ADS)

Meng, K. C.; Oremus, K. L.; Gaines, S.

2014-12-01

Effective fisheries management requires forecasting population changes. We find a negative relationship between the North Atlantic Oscillation (NAO) index and subsequently surveyed biomass and catch of Atlantic cod, Gadus morhua, off the New England coast. A 1-unit NAO increase is associated with a 17% decrease in surveyed biomass of age-1 cod the following year. This relationship persists as the cod mature, such that observed NAO can be used to forecast future adult biomass. We also document that an NAO event lowers catch for up to 15 years afterward. In contrast to forecasts by existing stock assessment models, our NAO-driven statistical model successfully hindcasts the recent collapse of New England cod fisheries following strong NAO events in 2007 and 2008 (see figure). This finding can serve as a template for forecasting other fisheries affected by climatic conditions.

Use of High-Resolution WRF Simulations to Forecast Lightning Threat

NASA Technical Reports Server (NTRS)

McCaul, E. W., Jr.; LaCasse, K.; Goodman, S. J.; Cecil, D. J.

2008-01-01

Recent observational studies have confirmed the existence of a robust statistical relationship between lightning flash rates and the amount of large precipitating ice hydrometeors aloft in storms. This relationship is exploited, in conjunction with the capabilities of cloud-resolving forecast models such as WRF, to forecast explicitly the threat of lightning from convective storms using selected output fields from the model forecasts. The simulated vertical flux of graupel at -15C and the shape of the simulated reflectivity profile are tested in this study as proxies for charge separation processes and their associated lightning risk. Our lightning forecast method differs from others in that it is entirely based on high-resolution simulation output, without reliance on any climatological data. short [6-8 h) simulations are conducted for a number of case studies for which three-dmmensional lightning validation data from the North Alabama Lightning Mapping Array are available. Experiments indicate that initialization of the WRF model on a 2 km grid using Eta boundary conditions, Doppler radar radial velocity fields, and METAR and ACARS data y&eld satisfactory simulations. __nalyses of the lightning threat fields suggests that both the graupel flux and reflectivity profile approaches, when properly calibrated, can yield reasonable lightning threat forecasts, although an ensemble approach is probably desirable in order to reduce the tendency for misplacement of modeled storms to hurt the accuracy of the forecasts. Our lightning threat forecasts are also compared to other more traditional means of forecasting thunderstorms, such as those based on inspection of the convective available potential energy field.

Prospective forecasts of annual dengue hemorrhagic fever incidence in Thailand, 2010–2014

PubMed Central

Sakrejda, Krzysztof; Ray, Evan L.; Keegan, Lindsay T.; Bi, Qifang; Suangtho, Paphanij; Hinjoy, Soawapak; Iamsirithaworn, Sopon; Suthachana, Suthanun; Laosiritaworn, Yongjua; Lessler, Justin; Reich, Nicholas G.

2018-01-01

Dengue hemorrhagic fever (DHF), a severe manifestation of dengue viral infection that can cause severe bleeding, organ impairment, and even death, affects between 15,000 and 105,000 people each year in Thailand. While all Thai provinces experience at least one DHF case most years, the distribution of cases shifts regionally from year to year. Accurately forecasting where DHF outbreaks occur before the dengue season could help public health officials prioritize public health activities. We develop statistical models that use biologically plausible covariates, observed by April each year, to forecast the cumulative DHF incidence for the remainder of the year. We perform cross-validation during the training phase (2000–2009) to select the covariates for these models. A parsimonious model based on preseason incidence outperforms the 10-y median for 65% of province-level annual forecasts, reduces the mean absolute error by 19%, and successfully forecasts outbreaks (area under the receiver operating characteristic curve = 0.84) over the testing period (2010–2014). We find that functions of past incidence contribute most strongly to model performance, whereas the importance of environmental covariates varies regionally. This work illustrates that accurate forecasts of dengue risk are possible in a policy-relevant timeframe. PMID:29463757

Prospective forecasts of annual dengue hemorrhagic fever incidence in Thailand, 2010-2014.

PubMed

Lauer, Stephen A; Sakrejda, Krzysztof; Ray, Evan L; Keegan, Lindsay T; Bi, Qifang; Suangtho, Paphanij; Hinjoy, Soawapak; Iamsirithaworn, Sopon; Suthachana, Suthanun; Laosiritaworn, Yongjua; Cummings, Derek A T; Lessler, Justin; Reich, Nicholas G

2018-03-06

Dengue hemorrhagic fever (DHF), a severe manifestation of dengue viral infection that can cause severe bleeding, organ impairment, and even death, affects between 15,000 and 105,000 people each year in Thailand. While all Thai provinces experience at least one DHF case most years, the distribution of cases shifts regionally from year to year. Accurately forecasting where DHF outbreaks occur before the dengue season could help public health officials prioritize public health activities. We develop statistical models that use biologically plausible covariates, observed by April each year, to forecast the cumulative DHF incidence for the remainder of the year. We perform cross-validation during the training phase (2000-2009) to select the covariates for these models. A parsimonious model based on preseason incidence outperforms the 10-y median for 65% of province-level annual forecasts, reduces the mean absolute error by 19%, and successfully forecasts outbreaks (area under the receiver operating characteristic curve = 0.84) over the testing period (2010-2014). We find that functions of past incidence contribute most strongly to model performance, whereas the importance of environmental covariates varies regionally. This work illustrates that accurate forecasts of dengue risk are possible in a policy-relevant timeframe. Copyright © 2018 the Author(s). Published by PNAS.

Daily water level forecasting using wavelet decomposition and artificial intelligence techniques

NASA Astrophysics Data System (ADS)

Seo, Youngmin; Kim, Sungwon; Kisi, Ozgur; Singh, Vijay P.

2015-01-01

Reliable water level forecasting for reservoir inflow is essential for reservoir operation. The objective of this paper is to develop and apply two hybrid models for daily water level forecasting and investigate their accuracy. These two hybrid models are wavelet-based artificial neural network (WANN) and wavelet-based adaptive neuro-fuzzy inference system (WANFIS). Wavelet decomposition is employed to decompose an input time series into approximation and detail components. The decomposed time series are used as inputs to artificial neural networks (ANN) and adaptive neuro-fuzzy inference system (ANFIS) for WANN and WANFIS models, respectively. Based on statistical performance indexes, the WANN and WANFIS models are found to produce better efficiency than the ANN and ANFIS models. WANFIS7-sym10 yields the best performance among all other models. It is found that wavelet decomposition improves the accuracy of ANN and ANFIS. This study evaluates the accuracy of the WANN and WANFIS models for different mother wavelets, including Daubechies, Symmlet and Coiflet wavelets. It is found that the model performance is dependent on input sets and mother wavelets, and the wavelet decomposition using mother wavelet, db10, can further improve the efficiency of ANN and ANFIS models. Results obtained from this study indicate that the conjunction of wavelet decomposition and artificial intelligence models can be a useful tool for accurate forecasting daily water level and can yield better efficiency than the conventional forecasting models.

Ensemble Streamflow Forecast Improvements in NYC's Operations Support Tool

NASA Astrophysics Data System (ADS)

Wang, L.; Weiss, W. J.; Porter, J.; Schaake, J. C.; Day, G. N.; Sheer, D. P.

2013-12-01

Like most other water supply utilities, New York City's Department of Environmental Protection (DEP) has operational challenges associated with drought and wet weather events. During drought conditions, DEP must maintain water supply reliability to 9 million customers as well as meet environmental release requirements downstream of its reservoirs. During and after wet weather events, DEP must maintain turbidity compliance in its unfiltered Catskill and Delaware reservoir systems and minimize spills to mitigate downstream flooding. Proactive reservoir management - such as release restrictions to prepare for a drought or preventative drawdown in advance of a large storm - can alleviate negative impacts associated with extreme events. It is important for water managers to understand the risks associated with proactive operations so unintended consequences such as endangering water supply reliability with excessive drawdown prior to a storm event are minimized. Probabilistic hydrologic forecasts are a critical tool in quantifying these risks and allow water managers to make more informed operational decisions. DEP has recently completed development of an Operations Support Tool (OST) that integrates ensemble streamflow forecasts, real-time observations, and a reservoir system operations model into a user-friendly graphical interface that allows its water managers to take robust and defensible proactive measures in the face of challenging system conditions. Since initial development of OST was first presented at the 2011 AGU Fall Meeting, significant improvements have been made to the forecast system. First, the monthly AR1 forecasts ('Hirsch method') were upgraded with a generalized linear model (GLM) utilizing historical daily correlations ('Extended Hirsch method' or 'eHirsch'). The development of eHirsch forecasts improved predictive skill over the Hirsch method in the first week to a month from the forecast date and produced more realistic hydrographs on the tail end of high flow periods. These improvements allowed DEP to more effectively manage water quality control and spill mitigation operations immediately after storm events. Later on, post-processed hydrologic forecasts from the National Weather Service (NWS) including the Advanced Hydrologic Prediction Service (AHPS) and the Hydrologic Ensemble Forecast Service (HEFS) were implemented into OST. These forecasts further increased the predictive skill over the initial statistical models as current basin conditions (e.g. soil moisture, snowpack) and meteorological forecasts (with HEFS) are now explicitly represented. With the post-processed HEFS forecasts, DEP may now truly quantify impacts associated with wet weather events on the horizon, rather than relying on statistical representations of current hydrologic trends. This presentation will highlight the benefits of the improved forecasts using examples from actual system operations.

An interdisciplinary approach for earthquake modelling and forecasting

NASA Astrophysics Data System (ADS)

Han, P.; Zhuang, J.; Hattori, K.; Ogata, Y.

2016-12-01

Earthquake is one of the most serious disasters, which may cause heavy casualties and economic losses. Especially in the past two decades, huge/mega earthquakes have hit many countries. Effective earthquake forecasting (including time, location, and magnitude) becomes extremely important and urgent. To date, various heuristically derived algorithms have been developed for forecasting earthquakes. Generally, they can be classified into two types: catalog-based approaches and non-catalog-based approaches. Thanks to the rapid development of statistical seismology in the past 30 years, now we are able to evaluate the performances of these earthquake forecast approaches quantitatively. Although a certain amount of precursory information is available in both earthquake catalogs and non-catalog observations, the earthquake forecast is still far from satisfactory. In most case, the precursory phenomena were studied individually. An earthquake model that combines self-exciting and mutually exciting elements was developed by Ogata and Utsu from the Hawkes process. The core idea of this combined model is that the status of the event at present is controlled by the event itself (self-exciting) and all the external factors (mutually exciting) in the past. In essence, the conditional intensity function is a time-varying Poisson process with rate λ(t), which is composed of the background rate, the self-exciting term (the information from past seismic events), and the external excitation term (the information from past non-seismic observations). This model shows us a way to integrate the catalog-based forecast and non-catalog-based forecast. Against this background, we are trying to develop a new earthquake forecast model which combines catalog-based and non-catalog-based approaches.

Client-Friendly Forecasting: Seasonal Runoff Predictions Using Out-of-the-Box Indices

NASA Astrophysics Data System (ADS)

Weil, P.

2013-12-01

For more than a century, statistical relationships have been recognized between atmospheric conditions at locations separated by thousands of miles, referred to as teleconnections. Some of the recognized teleconnections provide useful information about expected hydrologic conditions, so certain records of atmospheric conditions are quantified and published as hydroclimate indices. Certain hydroclimate indices can serve as strong leading indicators of climate patterns over North America and can be used to make skillful forecasts of seasonal runoff. The methodology described here creates a simple-to-use model that utilizes easily accessed data to make forecasts of April through September runoff months before the runoff season begins. For this project, forecasting models were developed for two snowmelt-driven river systems in Colorado and Wyoming. In addition to the global hydroclimate indices, the methodology uses several local hydrologic variables including the previous year's drought severity, headwater snow water equivalent and the reservoir contents for the major reservoirs in each basin. To improve the skill of the forecasts, logistic regression is used to develop a model that provides the likelihood that a year will fall into the upper, middle or lower tercile of historical flows. Categorical forecasting has two major advantages over modeling of specific flow amounts: (1) with less prediction outcomes models tend to have better predictive skill and (2) categorical models are very useful to clients and agencies with specific flow thresholds that dictate major changes in water resources management. The resulting methodology and functional forecasting model product is highly portable, applicable to many major river systems and easily explained to a non-technical audience.

Similarity-based multi-model ensemble approach for 1-15-day advance prediction of monsoon rainfall over India

NASA Astrophysics Data System (ADS)

Jaiswal, Neeru; Kishtawal, C. M.; Bhomia, Swati

2018-04-01

The southwest (SW) monsoon season (June, July, August and September) is the major period of rainfall over the Indian region. The present study focuses on the development of a new multi-model ensemble approach based on the similarity criterion (SMME) for the prediction of SW monsoon rainfall in the extended range. This approach is based on the assumption that training with the similar type of conditions may provide the better forecasts in spite of the sequential training which is being used in the conventional MME approaches. In this approach, the training dataset has been selected by matching the present day condition to the archived dataset and days with the most similar conditions were identified and used for training the model. The coefficients thus generated were used for the rainfall prediction. The precipitation forecasts from four general circulation models (GCMs), viz. European Centre for Medium-Range Weather Forecasts (ECMWF), United Kingdom Meteorological Office (UKMO), National Centre for Environment Prediction (NCEP) and China Meteorological Administration (CMA) have been used for developing the SMME forecasts. The forecasts of 1-5, 6-10 and 11-15 days were generated using the newly developed approach for each pentad of June-September during the years 2008-2013 and the skill of the model was analysed using verification scores, viz. equitable skill score (ETS), mean absolute error (MAE), Pearson's correlation coefficient and Nash-Sutcliffe model efficiency index. Statistical analysis of SMME forecasts shows superior forecast skill compared to the conventional MME and the individual models for all the pentads, viz. 1-5, 6-10 and 11-15 days.

Towards seasonal forecasting of malaria in India.

PubMed

Lauderdale, Jonathan M; Caminade, Cyril; Heath, Andrew E; Jones, Anne E; MacLeod, David A; Gouda, Krushna C; Murty, Upadhyayula Suryanarayana; Goswami, Prashant; Mutheneni, Srinivasa R; Morse, Andrew P

2014-08-10

Malaria presents public health challenge despite extensive intervention campaigns. A 30-year hindcast of the climatic suitability for malaria transmission in India is presented, using meteorological variables from a state of the art seasonal forecast model to drive a process-based, dynamic disease model. The spatial distribution and seasonal cycles of temperature and precipitation from the forecast model are compared to three observationally-based meteorological datasets. These time series are then used to drive the disease model, producing a simulated forecast of malaria and three synthetic malaria time series that are qualitatively compared to contemporary and pre-intervention malaria estimates. The area under the Relative Operator Characteristic (ROC) curve is calculated as a quantitative metric of forecast skill, comparing the forecast to the meteorologically-driven synthetic malaria time series. The forecast shows probabilistic skill in predicting the spatial distribution of Plasmodium falciparum incidence when compared to the simulated meteorologically-driven malaria time series, particularly where modelled incidence shows high seasonal and interannual variability such as in Orissa, West Bengal, and Jharkhand (North-east India), and Gujarat, Rajastan, Madhya Pradesh and Maharashtra (North-west India). Focusing on these two regions, the malaria forecast is able to distinguish between years of "high", "above average" and "low" malaria incidence in the peak malaria transmission seasons, with more than 70% sensitivity and a statistically significant area under the ROC curve. These results are encouraging given that the three month forecast lead time used is well in excess of the target for early warning systems adopted by the World Health Organization. This approach could form the basis of an operational system to identify the probability of regional malaria epidemics, allowing advanced and targeted allocation of resources for combatting malaria in India.

A New Statistical Model for Eruption Forecasting at Open Conduit Volcanoes: an Application to Mt Etna and Kilauea Volcanoes

NASA Astrophysics Data System (ADS)

Passarelli, Luigi; Sanso, Bruno; Laura, Sandri; Marzocchi, Warner

2010-05-01

One of the main goals in volcanology is to forecast volcanic eruptions. A trenchant forecast should be made before the onset of a volcanic eruption, using the data available at that time, with the aim of mitigating the volcanic risk associated to the volcanic event. In other words, models implemented with forecast purposes have to take into account the possibility to provide "forward" forecasts and should avoid the idea of a merely "retrospective" fitting of the data available. In this perspective, the main idea of the present model is to forecast the next volcanic eruption after the end of the last one, using only the data available at that time. We focus our attention on volcanoes with open conduit regime and high eruption frequency. We assume a generalization of the classical time predictable model to describe the eruptive behavior of open conduit volcanoes and we use a Bayesian hierarchical model to make probabilistic forecast. We apply the model to Kilauea volcano eruptive data and Mt. Etna volcano flank eruption data. The aims of this model are: 1) to test whether or not the Kilauea and Mt Etna volcanoes follow a time predictable behavior; 2) to discuss the volcanological implications of the time predictable model parameters inferred; 3) to compare the forecast capabilities of this model with other models present in literature. The results obtained using the MCMC sampling algorithm show that both volcanoes follow a time predictable behavior. The numerical values of the time predictable model parameters inferred suggest that the amount of the erupted volume could change the dynamics of the magma chamber refilling process during the repose period. The probability gain of this model compared with other models already present in literature is appreciably greater than zero. This means that our model performs better forecast than previous models and it could be used in a probabilistic volcanic hazard assessment scheme. In this perspective, the probability of eruptions given by our model for Mt Etna volcano flank eruption are published on a internet website and are updated after any change in the eruptive activity.

Prediction of Chemical Function: Model Development and Application

EPA Science Inventory

The United States Environmental Protection Agency’s Exposure Forecaster (ExpoCast) project is developing both statistical and mechanism-based computational models for predicting exposures to thousands of chemicals, including those in consumer products. The high-throughput (...

The Schaake shuffle: A method for reconstructing space-time variability in forecasted precipitation and temperature fields

USGS Publications Warehouse

Clark, M.R.; Gangopadhyay, S.; Hay, L.; Rajagopalan, B.; Wilby, R.

2004-01-01

A number of statistical methods that are used to provide local-scale ensemble forecasts of precipitation and temperature do not contain realistic spatial covariability between neighboring stations or realistic temporal persistence for subsequent forecast lead times. To demonstrate this point, output from a global-scale numerical weather prediction model is used in a stepwise multiple linear regression approach to downscale precipitation and temperature to individual stations located in and around four study basins in the United States. Output from the forecast model is downscaled for lead times up to 14 days. Residuals in the regression equation are modeled stochastically to provide 100 ensemble forecasts. The precipitation and temperature ensembles from this approach have a poor representation of the spatial variability and temporal persistence. The spatial correlations for downscaled output are considerably lower than observed spatial correlations at short forecast lead times (e.g., less than 5 days) when there is high accuracy in the forecasts. At longer forecast lead times, the downscaled spatial correlations are close to zero. Similarly, the observed temporal persistence is only partly present at short forecast lead times. A method is presented for reordering the ensemble output in order to recover the space-time variability in precipitation and temperature fields. In this approach, the ensemble members for a given forecast day are ranked and matched with the rank of precipitation and temperature data from days randomly selected from similar dates in the historical record. The ensembles are then reordered to correspond to the original order of the selection of historical data. Using this approach, the observed intersite correlations, intervariable correlations, and the observed temporal persistence are almost entirely recovered. This reordering methodology also has applications for recovering the space-time variability in modeled streamflow. ?? 2004 American Meteorological Society.

Statistical prediction of seasonal discharge in Central Asia for water resources management: development of a generic (pre-)operational modeling tool

NASA Astrophysics Data System (ADS)

Apel, Heiko; Baimaganbetov, Azamat; Kalashnikova, Olga; Gavrilenko, Nadejda; Abdykerimova, Zharkinay; Agalhanova, Marina; Gerlitz, Lars; Unger-Shayesteh, Katy; Vorogushyn, Sergiy; Gafurov, Abror

2017-04-01

The semi-arid regions of Central Asia crucially depend on the water resources supplied by the mountainous areas of the Tien-Shan and Pamirs. During the summer months the snow and glacier melt dominated river discharge originating in the mountains provides the main water resource available for agricultural production, but also for storage in reservoirs for energy generation during the winter months. Thus a reliable seasonal forecast of the water resources is crucial for a sustainable management and planning of water resources. In fact, seasonal forecasts are mandatory tasks of all national hydro-meteorological services in the region. In order to support the operational seasonal forecast procedures of hydromet services, this study aims at the development of a generic tool for deriving statistical forecast models of seasonal river discharge. The generic model is kept as simple as possible in order to be driven by available hydrological and meteorological data, and be applicable for all catchments with their often limited data availability in the region. As snowmelt dominates summer runoff, the main meteorological predictors for the forecast models are monthly values of winter precipitation and temperature as recorded by climatological stations in the catchments. These data sets are accompanied by snow cover predictors derived from the operational ModSnow tool, which provides cloud free snow cover data for the selected catchments based on MODIS satellite images. In addition to the meteorological data antecedent streamflow is used as a predictor variable. This basic predictor set was further extended by multi-monthly means of the individual predictors, as well as composites of the predictors. Forecast models are derived based on these predictors as linear combinations of up to 3 or 4 predictors. A user selectable number of best models according to pre-defined performance criteria is extracted automatically by the developed model fitting algorithm, which includes a test for robustness by a leave-one-out cross validation. Based on the cross validation the predictive uncertainty was quantified for every prediction model. According to the official procedures of the hydromet services forecasts of the mean seasonal discharge of the period April to September are derived every month starting from January until June. The application of the model for several catchments in Central Asia - ranging from small to the largest rivers - for the period 2000-2015 provided skillful forecasts for most catchments already in January. The skill of the prediction increased every month, with R2 values often in the range 0.8 - 0.9 in April just before the prediction period. The forecasts further improve in the following months, most likely due to the integration of spring precipitation, which is not included in the predictors before May, or spring discharge, which contains indicative information for the overall seasonal discharge. In summary, the proposed generic automatic forecast model development tool provides robust predictions for seasonal water availability in Central Asia, which will be tested against the official forecasts in the upcoming years, with the vision of eventual operational implementation.

Stochastic simulation of predictive space–time scenarios of wind speed using observations and physical model outputs

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

Bessac, Julie; Constantinescu, Emil; Anitescu, Mihai

We propose a statistical space-time model for predicting atmospheric wind speed based on deterministic numerical weather predictions and historical measurements. We consider a Gaussian multivariate space-time framework that combines multiple sources of past physical model outputs and measurements in order to produce a probabilistic wind speed forecast within the prediction window. We illustrate this strategy on wind speed forecasts during several months in 2012 for a region near the Great Lakes in the United States. The results show that the prediction is improved in the mean-squared sense relative to the numerical forecasts as well as in probabilistic scores. Moreover, themore » samples are shown to produce realistic wind scenarios based on sample spectra and space-time correlation structure.« less

Landslide early warning based on failure forecast models: the example of the Mt. de La Saxe rockslide, northern Italy

NASA Astrophysics Data System (ADS)

Manconi, A.; Giordan, D.

2015-07-01

We apply failure forecast models by exploiting near-real-time monitoring data for the La Saxe rockslide, a large unstable slope threatening Aosta Valley in northern Italy. Starting from the inverse velocity theory, we analyze landslide surface displacements automatically and in near real time on different temporal windows and apply straightforward statistical methods to obtain confidence intervals on the estimated time of failure. Here, we present the result obtained for the La Saxe rockslide, a large unstable slope located in Aosta Valley, northern Italy. Based on this case study, we identify operational thresholds that are established on the reliability of the forecast models. Our approach is aimed at supporting the management of early warning systems in the most critical phases of the landslide emergency.

Stochastic simulation of predictive space–time scenarios of wind speed using observations and physical model outputs

DOE PAGES

Bessac, Julie; Constantinescu, Emil; Anitescu, Mihai

2018-03-01

We propose a statistical space-time model for predicting atmospheric wind speed based on deterministic numerical weather predictions and historical measurements. We consider a Gaussian multivariate space-time framework that combines multiple sources of past physical model outputs and measurements in order to produce a probabilistic wind speed forecast within the prediction window. We illustrate this strategy on wind speed forecasts during several months in 2012 for a region near the Great Lakes in the United States. The results show that the prediction is improved in the mean-squared sense relative to the numerical forecasts as well as in probabilistic scores. Moreover, themore » samples are shown to produce realistic wind scenarios based on sample spectra and space-time correlation structure.« less

Statistical prediction of seasonal discharge in the Naryn basin for water resources planning in Central Asia

NASA Astrophysics Data System (ADS)

Apel, Heiko; Gafurov, Abror; Gerlitz, Lars; Unger-Shayesteh, Katy; Vorogushyn, Sergiy; Merkushkin, Aleksandr; Merz, Bruno

2016-04-01

The semi-arid regions of Central Asia crucially depend on the water resources supplied by the mountainous areas of the Tien-Shan and Pamirs. During the summer months the snow and glacier melt water of the rivers originating in the mountains provides the only water resource available for agricultural production but also for water collection in reservoirs for energy production in winter months. Thus a reliable seasonal forecast of the water resources is crucial for a sustainable management and planning of water resources.. In fact, seasonal forecasts are mandatory tasks of national hydro-meteorological services in the region. Thus this study aims at a statistical forecast of the seasonal water availability, whereas the focus is put on the usage of freely available data in order to facilitate an operational use without data access limitations. The study takes the Naryn basin as a test case, at which outlet the Toktogul reservoir stores the discharge of the Naryn River. As most of the water originates form snow and glacier melt, a statistical forecast model should use data sets that can serve as proxy data for the snow masses and snow water equivalent in late spring, which essentially determines the bulk of the seasonal discharge. CRU climate data describing the precipitation and temperature in the basin during winter and spring was used as base information, which was complemented by MODIS snow cover data processed through ModSnow tool, discharge during the spring and also GRACE gravimetry anomalies. For the construction of linear forecast models monthly as well as multi-monthly means over the period January to April were used to predict the seasonal mean discharge of May-September at the station Uchterek. An automatic model selection was performed in multiple steps, whereas the best models were selected according to several performance measures and their robustness in a leave-one-out cross validation. It could be shown that the seasonal discharge can be predicted with exceptionally high skill reaching explained variances of 86% in the cross validation using ModSnow processed snow cover data and CRU temperature and precipitation data, i.e. freely available data only. Using antecedent discharge information from the Uchterek station over the period January to April the skill can be improved even further. Also the addition of latest EGSIEM GRACE products can improve this skill to > 90% explained variance by replacing the CRU temperature data in the forecast model. From all variables the ModSnow processed MODIS snow cover data proved to be the most important predictor. However, although the prediction models proved to be robust in the cross validation, it has to be mentioned that the models are based on a limited time spanning the period 2000-2012 only. Nevertheless it is believed that the models are reliable, as this time period shows a high variability in seasonal water availability spanning from exceptionally dry to wet years. In summary, the developed forecast model may be a valuable complementary tool for the seasonal discharge prediction in Central Asia for water resources planning, that does not suffer from limited data access required for other forecast methods.

Towards uncertainty estimation for operational forecast products - a multi-model-ensemble approach for the North Sea and the Baltic Sea

NASA Astrophysics Data System (ADS)

Golbeck, Inga; Li, Xin; Janssen, Frank

2014-05-01

Several independent operational ocean models provide forecasts of the ocean state (e.g. sea level, temperature, salinity and ice cover) in the North Sea and the Baltic Sea on a daily basis. These forecasts are the primary source of information for a variety of information and emergency response systems used e.g. to issue sea level warnings or carry out oil drift forecast. The forecasts are of course highly valuable as such, but often suffer from a lack of information on their uncertainty. With the aim of augmenting the existing operational ocean forecasts in the North Sea and the Baltic Sea by a measure of uncertainty a multi-model-ensemble (MME) system for sea surface temperature (SST), sea surface salinity (SSS) and water transports has been set up in the framework of the MyOcean-2 project. Members of MyOcean-2, the NOOS² and HIROMB/BOOS³ communities provide 48h-forecasts serving as inputs. Different variables are processed separately due to their different physical characteristics. Based on the so far collected daily MME products of SST and SSS, a statistical method, Empirical Orthogonal Function (EOF) analysis is applied to assess their spatial and temporal variability. For sea surface currents, progressive vector diagrams at specific points are consulted to estimate the performance of the circulation models especially in hydrodynamic important areas, e.g. inflow/outflow of the Baltic Sea, Norwegian trench and English Channel. For further versions of the MME system, it is planned to extend the MME to other variables like e.g. sea level, ocean currents or ice cover based on the needs of the model providers and their customers. It is also planned to include in-situ data to augment the uncertainty information and for validation purposes. Additionally, weighting methods will be implemented into the MME system to develop more complex uncertainty measures. The methodology used to create the MME will be outlined and different ensemble products will be presented. In addition, some preliminary results based on the statistical analysis of the uncertainty measures provide first estimates of the regional and temporal performance of the ocean models for each parameter. ²Northwest European Shelf Operational Oceanography System ³High-resolution Operational Model of the Baltic / Baltic Operational Oceanographic System

Assessing probabilistic predictions of ENSO phase and intensity from the North American Multimodel Ensemble

NASA Astrophysics Data System (ADS)

Tippett, Michael K.; Ranganathan, Meghana; L'Heureux, Michelle; Barnston, Anthony G.; DelSole, Timothy

2017-05-01

Here we examine the skill of three, five, and seven-category monthly ENSO probability forecasts (1982-2015) from single and multi-model ensemble integrations of the North American Multimodel Ensemble (NMME) project. Three-category forecasts are typical and provide probabilities for the ENSO phase (El Niño, La Niña or neutral). Additional forecast categories indicate the likelihood of ENSO conditions being weak, moderate or strong. The level of skill observed for differing numbers of forecast categories can help to determine the appropriate degree of forecast precision. However, the dependence of the skill score itself on the number of forecast categories must be taken into account. For reliable forecasts with same quality, the ranked probability skill score (RPSS) is fairly insensitive to the number of categories, while the logarithmic skill score (LSS) is an information measure and increases as categories are added. The ignorance skill score decreases to zero as forecast categories are added, regardless of skill level. For all models, forecast formats and skill scores, the northern spring predictability barrier explains much of the dependence of skill on target month and forecast lead. RPSS values for monthly ENSO forecasts show little dependence on the number of categories. However, the LSS of multimodel ensemble forecasts with five and seven categories show statistically significant advantages over the three-category forecasts for the targets and leads that are least affected by the spring predictability barrier. These findings indicate that current prediction systems are capable of providing more detailed probabilistic forecasts of ENSO phase and amplitude than are typically provided.

Performance of an Advanced MOS System in the 1996-97 National Collegiate Weather Forecasting Contest.

NASA Astrophysics Data System (ADS)

Vislocky, Robert L.; Fritsch, J. Michael

1997-12-01

A prototype advanced model output statistics (MOS) forecast system that was entered in the 1996-97 National Collegiate Weather Forecast Contest is described and its performance compared to that of widely available objective guidance and to contest participants. The prototype system uses an optimal blend of aviation (AVN) and nested grid model (NGM) MOS forecasts, explicit output from the NGM and Eta guidance, and the latest surface weather observations from the forecast site. The forecasts are totally objective and can be generated quickly on a personal computer. Other "objective" forms of guidance tracked in the contest are 1) the consensus forecast (i.e., the average of the forecasts from all of the human participants), 2) the combination of NGM raw output (for precipitation forecasts) and NGM MOS guidance (for temperature forecasts), and 3) the combination of Eta Model raw output (for precipitation forecasts) and AVN MOS guidance (for temperature forecasts).Results show that the advanced MOS system finished in 20th place out of 737 original entrants, or better than approximately 97% of the human forecasters who entered the contest. Moreover, the advanced MOS system was slightly better than consensus (23d place). The fact that an objective forecast system finished ahead of consensus is a significant accomplishment since consensus is traditionally a very formidable "opponent" in forecast competitions. Equally significant is that the advanced MOS system was superior to the traditional guidance products available from the National Centers for Environmental Prediction (NCEP). Specifically, the combination of NGM raw output and NGM MOS guidance finished in 175th place, and the combination of Eta Model raw output and AVN MOS guidance finished in 266th place. The latter result is most intriguing since the proposed elimination of all NGM products would likely result in a serious degradation of objective products disseminated by NCEP, unless they are replaced with equal or better substitutes. On the other hand, the positive performance of the prototype advanced MOS system shows that it is possible to create a single objective product that is not only superior to currently available objective guidance products, but is also on par with some of the better human forecasters.

Operational 0-3 h probabilistic quantitative precipitation forecasts: Recent performance and potential enhancements

NASA Astrophysics Data System (ADS)

Sokol, Z.; Kitzmiller, D.; Pešice, P.; Guan, S.

2009-05-01

The NOAA National Weather Service has maintained an automated, centralized 0-3 h prediction system for probabilistic quantitative precipitation forecasts since 2001. This advective-statistical system (ADSTAT) produces probabilities that rainfall will exceed multiple threshold values up to 50 mm at some location within a 40-km grid box. Operational characteristics and development methods for the system are described. Although development data were stratified by season and time of day, ADSTAT utilizes only a single set of nation-wide equations that relate predictor variables derived from radar reflectivity, lightning, satellite infrared temperatures, and numerical prediction model output to rainfall occurrence. A verification study documented herein showed that the operational ADSTAT reliably models regional variations in the relative frequency of heavy rain events. This was true even in the western United States, where no regional-scale, gridded hourly precipitation data were available during the development period in the 1990s. An effort was recently launched to improve the quality of ADSTAT forecasts by regionalizing the prediction equations and to adapt the model for application in the Czech Republic. We have experimented with incorporating various levels of regional specificity in the probability equations. The geographic localization study showed that in the warm season, regional climate differences and variations in the diurnal temperature cycle have a marked effect on the predictor-predictand relationships, and thus regionalization would lead to better statistical reliability in the forecasts.

Evaluation of the synoptic and mesoscale predictive capabilities of a mesoscale atmospheric simulation system

NASA Technical Reports Server (NTRS)

Koch, S. E.; Skillman, W. C.; Kocin, P. J.; Wetzel, P. J.; Brill, K.; Keyser, D. A.; Mccumber, M. C.

1983-01-01

The overall performance characteristics of a limited area, hydrostatic, fine (52 km) mesh, primitive equation, numerical weather prediction model are determined in anticipation of satellite data assimilations with the model. The synoptic and mesoscale predictive capabilities of version 2.0 of this model, the Mesoscale Atmospheric Simulation System (MASS 2.0), were evaluated. The two part study is based on a sample of approximately thirty 12h and 24h forecasts of atmospheric flow patterns during spring and early summer. The synoptic scale evaluation results benchmark the performance of MASS 2.0 against that of an operational, synoptic scale weather prediction model, the Limited area Fine Mesh (LFM). The large sample allows for the calculation of statistically significant measures of forecast accuracy and the determination of systematic model errors. The synoptic scale benchmark is required before unsmoothed mesoscale forecast fields can be seriously considered.

Integrated Wind Power Planning Tool

NASA Astrophysics Data System (ADS)

Rosgaard, M. H.; Giebel, G.; Nielsen, T. S.; Hahmann, A.; Sørensen, P.; Madsen, H.

2012-04-01

This poster presents the current state of the public service obligation (PSO) funded project PSO 10464, with the working title "Integrated Wind Power Planning Tool". The project commenced October 1, 2011, and the goal is to integrate a numerical weather prediction (NWP) model with purely statistical tools in order to assess wind power fluctuations, with focus on long term power system planning for future wind farms as well as short term forecasting for existing wind farms. Currently, wind power fluctuation models are either purely statistical or integrated with NWP models of limited resolution. With regard to the latter, one such simulation tool has been developed at the Wind Energy Division, Risø DTU, intended for long term power system planning. As part of the PSO project the inferior NWP model used at present will be replaced by the state-of-the-art Weather Research & Forecasting (WRF) model. Furthermore, the integrated simulation tool will be improved so it can handle simultaneously 10-50 times more turbines than the present ~ 300, as well as additional atmospheric parameters will be included in the model. The WRF data will also be input for a statistical short term prediction model to be developed in collaboration with ENFOR A/S; a danish company that specialises in forecasting and optimisation for the energy sector. This integrated prediction model will allow for the description of the expected variability in wind power production in the coming hours to days, accounting for its spatio-temporal dependencies, and depending on the prevailing weather conditions defined by the WRF output. The output from the integrated prediction tool constitute scenario forecasts for the coming period, which can then be fed into any type of system model or decision making problem to be solved. The high resolution of the WRF results loaded into the integrated prediction model will ensure a high accuracy data basis is available for use in the decision making process of the Danish transmission system operator, and the need for high accuracy predictions will only increase over the next decade as Denmark approaches the goal of 50% wind power based electricity in 2020, from the current 20%.

Distortion Representation of Forecast Errors for Model Skill Assessment and Objective Analysis

NASA Technical Reports Server (NTRS)

Hoffman, Ross N.; Nehrkorn, Thomas; Grassotti, Christopher

1998-01-01

We proposed a novel characterization of errors for numerical weather predictions. A general distortion representation allows for the displacement and amplification or bias correction of forecast anomalies. Characterizing and decomposing forecast error in this way has several important applications, including the model assessment application and the objective analysis application. In this project, we have focused on the assessment application, restricted to a realistic but univariate 2-dimensional situation. Specifically, we study the forecast errors of the sea level pressure (SLP), the 500 hPa geopotential height, and the 315 K potential vorticity fields for forecasts of the short and medium range. The forecasts are generated by the Goddard Earth Observing System (GEOS) data assimilation system with and without ERS-1 scatterometer data. A great deal of novel work has been accomplished under the current contract. In broad terms, we have developed and tested an efficient algorithm for determining distortions. The algorithm and constraints are now ready for application to larger data sets to be used to determine the statistics of the distortion as outlined above, and to be applied in data analysis by using GEOS water vapor imagery to correct short-term forecast errors.

Improving Global Forecast System of extreme precipitation events with regional statistical model: Application of quantile-based probabilistic forecasts

NASA Astrophysics Data System (ADS)

Shastri, Hiteshri; Ghosh, Subimal; Karmakar, Subhankar

2017-02-01

Forecasting of extreme precipitation events at a regional scale is of high importance due to their severe impacts on society. The impacts are stronger in urban regions due to high flood potential as well high population density leading to high vulnerability. Although significant scientific improvements took place in the global models for weather forecasting, they are still not adequate at a regional scale (e.g., for an urban region) with high false alarms and low detection. There has been a need to improve the weather forecast skill at a local scale with probabilistic outcome. Here we develop a methodology with quantile regression, where the reliably simulated variables from Global Forecast System are used as predictors and different quantiles of rainfall are generated corresponding to that set of predictors. We apply this method to a flood-prone coastal city of India, Mumbai, which has experienced severe floods in recent years. We find significant improvements in the forecast with high detection and skill scores. We apply the methodology to 10 ensemble members of Global Ensemble Forecast System and find a reduction in ensemble uncertainty of precipitation across realizations with respect to that of original precipitation forecasts. We validate our model for the monsoon season of 2006 and 2007, which are independent of the training/calibration data set used in the study. We find promising results and emphasize to implement such data-driven methods for a better probabilistic forecast at an urban scale primarily for an early flood warning.

Statistical Modeling and Prediction for Tourism Economy Using Dendritic Neural Network

PubMed Central

Yu, Ying; Wang, Yirui; Tang, Zheng

2017-01-01

With the impact of global internationalization, tourism economy has also been a rapid development. The increasing interest aroused by more advanced forecasting methods leads us to innovate forecasting methods. In this paper, the seasonal trend autoregressive integrated moving averages with dendritic neural network model (SA-D model) is proposed to perform the tourism demand forecasting. First, we use the seasonal trend autoregressive integrated moving averages model (SARIMA model) to exclude the long-term linear trend and then train the residual data by the dendritic neural network model and make a short-term prediction. As the result showed in this paper, the SA-D model can achieve considerably better predictive performances. In order to demonstrate the effectiveness of the SA-D model, we also use the data that other authors used in the other models and compare the results. It also proved that the SA-D model achieved good predictive performances in terms of the normalized mean square error, absolute percentage of error, and correlation coefficient. PMID:28246527

Statistical Modeling and Prediction for Tourism Economy Using Dendritic Neural Network.

PubMed

Yu, Ying; Wang, Yirui; Gao, Shangce; Tang, Zheng

2017-01-01

With the impact of global internationalization, tourism economy has also been a rapid development. The increasing interest aroused by more advanced forecasting methods leads us to innovate forecasting methods. In this paper, the seasonal trend autoregressive integrated moving averages with dendritic neural network model (SA-D model) is proposed to perform the tourism demand forecasting. First, we use the seasonal trend autoregressive integrated moving averages model (SARIMA model) to exclude the long-term linear trend and then train the residual data by the dendritic neural network model and make a short-term prediction. As the result showed in this paper, the SA-D model can achieve considerably better predictive performances. In order to demonstrate the effectiveness of the SA-D model, we also use the data that other authors used in the other models and compare the results. It also proved that the SA-D model achieved good predictive performances in terms of the normalized mean square error, absolute percentage of error, and correlation coefficient.

Moisture Forecast Bias Correction in GEOS DAS

NASA Technical Reports Server (NTRS)

Dee, D.

1999-01-01

Data assimilation methods rely on numerous assumptions about the errors involved in measuring and forecasting atmospheric fields. One of the more disturbing of these is that short-term model forecasts are assumed to be unbiased. In case of atmospheric moisture, for example, observational evidence shows that the systematic component of errors in forecasts and analyses is often of the same order of magnitude as the random component. we have implemented a sequential algorithm for estimating forecast moisture bias from rawinsonde data in the Goddard Earth Observing System Data Assimilation System (GEOS DAS). The algorithm is designed to remove the systematic component of analysis errors and can be easily incorporated in an existing statistical data assimilation system. We will present results of initial experiments that show a significant reduction of bias in the GEOS DAS moisture analyses.

High resolution statistical downscaling of the EUROSIP seasonal prediction. Application for southeastern Romania

NASA Astrophysics Data System (ADS)

Busuioc, Aristita; Dumitrescu, Alexandru; Dumitrache, Rodica; Iriza, Amalia

2017-04-01

Seasonal climate forecasts in Europe are currently issued at the European Centre for Medium-Range Weather Forecasts (ECMWF) in the form of multi-model ensemble predictions available within the "EUROSIP" system. Different statistical techniques to calibrate, downscale and combine the EUROSIP direct model output are used to optimize the quality of the final probabilistic forecasts. In this study, a statistical downscaling model (SDM) based on canonical correlation analysis (CCA) is used to downscale the EUROSIP seasonal forecast at a spatial resolution of 1km x 1km over the Movila farm placed in southeastern Romania. This application is achieved in the framework of the H2020 MOSES project (http://www.moses-project.eu). The combination between monthly standardized values of three climate variables (maximum/minimum temperatures-Tmax/Tmin, total precipitation-Prec) is used as predictand while combinations of various large-scale predictors are tested in terms of their availability as outputs in the seasonal EUROSIP probabilistic forecasting (sea level pressure, temperature at 850 hPa and geopotential height at 500 hPa). The predictors are taken from the ECMWF system considering 15 members of the ensemble, for which the hindcasts since 1991 until present are available. The model was calibrated over the period 1991-2014 and predictions for summers 2015 and 2016 were achieved. The calibration was made for the ensemble average as well as for each ensemble member. The model was developed for each lead time: one month anticipation for June, two months anticipation for July and three months anticipation for August. The main conclusions from these preliminary results are: best predictions (in terms of the anomaly sign) for Tmax (July-2 months anticipation, August-3 months anticipation) for both years (2015, 2016); for Tmin - good predictions only for August (3 months anticipation ) for both years; for precipitation, good predictions for July (2 months anticipation) in 2015 and August (3 months anticipation) in 2016; failed prediction for June (1-month anticipation) for all parameters. To see if the results obtained for 2015 and 2016 summers are in agreement with the general ECMWF model performance in forecast of the three predictors used in the CCA SDM calibration, the mean bias and root mean square errors (RMSE) calculated over the entire period in each grid point, for each ensemble member and ensemble average were computed. The obtained results are confirmed, showing highest ECMWF performance in forecasting of the three predictors for 3 months anticipation (August) and lowest performance for one month anticipation (June). The added value of the CCA SDM in forecasting local Tmax/Tmin and total precipitation was compared to the ECMWF performance using nearest grid point method. Comparisons were performed for the 1991-2014 period, taking into account the forecast made in May for July. An important improvement was found for the CCA SDM predictions in terms of the RMSE value (computed against observations) for Tmax/Tmin and less for precipitation. The tests are in progress for the other summer months (June, July).

Statistical physics approach to earthquake occurrence and forecasting

NASA Astrophysics Data System (ADS)

de Arcangelis, Lucilla; Godano, Cataldo; Grasso, Jean Robert; Lippiello, Eugenio

2016-04-01

There is striking evidence that the dynamics of the Earth crust is controlled by a wide variety of mutually dependent mechanisms acting at different spatial and temporal scales. The interplay of these mechanisms produces instabilities in the stress field, leading to abrupt energy releases, i.e., earthquakes. As a consequence, the evolution towards instability before a single event is very difficult to monitor. On the other hand, collective behavior in stress transfer and relaxation within the Earth crust leads to emergent properties described by stable phenomenological laws for a population of many earthquakes in size, time and space domains. This observation has stimulated a statistical mechanics approach to earthquake occurrence, applying ideas and methods as scaling laws, universality, fractal dimension, renormalization group, to characterize the physics of earthquakes. In this review we first present a description of the phenomenological laws of earthquake occurrence which represent the frame of reference for a variety of statistical mechanical models, ranging from the spring-block to more complex fault models. Next, we discuss the problem of seismic forecasting in the general framework of stochastic processes, where seismic occurrence can be described as a branching process implementing space-time-energy correlations between earthquakes. In this context we show how correlations originate from dynamical scaling relations between time and energy, able to account for universality and provide a unifying description for the phenomenological power laws. Then we discuss how branching models can be implemented to forecast the temporal evolution of the earthquake occurrence probability and allow to discriminate among different physical mechanisms responsible for earthquake triggering. In particular, the forecasting problem will be presented in a rigorous mathematical framework, discussing the relevance of the processes acting at different temporal scales for different levels of prediction. In this review we also briefly discuss how the statistical mechanics approach can be applied to non-tectonic earthquakes and to other natural stochastic processes, such as volcanic eruptions and solar flares.

Impact of Assimilation on Heavy Rainfall Simulations Using WRF Model: Sensitivity of Assimilation Results to Background Error Statistics

NASA Astrophysics Data System (ADS)

Rakesh, V.; Kantharao, B.

2017-03-01

Data assimilation is considered as one of the effective tools for improving forecast skill of mesoscale models. However, for optimum utilization and effective assimilation of observations, many factors need to be taken into account while designing data assimilation methodology. One of the critical components that determines the amount and propagation observation information into the analysis, is model background error statistics (BES). The objective of this study is to quantify how BES in data assimilation impacts on simulation of heavy rainfall events over a southern state in India, Karnataka. Simulations of 40 heavy rainfall events were carried out using Weather Research and Forecasting Model with and without data assimilation. The assimilation experiments were conducted using global and regional BES while the experiment with no assimilation was used as the baseline for assessing the impact of data assimilation. The simulated rainfall is verified against high-resolution rain-gage observations over Karnataka. Statistical evaluation using several accuracy and skill measures shows that data assimilation has improved the heavy rainfall simulation. Our results showed that the experiment using regional BES outperformed the one which used global BES. Critical thermo-dynamic variables conducive for heavy rainfall like convective available potential energy simulated using regional BES is more realistic compared to global BES. It is pointed out that these results have important practical implications in design of forecast platforms while decision-making during extreme weather events

Multifractality and value-at-risk forecasting of exchange rates

NASA Astrophysics Data System (ADS)

Batten, Jonathan A.; Kinateder, Harald; Wagner, Niklas

2014-05-01

This paper addresses market risk prediction for high frequency foreign exchange rates under nonlinear risk scaling behaviour. We use a modified version of the multifractal model of asset returns (MMAR) where trading time is represented by the series of volume ticks. Our dataset consists of 138,418 5-min round-the-clock observations of EUR/USD spot quotes and trading ticks during the period January 5, 2006 to December 31, 2007. Considering fat-tails, long-range dependence as well as scale inconsistency with the MMAR, we derive out-of-sample value-at-risk (VaR) forecasts and compare our approach to historical simulation as well as a benchmark GARCH(1,1) location-scale VaR model. Our findings underline that the multifractal properties in EUR/USD returns in fact have notable risk management implications. The MMAR approach is a parsimonious model which produces admissible VaR forecasts at the 12-h forecast horizon. For the daily horizon, the MMAR outperforms both alternatives based on conditional as well as unconditional coverage statistics.

A simplified real time method to forecast semi-enclosed basins storm surge

NASA Astrophysics Data System (ADS)

Pasquali, D.; Di Risio, M.; De Girolamo, P.

2015-11-01

Semi-enclosed basins are often prone to storm surge events. Indeed, their meteorological exposition, the presence of large continental shelf and their shape can lead to strong sea level set-up. A real time system aimed at forecasting storm surge may be of great help to protect human activities (i.e. to forecast flooding due to storm surge events), to manage ports and to safeguard coasts safety. This paper aims at illustrating a simple method able to forecast storm surge events in semi-enclosed basins in real time. The method is based on a mixed approach in which the results obtained by means of a simplified physics based model with low computational costs are corrected by means of statistical techniques. The proposed method is applied to a point of interest located in the Northern part of the Adriatic Sea. The comparison of forecasted levels against observed values shows the satisfactory reliability of the forecasts.

Financial forecasts accuracy in Brazil's social security system.

PubMed

Silva, Carlos Patrick Alves da; Puty, Claudio Alberto Castelo Branco; Silva, Marcelino Silva da; Carvalho, Solon Venâncio de; Francês, Carlos Renato Lisboa

2017-01-01

Long-term social security statistical forecasts produced and disseminated by the Brazilian government aim to provide accurate results that would serve as background information for optimal policy decisions. These forecasts are being used as support for the government's proposed pension reform that plans to radically change the Brazilian Constitution insofar as Social Security is concerned. However, the reliability of official results is uncertain since no systematic evaluation of these forecasts has ever been published by the Brazilian government or anyone else. This paper aims to present a study of the accuracy and methodology of the instruments used by the Brazilian government to carry out long-term actuarial forecasts. We base our research on an empirical and probabilistic analysis of the official models. Our empirical analysis shows that the long-term Social Security forecasts are systematically biased in the short term and have significant errors that render them meaningless in the long run. Moreover, the low level of transparency in the methods impaired the replication of results published by the Brazilian Government and the use of outdated data compromises forecast results. In the theoretical analysis, based on a mathematical modeling approach, we discuss the complexity and limitations of the macroeconomic forecast through the computation of confidence intervals. We demonstrate the problems related to error measurement inherent to any forecasting process. We then extend this exercise to the computation of confidence intervals for Social Security forecasts. This mathematical exercise raises questions about the degree of reliability of the Social Security forecasts.

Financial forecasts accuracy in Brazil’s social security system

PubMed Central

2017-01-01

Long-term social security statistical forecasts produced and disseminated by the Brazilian government aim to provide accurate results that would serve as background information for optimal policy decisions. These forecasts are being used as support for the government’s proposed pension reform that plans to radically change the Brazilian Constitution insofar as Social Security is concerned. However, the reliability of official results is uncertain since no systematic evaluation of these forecasts has ever been published by the Brazilian government or anyone else. This paper aims to present a study of the accuracy and methodology of the instruments used by the Brazilian government to carry out long-term actuarial forecasts. We base our research on an empirical and probabilistic analysis of the official models. Our empirical analysis shows that the long-term Social Security forecasts are systematically biased in the short term and have significant errors that render them meaningless in the long run. Moreover, the low level of transparency in the methods impaired the replication of results published by the Brazilian Government and the use of outdated data compromises forecast results. In the theoretical analysis, based on a mathematical modeling approach, we discuss the complexity and limitations of the macroeconomic forecast through the computation of confidence intervals. We demonstrate the problems related to error measurement inherent to any forecasting process. We then extend this exercise to the computation of confidence intervals for Social Security forecasts. This mathematical exercise raises questions about the degree of reliability of the Social Security forecasts. PMID:28859172

Seasonal drought predictability in Portugal using statistical-dynamical techniques

NASA Astrophysics Data System (ADS)

Ribeiro, A. F. S.; Pires, C. A. L.

2016-08-01

Atmospheric forecasting and predictability are important to promote adaption and mitigation measures in order to minimize drought impacts. This study estimates hybrid (statistical-dynamical) long-range forecasts of the regional drought index SPI (3-months) over homogeneous regions from mainland Portugal, based on forecasts from the UKMO operational forecasting system, with lead-times up to 6 months. ERA-Interim reanalysis data is used for the purpose of building a set of SPI predictors integrating recent past information prior to the forecast launching. Then, the advantage of combining predictors with both dynamical and statistical background in the prediction of drought conditions at different lags is evaluated. A two-step hybridization procedure is performed, in which both forecasted and observed 500 hPa geopotential height fields are subjected to a PCA in order to use forecasted PCs and persistent PCs as predictors. A second hybridization step consists on a statistical/hybrid downscaling to the regional SPI, based on regression techniques, after the pre-selection of the statistically significant predictors. The SPI forecasts and the added value of combining dynamical and statistical methods are evaluated in cross-validation mode, using the R2 and binary event scores. Results are obtained for the four seasons and it was found that winter is the most predictable season, and that most of the predictive power is on the large-scale fields from past observations. The hybridization improves the downscaling based on the forecasted PCs, since they provide complementary information (though modest) beyond that of persistent PCs. These findings provide clues about the predictability of the SPI, particularly in Portugal, and may contribute to the predictability of crops yields and to some guidance on users (such as farmers) decision making process.

Seasonal ENSO forecasting: Where does a simple model stand amongst other operational ENSO models?

NASA Astrophysics Data System (ADS)

Halide, Halmar

2017-01-01

We apply a simple linear multiple regression model called IndOzy for predicting ENSO up to 7 seasonal lead times. The model still used 5 (five) predictors of the past seasonal Niño 3.4 ENSO indices derived from chaos theory and it was rolling-validated to give a one-step ahead forecast. The model skill was evaluated against data from the season of May-June-July (MJJ) 2003 to November-December-January (NDJ) 2015/2016. There were three skill measures such as: Pearson correlation, RMSE, and Euclidean distance were used for forecast verification. The skill of this simple model was than compared to those of combined Statistical and Dynamical models compiled at the IRI (International Research Institute) website. It was found that the simple model was only capable of producing a useful ENSO prediction only up to 3 seasonal leads, while the IRI statistical and Dynamical model skill were still useful up to 4 and 6 seasonal leads, respectively. Even with its short-range seasonal prediction skills, however, the simple model still has a potential to give ENSO-derived tailored products such as probabilistic measures of precipitation and air temperature. Both meteorological conditions affect the presence of wild-land fire hot-spots in Sumatera and Kalimantan. It is suggested that to improve its long-range skill, the simple INDOZY model needs to incorporate a nonlinear model such as an artificial neural network technique.

A New Approach in Generating Meteorological Forecasts for Ensemble Streamflow Forecasting using Multivariate Functions

NASA Astrophysics Data System (ADS)

Khajehei, S.; Madadgar, S.; Moradkhani, H.

2014-12-01

The reliability and accuracy of hydrological predictions are subject to various sources of uncertainty, including meteorological forcing, initial conditions, model parameters and model structure. To reduce the total uncertainty in hydrological applications, one approach is to reduce the uncertainty in meteorological forcing by using the statistical methods based on the conditional probability density functions (pdf). However, one of the requirements for current methods is to assume the Gaussian distribution for the marginal distribution of the observed and modeled meteorology. Here we propose a Bayesian approach based on Copula functions to develop the conditional distribution of precipitation forecast needed in deriving a hydrologic model for a sub-basin in the Columbia River Basin. Copula functions are introduced as an alternative approach in capturing the uncertainties related to meteorological forcing. Copulas are multivariate joint distribution of univariate marginal distributions, which are capable to model the joint behavior of variables with any level of correlation and dependency. The method is applied to the monthly forecast of CPC with 0.25x0.25 degree resolution to reproduce the PRISM dataset over 1970-2000. Results are compared with Ensemble Pre-Processor approach as a common procedure used by National Weather Service River forecast centers in reproducing observed climatology during a ten-year verification period (2000-2010).

Time series modelling of global mean temperature for managerial decision-making.

PubMed

Romilly, Peter

2005-07-01

Climate change has important implications for business and economic activity. Effective management of climate change impacts will depend on the availability of accurate and cost-effective forecasts. This paper uses univariate time series techniques to model the properties of a global mean temperature dataset in order to develop a parsimonious forecasting model for managerial decision-making over the short-term horizon. Although the model is estimated on global temperature data, the methodology could also be applied to temperature data at more localised levels. The statistical techniques include seasonal and non-seasonal unit root testing with and without structural breaks, as well as ARIMA and GARCH modelling. A forecasting evaluation shows that the chosen model performs well against rival models. The estimation results confirm the findings of a number of previous studies, namely that global mean temperatures increased significantly throughout the 20th century. The use of GARCH modelling also shows the presence of volatility clustering in the temperature data, and a positive association between volatility and global mean temperature.

Improving Timeliness of Winter Wheat Production Forecast in United States of America, Ukraine and China Using MODIS Data and NCAR Growing Degree Day

NASA Astrophysics Data System (ADS)

Vermote, E.; Franch, B.; Becker-Reshef, I.; Claverie, M.; Huang, J.; Zhang, J.; Sobrino, J. A.

2014-12-01

Wheat is the most important cereal crop traded on international markets and winter wheat constitutes approximately 80% of global wheat production. Thus, accurate and timely forecasts of its production are critical for informing agricultural policies and investments, as well as increasing market efficiency and stability. Becker-Reshef et al. (2010) used an empirical generalized model for forecasting winter wheat production. Their approach combined BRDF-corrected daily surface reflectance from Moderate resolution Imaging Spectroradiometer (MODIS) Climate Modeling Grid (CMG) with detailed official crop statistics and crop type masks. It is based on the relationship between the Normalized Difference Vegetation Index (NDVI) at the peak of the growing season, percent wheat within the CMG pixel, and the final yields. This method predicts the yield approximately one month to six weeks prior to harvest. In this study, we include the Growing Degree Day (GDD) information extracted from NCEP/NCAR reanalysis data in order to improve the winter wheat production forecast by increasing the timeliness of the forecasts while conserving the accuracy of the original model. We apply this modified model to three major wheat-producing countries: United States of America, Ukraine and China from 2001 to 2012. We show that a reliable forecast can be made between one month to a month and a half prior to the peak NDVI (meaning two months to two and a half months prior to harvest) while conserving an accuracy of 10% in the production forecast.

An Analysis on the Unemployment Rate in the Philippines: A Time Series Data Approach

NASA Astrophysics Data System (ADS)

Urrutia, J. D.; Tampis, R. L.; E Atienza, JB

2017-03-01

This study aims to formulate a mathematical model for forecasting and estimating unemployment rate in the Philippines. Also, factors which can predict the unemployment is to be determined among the considered variables namely Labor Force Rate, Population, Inflation Rate, Gross Domestic Product, and Gross National Income. Granger-causal relationship and integration among the dependent and independent variables are also examined using Pairwise Granger-causality test and Johansen Cointegration Test. The data used were acquired from the Philippine Statistics Authority, National Statistics Office, and Bangko Sentral ng Pilipinas. Following the Box-Jenkins method, the formulated model for forecasting the unemployment rate is SARIMA (6, 1, 5) × (0, 1, 1)4 with a coefficient of determination of 0.79. The actual values are 99 percent identical to the predicted values obtained through the model, and are 72 percent closely relative to the forecasted ones. According to the results of the regression analysis, Labor Force Rate and Population are the significant factors of unemployment rate. Among the independent variables, Population, GDP, and GNI showed to have a granger-causal relationship with unemployment. It is also found that there are at least four cointegrating relations between the dependent and independent variables.

Technical Note: Initial assessment of a multi-method approach to spring-flood forecasting in Sweden

NASA Astrophysics Data System (ADS)

Olsson, J.; Uvo, C. B.; Foster, K.; Yang, W.

2016-02-01

Hydropower is a major energy source in Sweden, and proper reservoir management prior to the spring-flood onset is crucial for optimal production. This requires accurate forecasts of the accumulated discharge in the spring-flood period (i.e. the spring-flood volume, SFV). Today's SFV forecasts are generated using a model-based climatological ensemble approach, where time series of precipitation and temperature from historical years are used to force a calibrated and initialized set-up of the HBV model. In this study, a number of new approaches to spring-flood forecasting that reflect the latest developments with respect to analysis and modelling on seasonal timescales are presented and evaluated. Three main approaches, represented by specific methods, are evaluated in SFV hindcasts for the Swedish river Vindelälven over a 10-year period with lead times between 0 and 4 months. In the first approach, historically analogue years with respect to the climate in the period preceding the spring flood are identified and used to compose a reduced ensemble. In the second, seasonal meteorological ensemble forecasts are used to drive the HBV model over the spring-flood period. In the third approach, statistical relationships between SFV and the large-sale atmospheric circulation are used to build forecast models. None of the new approaches consistently outperform the climatological ensemble approach, but for early forecasts improvements of up to 25 % are found. This potential is reasonably well realized in a multi-method system, which over all forecast dates reduced the error in SFV by ˜ 4 %. This improvement is limited but potentially significant for e.g. energy trading.

Intercomparison of air quality data using principal component analysis, and forecasting of PM₁₀ and PM₂.₅ concentrations using artificial neural networks, in Thessaloniki and Helsinki.

PubMed

Voukantsis, Dimitris; Karatzas, Kostas; Kukkonen, Jaakko; Räsänen, Teemu; Karppinen, Ari; Kolehmainen, Mikko

2011-03-01

In this paper we propose a methodology consisting of specific computational intelligence methods, i.e. principal component analysis and artificial neural networks, in order to inter-compare air quality and meteorological data, and to forecast the concentration levels for environmental parameters of interest (air pollutants). We demonstrate these methods to data monitored in the urban areas of Thessaloniki and Helsinki in Greece and Finland, respectively. For this purpose, we applied the principal component analysis method in order to inter-compare the patterns of air pollution in the two selected cities. Then, we proceeded with the development of air quality forecasting models for both studied areas. On this basis, we formulated and employed a novel hybrid scheme in the selection process of input variables for the forecasting models, involving a combination of linear regression and artificial neural networks (multi-layer perceptron) models. The latter ones were used for the forecasting of the daily mean concentrations of PM₁₀ and PM₂.₅ for the next day. Results demonstrated an index of agreement between measured and modelled daily averaged PM₁₀ concentrations, between 0.80 and 0.85, while the kappa index for the forecasting of the daily averaged PM₁₀ concentrations reached 60% for both cities. Compared with previous corresponding studies, these statistical parameters indicate an improved performance of air quality parameters forecasting. It was also found that the performance of the models for the forecasting of the daily mean concentrations of PM₁₀ was not substantially different for both cities, despite the major differences of the two urban environments under consideration. Copyright © 2011 Elsevier B.V. All rights reserved.

Forecasting defoliation by the gypsy moth in oak stands

Treesearch

Robert W. Campbell; Joseph P. Standaert

1974-01-01

A multiple-regression model is presented that reflects statistically significant correlations between defoliation by the gypsy moth, the dependent variable, and a series of biotic and physical independent variables. Both possible uses and shortcomings of this model are discussed.

Initial assessment of a multi-model approach to spring flood forecasting in Sweden

NASA Astrophysics Data System (ADS)

Olsson, J.; Uvo, C. B.; Foster, K.; Yang, W.

2015-06-01

Hydropower is a major energy source in Sweden and proper reservoir management prior to the spring flood onset is crucial for optimal production. This requires useful forecasts of the accumulated discharge in the spring flood period (i.e. the spring-flood volume, SFV). Today's SFV forecasts are generated using a model-based climatological ensemble approach, where time series of precipitation and temperature from historical years are used to force a calibrated and initialised set-up of the HBV model. In this study, a number of new approaches to spring flood forecasting, that reflect the latest developments with respect to analysis and modelling on seasonal time scales, are presented and evaluated. Three main approaches, represented by specific methods, are evaluated in SFV hindcasts for three main Swedish rivers over a 10-year period with lead times between 0 and 4 months. In the first approach, historically analogue years with respect to the climate in the period preceding the spring flood are identified and used to compose a reduced ensemble. In the second, seasonal meteorological ensemble forecasts are used to drive the HBV model over the spring flood period. In the third approach, statistical relationships between SFV and the large-sale atmospheric circulation are used to build forecast models. None of the new approaches consistently outperform the climatological ensemble approach, but for specific locations and lead times improvements of 20-30 % are found. When combining all forecasts in a weighted multi-model approach, a mean improvement over all locations and lead times of nearly 10 % was indicated. This demonstrates the potential of the approach and further development and optimisation into an operational system is ongoing.

Forecasting longitudinal changes in oropharyngeal tumor morphology throughout the course of head and neck radiation therapy

PubMed Central

Yock, Adam D.; Rao, Arvind; Dong, Lei; Beadle, Beth M.; Garden, Adam S.; Kudchadker, Rajat J.; Court, Laurence E.

2014-01-01

Purpose: To create models that forecast longitudinal trends in changing tumor morphology and to evaluate and compare their predictive potential throughout the course of radiation therapy. Methods: Two morphology feature vectors were used to describe 35 gross tumor volumes (GTVs) throughout the course of intensity-modulated radiation therapy for oropharyngeal tumors. The feature vectors comprised the coordinates of the GTV centroids and a description of GTV shape using either interlandmark distances or a spherical harmonic decomposition of these distances. The change in the morphology feature vector observed at 33 time points throughout the course of treatment was described using static, linear, and mean models. Models were adjusted at 0, 1, 2, 3, or 5 different time points (adjustment points) to improve prediction accuracy. The potential of these models to forecast GTV morphology was evaluated using leave-one-out cross-validation, and the accuracy of the models was compared using Wilcoxon signed-rank tests. Results: Adding a single adjustment point to the static model without any adjustment points decreased the median error in forecasting the position of GTV surface landmarks by the largest amount (1.2 mm). Additional adjustment points further decreased the forecast error by about 0.4 mm each. Selection of the linear model decreased the forecast error for both the distance-based and spherical harmonic morphology descriptors (0.2 mm), while the mean model decreased the forecast error for the distance-based descriptor only (0.2 mm). The magnitude and statistical significance of these improvements decreased with each additional adjustment point, and the effect from model selection was not as large as that from adding the initial points. Conclusions: The authors present models that anticipate longitudinal changes in tumor morphology using various models and model adjustment schemes. The accuracy of these models depended on their form, and the utility of these models includes the characterization of patient-specific response with implications for treatment management and research study design. PMID:25086518

Forecasting longitudinal changes in oropharyngeal tumor morphology throughout the course of head and neck radiation therapy

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

Yock, Adam D.; Kudchadker, Rajat J.; Rao, Arvind

2014-08-15

Purpose: To create models that forecast longitudinal trends in changing tumor morphology and to evaluate and compare their predictive potential throughout the course of radiation therapy. Methods: Two morphology feature vectors were used to describe 35 gross tumor volumes (GTVs) throughout the course of intensity-modulated radiation therapy for oropharyngeal tumors. The feature vectors comprised the coordinates of the GTV centroids and a description of GTV shape using either interlandmark distances or a spherical harmonic decomposition of these distances. The change in the morphology feature vector observed at 33 time points throughout the course of treatment was described using static, linear,more » and mean models. Models were adjusted at 0, 1, 2, 3, or 5 different time points (adjustment points) to improve prediction accuracy. The potential of these models to forecast GTV morphology was evaluated using leave-one-out cross-validation, and the accuracy of the models was compared using Wilcoxon signed-rank tests. Results: Adding a single adjustment point to the static model without any adjustment points decreased the median error in forecasting the position of GTV surface landmarks by the largest amount (1.2 mm). Additional adjustment points further decreased the forecast error by about 0.4 mm each. Selection of the linear model decreased the forecast error for both the distance-based and spherical harmonic morphology descriptors (0.2 mm), while the mean model decreased the forecast error for the distance-based descriptor only (0.2 mm). The magnitude and statistical significance of these improvements decreased with each additional adjustment point, and the effect from model selection was not as large as that from adding the initial points. Conclusions: The authors present models that anticipate longitudinal changes in tumor morphology using various models and model adjustment schemes. The accuracy of these models depended on their form, and the utility of these models includes the characterization of patient-specific response with implications for treatment management and research study design.« less

United States geological survey's reserve-growth models and their implementation

USGS Publications Warehouse

Klett, T.R.

2005-01-01

The USGS has developed several mathematical models to forecast reserve growth of fields both in the United States (U.S.) and the world. The models are based on historical reserve growth patterns of fields in the U.S. The patterns of past reserve growth are extrapolated to forecast future reserve growth. Changes of individual field sizes through time are extremely variable, therefore, the reserve growth models take on a statistical approach whereby volumetric changes for populations of fields are used in the models. Field age serves as a measure of the field-development effort that is applied to promote reserve growth. At the time of the USGS World Petroleum Assessment 2000, a reserve growth model for discovered fields of the world was not available. Reserve growth forecasts, therefore, were made based on a model of historical reserve growth of fields of the U.S. To test the feasibility of such an application, reserve growth forecasts were made of 186 giant oil fields of the world (excluding the U.S. and Canada). In addition, forecasts were made for these giant oil fields subdivided into those located in and outside of Organization of Petroleum Exporting Countries (OPEC). The model provided a reserve-growth forecast that closely matched the actual reserve growth that occurred from 1981 through 1996 for the 186 fields as a whole, as well as for both OPEC and non-OPEC subdivisions, despite the differences in reserves definition among the fields of the U.S. and the rest of the world. ?? 2005 International Association for Mathematical Geology.

Future mission studies: Preliminary comparisons of solar flux models

NASA Technical Reports Server (NTRS)

Ashrafi, S.

1991-01-01

The results of comparisons of the solar flux models are presented. (The wavelength lambda = 10.7 cm radio flux is the best indicator of the strength of the ionizing radiations such as solar ultraviolet and x-ray emissions that directly affect the atmospheric density thereby changing the orbit lifetime of satellites. Thus, accurate forecasting of solar flux F sub 10.7 is crucial for orbit determination of spacecrafts.) The measured solar flux recorded by National Oceanic and Atmospheric Administration (NOAA) is compared against the forecasts made by Schatten, MSFC, and NOAA itself. The possibility of a combined linear, unbiased minimum-variance estimation that properly combines all three models into one that minimizes the variance is also discussed. All the physics inherent in each model are combined. This is considered to be the dead-end statistical approach to solar flux forecasting before any nonlinear chaotic approach.

How well can the observed Arctic sea ice summer retreat and winter advance be represented in the NCEP Climate Forecast System version 2?

NASA Astrophysics Data System (ADS)

Collow, Thomas W.; Wang, Wanqiu; Kumar, Arun; Zhang, Jinlun

2017-09-01

The capability of a numerical model to simulate the statistical characteristics of the summer sea ice date of retreat (DOR) and the winter date of advance (DOA) is investigated using sea ice concentration output from the Climate Forecast System Version 2 model (CFSv2). Two model configurations are tested, the operational setting (CFSv2CFSR) which uses initial data from the Climate Forecast System Reanalysis, and a modified version (CFSv2PIOMp) which ingests sea ice thickness initialization data from the Pan-Arctic Ice Ocean Modeling and Assimilation System (PIOMAS) and includes physics modifications for a more realistic representation of heat fluxes at the sea ice top and bottom. First, a method to define DOR and DOA is presented. Then, DOR and DOA are determined from the model simulations and observational sea ice concentration from the National Aeronautics and Space Administration (NASA). Means, trends, and detrended standard deviations of DOR and DOA are compared, along with DOR/DOA rates in the Arctic Ocean. It is found that the statistics are generally similar between the model and observations, although some regional biases exist. In addition, regions of new ice retreat in recent years are represented well in CFSv2PIOMp over the Arctic Ocean, in terms of both spatial extent and timing. Overall, CFSv2PIOMp shows a reduction in error throughout the Arctic. Based on results, it is concluded that the model produces a reasonable representation of the climatology and variability statistics of DOR and DOA in most regions. This assessment serves as a prerequisite for future predictability experiments.

Social Indicators and Social Forecasting.

ERIC Educational Resources Information Center

Johnston, Denis F.

The paper identifies major types of social indicators and explains how they can be used in social forecasting. Social indicators are defined as statistical measures relating to major areas of social concern and/or individual well being. Examples of social indicators are projections, forecasts, outlook statements, time-series statistics, and…

Statistical models of temperature in the Sacramento-San Joaquin delta under climate-change scenarios and ecological implications

USGS Publications Warehouse

Wagner, R.W.; Stacey, M.; Brown, L.R.; Dettinger, M.

2011-01-01

Changes in water temperatures caused by climate change in California's Sacramento-San Joaquin Delta will affect the ecosystem through physiological rates of fishes and invertebrates. This study presents statistical models that can be used to forecast water temperature within the Delta as a response to atmospheric conditions. The daily average model performed well (R2 values greater than 0.93 during verification periods) for all stations within the Delta and San Francisco Bay provided there was at least 1 year of calibration data. To provide long-term projections of Delta water temperature, we forced the model with downscaled data from climate scenarios. Based on these projections, the ecological implications for the delta smelt, a key species, were assessed based on temperature thresholds. The model forecasts increases in the number of days above temperatures causing high mortality (especially along the Sacramento River) and a shift in thermal conditions for spawning to earlier in the year. ?? 2011 The Author(s).

Precipitation forecast using artificial neural networks. An application to the Guadalupe Valley, Baja California, Mexico

NASA Astrophysics Data System (ADS)

Herrera-Oliva, C. S.

2013-05-01

In this work we design and implement a method for the determination of precipitation forecast through the application of an elementary neuronal network (perceptron) to the statistical analysis of the precipitation reported in catalogues. The method is limited mainly by the catalogue length (and, in a smaller degree, by its accuracy). The method performance is measured using grading functions that evaluate a tradeoff between positive and negative aspects of performance. The method is applied to the Guadalupe Valley, Baja California, Mexico. Using consecutive intervals of dt=0.1 year, employing the data of several climatological stations situated in and surrounding this important wine industries zone. We evaluated the performance of different models of ANN, whose variables of entrance are the heights of precipitation. The results obtained were satisfactory, except for exceptional values of rain. Key words: precipitation forecast, artificial neural networks, statistical analysis

Chaos and Forecasting - Proceedings of the Royal Society Discussion Meeting

NASA Astrophysics Data System (ADS)

Tong, Howell

1995-04-01

The Table of Contents for the full book PDF is as follows: * Preface * Orthogonal Projection, Embedding Dimension and Sample Size in Chaotic Time Series from a Statistical Perspective * A Theory of Correlation Dimension for Stationary Time Series * On Prediction and Chaos in Stochastic Systems * Locally Optimized Prediction of Nonlinear Systems: Stochastic and Deterministic * A Poisson Distribution for the BDS Test Statistic for Independence in a Time Series * Chaos and Nonlinear Forecastability in Economics and Finance * Paradigm Change in Prediction * Predicting Nonuniform Chaotic Attractors in an Enzyme Reaction * Chaos in Geophysical Fluids * Chaotic Modulation of the Solar Cycle * Fractal Nature in Earthquake Phenomena and its Simple Models * Singular Vectors and the Predictability of Weather and Climate * Prediction as a Criterion for Classifying Natural Time Series * Measuring and Characterising Spatial Patterns, Dynamics and Chaos in Spatially-Extended Dynamical Systems and Ecologies * Non-Linear Forecasting and Chaos in Ecology and Epidemiology: Measles as a Case Study

Considering inventory distributions in a stochastic periodic inventory routing system

NASA Astrophysics Data System (ADS)

Yadollahi, Ehsan; Aghezzaf, El-Houssaine

2017-07-01

Dealing with the stochasticity of parameters is one of the critical issues in business and industry nowadays. Supply chain planners have difficulties in forecasting stochastic parameters of a distribution system. Demand rates of customers during their lead time are one of these parameters. In addition, holding a huge level of inventory at the retailers is costly and inefficient. To cover the uncertainty of forecasting demand rates, researchers have proposed the usage of safety stock to avoid stock-out. However, finding the precise level of safety stock depends on forecasting the statistical distribution of demand rates and their variations in different settings among the planning horizon. In this paper the demand rate distributions and its parameters are taken into account for each time period in a stochastic periodic IRP. An analysis of the achieved statistical distribution of the inventory and safety stock level is provided to measure the effects of input parameters on the output indicators. Different values for coefficient of variation are applied to the customers' demand rate in the optimization model. The outcome of the deterministic equivalent model of SPIRP is simulated in form of an illustrative case.

Impacts of high resolution model downscaling in coastal regions

NASA Astrophysics Data System (ADS)

Bricheno, Lucy; Wolf, Judith

2013-04-01

With model development and cheaper computational resources ocean forecasts are becoming readily available, high resolution coastal forecasting is now a reality. This can only be achieved, however, by downscaling global or basin-scale products such as the MyOcean reanalyses and forecasts. These model products have resolution ranging from 1/16th - 1/4 degree, which are often insufficient for coastal scales, but can provide initialisation and boundary data. We present applications of downscaling the MyOcean products for use in shelf-seas and the nearshore. We will address the question 'Do coastal predictions improve with higher resolution modelling?' with a few focused examples, while also discussing what is meant by an improved result. Increasing resolution appears to be an obvious route for getting more accurate forecasts in operational coastal models. However, when models resolve finer scales, this may lead to the introduction of high-frequency variability which is not necessarily deterministic. Thus a flow may appear more realistic by generating eddies but the simple statistics like rms error and correlation may become less good because the model variability is not exactly in phase with the observations (Hoffman et al., 1995). By deciding on a specific process to simulate (rather than concentrating on reducing rms error) we can better assess the improvements gained by downscaling. In this work we will select two processes which are dominant in our case-study site: Liverpool Bay. Firstly we consider the magnitude and timing of a peak in tide-surge elevations, by separating out the event into timing (or displacement) and intensity (or amplitude) errors. The model can thus be evaluated on how well it predicts the timing and magnitude of the surge. The second important characteristic of Liverpool Bay is the position of the freshwater front. To evaluate model performance in this case, the location, sharpness, and temperature difference across the front will be considered. We will show that by using intelligent metrics designed with a physical process in mind, we can learn more about model performance than by considering 'bulk' statistics alone. R. M. Hoffman and Z. Liu and J-F. Louic and C. Grassotti (1995) 'Distortion Representation of Forecast Errors' Monthly Weather Review 123: 2758-2770

Season-ahead Drought Forecast Models for the Lower Colorado River Authority in Texas

NASA Astrophysics Data System (ADS)

Block, P. J.; Zimmerman, B.; Grzegorzewski, M.; Watkins, D. W., Jr.; Anderson, R.

2014-12-01

The Lower Colorado River Authority (LCRA) in Austin, Texas, manages the Highland Lakes reservoir system in Central Texas, a series of six lakes on the Lower Colorado River. This system provides water to approximately 1.1 million people in Central Texas, supplies hydropower to a 55-county area, supports rice farming along the Texas Gulf Coast, and sustains in-stream flows in the Lower Colorado River and freshwater inflows to Matagorda Bay. The current, prolonged drought conditions are severely taxing the LCRA's system, making allocation and management decisions exceptionally challenging, and affecting the ability of constituents to conduct proper planning. In this work, we further develop and evaluate season-ahead statistical streamflow and precipitation forecast models for integration into LCRA decision support models. Optimal forecast lead time, predictive skill, form, and communication are all considered.

Physics-based and statistical earthquake forecasting in a continental rift zone: the case study of Corinth Gulf (Greece)

NASA Astrophysics Data System (ADS)

Segou, Margarita

2016-01-01

I perform a retrospective forecast experiment in the most rapid extensive continental rift worldwide, the western Corinth Gulf (wCG, Greece), aiming to predict shallow seismicity (depth <15 km) with magnitude M ≥ 3.0 for the time period between 1995 and 2013. I compare two short-term earthquake clustering models, based on epidemic-type aftershock sequence (ETAS) statistics, four physics-based (CRS) models, combining static stress change estimations and the rate-and-state laboratory law and one hybrid model. For the latter models, I incorporate the stress changes imparted from 31 earthquakes with magnitude M ≥ 4.5 at the extended area of wCG. Special attention is given on the 3-D representation of active faults, acting as potential receiver planes for the estimation of static stress changes. I use reference seismicity between 1990 and 1995, corresponding to the learning phase of physics-based models, and I evaluate the forecasts for six months following the 1995 M = 6.4 Aigio earthquake using log-likelihood performance metrics. For the ETAS realizations, I use seismic events with magnitude M ≥ 2.5 within daily update intervals to enhance their predictive power. For assessing the role of background seismicity, I implement a stochastic reconstruction (aka declustering) aiming to answer whether M > 4.5 earthquakes correspond to spontaneous events and identify, if possible, different triggering characteristics between aftershock sequences and swarm-type seismicity periods. I find that: (1) ETAS models outperform CRS models in most time intervals achieving very low rejection ratio RN = 6 per cent, when I test their efficiency to forecast the total number of events inside the study area, (2) the best rejection ratio for CRS models reaches RN = 17 per cent, when I use varying target depths and receiver plane geometry, (3) 75 per cent of the 1995 Aigio aftershocks that occurred within the first month can be explained by static stress changes, (4) highly variable performance on behalf of both statistical and physical models is suggested by large confidence intervals of information gain per earthquake and (5) generic ETAS models can adequately predict the temporal evolution of seismicity during swarms. Furthermore, stochastic reconstruction of seismicity makes possible the identification of different triggering processes between specific seismic crises (2001, 2003-04, 2006-07) and the 1995 aftershock sequence. I find that: (1) seismic events with M ≥ 5.0 are not a part of a preceding earthquake cascade, since they are characterized by high probability being a background event (average Pback > 0.8) and (2) triggered seismicity within swarms is characterized by lower event productivity when compared with the corresponding value during aftershock sequences. I conclude that physics-based models contribute on the determination of the `new-normal' seismicity rate at longer time intervals and that their joint implementation with statistical models is beneficial for future operational forecast systems.

Time series analysis of temporal trends in the pertussis incidence in Mainland China from 2005 to 2016.

PubMed

Zeng, Qianglin; Li, Dandan; Huang, Gui; Xia, Jin; Wang, Xiaoming; Zhang, Yamei; Tang, Wanping; Zhou, Hui

2016-08-31

Short-term forecast of pertussis incidence is helpful for advanced warning and planning resource needs for future epidemics. By utilizing the Auto-Regressive Integrated Moving Average (ARIMA) model and Exponential Smoothing (ETS) model as alterative models with R software, this paper analyzed data from Chinese Center for Disease Control and Prevention (China CDC) between January 2005 and June 2016. The ARIMA (0,1,0)(1,1,1)12 model (AICc = 1342.2 BIC = 1350.3) was selected as the best performing ARIMA model and the ETS (M,N,M) model (AICc = 1678.6, BIC = 1715.4) was selected as the best performing ETS model, and the ETS (M,N,M) model with the minimum RMSE was finally selected for in-sample-simulation and out-of-sample forecasting. Descriptive statistics showed that the reported number of pertussis cases by China CDC increased by 66.20% from 2005 (4058 cases) to 2015 (6744 cases). According to Hodrick-Prescott filter, there was an apparent cyclicity and seasonality in the pertussis reports. In out of sample forecasting, the model forecasted a relatively high incidence cases in 2016, which predicates an increasing risk of ongoing pertussis resurgence in the near future. In this regard, the ETS model would be a useful tool in simulating and forecasting the incidence of pertussis, and helping decision makers to take efficient decisions based on the advanced warning of disease incidence.

Intelligent Soft Computing on Forex: Exchange Rates Forecasting with Hybrid Radial Basis Neural Network

PubMed Central

Marcek, Dusan; Durisova, Maria

2016-01-01

This paper deals with application of quantitative soft computing prediction models into financial area as reliable and accurate prediction models can be very helpful in management decision-making process. The authors suggest a new hybrid neural network which is a combination of the standard RBF neural network, a genetic algorithm, and a moving average. The moving average is supposed to enhance the outputs of the network using the error part of the original neural network. Authors test the suggested model on high-frequency time series data of USD/CAD and examine the ability to forecast exchange rate values for the horizon of one day. To determine the forecasting efficiency, they perform a comparative statistical out-of-sample analysis of the tested model with autoregressive models and the standard neural network. They also incorporate genetic algorithm as an optimizing technique for adapting parameters of ANN which is then compared with standard backpropagation and backpropagation combined with K-means clustering algorithm. Finally, the authors find out that their suggested hybrid neural network is able to produce more accurate forecasts than the standard models and can be helpful in eliminating the risk of making the bad decision in decision-making process. PMID:26977450

Intelligent Soft Computing on Forex: Exchange Rates Forecasting with Hybrid Radial Basis Neural Network.

PubMed

Falat, Lukas; Marcek, Dusan; Durisova, Maria

2016-01-01

This paper deals with application of quantitative soft computing prediction models into financial area as reliable and accurate prediction models can be very helpful in management decision-making process. The authors suggest a new hybrid neural network which is a combination of the standard RBF neural network, a genetic algorithm, and a moving average. The moving average is supposed to enhance the outputs of the network using the error part of the original neural network. Authors test the suggested model on high-frequency time series data of USD/CAD and examine the ability to forecast exchange rate values for the horizon of one day. To determine the forecasting efficiency, they perform a comparative statistical out-of-sample analysis of the tested model with autoregressive models and the standard neural network. They also incorporate genetic algorithm as an optimizing technique for adapting parameters of ANN which is then compared with standard backpropagation and backpropagation combined with K-means clustering algorithm. Finally, the authors find out that their suggested hybrid neural network is able to produce more accurate forecasts than the standard models and can be helpful in eliminating the risk of making the bad decision in decision-making process.

Valuing year-to-go hydrologic forecast improvements for a peaking hydropower system in the Sierra Nevada

NASA Astrophysics Data System (ADS)

Rheinheimer, David E.; Bales, Roger C.; Oroza, Carlos A.; Lund, Jay R.; Viers, Joshua H.

2016-05-01

We assessed the potential value of hydrologic forecasting improvements for a snow-dominated high-elevation hydropower system in the Sierra Nevada of California, using a hydropower optimization model. To mimic different forecasting skill levels for inflow time series, rest-of-year inflows from regression-based forecasts were blended in different proportions with representative inflows from a spatially distributed hydrologic model. The statistical approach mimics the simpler, historical forecasting approach that is still widely used. Revenue was calculated using historical electricity prices, with perfect price foresight assumed. With current infrastructure and operations, perfect hydrologic forecasts increased annual hydropower revenue by 0.14 to 1.6 million, with lower values in dry years and higher values in wet years, or about $0.8 million (1.2%) on average, representing overall willingness-to-pay for perfect information. A second sensitivity analysis found a wider range of annual revenue gain or loss using different skill levels in snow measurement in the regression-based forecast, mimicking expected declines in skill as the climate warms and historical snow measurements no longer represent current conditions. The value of perfect forecasts was insensitive to storage capacity for small and large reservoirs, relative to average inflow, and modestly sensitive to storage capacity with medium (current) reservoir storage. The value of forecasts was highly sensitive to powerhouse capacity, particularly for the range of capacities in the northern Sierra Nevada. The approach can be extended to multireservoir, multipurpose systems to help guide investments in forecasting.

Seasonal forecasting for water resource management: the example of CNR Genissiat dam on the Rhone River in France

NASA Astrophysics Data System (ADS)

Dommanget, Etienne; Bellier, Joseph; Ben Daoud, Aurélien; Graff, Benjamin

2014-05-01

Compagnie Nationale du Rhône (CNR) has been granted the concession to operate the Rhone River from the Swiss border to the Mediterranean Sea since 1933 and carries out three interdependent missions: navigation, irrigation and hydropower production. Nowadays, CNR generates one quarter of France's hydropower electricity. The convergence of public and private interests around optimizing the management of water resources throughout the French Rhone valley led CNR to develop hydrological models dedicated to discharge seasonal forecasting. Indeed, seasonal forecasting is a major issue for CNR and water resource management, in order to optimize long-term investments of the produced electricity, plan dam maintenance operations and anticipate low water period. Seasonal forecasting models have been developed on the Genissiat dam. With an installed capacity of 420MW, Genissiat dam is the first of the 19 CNR's hydropower plants. Discharge forecasting at Genissiat dam is strategic since its inflows contributes to 20% of the total Rhone average discharge and consequently to 40% of the total Rhone hydropower production. Forecasts are based on hydrological statistical models. Discharge on the main Rhone River tributaries upstream Genissiat dam are forecasted from 1 to 6 months ahead thanks to multiple linear regressions. Inputs data of these regressions are identified depending on river hydrological regimes and periods of the year. For the melting season, from spring to summer, snow water equivalent (SWE) data are of major importance. SWE data are calculated from Crocus model (Météo France) and SLF's model (Switzerland). CNR hydro-meteorological forecasters assessed meteorological trends regarding precipitations for the next coming months. These trends are used to generate stochastically precipitation scenarios in order to complement regression data set. This probabilistic approach build a decision-making supports for CNR's water resource management team and provides them with seasonal forecasts and their confidence interval. After a presentation of CNR methodology, results for the years 2011 and 2013 will illustrate CNR's seasonal forecasting models ability. These years are of particular interest regarding water resource management seeing that they are, respectively, unusually dry and snowy. Model performances will be assessed in comparison with historical climatology thanks to CRPS skill score.

Probabilistic empirical prediction of seasonal climate: evaluation and potential applications

NASA Astrophysics Data System (ADS)

Dieppois, B.; Eden, J.; van Oldenborgh, G. J.

2017-12-01

Preparing for episodes with risks of anomalous weather a month to a year ahead is an important challenge for governments, non-governmental organisations, and private companies and is dependent on the availability of reliable forecasts. The majority of operational seasonal forecasts are made using process-based dynamical models, which are complex, computationally challenging and prone to biases. Empirical forecast approaches built on statistical models to represent physical processes offer an alternative to dynamical systems and can provide either a benchmark for comparison or independent supplementary forecasts. Here, we present a new evaluation of an established empirical system used to predict seasonal climate across the globe. Forecasts for surface air temperature, precipitation and sea level pressure are produced by the KNMI Probabilistic Empirical Prediction (K-PREP) system every month and disseminated via the KNMI Climate Explorer (climexp.knmi.nl). K-PREP is based on multiple linear regression and built on physical principles to the fullest extent with predictive information taken from the global CO2-equivalent concentration, large-scale modes of variability in the climate system and regional-scale information. K-PREP seasonal forecasts for the period 1981-2016 will be compared with corresponding dynamically generated forecasts produced by operational forecast systems. While there are many regions of the world where empirical forecast skill is extremely limited, several areas are identified where K-PREP offers comparable skill to dynamical systems. We discuss two key points in the future development and application of the K-PREP system: (a) the potential for K-PREP to provide a more useful basis for reference forecasts than those based on persistence or climatology, and (b) the added value of including K-PREP forecast information in multi-model forecast products, at least for known regions of good skill. We also discuss the potential development of stakeholder-driven applications of the K-PREP system, including empirical forecasts for circumboreal fire activity.

School District Enrollment Projections: A Comparison of Three Methods.

ERIC Educational Resources Information Center

Pettibone, Timothy J.; Bushan, Latha

This study assesses three methods of forecasting school enrollments: the cohort-sruvival method (grade progression), the statistical forecasting procedure developed by the Statistical Analysis System (SAS) Institute, and a simple ratio computation. The three methods were used to forecast school enrollments for kindergarten through grade 12 in a…

Forecasting distributions of large federal-lands fires utilizing satellite and gridded weather information

USGS Publications Warehouse

Preisler, H.K.; Burgan, R.E.; Eidenshink, J.C.; Klaver, Jacqueline M.; Klaver, R.W.

2009-01-01

The current study presents a statistical model for assessing the skill of fire danger indices and for forecasting the distribution of the expected numbers of large fires over a given region and for the upcoming week. The procedure permits development of daily maps that forecast, for the forthcoming week and within federal lands, percentiles of the distributions of (i) number of ignitions; (ii) number of fires above a given size; (iii) conditional probabilities of fires greater than a specified size, given ignition. As an illustration, we used the methods to study the skill of the Fire Potential Index an index that incorporates satellite and surface observations to map fire potential at a national scale in forecasting distributions of large fires. ?? 2009 IAWF.

Generating synthetic daily precipitation realizations for seasonal precipitation forecasts

USDA-ARS?s Scientific Manuscript database

Synthetic weather generation models that depend on statistics of past weather observations are often limited in their applications to issues that depend upon historical weather characteristics. Enhancing these models to take advantage of increasingly available and skillful seasonal climate outlook p...

Use of medium-range numerical weather prediction model output to produce forecasts of streamflow

USGS Publications Warehouse

Clark, M.P.; Hay, L.E.

2004-01-01

This paper examines an archive containing over 40 years of 8-day atmospheric forecasts over the contiguous United States from the NCEP reanalysis project to assess the possibilities for using medium-range numerical weather prediction model output for predictions of streamflow. This analysis shows the biases in the NCEP forecasts to be quite extreme. In many regions, systematic precipitation biases exceed 100% of the mean, with temperature biases exceeding 3??C. In some locations, biases are even higher. The accuracy of NCEP precipitation and 2-m maximum temperature forecasts is computed by interpolating the NCEP model output for each forecast day to the location of each station in the NWS cooperative network and computing the correlation with station observations. Results show that the accuracy of the NCEP forecasts is rather low in many areas of the country. Most apparent is the generally low skill in precipitation forecasts (particularly in July) and low skill in temperature forecasts in the western United States, the eastern seaboard, and the southern tier of states. These results outline a clear need for additional processing of the NCEP Medium-Range Forecast Model (MRF) output before it is used for hydrologic predictions. Techniques of model output statistics (MOS) are used in this paper to downscale the NCEP forecasts to station locations. Forecasted atmospheric variables (e.g., total column precipitable water, 2-m air temperature) are used as predictors in a forward screening multiple linear regression model to improve forecasts of precipitation and temperature for stations in the National Weather Service cooperative network. This procedure effectively removes all systematic biases in the raw NCEP precipitation and temperature forecasts. MOS guidance also results in substantial improvements in the accuracy of maximum and minimum temperature forecasts throughout the country. For precipitation, forecast improvements were less impressive. MOS guidance increases he accuracy of precipitation forecasts over the northeastern United States, but overall, the accuracy of MOS-based precipitation forecasts is slightly lower than the raw NCEP forecasts. Four basins in the United States were chosen as case studies to evaluate the value of MRF output for predictions of streamflow. Streamflow forecasts using MRF output were generated for one rainfall-dominated basin (Alapaha River at Statenville, Georgia) and three snowmelt-dominated basins (Animas River at Durango, Colorado: East Fork of the Carson River near Gardnerville, Nevada: and Cle Elum River near Roslyn, Washington). Hydrologic model output forced with measured-station data were used as "truth" to focus attention on the hydrologic effects of errors in the MRF forecasts. Eight-day streamflow forecasts produced using the MOS-corrected MRF output as input (MOS) were compared with those produced using the climatic Ensemble Streamflow Prediction (ESP) technique. MOS-based streamflow forecasts showed increased skill in the snowmelt-dominated river basins, where daily variations in streamflow are strongly forced by temperature. In contrast, the skill of MOS forecasts in the rainfall-dominated basin (the Alapaha River) were equivalent to the skill of the ESP forecasts. Further improvements in streamflow forecasts require more accurate local-scale forecasts of precipitation and temperature, more accurate specification of basin initial conditions, and more accurate model simulations of streamflow. ?? 2004 American Meteorological Society.

Development of extended WRF variational data assimilation system (WRFDA) for WRF non-hydrostatic mesoscale model

NASA Astrophysics Data System (ADS)

Pattanayak, Sujata; Mohanty, U. C.

2018-06-01

The paper intends to present the development of the extended weather research forecasting data assimilation (WRFDA) system in the framework of the non-hydrostatic mesoscale model core of weather research forecasting system (WRF-NMM), as an imperative aspect of numerical modeling studies. Though originally the WRFDA provides improved initial conditions for advanced research WRF, we have successfully developed a unified WRFDA utility that can be used by the WRF-NMM core, as well. After critical evaluation, it has been strategized to develop a code to merge WRFDA framework and WRF-NMM output. In this paper, we have provided a few selected implementations and initial results through single observation test, and background error statistics like eigenvalues, eigenvector and length scale among others, which showcase the successful development of extended WRFDA code for WRF-NMM model. Furthermore, the extended WRFDA system is applied for the forecast of three severe cyclonic storms: Nargis (27 April-3 May 2008), Aila (23-26 May 2009) and Jal (4-8 November 2010) formed over the Bay of Bengal. Model results are compared and contrasted within the analysis fields and later on with high-resolution model forecasts. The mean initial position error is reduced by 33% with WRFDA as compared to GFS analysis. The vector displacement errors in track forecast are reduced by 33, 31, 30 and 20% to 24, 48, 72 and 96 hr forecasts respectively, in data assimilation experiments as compared to control run. The model diagnostics indicates successful implementation of WRFDA within the WRF-NMM system.

A Real-time 1/16° Global Ocean Nowcast/Forecast System

NASA Astrophysics Data System (ADS)

Shriver, J. F.; Rhodes, R. C.; Hurlburt, H. E.; Wallcraft, A. J.; Metzger, E. J.; Smedstad, O. M.; Kara, A. B.

2001-05-01

A 1/16° eddy-resolving global ocean prediction system that uses the NRL Layered Ocean Model (NLOM) has been transitioned to the Naval Oceanographic Office (NAVO), Stennis Space Center, MS. The system gives a real time view of the ocean down to the 50-100 mile scale of ocean eddies and the meandering of ocean currents and fronts, a view with unprecedented resolution and clarity, and demonstrated forecast skill for a month or more for many ocean features. It has been running in real time at NAVO since 19 Oct 2000 with assimilation of real-time altimeter sea surface height (SSH) data (currently ERS-2, GFO and TOPEX/POSEIDON) and sea surface temperature (SST). The model is updated daily and 4-day forecasts are made daily. 30-day forecasts are made once a week. Nowcasts and forecasts using this model are viewable on the web, including SSH, SST and 30-day forecast verification statistics for many zoom regions. The NRL web address is http://www7320.nrlssc.navy.mil/global_nlom/index.html. The NAVO web address is: http://www.navo.navy.mil. Click on "Operational Products", then "Product Search Form", then "Product Type View", then select "Model Navy Layered Ocean Model" and a region and click on "Submit Query". This system is used at NAVO for ocean front and eddy analyses and predictions and to provide accurate sea surface height for use in computing synthetic temperature and salinity profiles, among other applications.

Spatial occupancy models for predicting metapopulation dynamics and viability following reintroduction

USGS Publications Warehouse

Chandler, Richard B.; Muths, Erin L.; Sigafus, Brent H.; Schwalbe, Cecil R.; Jarchow, Christopher J.; Hossack, Blake R.

2015-01-01

Synthesis and applications. This work demonstrates how spatio-temporal statistical models based on ecological theory can be applied to forecast the outcomes of conservation actions such as reintroduction. Our spatial occupancy model should be particularly useful when management agencies lack the funds to collect intensive individual-level data.

Statistical evaluation of forecasts

NASA Astrophysics Data System (ADS)

Mader, Malenka; Mader, Wolfgang; Gluckman, Bruce J.; Timmer, Jens; Schelter, Björn

2014-08-01

Reliable forecasts of extreme but rare events, such as earthquakes, financial crashes, and epileptic seizures, would render interventions and precautions possible. Therefore, forecasting methods have been developed which intend to raise an alarm if an extreme event is about to occur. In order to statistically validate the performance of a prediction system, it must be compared to the performance of a random predictor, which raises alarms independent of the events. Such a random predictor can be obtained by bootstrapping or analytically. We propose an analytic statistical framework which, in contrast to conventional methods, allows for validating independently the sensitivity and specificity of a forecasting method. Moreover, our method accounts for the periods during which an event has to remain absent or occur after a respective forecast.

Identifying trends in climate: an application to the cenozoic

NASA Astrophysics Data System (ADS)

Richards, Gordon R.

1998-05-01

The recent literature on trending in climate has raised several issues, whether trends should be modeled as deterministic or stochastic, whether trends are nonlinear, and the relative merits of statistical models versus models based on physics. This article models trending since the late Cretaceous. This 68 million-year interval is selected because the reliability of tests for trending is critically dependent on the length of time spanned by the data. Two main hypotheses are tested, that the trend has been caused primarily by CO2 forcing, and that it reflects a variety of forcing factors which can be approximated by statistical methods. The CO2 data is obtained from model simulations. Several widely-used statistical models are found to be inadequate. ARIMA methods parameterize too much of the short-term variation, and do not identify low frequency movements. Further, the unit root in the ARIMA process does not predict the long-term path of temperature. Spectral methods also have little ability to predict temperature at long horizons. Instead, the statistical trend is estimated using a nonlinear smoothing filter. Both of these paradigms make it possible to model climate as a cointegrated process, in which temperature can wander quite far from the trend path in the intermediate term, but converges back over longer horizons. Comparing the forecasting properties of the two trend models demonstrates that the optimal forecasting model includes CO2 forcing and a parametric representation of the nonlinear variability in climate.

Impact of scatterometer wind (ASCAT-A/B) data assimilation on semi real-time forecast system at KIAPS

NASA Astrophysics Data System (ADS)

Han, H. J.; Kang, J. H.

2016-12-01

Since Jul. 2015, KIAPS (Korea Institute of Atmospheric Prediction Systems) has been performing the semi real-time forecast system to assess the performance of their forecast system as a NWP model. KPOP (KIAPS Protocol for Observation Processing) is a part of KIAPS data assimilation system and has been performing well in KIAPS semi real-time forecast system. In this study, due to the fact that KPOP would be able to treat the scatterometer wind data, we analyze the effect of scatterometer wind (ASCAT-A/B) on KIAPS semi real-time forecast system. O-B global distribution and statistics of scatterometer wind give use two information which are the difference between background field and observation is not too large and KPOP processed the scatterometer wind data well. The changes of analysis increment because of O-B global distribution appear remarkably at the bottom of atmospheric field. It also shows that scatterometer wind data cover wide ocean where data would be able to short. Performance of scatterometer wind data can be checked through the vertical error reduction against IFS between background and analysis field and vertical statistics of O-A. By these analysis result, we can notice that scatterometer wind data will influence the positive effect on lower level performance of semi real-time forecast system at KIAPS. After, long-term result based on effect of scatterometer wind data will be analyzed.

An Extended Objective Evaluation of the 29-km Eta Model for Weather Support to the United States Space Program

NASA Technical Reports Server (NTRS)

Nutter, Paul; Manobianco, John

1998-01-01

This report describes the Applied Meteorology Unit's objective verification of the National Centers for Environmental Prediction 29-km eta model during separate warm and cool season periods from May 1996 through January 1998. The verification of surface and upper-air point forecasts was performed at three selected stations important for 45th Weather Squadron, Spaceflight Meteorology Group, and National Weather Service, Melbourne operational weather concerns. The statistical evaluation identified model biases that may result from inadequate parameterization of physical processes. Since model biases are relatively small compared to the random error component, most of the total model error results from day-to-day variability in the forecasts and/or observations. To some extent, these nonsystematic errors reflect the variability in point observations that sample spatial and temporal scales of atmospheric phenomena that cannot be resolved by the model. On average, Meso-Eta point forecasts provide useful guidance for predicting the evolution of the larger scale environment. A more substantial challenge facing model users in real time is the discrimination of nonsystematic errors that tend to inflate the total forecast error. It is important that model users maintain awareness of ongoing model changes. Such changes are likely to modify the basic error characteristics, particularly near the surface.

A comparative study of artificial neural network, adaptive neuro fuzzy inference system and support vector machine for forecasting river flow in the semiarid mountain region

NASA Astrophysics Data System (ADS)

He, Zhibin; Wen, Xiaohu; Liu, Hu; Du, Jun

2014-02-01

Data driven models are very useful for river flow forecasting when the underlying physical relationships are not fully understand, but it is not clear whether these data driven models still have a good performance in the small river basin of semiarid mountain regions where have complicated topography. In this study, the potential of three different data driven methods, artificial neural network (ANN), adaptive neuro fuzzy inference system (ANFIS) and support vector machine (SVM) were used for forecasting river flow in the semiarid mountain region, northwestern China. The models analyzed different combinations of antecedent river flow values and the appropriate input vector has been selected based on the analysis of residuals. The performance of the ANN, ANFIS and SVM models in training and validation sets are compared with the observed data. The model which consists of three antecedent values of flow has been selected as the best fit model for river flow forecasting. To get more accurate evaluation of the results of ANN, ANFIS and SVM models, the four quantitative standard statistical performance evaluation measures, the coefficient of correlation (R), root mean squared error (RMSE), Nash-Sutcliffe efficiency coefficient (NS) and mean absolute relative error (MARE), were employed to evaluate the performances of various models developed. The results indicate that the performance obtained by ANN, ANFIS and SVM in terms of different evaluation criteria during the training and validation period does not vary substantially; the performance of the ANN, ANFIS and SVM models in river flow forecasting was satisfactory. A detailed comparison of the overall performance indicated that the SVM model performed better than ANN and ANFIS in river flow forecasting for the validation data sets. The results also suggest that ANN, ANFIS and SVM method can be successfully applied to establish river flow with complicated topography forecasting models in the semiarid mountain regions.

Seamless hydrological predictions for a monsoon driven catchment in North-East India

NASA Astrophysics Data System (ADS)

Köhn, Lisei; Bürger, Gerd; Bronstert, Axel

2016-04-01

Improving hydrological forecasting systems on different time scales is interesting and challenging with regards to humanitarian as well as scientific aspects. In meteorological research, short-, medium-, and long-term forecasts are now being merged to form a system of seamless weather and climate predictions. Coupling of these meteorological forecasts with a hydrological model leads to seamless predictions of streamflow, ranging from one day to a season. While there are big efforts made to analyse the uncertainties of probabilistic streamflow forecasts, knowledge of the single uncertainty contributions from meteorological and hydrological modeling is still limited. The overarching goal of this project is to gain knowledge in this subject by decomposing and quantifying the overall predictive uncertainty into its single factors for the entire seamless forecast horizon. Our study area is the Mahanadi River Basin in North-East India, which is prone to severe floods and droughts. Improved streamflow forecasts on different time scales would contribute to early flood warning as well as better water management operations in the agricultural sector. Because of strong inter-annual monsoon variations in this region, which are, unlike the mid-latitudes, partly predictable from long-term atmospheric-oceanic oscillations, the Mahanadi catchment represents an ideal study site. Regionalized precipitation forecasts are obtained by applying the method of expanded downscaling to the ensemble prediction systems of ECMWF and NCEP. The semi-distributed hydrological model HYPSO-RR, which was developed in the Eco-Hydrological Simulation Environment ECHSE, is set up for several sub-catchments of the Mahanadi River Basin. The model is calibrated automatically using the Dynamically Dimensioned Search algorithm, with a modified Nash-Sutcliff efficiency as objective function. Meteorological uncertainty is estimated from the existing ensemble simulations, while the hydrological uncertainty is derived from a statistical post-processor. After running the hydrological model with the precipitation forecasts and applying the hydrological post-processor, the predictive uncertainty of the streamflow forecast can be analysed. The decomposition of total uncertainty is done using a two-way analysis of variance. In this contribution we present the model set-up and the first results of our hydrological forecasts with up to a 180 days lead time, which are derived by using 15 downscaled members of the ECMWF multi-model seasonal forecast ensemble as model input.

Decomposition of Sources of Errors in Seasonal Streamflow Forecasting over the U.S. Sunbelt

NASA Technical Reports Server (NTRS)

Mazrooei, Amirhossein; Sinah, Tusshar; Sankarasubramanian, A.; Kumar, Sujay V.; Peters-Lidard, Christa D.

2015-01-01

Seasonal streamflow forecasts, contingent on climate information, can be utilized to ensure water supply for multiple uses including municipal demands, hydroelectric power generation, and for planning agricultural operations. However, uncertainties in the streamflow forecasts pose significant challenges in their utilization in real-time operations. In this study, we systematically decompose various sources of errors in developing seasonal streamflow forecasts from two Land Surface Models (LSMs) (Noah3.2 and CLM2), which are forced with downscaled and disaggregated climate forecasts. In particular, the study quantifies the relative contributions of the sources of errors from LSMs, climate forecasts, and downscaling/disaggregation techniques in developing seasonal streamflow forecast. For this purpose, three month ahead seasonal precipitation forecasts from the ECHAM4.5 general circulation model (GCM) were statistically downscaled from 2.8deg to 1/8deg spatial resolution using principal component regression (PCR) and then temporally disaggregated from monthly to daily time step using kernel-nearest neighbor (K-NN) approach. For other climatic forcings, excluding precipitation, we considered the North American Land Data Assimilation System version 2 (NLDAS-2) hourly climatology over the years 1979 to 2010. Then the selected LSMs were forced with precipitation forecasts and NLDAS-2 hourly climatology to develop retrospective seasonal streamflow forecasts over a period of 20 years (1991-2010). Finally, the performance of LSMs in forecasting streamflow under different schemes was analyzed to quantify the relative contribution of various sources of errors in developing seasonal streamflow forecast. Our results indicate that the most dominant source of errors during winter and fall seasons is the errors due to ECHAM4.5 precipitation forecasts, while temporal disaggregation scheme contributes to maximum errors during summer season.

Forecasting Natural Rubber Price In Malaysia Using Arima

NASA Astrophysics Data System (ADS)

Zahari, Fatin Z.; Khalid, Kamil; Roslan, Rozaini; Sufahani, Suliadi; Mohamad, Mahathir; Saifullah Rusiman, Mohd; Ali, Maselan

2018-04-01

This paper contains introduction, materials and methods, results and discussions, conclusions and references. Based on the title mentioned, high volatility of the price of natural rubber nowadays will give the significant risk to the producers, traders, consumers, and others parties involved in the production of natural rubber. To help them in making decisions, forecasting is needed to predict the price of natural rubber. The main objective of the research is to forecast the upcoming price of natural rubber by using the reliable statistical method. The data are gathered from Malaysia Rubber Board which the data are from January 2000 until December 2015. In this research, average monthly price of Standard Malaysia Rubber 20 (SMR20) will be forecast by using Box-Jenkins approach. Time series plot is used to determine the pattern of the data. The data have trend pattern which indicates the data is non-stationary data and the data need to be transformed. By using the Box-Jenkins method, the best fit model for the time series data is ARIMA (1, 1, 0) which this model satisfy all the criteria needed. Hence, ARIMA (1, 1, 0) is the best fitted model and the model will be used to forecast the average monthly price of Standard Malaysia Rubber 20 (SMR20) for twelve months ahead.

Forecasting obesity prevalence in Korean adults for the years 2020 and 2030 by the analysis of contributing factors.

PubMed

Baik, Inkyung

2018-06-01

There are few studies that forecast the future prevalence of obesity based on the predicted prevalence model including contributing factors. The present study aimed to identify factors associated with obesity and construct forecasting models including significant contributing factors to estimate the 2020 and 2030 prevalence of obesity and abdominal obesity. Panel data from the Korea National Health and Nutrition Examination Survey and national statistics from the Korean Statistical Information Service were used for the analysis. The study subjects were 17,685 male and 24,899 female adults aged 19 years or older. The outcome variables were the prevalence of obesity (body mass index ≥ 25 kg/m 2 ) and abdominal obesity (waist circumference ≥ 90 cm for men and ≥ 85 cm for women). Stepwise logistic regression analysis was used to select significant variables from potential exposures. The survey year, age, marital status, job status, income status, smoking, alcohol consumption, sleep duration, psychological factors, dietary intake, and fertility rate were found to contribute to the prevalence of obesity and abdominal obesity. Based on the forecasting models including these variables, the 2020 and 2030 estimates for obesity prevalence were 47% and 62% for men and 32% and 37% for women, respectively. The present study suggested an increased prevalence of obesity and abdominal obesity in 2020 and 2030. Lifestyle factors were found to be significantly associated with the increasing trend in obesity prevalence and, therefore, they may require modification to prevent the rising trend.

What Fraction of Global Fire Activity Can Be Forecast Using Sea Surface Temperatures?

NASA Astrophysics Data System (ADS)

Chen, Y.; Randerson, J. T.; Morton, D. C.; Andela, N.; Giglio, L.

2015-12-01

Variations in sea surface temperatures (SSTs) can influence climate dynamics in local and remote land areas, and thus influence fire-climate interactions that govern burned area. SST information has been recently used in statistical models to create seasonal outlooks of fire season severity in South America and as the initial condition for dynamical model predictions of fire activity in Indonesia. However, the degree to which large-scale ocean-atmosphere interactions can influence burned area in other continental regions has not been systematically explored. Here we quantified the amount of global burned area that can be predicted using SSTs in 14 different oceans regions as statistical predictors. We first examined lagged correlations between GFED4s burned area and the 14 ocean climate indices (OCIs) individually. The maximum correlations from different OCIs were used to construct a global map of fire predictability. About half of the global burned area can be forecast by this approach 3 months before the peak burning month (with a Pearson's r of 0.5 or higher), with the highest levels of predictability in Central America and Equatorial Asia. Several hotspots of predictability were identified using k-means cluster analysis. Within these regions, we tested the improvements of the forecast by using two OCIs from different oceans. Our forecast models were based on near-real-time SST data and may therefore support the development of new seasonal outlooks for fire activity that can aid the sustainable management of these fire-prone ecosystems.

High-Resolution WRF Forecasts of Lightning Threat

NASA Technical Reports Server (NTRS)

Goodman, S. J.; McCaul, E. W., Jr.; LaCasse, K.

2007-01-01

Tropical Rainfall Measuring Mission (TRMM)lightning and precipitation observations have confirmed the existence of a robust relationship between lightning flash rates and the amount of large precipitating ice hydrometeors in storms. This relationship is exploited, in conjunction with the capabilities of the Weather Research and Forecast (WRF) model, to forecast the threat of lightning from convective storms using the output fields from the model forecasts. The simulated vertical flux of graupel at -15C is used in this study as a proxy for charge separation processes and their associated lightning risk. Initial experiments using 6-h simulations are conducted for a number of case studies for which three-dimensional lightning validation data from the North Alabama Lightning Mapping Array are available. The WRF has been initialized on a 2 km grid using Eta boundary conditions, Doppler radar radial velocity and reflectivity fields, and METAR and ACARS data. An array of subjective and objective statistical metrics is employed to document the utility of the WRF forecasts. The simulation results are also compared to other more traditional means of forecasting convective storms, such as those based on inspection of the convective available potential energy field.

Development of a short-term irradiance prediction system using post-processing tools on WRF-ARW meteorological forecasts in Spain

NASA Astrophysics Data System (ADS)

Rincón, A.; Jorba, O.; Baldasano, J. M.

2010-09-01

The increased contribution of solar energy in power generation sources requires an accurate estimation of surface solar irradiance conditioned by geographical, temporal and meteorological conditions. The knowledge of the variability of these factors is essential to estimate the expected energy production and therefore help stabilizing the electricity grid and increase the reliability of available solar energy. The use of numerical meteorological models in combination with statistical post-processing tools may have the potential to satisfy the requirements for short-term forecasting of solar irradiance for up to several days ahead and its application in solar devices. In this contribution, we present an assessment of a short-term irradiance prediction system based on the WRF-ARW mesoscale meteorological model (Skamarock et al., 2005) and several post-processing tools in order to improve the overall skills of the system in an annual simulation of the year 2004 in Spain. The WRF-ARW model is applied with 4 km x 4 km horizontal resolution and 38 vertical layers over the Iberian Peninsula. The hourly model irradiance is evaluated against more than 90 surface stations. The stations are used to assess the temporal and spatial fluctuations and trends of the system evaluating three different post-processes: Model Output Statistics technique (MOS; Glahn and Lowry, 1972), Recursive statistical method (REC; Boi, 2004) and Kalman Filter Predictor (KFP, Bozic, 1994; Roeger et al., 2003). A first evaluation of the system without post-processing tools shows an overestimation of the surface irradiance, due to the lack of atmospheric absorbers attenuation different than clouds not included in the meteorological model. This produces an annual BIAS of 16 W m-2 h-1, annual RMSE of 106 W m-2 h-1 and annual NMAE of 42%. The largest errors are observed in spring and summer, reaching RMSE of 350 W m-2 h-1. Results using Kalman Filter Predictor show a reduction of 8% of RMSE, 83% of BIAS, and NMAE decreases down to 32%. The REC method shows a reduction of 6% of RMSE, 79% of BIAS, and NMAE decreases down to 28%. When comparing stations at different altitudes, the overestimation is enhanced at coastal stations (less than 200m) up to 900 W m-2 h-1. The results allow us to analyze strengths and drawbacks of the irradiance prediction system and its application in the estimation of energy production from photovoltaic system cells. References Boi, P.: A statistical method for forecasting extreme daily temperatures using ECMWF 2-m temperatures and ground station measurements, Meteorol. Appl., 11, 245-251, 2004. Bozic, S.: Digital and Kalman filtering, John Wiley, Hoboken, New Jersey, 2nd edn., 1994. Glahn, H. and Lowry, D.: The use of Model Output Statistics (MOS) in Objective Weather Forecasting, Applied Meteorology, 11, 1203-1211, 1972. Roeger, C., Stull, R., McClung, D., Hacker, J., Deng, X., and Modzelewski, H.: Verification of Mesoscale Numerical Weather Forecasts in Mountainous Terrain for Application to Avalanche Prediction, Weather and forecasting, 18, 1140-1160, 2003. Skamarock, W., Klemp, J., Dudhia, J., Gill, D., Barker, D. M., Wang, W., and Powers, J. G.: A Description of the Advanced Research WRF Version 2, Tech. Rep. NCAR/TN-468+STR, NCAR Technical note, 2005.

Forecasting methodologies for Ganoderma spore concentration using combined statistical approaches and model evaluations

NASA Astrophysics Data System (ADS)

Sadyś, Magdalena; Skjøth, Carsten Ambelas; Kennedy, Roy

2016-04-01

High concentration levels of Ganoderma spp. spores were observed in Worcester, UK, during 2006-2010. These basidiospores are known to cause sensitization due to the allergen content and their small dimensions. This enables them to penetrate the lower part of the respiratory tract in humans. Establishment of a link between occurring symptoms of sensitization to Ganoderma spp. and other basidiospores is challenging due to lack of information regarding spore concentration in the air. Hence, aerobiological monitoring should be conducted, and if possible extended with the construction of forecast models. Daily mean concentration of allergenic Ganoderma spp. spores in the atmosphere of Worcester was measured using 7-day volumetric spore sampler through five consecutive years. The relationships between the presence of spores in the air and the weather parameters were examined. Forecast models were constructed for Ganoderma spp. spores using advanced statistical techniques, i.e. multivariate regression trees and artificial neural networks. Dew point temperature along with maximum temperature was the most important factor influencing the presence of spores in the air of Worcester. Based on these two major factors and several others of lesser importance, thresholds for certain levels of fungal spore concentration, i.e. low (0-49 s m-3), moderate (50-99 s m-3), high (100-149 s m-3) and very high (150 < n s m-3), could be designated. Despite some deviation in results obtained by artificial neural networks, authors have achieved a forecasting model, which was accurate (correlation between observed and predicted values varied from r s = 0.57 to r s = 0.68).

Forecasting methodologies for Ganoderma spore concentration using combined statistical approaches and model evaluations.

PubMed

Sadyś, Magdalena; Skjøth, Carsten Ambelas; Kennedy, Roy

2016-04-01

High concentration levels of Ganoderma spp. spores were observed in Worcester, UK, during 2006-2010. These basidiospores are known to cause sensitization due to the allergen content and their small dimensions. This enables them to penetrate the lower part of the respiratory tract in humans. Establishment of a link between occurring symptoms of sensitization to Ganoderma spp. and other basidiospores is challenging due to lack of information regarding spore concentration in the air. Hence, aerobiological monitoring should be conducted, and if possible extended with the construction of forecast models. Daily mean concentration of allergenic Ganoderma spp. spores in the atmosphere of Worcester was measured using 7-day volumetric spore sampler through five consecutive years. The relationships between the presence of spores in the air and the weather parameters were examined. Forecast models were constructed for Ganoderma spp. spores using advanced statistical techniques, i.e. multivariate regression trees and artificial neural networks. Dew point temperature along with maximum temperature was the most important factor influencing the presence of spores in the air of Worcester. Based on these two major factors and several others of lesser importance, thresholds for certain levels of fungal spore concentration, i.e. low (0-49 s m(-3)), moderate (50-99 s m(-3)), high (100-149 s m(-3)) and very high (150 < n s m(-3)), could be designated. Despite some deviation in results obtained by artificial neural networks, authors have achieved a forecasting model, which was accurate (correlation between observed and predicted values varied from r s = 0.57 to r s = 0.68).

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

Mendes, J.; Bessa, R.J.; Keko, H.

Wind power forecasting (WPF) provides important inputs to power system operators and electricity market participants. It is therefore not surprising that WPF has attracted increasing interest within the electric power industry. In this report, we document our research on improving statistical WPF algorithms for point, uncertainty, and ramp forecasting. Below, we provide a brief introduction to the research presented in the following chapters. For a detailed overview of the state-of-the-art in wind power forecasting, we refer to [1]. Our related work on the application of WPF in operational decisions is documented in [2]. Point forecasts of wind power are highlymore » dependent on the training criteria used in the statistical algorithms that are used to convert weather forecasts and observational data to a power forecast. In Chapter 2, we explore the application of information theoretic learning (ITL) as opposed to the classical minimum square error (MSE) criterion for point forecasting. In contrast to the MSE criterion, ITL criteria do not assume a Gaussian distribution of the forecasting errors. We investigate to what extent ITL criteria yield better results. In addition, we analyze time-adaptive training algorithms and how they enable WPF algorithms to cope with non-stationary data and, thus, to adapt to new situations without requiring additional offline training of the model. We test the new point forecasting algorithms on two wind farms located in the U.S. Midwest. Although there have been advancements in deterministic WPF, a single-valued forecast cannot provide information on the dispersion of observations around the predicted value. We argue that it is essential to generate, together with (or as an alternative to) point forecasts, a representation of the wind power uncertainty. Wind power uncertainty representation can take the form of probabilistic forecasts (e.g., probability density function, quantiles), risk indices (e.g., prediction risk index) or scenarios (with spatial and/or temporal dependence). Statistical approaches to uncertainty forecasting basically consist of estimating the uncertainty based on observed forecasting errors. Quantile regression (QR) is currently a commonly used approach in uncertainty forecasting. In Chapter 3, we propose new statistical approaches to the uncertainty estimation problem by employing kernel density forecast (KDF) methods. We use two estimators in both offline and time-adaptive modes, namely, the Nadaraya-Watson (NW) and Quantilecopula (QC) estimators. We conduct detailed tests of the new approaches using QR as a benchmark. One of the major issues in wind power generation are sudden and large changes of wind power output over a short period of time, namely ramping events. In Chapter 4, we perform a comparative study of existing definitions and methodologies for ramp forecasting. We also introduce a new probabilistic method for ramp event detection. The method starts with a stochastic algorithm that generates wind power scenarios, which are passed through a high-pass filter for ramp detection and estimation of the likelihood of ramp events to happen. The report is organized as follows: Chapter 2 presents the results of the application of ITL training criteria to deterministic WPF; Chapter 3 reports the study on probabilistic WPF, including new contributions to wind power uncertainty forecasting; Chapter 4 presents a new method to predict and visualize ramp events, comparing it with state-of-the-art methodologies; Chapter 5 briefly summarizes the main findings and contributions of this report.« less

Data sensitivity in a hybrid STEP/Coulomb model for aftershock forecasting

NASA Astrophysics Data System (ADS)

Steacy, S.; Jimenez Lloret, A.; Gerstenberger, M.

2014-12-01

Operational earthquake forecasting is rapidly becoming a 'hot topic' as civil protection authorities seek quantitative information on likely near future earthquake distributions during seismic crises. At present, most of the models in public domain are statistical and use information about past and present seismicity as well as b-value and Omori's law to forecast future rates. A limited number of researchers, however, are developing hybrid models which add spatial constraints from Coulomb stress modeling to existing statistical approaches. Steacy et al. (2013), for instance, recently tested a model that combines Coulomb stress patterns with the STEP (short-term earthquake probability) approach against seismicity observed during the 2010-2012 Canterbury earthquake sequence. They found that the new model performed at least as well as, and often better than, STEP when tested against retrospective data but that STEP was generally better in pseudo-prospective tests that involved data actually available within the first 10 days of each event of interest. They suggested that the major reason for this discrepancy was uncertainty in the slip models and, in particular, in the geometries of the faults involved in each complex major event. Here we test this hypothesis by developing a number of retrospective forecasts for the Landers earthquake using hypothetical slip distributions developed by Steacy et al. (2004) to investigate the sensitivity of Coulomb stress models to fault geometry and earthquake slip, and we also examine how the choice of receiver plane geometry affects the results. We find that the results are strongly sensitive to the slip models and moderately sensitive to the choice of receiver orientation. We further find that comparison of the stress fields (resulting from the slip models) with the location of events in the learning period provides advance information on whether or not a particular hybrid model will perform better than STEP.

Natural Gas Prices Forecast Comparison--AEO vs. Natural Gas Markets

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

Wong-Parodi, Gabrielle; Lekov, Alex; Dale, Larry

This paper evaluates the accuracy of two methods to forecast natural gas prices: using the Energy Information Administration's ''Annual Energy Outlook'' forecasted price (AEO) and the ''Henry Hub'' compared to U.S. Wellhead futures price. A statistical analysis is performed to determine the relative accuracy of the two measures in the recent past. A statistical analysis suggests that the Henry Hub futures price provides a more accurate average forecast of natural gas prices than the AEO. For example, the Henry Hub futures price underestimated the natural gas price by 35 cents per thousand cubic feet (11.5 percent) between 1996 and 2003more » and the AEO underestimated by 71 cents per thousand cubic feet (23.4 percent). Upon closer inspection, a liner regression analysis reveals that two distinct time periods exist, the period between 1996 to 1999 and the period between 2000 to 2003. For the time period between 1996 to 1999, AEO showed a weak negative correlation (R-square = 0.19) between forecast price by actual U.S. Wellhead natural gas price versus the Henry Hub with a weak positive correlation (R-square = 0.20) between forecasted price and U.S. Wellhead natural gas price. During the time period between 2000 to 2003, AEO shows a moderate positive correlation (R-square = 0.37) between forecasted natural gas price and U.S. Wellhead natural gas price versus the Henry Hub that show a moderate positive correlation (R-square = 0.36) between forecast price and U.S. Wellhead natural gas price. These results suggest that agencies forecasting natural gas prices should consider incorporating the Henry Hub natural gas futures price into their forecasting models along with the AEO forecast. Our analysis is very preliminary and is based on a very small data set. Naturally the results of the analysis may change, as more data is made available.« less

Temporal Characteristics of Electron Flux Events at Geosynchronous Orbit

NASA Astrophysics Data System (ADS)

Olson, D. K.; Larsen, B.; Henderson, M. G.

2017-12-01

Geosynchronous satellites such as the LANL-GEO fleet are exposed to hazardous conditions when they encounter regions of hot, intense plasma such as that from the plasma sheet. These conditions can lead to the build-up of charge on the surface of a spacecraft, with undesired, and often dangerous, side effects. Observation of electron flux levels at geosynchronous orbit (GEO) with multiple satellites provides a unique view of plasma sheet access to that region. Flux "events", or periods when fluxes are elevated continuously above the LANL-GEO spacecraft charging threshold, can be characterized by duration in two dimensions: a spatial dimension of local time, describing the duration of an event from the perspective of a single spacecraft, and a temporal dimension describing the duration in time in which high energy plasma sheet particles have access to geosynchronous orbit. We examine the statistical properties of the temporal duration of 8 keV electron flux events at geosynchronous orbit over a twelve-year period. These results, coupled with the spatial duration characteristics, provide the key information needed to formulate a statistical model for forecasting the electron flux conditions at GEO that are correlated with LANL-GEO surface charging. Forecasting models are an essential component to understanding space weather and mitigating the dangers of surface charging on our satellites. We also examine the correlation of flux event durations with solar wind parameters and geomagnetic indices, identifying the data needed to improve upon a statistical forecasting model

[Development of forecasting models for fatal road traffic injuries].

PubMed

Tan, Aichun; Tian, Danping; Huang, Yuanxiu; Gao, Lin; Deng, Xin; Li, Li; He, Qiong; Chen, Tianmu; Hu, Guoqing; Wu, Jing

2014-02-01

To develop the forecasting models for fatal road traffic injuries and to provide evidence for predicting the future trends on road traffic injuries. Data on the mortality of road traffic injury including factors as gender and age in different countries, were obtained from the World Health Organization Mortality Database. Other information on GDP per capita, urbanization, motorization and education were collected from online resources of World Bank, WHO, the United Nations Population Division and other agencies. We fitted logarithmic models of road traffic injury mortality by gender and age group, including predictors of GDP per capita, urbanization, motorization and education. Sex- and age-specific forecasting models developed by WHO that including GDP per capita, education and time etc. were also fitted. Coefficient of determination(R(2)) was used to compare the performance between our modes and WHO models. 2 626 sets of data were collected from 153 countries/regions for both genders, between 1965 and 2010. The forecasting models of road traffic injury mortality based on GDP per capita, motorization, urbanization and education appeared to be statistically significant(P < 0.001), and the coefficients of determination for males at the age groups of 0-4, 5-14, 15-24, 25-34, 35-44, 45-54, 55-64, 65+ were 22.7% , 31.1%, 51.8%, 52.3%, 44.9%, 41.8%, 40.1%, 25.5%, respectively while the coefficients for these age groups in women were 22.9%, 32.6%, 51.1%, 49.3%, 41.3%, 35.9%, 30.7%, 20.1%, respectively. The WHO models that were based on the GDP per capita, education and time variables were statistically significant (P < 0.001)and the coefficients of determination were 14.9% , 22.0%, 31.5%, 33.1% , 30.7%, 28.5%, 27.7% and 17.8% for males, but 14.1%, 20.6%, 30.4%, 31.8%, 26.7%, 24.3%, 17.3% and 8.8% for females, respectively. The forecasting models that we developed seemed to be better than those developed by WHO.

Short-term load forecasting of power system

NASA Astrophysics Data System (ADS)

Xu, Xiaobin

2017-05-01

In order to ensure the scientific nature of optimization about power system, it is necessary to improve the load forecasting accuracy. Power system load forecasting is based on accurate statistical data and survey data, starting from the history and current situation of electricity consumption, with a scientific method to predict the future development trend of power load and change the law of science. Short-term load forecasting is the basis of power system operation and analysis, which is of great significance to unit combination, economic dispatch and safety check. Therefore, the load forecasting of the power system is explained in detail in this paper. First, we use the data from 2012 to 2014 to establish the partial least squares model to regression analysis the relationship between daily maximum load, daily minimum load, daily average load and each meteorological factor, and select the highest peak by observing the regression coefficient histogram Day maximum temperature, daily minimum temperature and daily average temperature as the meteorological factors to improve the accuracy of load forecasting indicators. Secondly, in the case of uncertain climate impact, we use the time series model to predict the load data for 2015, respectively, the 2009-2014 load data were sorted out, through the previous six years of the data to forecast the data for this time in 2015. The criterion for the accuracy of the prediction is the average of the standard deviations for the prediction results and average load for the previous six years. Finally, considering the climate effect, we use the BP neural network model to predict the data in 2015, and optimize the forecast results on the basis of the time series model.

Forecasting disease risk for increased epidemic preparedness in public health

NASA Technical Reports Server (NTRS)

Myers, M. F.; Rogers, D. J.; Cox, J.; Flahault, A.; Hay, S. I.

2000-01-01

Emerging infectious diseases pose a growing threat to human populations. Many of the world's epidemic diseases (particularly those transmitted by intermediate hosts) are known to be highly sensitive to long-term changes in climate and short-term fluctuations in the weather. The application of environmental data to the study of disease offers the capability to demonstrate vector-environment relationships and potentially forecast the risk of disease outbreaks or epidemics. Accurate disease forecasting models would markedly improve epidemic prevention and control capabilities. This chapter examines the potential for epidemic forecasting and discusses the issues associated with the development of global networks for surveillance and prediction. Existing global systems for epidemic preparedness focus on disease surveillance using either expert knowledge or statistical modelling of disease activity and thresholds to identify times and areas of risk. Predictive health information systems would use monitored environmental variables, linked to a disease system, to be observed and provide prior information of outbreaks. The components and varieties of forecasting systems are discussed with selected examples, along with issues relating to further development.

Forecasting Disease Risk for Increased Epidemic Preparedness in Public Health

PubMed Central

Myers, M.F.; Rogers, D.J.; Cox, J.; Flahault, A.; Hay, S.I.

2011-01-01

Emerging infectious diseases pose a growing threat to human populations. Many of the world’s epidemic diseases (particularly those transmitted by intermediate hosts) are known to be highly sensitive to long-term changes in climate and short-term fluctuations in the weather. The application of environmental data to the study of disease offers the capability to demonstrate vector–environment relationships and potentially forecast the risk of disease outbreaks or epidemics. Accurate disease forecasting models would markedly improve epidemic prevention and control capabilities. This chapter examines the potential for epidemic forecasting and discusses the issues associated with the development of global networks for surveillance and prediction. Existing global systems for epidemic preparedness focus on disease surveillance using either expert knowledge or statistical modelling of disease activity and thresholds to identify times and areas of risk. Predictive health information systems would use monitored environmental variables, linked to a disease system, to be observed and provide prior information of outbreaks. The components and varieties of forecasting systems are discussed with selected examples, along with issues relating to further development. PMID:10997211

Model documentation report: Transportation sector model of the National Energy Modeling System

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

Not Available

1994-03-01

This report documents the objectives, analytical approach and development of the National Energy Modeling System (NEMS) Transportation Model (TRAN). The report catalogues and describes the model assumptions, computational methodology, parameter estimation techniques, model source code, and forecast results generated by the model. This document serves three purposes. First, it is a reference document providing a detailed description of TRAN for model analysts, users, and the public. Second, this report meets the legal requirements of the Energy Information Administration (EIA) to provide adequate documentation in support of its statistical and forecast reports (Public Law 93-275, 57(b)(1)). Third, it permits continuity inmore » model development by providing documentation from which energy analysts can undertake model enhancements, data updates, and parameter refinements.« less

Predictive data modeling of human type II diabetes related statistics

NASA Astrophysics Data System (ADS)

Jaenisch, Kristina L.; Jaenisch, Holger M.; Handley, James W.; Albritton, Nathaniel G.

2009-04-01

During the course of routine Type II treatment of one of the authors, it was decided to derive predictive analytical Data Models of the daily sampled vital statistics: namely weight, blood pressure, and blood sugar, to determine if the covariance among the observed variables could yield a descriptive equation based model, or better still, a predictive analytical model that could forecast the expected future trend of the variables and possibly eliminate the number of finger stickings required to montior blood sugar levels. The personal history and analysis with resulting models are presented.

Towards an automatic statistical model for seasonal precipitation prediction and its application to Central and South Asian headwater catchments

NASA Astrophysics Data System (ADS)

Gerlitz, Lars; Gafurov, Abror; Apel, Heiko; Unger-Sayesteh, Katy; Vorogushyn, Sergiy; Merz, Bruno

2016-04-01

Statistical climate forecast applications typically utilize a small set of large scale SST or climate indices, such as ENSO, PDO or AMO as predictor variables. If the predictive skill of these large scale modes is insufficient, specific predictor variables such as customized SST patterns are frequently included. Hence statistically based climate forecast models are either based on a fixed number of climate indices (and thus might not consider important predictor variables) or are highly site specific and barely transferable to other regions. With the aim of developing an operational seasonal forecast model, which is easily transferable to any region in the world, we present a generic data mining approach which automatically selects potential predictors from gridded SST observations and reanalysis derived large scale atmospheric circulation patterns and generates robust statistical relationships with posterior precipitation anomalies for user selected target regions. Potential predictor variables are derived by means of a cellwise correlation analysis of precipitation anomalies with gridded global climate variables under consideration of varying lead times. Significantly correlated grid cells are subsequently aggregated to predictor regions by means of a variability based cluster analysis. Finally for every month and lead time, an individual random forest based forecast model is automatically calibrated and evaluated by means of the preliminary generated predictor variables. The model is exemplarily applied and evaluated for selected headwater catchments in Central and South Asia. Particularly the for winter and spring precipitation (which is associated with westerly disturbances in the entire target domain) the model shows solid results with correlation coefficients up to 0.7, although the variability of precipitation rates is highly underestimated. Likewise for the monsoonal precipitation amounts in the South Asian target areas a certain skill of the model could be detected. The skill of the model for the dry summer season in Central Asia and the transition seasons over South Asia is found to be low. A sensitivity analysis by means on well known climate indices reveals the major large scale controlling mechanisms for the seasonal precipitation climate of each target area. For the Central Asian target areas, both, the El Nino Southern Oscillation and the North Atlantic Oscillation are identified as important controlling factors for precipitation totals during moist spring season. Drought conditions are found to be triggered by a warm ENSO phase in combination with a positive phase of the NAO. For the monsoonal summer precipitation amounts over Southern Asia, the model suggests a distinct negative response to El Nino events.

Using Google Flu Trends data in forecasting influenza-like-illness related ED visits in Omaha, Nebraska.

PubMed

Araz, Ozgur M; Bentley, Dan; Muelleman, Robert L

2014-09-01

Emergency department (ED) visits increase during the influenza seasons. It is essential to identify statistically significant correlates in order to develop an accurate forecasting model for ED visits. Forecasting influenza-like-illness (ILI)-related ED visits can significantly help in developing robust resource management strategies at the EDs. We first performed correlation analyses to understand temporal correlations between several predictors of ILI-related ED visits. We used the data available for Douglas County, the biggest county in Nebraska, for Omaha, the biggest city in the state, and for a major hospital in Omaha. The data set included total and positive influenza test results from the hospital (ie, Antigen rapid (Ag) and Respiratory Syncytial Virus Infection (RSV) tests); an Internet-based influenza surveillance system data, that is, Google Flu Trends, for both Nebraska and Omaha; total ED visits in Douglas County attributable to ILI; and ILI surveillance network data for Douglas County and Nebraska as the predictors and data for the hospital's ILI-related ED visits as the dependent variable. We used Seasonal Autoregressive Integrated Moving Average and Holt Winters methods with3 linear regression models to forecast ILI-related ED visits at the hospital and evaluated model performances by comparing the root means square errors (RMSEs). Because of strong positive correlations with ILI-related ED visits between 2008 and 2012, we validated the use of Google Flu Trends data as a predictor in an ED influenza surveillance tool. Of the 5 forecasting models we have tested, linear regression models performed significantly better when Google Flu Trends data were included as a predictor. Regression models including Google Flu Trends data as a predictor variable have lower RMSE, and the lowest is achieved when all other variables are also included in the model in our forecasting experiments for the first 5 weeks of 2013 (with RMSE = 57.61). Google Flu Trends data statistically improve the performance of predicting ILI-related ED visits in Douglas County, and this result can be generalized to other communities. Timely and accurate estimates of ED volume during the influenza season, as well as during pandemic outbreaks, can help hospitals plan their ED resources accordingly and lower their costs by optimizing supplies and staffing and can improve service quality by decreasing ED wait times and overcrowding. Copyright © 2014 Elsevier Inc. All rights reserved.

Forecasting Space Weather-Induced GPS Performance Degradation Using Random Forest

NASA Astrophysics Data System (ADS)

Filjar, R.; Filic, M.; Milinkovic, F.

2017-12-01

Space weather and ionospheric dynamics have a profound effect on positioning performance of the Global Satellite Navigation System (GNSS). However, the quantification of that effect is still the subject of scientific activities around the world. In the latest contribution to the understanding of the space weather and ionospheric effects on satellite-based positioning performance, we conducted a study of several candidates for forecasting method for space weather-induced GPS positioning performance deterioration. First, a 5-days set of experimentally collected data was established, encompassing the space weather and ionospheric activity indices (including: the readings of the Sudden Ionospheric Disturbance (SID) monitors, components of geomagnetic field strength, global Kp index, Dst index, GPS-derived Total Electron Content (TEC) samples, standard deviation of TEC samples, and sunspot number) and observations of GPS positioning error components (northing, easting, and height positioning error) derived from the Adriatic Sea IGS reference stations' RINEX raw pseudorange files in quiet space weather periods. This data set was split into the training and test sub-sets. Then, a selected set of supervised machine learning methods based on Random Forest was applied to the experimentally collected data set in order to establish the appropriate regional (the Adriatic Sea) forecasting models for space weather-induced GPS positioning performance deterioration. The forecasting models were developed in the R/rattle statistical programming environment. The forecasting quality of the regional forecasting models developed was assessed, and the conclusions drawn on the advantages and shortcomings of the regional forecasting models for space weather-caused GNSS positioning performance deterioration.

Application of a Fuzzy Verification Technique for Assessment of the Weather Running Estimate-Nowcast (WRE-N) Model

DTIC Science & Technology

2016-10-01

comes when considering numerous scores and statistics during a preliminary evaluation of the applicability of the fuzzy- verification minimum coverage...The selection of thresholds with which to generate categorical-verification scores and statistics from the application of both traditional and...of statistically significant numbers of cases; the latter presents a challenge of limited application for assessment of the forecast models’ ability

Stream-flow forecasting using extreme learning machines: A case study in a semi-arid region in Iraq

NASA Astrophysics Data System (ADS)

Yaseen, Zaher Mundher; Jaafar, Othman; Deo, Ravinesh C.; Kisi, Ozgur; Adamowski, Jan; Quilty, John; El-Shafie, Ahmed

2016-11-01

Monthly stream-flow forecasting can yield important information for hydrological applications including sustainable design of rural and urban water management systems, optimization of water resource allocations, water use, pricing and water quality assessment, and agriculture and irrigation operations. The motivation for exploring and developing expert predictive models is an ongoing endeavor for hydrological applications. In this study, the potential of a relatively new data-driven method, namely the extreme learning machine (ELM) method, was explored for forecasting monthly stream-flow discharge rates in the Tigris River, Iraq. The ELM algorithm is a single-layer feedforward neural network (SLFNs) which randomly selects the input weights, hidden layer biases and analytically determines the output weights of the SLFNs. Based on the partial autocorrelation functions of historical stream-flow data, a set of five input combinations with lagged stream-flow values are employed to establish the best forecasting model. A comparative investigation is conducted to evaluate the performance of the ELM compared to other data-driven models: support vector regression (SVR) and generalized regression neural network (GRNN). The forecasting metrics defined as the correlation coefficient (r), Nash-Sutcliffe efficiency (ENS), Willmott's Index (WI), root-mean-square error (RMSE) and mean absolute error (MAE) computed between the observed and forecasted stream-flow data are employed to assess the ELM model's effectiveness. The results revealed that the ELM model outperformed the SVR and the GRNN models across a number of statistical measures. In quantitative terms, superiority of ELM over SVR and GRNN models was exhibited by ENS = 0.578, 0.378 and 0.144, r = 0.799, 0.761 and 0.468 and WI = 0.853, 0.802 and 0.689, respectively and the ELM model attained lower RMSE value by approximately 21.3% (relative to SVR) and by approximately 44.7% (relative to GRNN). Based on the findings of this study, several recommendations were suggested for further exploration of the ELM model in hydrological forecasting problems.

On forecasting mortality.

PubMed

Olshansky, S J

1988-01-01

Official forecasts of mortality made by the U.S. Office of the Actuary throughout this century have consistently underestimated observed mortality declines. This is due, in part, to their reliance on the static extrapolation of past trends, an atheoretical statistical method that pays scant attention to the behavioral, medical, and social factors contributing to mortality change. A "multiple cause-delay model" more realistically portrays the effects on mortality of the presence of more favorable risk factors at the population level. Such revised assumptions produce large increases in forecasts of the size of the elderly population, and have a dramatic impact on related estimates of population morbidity, disability, and health care costs.

Operational Earthquake Forecasting of Aftershocks for New England

NASA Astrophysics Data System (ADS)

Ebel, J.; Fadugba, O. I.

2015-12-01

Although the forecasting of mainshocks is not possible, recent research demonstrates that probabilistic forecasts of expected aftershock activity following moderate and strong earthquakes is possible. Previous work has shown that aftershock sequences in intraplate regions behave similarly to those in California, and thus the operational aftershocks forecasting methods that are currently employed in California can be adopted for use in areas of the eastern U.S. such as New England. In our application, immediately after a felt earthquake in New England, a forecast of expected aftershock activity for the next 7 days will be generated based on a generic aftershock activity model. Approximately 24 hours after the mainshock, the parameters of the aftershock model will be updated using the observed aftershock activity observed to that point in time, and a new forecast of expected aftershock activity for the next 7 days will be issued. The forecast will estimate the average number of weak, felt aftershocks and the average expected number of aftershocks based on the aftershock statistics of past New England earthquakes. The forecast also will estimate the probability that an earthquake that is stronger than the mainshock will take place during the next 7 days. The aftershock forecast will specify the expected aftershocks locations as well as the areas over which aftershocks of different magnitudes could be felt. The system will use web pages, email and text messages to distribute the aftershock forecasts. For protracted aftershock sequences, new forecasts will be issued on a regular basis, such as weekly. Initially, the distribution system of the aftershock forecasts will be limited, but later it will be expanded as experience with and confidence in the system grows.

Recent Achievements of the Collaboratory for the Study of Earthquake Predictability

NASA Astrophysics Data System (ADS)

Jackson, D. D.; Liukis, M.; Werner, M. J.; Schorlemmer, D.; Yu, J.; Maechling, P. J.; Zechar, J. D.; Jordan, T. H.

2015-12-01

Maria Liukis, SCEC, USC; Maximilian Werner, University of Bristol; Danijel Schorlemmer, GFZ Potsdam; John Yu, SCEC, USC; Philip Maechling, SCEC, USC; Jeremy Zechar, Swiss Seismological Service, ETH; Thomas H. Jordan, SCEC, USC, and the CSEP Working Group The Collaboratory for the Study of Earthquake Predictability (CSEP) supports a global program to conduct prospective earthquake forecasting experiments. CSEP testing centers are now operational in California, New Zealand, Japan, China, and Europe with 435 models under evaluation. The California testing center, operated by SCEC, has been operational since Sept 1, 2007, and currently hosts 30-minute, 1-day, 3-month, 1-year and 5-year forecasts, both alarm-based and probabilistic, for California, the Western Pacific, and worldwide. We have reduced testing latency, implemented prototype evaluation of M8 forecasts, and are currently developing formats and procedures to evaluate externally-hosted forecasts and predictions. These efforts are related to CSEP support of the USGS program in operational earthquake forecasting and a DHS project to register and test external forecast procedures from experts outside seismology. A retrospective experiment for the 2010-2012 Canterbury earthquake sequence has been completed, and the results indicate that some physics-based and hybrid models outperform purely statistical (e.g., ETAS) models. The experiment also demonstrates the power of the CSEP cyberinfrastructure for retrospective testing. Our current development includes evaluation strategies that increase computational efficiency for high-resolution global experiments, such as the evaluation of the Global Earthquake Activity Rate (GEAR) model. We describe the open-source CSEP software that is available to researchers as they develop their forecast models (http://northridge.usc.edu/trac/csep/wiki/MiniCSEP). We also discuss applications of CSEP infrastructure to geodetic transient detection and how CSEP procedures are being adapted to ground motion prediction experiments.

Impact of AIRS Thermodynamic Profiles on Precipitation Forecasts for Atmospheric River Cases Affecting the Western United States

NASA Technical Reports Server (NTRS)

Zavodsky, Bradley T.; Jedlovec, Gary J.; Blakenship, Clay B.; Wick, Gary A.; Neiman, Paul J.

2013-01-01

This project is a collaborative activity between the NASA Short-term Prediction Research and Transition (SPoRT) Center and the NOAA Hydrometeorology Testbed (HMT) to evaluate a SPoRT Advanced Infrared Sounding Radiometer (AIRS: Aumann et al. 2003) enhanced moisture analysis product. We test the impact of assimilating AIRS temperature and humidity profiles above clouds and in partly cloudy regions, using the three-dimensional variational Gridpoint Statistical Interpolation (GSI) data assimilation (DA) system (Developmental Testbed Center 2012) to produce a new analysis. Forecasts of the Weather Research and Forecasting (WRF) model initialized from the new analysis are compared to control forecasts without the additional AIRS data. We focus on some cases where atmospheric rivers caused heavy precipitation on the US West Coast. We verify the forecasts by comparison with dropsondes and the Cooperative Institute for Research in the Atmosphere (CIRA) Blended Total Precipitable Water product.

Egg production forecasting: Determining efficient modeling approaches.

PubMed

Ahmad, H A

2011-12-01

Several mathematical or statistical and artificial intelligence models were developed to compare egg production forecasts in commercial layers. Initial data for these models were collected from a comparative layer trial on commercial strains conducted at the Poultry Research Farms, Auburn University. Simulated data were produced to represent new scenarios by using means and SD of egg production of the 22 commercial strains. From the simulated data, random examples were generated for neural network training and testing for the weekly egg production prediction from wk 22 to 36. Three neural network architectures-back-propagation-3, Ward-5, and the general regression neural network-were compared for their efficiency to forecast egg production, along with other traditional models. The general regression neural network gave the best-fitting line, which almost overlapped with the commercial egg production data, with an R(2) of 0.71. The general regression neural network-predicted curve was compared with original egg production data, the average curves of white-shelled and brown-shelled strains, linear regression predictions, and the Gompertz nonlinear model. The general regression neural network was superior in all these comparisons and may be the model of choice if the initial overprediction is managed efficiently. In general, neural network models are efficient, are easy to use, require fewer data, and are practical under farm management conditions to forecast egg production.

Predictability of a Coupled Model of ENSO Using Singular Vector Analysis: Optimal Growth and Forecast Skill.

NASA Astrophysics Data System (ADS)

Xue, Yan

The optimal growth and its relationship with the forecast skill of the Zebiak and Cane model are studied using a simple statistical model best fit to the original nonlinear model and local linear tangent models about idealized climatic states (the mean background and ENSO cycles in a long model run), and the actual forecast states, including two sets of runs using two different initialization procedures. The seasonally varying Markov model best fit to a suite of 3-year forecasts in a reduced EOF space (18 EOFs) fits the original nonlinear model reasonably well and has comparable or better forecast skill. The initial error growth in a linear evolution operator A is governed by the eigenvalues of A^{T}A, and the square roots of eigenvalues and eigenvectors of A^{T}A are named singular values and singular vectors. One dominant growing singular vector is found, and the optimal 6 month growth rate is largest for a (boreal) spring start and smallest for a fall start. Most of the variation in the optimal growth rate of the two forecasts is seasonal, attributable to the seasonal variations in the mean background, except that in the cold events it is substantially suppressed. It is found that the mean background (zero anomaly) is the most unstable state, and the "forecast IC states" are more unstable than the "coupled model states". One dominant growing singular vector is found, characterized by north-south and east -west dipoles, convergent winds on the equator in the eastern Pacific and a deepened thermocline in the whole equatorial belt. This singular vector is insensitive to initial time and optimization time, but its final pattern is a strong function of initial states. The ENSO system is inherently unpredictable for the dominant singular vector can amplify 5-fold to 24-fold in 6 months and evolve into the large scales characteristic of ENSO. However, the inherent ENSO predictability is only a secondary factor, while the mismatches between the model and data is a primary factor controlling the current forecast skill.

Forecasting daily source air quality using multivariate statistical analysis and radial basis function networks.

PubMed

Sun, Gang; Hoff, Steven J; Zelle, Brian C; Nelson, Minda A

2008-12-01

It is vital to forecast gas and particle matter concentrations and emission rates (GPCER) from livestock production facilities to assess the impact of airborne pollutants on human health, ecological environment, and global warming. Modeling source air quality is a complex process because of abundant nonlinear interactions between GPCER and other factors. The objective of this study was to introduce statistical methods and radial basis function (RBF) neural network to predict daily source air quality in Iowa swine deep-pit finishing buildings. The results show that four variables (outdoor and indoor temperature, animal units, and ventilation rates) were identified as relative important model inputs using statistical methods. It can be further demonstrated that only two factors, the environment factor and the animal factor, were capable of explaining more than 94% of the total variability after performing principal component analysis. The introduction of fewer uncorrelated variables to the neural network would result in the reduction of the model structure complexity, minimize computation cost, and eliminate model overfitting problems. The obtained results of RBF network prediction were in good agreement with the actual measurements, with values of the correlation coefficient between 0.741 and 0.995 and very low values of systemic performance indexes for all the models. The good results indicated the RBF network could be trained to model these highly nonlinear relationships. Thus, the RBF neural network technology combined with multivariate statistical methods is a promising tool for air pollutant emissions modeling.

Time Series Analysis and Forecasting of Wastewater Inflow into Bandar Tun Razak Sewage Treatment Plant in Selangor, Malaysia

NASA Astrophysics Data System (ADS)

Abunama, Taher; Othman, Faridah

2017-06-01

Analysing the fluctuations of wastewater inflow rates in sewage treatment plants (STPs) is essential to guarantee a sufficient treatment of wastewater before discharging it to the environment. The main objectives of this study are to statistically analyze and forecast the wastewater inflow rates into the Bandar Tun Razak STP in Kuala Lumpur, Malaysia. A time series analysis of three years’ weekly influent data (156weeks) has been conducted using the Auto-Regressive Integrated Moving Average (ARIMA) model. Various combinations of ARIMA orders (p, d, q) have been tried to select the most fitted model, which was utilized to forecast the wastewater inflow rates. The linear regression analysis was applied to testify the correlation between the observed and predicted influents. ARIMA (3, 1, 3) model was selected with the highest significance R-square and lowest normalized Bayesian Information Criterion (BIC) value, and accordingly the wastewater inflow rates were forecasted to additional 52weeks. The linear regression analysis between the observed and predicted values of the wastewater inflow rates showed a positive linear correlation with a coefficient of 0.831.

The ARPAL operational high resolution Poor Man's Ensemble, description and validation

NASA Astrophysics Data System (ADS)

Corazza, Matteo; Sacchetti, Davide; Antonelli, Marta; Drofa, Oxana

2018-05-01

The Meteo Hydrological Functional Center for Civil Protection of the Environmental Protection Agency of the Liguria Region is responsible for issuing forecasts primarily aimed at the Civil Protection needs. Several deterministic high resolution models, run every 6 or 12 h, are regularly used in the Center to elaborate weather forecasts at short to medium range. The Region is frequently affected by severe flash floods over its very small basins, characterized by a steep orography close to the sea. These conditions led the Center in the past years to pay particular attention to the use and development of high resolution model chains for explicit simulation of convective phenomena. For years, the availability of several models has been used by the forecasters for subjective analyses of the potential evolution of the atmosphere and of its uncertainty. More recently, an Interactive Poor Man's Ensemble has been developed, aimed at providing statistical ensemble variables to help forecaster's evaluations. In this paper the structure of this system is described and results are validated using the regional dense ground observational network.

The joint space-time statistics of macroweather precipitation, space-time statistical factorization and macroweather models.

PubMed

Lovejoy, S; de Lima, M I P

2015-07-01

Over the range of time scales from about 10 days to 30-100 years, in addition to the familiar weather and climate regimes, there is an intermediate "macroweather" regime characterized by negative temporal fluctuation exponents: implying that fluctuations tend to cancel each other out so that averages tend to converge. We show theoretically and numerically that macroweather precipitation can be modeled by a stochastic weather-climate model (the Climate Extended Fractionally Integrated Flux, model, CEFIF) first proposed for macroweather temperatures and we show numerically that a four parameter space-time CEFIF model can approximately reproduce eight or so empirical space-time exponents. In spite of this success, CEFIF is theoretically and numerically difficult to manage. We therefore propose a simplified stochastic model in which the temporal behavior is modeled as a fractional Gaussian noise but the spatial behaviour as a multifractal (climate) cascade: a spatial extension of the recently introduced ScaLIng Macroweather Model, SLIMM. Both the CEFIF and this spatial SLIMM model have a property often implicitly assumed by climatologists that climate statistics can be "homogenized" by normalizing them with the standard deviation of the anomalies. Physically, it means that the spatial macroweather variability corresponds to different climate zones that multiplicatively modulate the local, temporal statistics. This simplified macroweather model provides a framework for macroweather forecasting that exploits the system's long range memory and spatial correlations; for it, the forecasting problem has been solved. We test this factorization property and the model with the help of three centennial, global scale precipitation products that we analyze jointly in space and in time.

Evaluation of the NCEP CFSv2 45-day Forecasts for Predictability of Intraseasonal Tropical Storm Activities

NASA Astrophysics Data System (ADS)

Schemm, J. E.; Long, L.; Baxter, S.

2013-12-01

Evaluation of the NCEP CFSv2 45-day Forecasts for Predictability of Intraseasonal Tropical Storm Activities Jae-Kyung E. Schemm, Lindsey Long and Stephen Baxter Climate Prediction Center, NCEP/NWS/NOAA Predictability of intraseasonal tropical storm (TS) activities is assessed using the 1999-2010 CFSv2 hindcast suite. Weekly TS activities in the CFSv2 45-day forecasts were determined using the TS detection and tracking method devised by Carmago and Zebiak (2002). The forecast periods are divided into weekly intervals for Week 1 through Week 6, and also the 30-day mean. The TS activities in those intervals are compared to the observed activities based on the NHC HURDAT and JTWC Best Track datasets. The CFSv2 45-day hindcast suite is made of forecast runs initialized at 00, 06, 12 and 18Z every day during the 1999 - 2010 period. For predictability evaluation, forecast TS activities are analyzed based on 20-member ensemble forecasts comprised of 45-day runs made during the most recent 5 days prior to the verification period. The forecast TS activities are evaluated in terms of the number of storms, genesis locations and storm tracks during the weekly periods. The CFSv2 forecasts are shown to have a fair level of skill in predicting the number of storms over the Atlantic Basin with the temporal correlation scores ranging from 0.73 for Week 1 forecasts to 0.63 for Week 6, and the average RMS errors ranging from 0.86 to 1.07 during the 1999-2010 hurricane season. Also, the forecast track density distribution and false alarm statistics are compiled using the hindcast analyses. In real-time applications of the intraseasonal TS activity forecasts, the climatological TS forecast statistics will be used to make the model bias corrections in terms of the storm counts, track distribution and removal of false alarms. An operational implementation of the weekly TS activity prediction is planned for early 2014 to provide an objective input for the CPC's Global Tropical Hazards Outlooks.

Automated flare forecasting using a statistical learning technique

NASA Astrophysics Data System (ADS)

Yuan, Yuan; Shih, Frank Y.; Jing, Ju; Wang, Hai-Min

2010-08-01

We present a new method for automatically forecasting the occurrence of solar flares based on photospheric magnetic measurements. The method is a cascading combination of an ordinal logistic regression model and a support vector machine classifier. The predictive variables are three photospheric magnetic parameters, i.e., the total unsigned magnetic flux, length of the strong-gradient magnetic polarity inversion line, and total magnetic energy dissipation. The output is true or false for the occurrence of a certain level of flares within 24 hours. Experimental results, from a sample of 230 active regions between 1996 and 2005, show the accuracies of a 24-hour flare forecast to be 0.86, 0.72, 0.65 and 0.84 respectively for the four different levels. Comparison shows an improvement in the accuracy of X-class flare forecasting.

Input selection and performance optimization of ANN-based streamflow forecasts in the drought-prone Murray Darling Basin region using IIS and MODWT algorithm

NASA Astrophysics Data System (ADS)

Prasad, Ramendra; Deo, Ravinesh C.; Li, Yan; Maraseni, Tek

2017-11-01

Forecasting streamflow is vital for strategically planning, utilizing and redistributing water resources. In this paper, a wavelet-hybrid artificial neural network (ANN) model integrated with iterative input selection (IIS) algorithm (IIS-W-ANN) is evaluated for its statistical preciseness in forecasting monthly streamflow, and it is then benchmarked against M5 Tree model. To develop hybrid IIS-W-ANN model, a global predictor matrix is constructed for three local hydrological sites (Richmond, Gwydir, and Darling River) in Australia's agricultural (Murray-Darling) Basin. Model inputs comprised of statistically significant lagged combination of streamflow water level, are supplemented by meteorological data (i.e., precipitation, maximum and minimum temperature, mean solar radiation, vapor pressure and evaporation) as the potential model inputs. To establish robust forecasting models, iterative input selection (IIS) algorithm is applied to screen the best data from the predictor matrix and is integrated with the non-decimated maximum overlap discrete wavelet transform (MODWT) applied on the IIS-selected variables. This resolved the frequencies contained in predictor data while constructing a wavelet-hybrid (i.e., IIS-W-ANN and IIS-W-M5 Tree) model. Forecasting ability of IIS-W-ANN is evaluated via correlation coefficient (r), Willmott's Index (WI), Nash-Sutcliffe Efficiency (ENS), root-mean-square-error (RMSE), and mean absolute error (MAE), including the percentage RMSE and MAE. While ANN models are seen to outperform M5 Tree executed for all hydrological sites, the IIS variable selector was efficient in determining the appropriate predictors, as stipulated by the better performance of the IIS coupled (ANN and M5 Tree) models relative to the models without IIS. When IIS-coupled models are integrated with MODWT, the wavelet-hybrid IIS-W-ANN and IIS-W-M5 Tree are seen to attain significantly accurate performance relative to their standalone counterparts. Importantly, IIS-W-ANN model accuracy outweighs IIS-ANN, as evidenced by a larger r and WI (by 7.5% and 3.8%, respectively) and a lower RMSE (by 21.3%). In comparison to the IIS-W-M5 Tree model, IIS-W-ANN model yielded larger values of WI = 0.936-0.979 and ENS = 0.770-0.920. Correspondingly, the errors (RMSE and MAE) ranged from 0.162-0.487 m and 0.139-0.390 m, respectively, with relative errors, RRMSE = (15.65-21.00) % and MAPE = (14.79-20.78) %. Distinct geographic signature is evident where the most and least accurately forecasted streamflow data is attained for the Gwydir and Darling River, respectively. Conclusively, this study advocates the efficacy of iterative input selection, allowing the proper screening of model predictors, and subsequently, its integration with MODWT resulting in enhanced performance of the models applied in streamflow forecasting.

A forecast experiment of earthquake activity in Japan under Collaboratory for the Study of Earthquake Predictability (CSEP)

NASA Astrophysics Data System (ADS)

Hirata, N.; Yokoi, S.; Nanjo, K. Z.; Tsuruoka, H.

2012-04-01

One major focus of the current Japanese earthquake prediction research program (2009-2013), which is now integrated with the research program for prediction of volcanic eruptions, is to move toward creating testable earthquake forecast models. For this purpose we started an experiment of forecasting earthquake activity in Japan under the framework of the Collaboratory for the Study of Earthquake Predictability (CSEP) through an international collaboration. We established the CSEP Testing Centre, an infrastructure to encourage researchers to develop testable models for Japan, and to conduct verifiable prospective tests of their model performance. We started the 1st earthquake forecast testing experiment in Japan within the CSEP framework. We use the earthquake catalogue maintained and provided by the Japan Meteorological Agency (JMA). The experiment consists of 12 categories, with 4 testing classes with different time spans (1 day, 3 months, 1 year, and 3 years) and 3 testing regions called "All Japan," "Mainland," and "Kanto." A total of 105 models were submitted, and are currently under the CSEP official suite of tests for evaluating the performance of forecasts. The experiments were completed for 92 rounds for 1-day, 6 rounds for 3-month, and 3 rounds for 1-year classes. For 1-day testing class all models passed all the CSEP's evaluation tests at more than 90% rounds. The results of the 3-month testing class also gave us new knowledge concerning statistical forecasting models. All models showed a good performance for magnitude forecasting. On the other hand, observation is hardly consistent in space distribution with most models when many earthquakes occurred at a spot. Now we prepare the 3-D forecasting experiment with a depth range of 0 to 100 km in Kanto region. The testing center is improving an evaluation system for 1-day class experiment to finish forecasting and testing results within one day. The special issue of 1st part titled Earthquake Forecast Testing Experiment in Japan was published on the Earth, Planets and Space Vol. 63, No.3, 2011 on March, 2011. The 2nd part of this issue, which is now on line, will be published soon. An outline of the experiment and activities of the Japanese Testing Center are published on our WEB site; http://wwweic.eri.u-tokyo.ac.jp/ZISINyosoku/wiki.en/wiki.cgi

Evaluation of a Revised Interplanetary Shock Prediction Model: 1D CESE-HD-2 Solar-Wind Model

NASA Astrophysics Data System (ADS)

Zhang, Y.; Du, A. M.; Du, D.; Sun, W.

2014-08-01

We modified the one-dimensional conservation element and solution element (CESE) hydrodynamic (HD) model into a new version [ 1D CESE-HD-2], by considering the direction of the shock propagation. The real-time performance of the 1D CESE-HD-2 model during Solar Cycle 23 (February 1997 - December 2006) is investigated and compared with those of the Shock Time of Arrival Model ( STOA), the Interplanetary-Shock-Propagation Model ( ISPM), and the Hakamada-Akasofu-Fry version 2 ( HAFv.2). Of the total of 584 flare events, 173 occurred during the rising phase, 166 events during the maximum phase, and 245 events during the declining phase. The statistical results show that the success rates of the predictions by the 1D CESE-HD-2 model for the rising, maximum, declining, and composite periods are 64 %, 62 %, 57 %, and 61 %, respectively, with a hit window of ± 24 hours. The results demonstrate that the 1D CESE-HD-2 model shows the highest success rates when the background solar-wind speed is relatively fast. Thus, when the background solar-wind speed at the time of shock initiation is enhanced, the forecasts will provide potential values to the customers. A high value (27.08) of χ 2 and low p-value (< 0.0001) for the 1D CESE-HD-2 model give considerable confidence for real-time forecasts by using this new model. Furthermore, the effects of various shock characteristics (initial speed, shock duration, background solar wind, longitude, etc.) and background solar wind on the forecast are also investigated statistically.

Evaluating the Contribution of NASA Remotely-Sensed Data Sets on a Convection-Allowing Forecast Model

NASA Technical Reports Server (NTRS)

Zavodsky, Bradley T.; Case, Jonathan L.; Molthan, Andrew L.

2012-01-01

The Short-term Prediction Research and Transition (SPoRT) Center is a collaborative partnership between NASA and operational forecasting partners, including a number of National Weather Service forecast offices. SPoRT provides real-time NASA products and capabilities to help its partners address specific operational forecast challenges. One challenge that forecasters face is using guidance from local and regional deterministic numerical models configured at convection-allowing resolution to help assess a variety of mesoscale/convective-scale phenomena such as sea-breezes, local wind circulations, and mesoscale convective weather potential on a given day. While guidance from convection-allowing models has proven valuable in many circumstances, the potential exists for model improvements by incorporating more representative land-water surface datasets, and by assimilating retrieved temperature and moisture profiles from hyper-spectral sounders. In order to help increase the accuracy of deterministic convection-allowing models, SPoRT produces real-time, 4-km CONUS forecasts using a configuration of the Weather Research and Forecasting (WRF) model (hereafter SPoRT-WRF) that includes unique NASA products and capabilities including 4-km resolution soil initialization data from the Land Information System (LIS), 2-km resolution SPoRT SST composites over oceans and large water bodies, high-resolution real-time Green Vegetation Fraction (GVF) composites derived from the Moderate-resolution Imaging Spectroradiometer (MODIS) instrument, and retrieved temperature and moisture profiles from the Atmospheric Infrared Sounder (AIRS) and Infrared Atmospheric Sounding Interferometer (IASI). NCAR's Model Evaluation Tools (MET) verification package is used to generate statistics of model performance compared to in situ observations and rainfall analyses for three months during the summer of 2012 (June-August). Detailed analyses of specific severe weather outbreaks during the summer will be presented to assess the potential added-value of the SPoRT datasets and data assimilation methodology compared to a WRF configuration without the unique datasets and data assimilation.

The impact of vertical resolution in mesoscale model AROME forecasting of radiation fog

NASA Astrophysics Data System (ADS)

Philip, Alexandre; Bergot, Thierry; Bouteloup, Yves; Bouyssel, François

2015-04-01

Airports short-term forecasting of fog has a security and economic impact. Numerical simulations have been performed with the mesoscale model AROME (Application of Research to Operations at Mesoscale) (Seity et al. 2011). Three vertical resolutions (60, 90 and 156 levels) are used to show the impact of radiation fog on numerical forecasting. Observations at Roissy Charles De Gaulle airport are compared to simulations. Significant differences in the onset, evolution and dissipation of fog were found. The high resolution simulation is in better agreement with observations than a coarser one. The surface boundary layer and incoming long-wave radiations are better represented. A more realistic behaviour of liquid water content evolution allows a better anticipation of low visibility procedures (ceiling < 60m and/or visibility < 600m). The case study of radiation fog shows that it is necessary to have a well defined vertical grid to better represent local phenomena. A statistical study over 6 months (October 2011 - March 2012 ) using different configurations was carried out. Statistically, results were the same as in the case study of radiation fog. Seity Y., P. Brousseau, S. Malardel, G. Hello, P. Bénard, F. Bouttier, C. Lac, V. Masson, 2011: The AROME-France convective scale operational model. Mon.Wea.Rev., 139, 976-991.

Comparison of Conventional and ANN Models for River Flow Forecasting

NASA Astrophysics Data System (ADS)

Jain, A.; Ganti, R.

2011-12-01

Hydrological models are useful in many water resources applications such as flood control, irrigation and drainage, hydro power generation, water supply, erosion and sediment control, etc. Estimates of runoff are needed in many water resources planning, design development, operation and maintenance activities. River flow is generally estimated using time series or rainfall-runoff models. Recently, soft artificial intelligence tools such as Artificial Neural Networks (ANNs) have become popular for research purposes but have not been extensively adopted in operational hydrological forecasts. There is a strong need to develop ANN models based on real catchment data and compare them with the conventional models. In this paper, a comparative study has been carried out for river flow forecasting using the conventional and ANN models. Among the conventional models, multiple linear, and non linear regression, and time series models of auto regressive (AR) type have been developed. Feed forward neural network model structure trained using the back propagation algorithm, a gradient search method, was adopted. The daily river flow data derived from Godavari Basin @ Polavaram, Andhra Pradesh, India have been employed to develop all the models included here. Two inputs, flows at two past time steps, (Q(t-1) and Q(t-2)) were selected using partial auto correlation analysis for forecasting flow at time t, Q(t). A wide range of error statistics have been used to evaluate the performance of all the models developed in this study. It has been found that the regression and AR models performed comparably, and the ANN model performed the best amongst all the models investigated in this study. It is concluded that ANN model should be adopted in real catchments for hydrological modeling and forecasting.

Prediction of PM10 grades in Seoul, Korea using a neural network model based on synoptic patterns

NASA Astrophysics Data System (ADS)

Hur, S. K.; Oh, H. R.; Ho, C. H.; Kim, J.; Song, C. K.; Chang, L. S.; Lee, J. B.

2016-12-01

As of November 2014, the Korean Ministry of Environment (KME) started forecasting the level of ambient particulate matter with diameters ≤ 10 μm (PM10) as four grades: low (PM10 ≤ 30 μg m-3), moderate (30 < PM10 ≤ 80 μg m-3), high (80 < PM10 ≤ 150 μg m-3), and very high (PM10 > 150 μg m-3). Due to short history of forecast, overall performance of the operational forecasting system and its hit rate for the four PM10 grades are difficult to evaluate. In attempt to provide a statistical reference for the current air quality forecasting system, we hindcasted the four PM10 grades for the cold seasons (October-March) of 2001-2014 in Seoul, Korea using a neural network model based on the synoptic patterns of meteorological fields such as geopotential height, air temperature, relative humidity, and wind. In the form of cosine similarity, the distinctive synoptic patterns for each PM10 grades are well quantified as predictors to train the neural network model. Using these fields as predictors and considering the PM10 concentration in Seoul from the day before prediction as an additional predictor, an overall hit rate of 69% was achieved; the hit rates for the low, moderate, high, and very high PM10 grades were 33%, 83%, 45%, and 33%, respectively. This study reveals that the synoptic patterns of meteorological fields are useful predictors for the identification of favorable conditions for each PM10 grade, and the associated transboundary transport and local accumulation of PM10 from the industrialized regions of China. Consequently, the assessments of predictability obtained from the neural network model in this study are reliable to use as a statistical reference for the current air quality forecasting system.

Applications systems verification and transfer project. Volume 4: Operational applications of satellite snow cover observations. Colorado Field Test Center

NASA Technical Reports Server (NTRS)

Shafer, B. A.; Leaf, C. F.; Danielson, J. A.; Moravec, G. F.

1981-01-01

The study was conducted on six watersheds ranging in size from 277 km to 3460 km in the Rio Grande and Arkansas River basins of southwestern Colorado. Six years of satellite data in the period 1973-78 were analyzed and snowcover maps prepared for all available image dates. Seven snowmapping techniques were explored; the photointerpretative method was selected as the most accurate. Three schemes to forecast snowmelt runoff employing satellite snowcover observations were investigated. They included a conceptual hydrologic model, a statistical model, and a graphical method. A reduction of 10% in the current average forecast error is estimated when snowcover data in snowmelt runoff forecasting is shown to be extremely promising. Inability to obtain repetitive coverage due to the 18 day cycle of LANDSAT, the occurrence of cloud cover and slow image delivery are obstacles to the immediate implementation of satellite derived snowcover in operational streamflow forecasting programs.

Chesapeake Bay Forecast System: Oxygen Prediction for the Sustainable Ecosystem Management

NASA Astrophysics Data System (ADS)

Mathukumalli, B.; Long, W.; Zhang, X.; Wood, R.; Murtugudde, R. G.

2010-12-01

The Chesapeake Bay Forecast System (CBFS) is a flexible, end-to-end expert prediction tool for decision makers that will provide customizable, user-specified predictions and projections of the region’s climate, air and water quality, local chemistry, and ecosystems at days to decades. As a part of CBFS, the long-term water quality data were collected and assembled to develop ecological models for the sustainable management of the Chesapeake Bay. Cultural eutrophication depletes oxygen levels in this ecosystem particularly in summer which has several negative implications on the structure and function of ecosystem. In order to understand dynamics and prediction of spatially-explicit oxygen levels in the Bay, an empirical process based ecological model is developed with long-term control variables (water temperature, salinity, nitrogen and phosphorus). Statistical validation methods were employed to demonstrate usability of predictions for management purposes and the predicted oxygen levels are quite faithful to observations. The predicted oxygen values and other physical outputs from downscaling of regional weather and climate predictions, or forecasts from hydrodynamic models can be used to forecast various ecological components. Such forecasts would be useful for both recreational and commercial users of the bay (for example, bass fishing). Furthermore, this work can also be used to predict extent of hypoxia/anoxia not only from anthropogenic nutrient pollution, but also from global warming. Some hindcasts and forecasts are discussed along with the ongoing efforts at a mechanistic ecosystem model to provide prognostic oxygen predictions and projections and upper trophic modeling using an energetics approach.

The surface drifter program for real time and off-line validation of ocean forecasts and reanalyses

NASA Astrophysics Data System (ADS)

Hernandez, Fabrice; Regnier, Charly; Drévillon, Marie

2017-04-01

As part of the Global Ocean Observing System, the Global Drifter Program (GDP) is comprised of an array of about 1250 drifting buoys spread over the global ocean, that provide operational, near-real time surface velocity, sea surface temperature (SST) and sea level pressure observations. This information is used mainly used for numerical weather forecasting, research, and in-situ calibration/verification of satellite observations. Since 2013 the drifting buoy SST measurements are used for near real time assessment of global forecasting systems from Canada, France, UK, USA, Australia in the frame of the GODAE OceanView Intercomparison and Validation Task. For most of these operational systems, these data are not used for assimilation, and offer an independent observation assessment. This approach mimics the validation performed for SST satellite products. More recently, validation procedures have been proposed in order to assess the surface dynamics of Mercator Océan global and regional forecast and reanalyses. Velocities deduced from drifter trajectories are used in two ways. First, the Eulerian approach where buoy and ocean model velocity values are compared at the position of drifters. Then, from discrepancies, statistics are computed and provide an evaluation of the ocean model's surface dynamics reliability. Second, the Lagrangian approach, where drifting trajectories are simulated at each location of the real drifter trajectory using the ocean model velocity fields. Then, on daily basis, real and simulated drifter trajectories are compared by analyzing the spread after one day, two days etc…. The cumulated statistics on specific geographical boxes are evaluated in term of dispersion properties of the "real ocean" as captured by drifters, and those properties in the ocean model. This approach allows to better evaluate forecasting score for surface dispersion applications, like Search and Rescue, oil spill forecast, drift of other objects or contaminant, larvae dispersion etc… These Eulerian and Lagrangian validation approach can be applied for real time or offline assessment of ocean velocity products. In real time, the main limitation is our capability to detect drifter drogue's loss, causing erroneous assessment. Several methods, by comparison to wind entrainment effect or other velocity estimates like from satellite altimetry, are used. These Eulerian and Lagrangian surface velocity validation methods are planned to be adopted by the GODAE OceanView operational community in order to offer independent verification of surface current forecast.

State estimation and prediction using clustered particle filters.

PubMed

Lee, Yoonsang; Majda, Andrew J

2016-12-20

Particle filtering is an essential tool to improve uncertain model predictions by incorporating noisy observational data from complex systems including non-Gaussian features. A class of particle filters, clustered particle filters, is introduced for high-dimensional nonlinear systems, which uses relatively few particles compared with the standard particle filter. The clustered particle filter captures non-Gaussian features of the true signal, which are typical in complex nonlinear dynamical systems such as geophysical systems. The method is also robust in the difficult regime of high-quality sparse and infrequent observations. The key features of the clustered particle filtering are coarse-grained localization through the clustering of the state variables and particle adjustment to stabilize the method; each observation affects only neighbor state variables through clustering and particles are adjusted to prevent particle collapse due to high-quality observations. The clustered particle filter is tested for the 40-dimensional Lorenz 96 model with several dynamical regimes including strongly non-Gaussian statistics. The clustered particle filter shows robust skill in both achieving accurate filter results and capturing non-Gaussian statistics of the true signal. It is further extended to multiscale data assimilation, which provides the large-scale estimation by combining a cheap reduced-order forecast model and mixed observations of the large- and small-scale variables. This approach enables the use of a larger number of particles due to the computational savings in the forecast model. The multiscale clustered particle filter is tested for one-dimensional dispersive wave turbulence using a forecast model with model errors.

State estimation and prediction using clustered particle filters

PubMed Central

Lee, Yoonsang; Majda, Andrew J.

2016-01-01

Particle filtering is an essential tool to improve uncertain model predictions by incorporating noisy observational data from complex systems including non-Gaussian features. A class of particle filters, clustered particle filters, is introduced for high-dimensional nonlinear systems, which uses relatively few particles compared with the standard particle filter. The clustered particle filter captures non-Gaussian features of the true signal, which are typical in complex nonlinear dynamical systems such as geophysical systems. The method is also robust in the difficult regime of high-quality sparse and infrequent observations. The key features of the clustered particle filtering are coarse-grained localization through the clustering of the state variables and particle adjustment to stabilize the method; each observation affects only neighbor state variables through clustering and particles are adjusted to prevent particle collapse due to high-quality observations. The clustered particle filter is tested for the 40-dimensional Lorenz 96 model with several dynamical regimes including strongly non-Gaussian statistics. The clustered particle filter shows robust skill in both achieving accurate filter results and capturing non-Gaussian statistics of the true signal. It is further extended to multiscale data assimilation, which provides the large-scale estimation by combining a cheap reduced-order forecast model and mixed observations of the large- and small-scale variables. This approach enables the use of a larger number of particles due to the computational savings in the forecast model. The multiscale clustered particle filter is tested for one-dimensional dispersive wave turbulence using a forecast model with model errors. PMID:27930332

Development of a monthly to seasonal forecast framework tailored to inland waterway transport in central Europe

NASA Astrophysics Data System (ADS)

Meißner, Dennis; Klein, Bastian; Ionita, Monica

2017-12-01

Traditionally, navigation-related forecasts in central Europe cover short- to medium-range lead times linked to the travel times of vessels to pass the main waterway bottlenecks leaving the loading ports. Without doubt, this aspect is still essential for navigational users, but in light of the growing political intention to use the free capacity of the inland waterway transport in Europe, additional lead time supporting strategic decisions is more and more in demand. However, no such predictions offering extended lead times of several weeks up to several months currently exist for considerable parts of the European waterway network. This paper describes the set-up of a monthly to seasonal forecasting system for the German stretches of the international waterways of the Rhine, Danube and Elbe rivers. Two competitive forecast approaches have been implemented: the dynamical set-up forces a hydrological model with post-processed outputs from ECMWF general circulation model System 4, whereas the statistical approach is based on the empirical relationship (teleconnection) of global oceanic, climate and regional hydro-meteorological data with river flows. The performance of both forecast methods is evaluated in relation to the climatological forecast (ensemble of historical streamflow) and the well-known ensemble streamflow prediction approach (ESP, ensemble based on historical meteorology) using common performance indicators (correlation coefficient; mean absolute error, skill score; mean squared error, skill score; and continuous ranked probability, skill score) and an impact-based evaluation quantifying the potential economic gain. The following four key findings result from this study: (1) as former studies for other regions of central Europe indicate, the accuracy and/or skill of the meteorological forcing used has a larger effect than the quality of initial hydrological conditions for relevant stations along the German waterways. (2) Despite the predictive limitations on longer lead times in central Europe, this study reveals the existence of a valuable predictability of streamflow on monthly up to seasonal timescales along the Rhine, upper Danube and Elbe waterways, and the Elbe achieves the highest skill and economic value. (3) The more physically based and the statistical approach are able to improve the predictive skills and economic value compared to climatology and the ESP approach. The specific forecast skill highly depends on the forecast location, the lead time and the season. (4) Currently, the statistical approach seems to be most skilful for the three waterways investigated. The lagged relationship between the monthly and/or seasonal streamflow and the climatic and/or oceanic variables vary between 1 month (e.g. local precipitation, temperature and soil moisture) up to 6 months (e.g. sea surface temperature). Besides focusing on improving the forecast methodology, especially by combining the individual approaches, the focus is on developing useful forecast products on monthly to seasonal timescales for waterway transport and to operationalize the related forecasting service.

Estimation Methods for Non-Homogeneous Regression - Minimum CRPS vs Maximum Likelihood

NASA Astrophysics Data System (ADS)

Gebetsberger, Manuel; Messner, Jakob W.; Mayr, Georg J.; Zeileis, Achim

2017-04-01

Non-homogeneous regression models are widely used to statistically post-process numerical weather prediction models. Such regression models correct for errors in mean and variance and are capable to forecast a full probability distribution. In order to estimate the corresponding regression coefficients, CRPS minimization is performed in many meteorological post-processing studies since the last decade. In contrast to maximum likelihood estimation, CRPS minimization is claimed to yield more calibrated forecasts. Theoretically, both scoring rules used as an optimization score should be able to locate a similar and unknown optimum. Discrepancies might result from a wrong distributional assumption of the observed quantity. To address this theoretical concept, this study compares maximum likelihood and minimum CRPS estimation for different distributional assumptions. First, a synthetic case study shows that, for an appropriate distributional assumption, both estimation methods yield to similar regression coefficients. The log-likelihood estimator is slightly more efficient. A real world case study for surface temperature forecasts at different sites in Europe confirms these results but shows that surface temperature does not always follow the classical assumption of a Gaussian distribution. KEYWORDS: ensemble post-processing, maximum likelihood estimation, CRPS minimization, probabilistic temperature forecasting, distributional regression models

Variational Continuous Assimilation of TMI and SSM/I Rain Rates: Impact on GEOS-3 Hurricane Analyses and Forecasts

NASA Technical Reports Server (NTRS)

Hou, Arthur Y.; Zhang, Sara Q.; Reale, Oreste

2003-01-01

We describe a variational continuous assimilation (VCA) algorithm for assimilating tropical rainfall data using moisture and temperature tendency corrections as the control variable to offset model deficiencies. For rainfall assimilation, model errors are of special concern since model-predicted precipitation is based on parameterized moist physics, which can have substantial systematic errors. This study examines whether a VCA scheme using the forecast model as a weak constraint offers an effective pathway to precipitation assimilation. The particular scheme we exarnine employs a '1+1' dimension precipitation observation operator based on a 6-h integration of a column model of moist physics from the Goddard Earth Observing System (GEOS) global data assimilation system DAS). In earlier studies, we tested a simplified version of this scheme and obtained improved monthly-mean analyses and better short-range forecast skills. This paper describes the full implementation ofthe 1+1D VCA scheme using background and observation error statistics, and examines how it may improve GEOS analyses and forecasts of prominent tropical weather systems such as hurricanes. Parallel assimilation experiments with and without rainfall data for Hurricanes Bonnie and Floyd show that assimilating 6-h TMI and SSM/I surfice rain rates leads to more realistic storm features in the analysis, which, in turn, provide better initial conditions for 5-day storm track prediction and precipitation forecast. These results provide evidence that addressing model deficiencies in moisture tendency may be crucial to making effective use of precipitation information in data assimilation.

Empirical-statistical downscaling of reanalysis data to high-resolution air temperature and specific humidity above a glacier surface (Cordillera Blanca, Peru)

NASA Astrophysics Data System (ADS)

Hofer, Marlis; MöLg, Thomas; Marzeion, Ben; Kaser, Georg

2010-06-01

Recently initiated observation networks in the Cordillera Blanca (Peru) provide temporally high-resolution, yet short-term, atmospheric data. The aim of this study is to extend the existing time series into the past. We present an empirical-statistical downscaling (ESD) model that links 6-hourly National Centers for Environmental Prediction (NCEP)/National Center for Atmospheric Research (NCAR) reanalysis data to air temperature and specific humidity, measured at the tropical glacier Artesonraju (northern Cordillera Blanca). The ESD modeling procedure includes combined empirical orthogonal function and multiple regression analyses and a double cross-validation scheme for model evaluation. Apart from the selection of predictor fields, the modeling procedure is automated and does not include subjective choices. We assess the ESD model sensitivity to the predictor choice using both single-field and mixed-field predictors. Statistical transfer functions are derived individually for different months and times of day. The forecast skill largely depends on month and time of day, ranging from 0 to 0.8. The mixed-field predictors perform better than the single-field predictors. The ESD model shows added value, at all time scales, against simpler reference models (e.g., the direct use of reanalysis grid point values). The ESD model forecast 1960-2008 clearly reflects interannual variability related to the El Niño/Southern Oscillation but is sensitive to the chosen predictor type.

Using remote sensing environmental data to forecast malaria incidence at a rural district hospital in Western Kenya.

PubMed

Sewe, Maquins Odhiambo; Tozan, Yesim; Ahlm, Clas; Rocklöv, Joacim

2017-06-01

Malaria surveillance data provide opportunity to develop forecasting models. Seasonal variability in environmental factors correlate with malaria transmission, thus the identification of transmission patterns is useful in developing prediction models. However, with changing seasonal transmission patterns, either due to interventions or shifting weather seasons, traditional modelling approaches may not yield adequate predictive skill. Two statistical models,a general additive model (GAM) and GAMBOOST model with boosted regression were contrasted by assessing their predictive accuracy in forecasting malaria admissions at lead times of one to three months. Monthly admission data for children under five years with confirmed malaria at the Siaya district hospital in Western Kenya for the period 2003 to 2013 were used together with satellite derived data on rainfall, average temperature and evapotranspiration(ET). There was a total of 8,476 confirmed malaria admissions. The peak of malaria season changed and malaria admissions reduced overtime. The GAMBOOST model at 1-month lead time had the highest predictive skill during both the training and test periods and thus can be utilized in a malaria early warning system.

How accurate are the weather forecasts for Bierun (southern Poland)?

NASA Astrophysics Data System (ADS)

Gawor, J.

2012-04-01

Weather forecast accuracy has increased in recent times mainly thanks to significant development of numerical weather prediction models. Despite the improvements, the forecasts should be verified to control their quality. The evaluation of forecast accuracy can also be an interesting learning activity for students. It joins natural curiosity about everyday weather and scientific process skills: problem solving, database technologies, graph construction and graphical analysis. The examination of the weather forecasts has been taken by a group of 14-year-old students from Bierun (southern Poland). They participate in the GLOBE program to develop inquiry-based investigations of the local environment. For the atmospheric research the automatic weather station is used. The observed data were compared with corresponding forecasts produced by two numerical weather prediction models, i.e. COAMPS (Coupled Ocean/Atmosphere Mesoscale Prediction System) developed by Naval Research Laboratory Monterey, USA; it runs operationally at the Interdisciplinary Centre for Mathematical and Computational Modelling in Warsaw, Poland and COSMO (The Consortium for Small-scale Modelling) used by the Polish Institute of Meteorology and Water Management. The analysed data included air temperature, precipitation, wind speed, wind chill and sea level pressure. The prediction periods from 0 to 24 hours (Day 1) and from 24 to 48 hours (Day 2) were considered. The verification statistics that are commonly used in meteorology have been applied: mean error, also known as bias, for continuous data and a 2x2 contingency table to get the hit rate and false alarm ratio for a few precipitation thresholds. The results of the aforementioned activity became an interesting basis for discussion. The most important topics are: 1) to what extent can we rely on the weather forecasts? 2) How accurate are the forecasts for two considered time ranges? 3) Which precipitation threshold is the most predictable? 4) Why are some weather elements easier to verify than others? 5) What factors may contribute to the quality of the weather forecast?

Comparison of the performance and reliability of 18 lumped hydrological models driven by ECMWF rainfall ensemble forecasts: a case study on 29 French catchments

NASA Astrophysics Data System (ADS)

Velázquez, Juan Alberto; Anctil, François; Ramos, Maria-Helena; Perrin, Charles

2010-05-01

An ensemble forecasting system seeks to assess and to communicate the uncertainty of hydrological predictions by proposing, at each time step, an ensemble of forecasts from which one can estimate the probability distribution of the predictant (the probabilistic forecast), in contrast with a single estimate of the flow, for which no distribution is obtainable (the deterministic forecast). In the past years, efforts towards the development of probabilistic hydrological prediction systems were made with the adoption of ensembles of numerical weather predictions (NWPs). The additional information provided by the different available Ensemble Prediction Systems (EPS) was evaluated in a hydrological context on various case studies (see the review by Cloke and Pappenberger, 2009). For example, the European ECMWF-EPS was explored in case studies by Roulin et al. (2005), Bartholmes et al. (2005), Jaun et al. (2008), and Renner et al. (2009). The Canadian EC-EPS was also evaluated by Velázquez et al. (2009). Most of these case studies investigate the ensemble predictions of a given hydrological model, set up over a limited number of catchments. Uncertainty from weather predictions is assessed through the use of meteorological ensembles. However, uncertainty from the tested hydrological model and statistical robustness of the forecasting system when coping with different hydro-meteorological conditions are less frequently evaluated. The aim of this study is to evaluate and compare the performance and the reliability of 18 lumped hydrological models applied to a large number of catchments in an operational ensemble forecasting context. Some of these models were evaluated in a previous study (Perrin et al. 2001) for their ability to simulate streamflow. Results demonstrated that very simple models can achieve a level of performance almost as high (sometimes higher) as models with more parameters. In the present study, we focus on the ability of the hydrological models to provide reliable probabilistic forecasts of streamflow, based on ensemble weather predictions. The models were therefore adapted to run in a forecasting mode, i.e., to update initial conditions according to the last observed discharge at the time of the forecast, and to cope with ensemble weather scenarios. All models are lumped, i.e., the hydrological behavior is integrated over the spatial scale of the catchment, and run at daily time steps. The complexity of tested models varies between 3 and 13 parameters. The models are tested on 29 French catchments. Daily streamflow time series extend over 17 months, from March 2005 to July 2006. Catchment areas range between 1470 km2 and 9390 km2, and represent a variety of hydrological and meteorological conditions. The 12 UTC 10-day ECMWF rainfall ensemble (51 members) was used, which led to daily streamflow forecasts for a 9-day lead time. In order to assess the performance and reliability of the hydrological ensemble predictions, we computed the Continuous Ranked probability Score (CRPS) (Matheson and Winkler, 1976), as well as the reliability diagram (e.g. Wilks, 1995) and the rank histogram (Talagrand et al., 1999). Since the ECMWF deterministic forecasts are also available, the performance of the hydrological forecasting systems was also evaluated by comparing the deterministic score (MAE) with the probabilistic score (CRPS). The results obtained for the 18 hydrological models and the 29 studied catchments are discussed in the perspective of improving the operational use of ensemble forecasting in hydrology. References Bartholmes, J. and Todini, E.: Coupling meteorological and hydrological models for flood forecasting, Hydrol. Earth Syst. Sci., 9, 333-346, 2005. Cloke, H. and Pappenberger, F.: Ensemble Flood Forecasting: A Review. Journal of Hydrology 375 (3-4): 613-626, 2009. Jaun, S., Ahrens, B., Walser, A., Ewen, T., and Schär, C.: A probabilistic view on the August 2005 floods in the upper Rhine catchment, Nat. Hazards Earth Syst. Sci., 8, 281-291, 2008. Matheson, J. E. and Winkler, R. L.: Scoring rules for continuous probability distributions, Manage Sci., 22, 1087-1096, 1976. Perrin, C., Michel C. and Andréassian,V. Does a large number of parameters enhance model performance? Comparative assessment of common catchment model structures on 429 catchments, J. Hydrol., 242, 275-301, 2001. Renner, M., Werner, M. G. F., Rademacher, S., and Sprokkereef, E.: Verification of ensemble flow forecast for the River Rhine, J. Hydrol., 376, 463-475, 2009. Roulin, E. and Vannitsem, S.: Skill of medium-range hydrological ensemble predictions, J. Hydrometeorol., 6, 729-744, 2005. Talagrand, O., Vautard, R., and Strauss, B.: Evaluation of the probabilistic prediction systems, in: Proceedings, ECMWF Workshop on Predictability, Shinfield Park, Reading, Berkshire, ECMWF, 1-25, 1999. Velázquez, J.A., Petit, T., Lavoie, A., Boucher M.-A., Turcotte R., Fortin V., and Anctil, F. : An evaluation of the Canadian global meteorological ensemble prediction system for short-term hydrological forecasting, Hydrol. Earth Syst. Sci., 13, 2221-2231, 2009. Wilks, D. S.: Statistical Methods in the Atmospheric Sciences, Academic Press, San Diego, CA, 465 pp., 1995.

Physics-based forecasting of induced seismicity at Groningen gas field, the Netherlands

NASA Astrophysics Data System (ADS)

Dempsey, David; Suckale, Jenny

2017-08-01

Earthquakes induced by natural gas extraction from the Groningen reservoir, the Netherlands, put local communities at risk. Responsible operation of a reservoir whose gas reserves are of strategic importance to the country requires understanding of the link between extraction and earthquakes. We synthesize observations and a model for Groningen seismicity to produce forecasts for felt seismicity (M > 2.5) in the period February 2017 to 2024. Our model accounts for poroelastic earthquake triggering and rupture on the 325 largest reservoir faults, using an ensemble approach to model unknown heterogeneity and replicate earthquake statistics. We calculate probability distributions for key model parameters using a Bayesian method that incorporates the earthquake observations with a nonhomogeneous Poisson process. Our analysis indicates that the Groningen reservoir was not critically stressed prior to the start of production. Epistemic uncertainty and aleatoric uncertainty are incorporated into forecasts for three different future extraction scenarios. The largest expected earthquake was similar for all scenarios, with a 5% likelihood of exceeding M 4.0.

Development of statistical models to forecast crossing times of commercial vehicles.

DOT National Transportation Integrated Search

2011-07-01

Border crossing time measurement systems for commercial vehicles are being implemented throughout : the U.S.-Mexico border. These systems are based on radio frequency identification (RFID) technology. : With funding from the Federal Highway Administr...

Uncertainty estimation of long-range ensemble forecasts of snowmelt flood characteristics

NASA Astrophysics Data System (ADS)

Kuchment, L.

2012-04-01

Long-range forecasts of snowmelt flood characteristics with the lead time of 2-3 months have important significance for regulation of flood runoff and mitigation of flood damages at almost all large Russian rivers At the same time, the application of current forecasting techniques based on regression relationships between the runoff volume and the indexes of river basin conditions can lead to serious errors in forecasting resulted in large economic losses caused by wrong flood regulation. The forecast errors can be caused by complicated processes of soil freezing and soil moisture redistribution, too high rate of snow melt, large liquid precipitation before snow melt. or by large difference of meteorological conditions during the lead-time periods from climatologic ones. Analysis of economic losses had shown that the largest damages could, to a significant extent, be avoided if the decision makers had an opportunity to take into account predictive uncertainty and could use more cautious strategies in runoff regulation. Development of methodology of long-range ensemble forecasting of spring/summer floods which is based on distributed physically-based runoff generation models has created, in principle, a new basis for improving hydrological predictions as well as for estimating their uncertainty. This approach is illustrated by forecasting of the spring-summer floods at the Vyatka River and the Seim River basins. The application of the physically - based models of snowmelt runoff generation give a essential improving of statistical estimates of the deterministic forecasts of the flood volume in comparison with the forecasts obtained from the regression relationships. These models had been used also for the probabilistic forecasts assigning meteorological inputs during lead time periods from the available historical daily series, and from the series simulated by using a weather generator and the Monte Carlo procedure. The weather generator consists of the stochastic models of daily temperature and precipitation. The performance of the probabilistic forecasts were estimated by the ranked probability skill scores. The application of Monte Carlo simulations using weather generator has given better results then using the historical meteorological series.

A Sensor Driven Probabilistic Method for Enabling Hyper Resolution Flood Simulations

NASA Astrophysics Data System (ADS)

Fries, K. J.; Salas, F.; Kerkez, B.

2016-12-01

A reduction in the cost of sensors and wireless communications is now enabling researchers and local governments to make flow, stage and rain measurements at locations that are not covered by existing USGS or state networks. We ask the question: how should these new sources of densified, street-level sensor measurements be used to make improved forecasts using the National Water Model (NWM)? Assimilating these data "into" the NWM can be challenging due to computational complexity, as well as heterogeneity of sensor and other input data. Instead, we introduce a machine learning and statistical framework that layers these data "on top" of the NWM outputs to improve high-resolution hydrologic and hydraulic forecasting. By generalizing our approach into a post-processing framework, a rapidly repeatable blueprint is generated for for decision makers who want to improve local forecasts by coupling sensor data with the NWM. We present preliminary results based on case studies in highly instrumented watersheds in the US. Through the use of statistical learning tools and hydrologic routing schemes, we demonstrate the ability of our approach to improve forecasts while simultaneously characterizing bias and uncertainty in the NWM.

The Influence Analysis of the Rainfall Meteorological Conditions on the Operation of the Balloon Borne Radar in Plateau

NASA Astrophysics Data System (ADS)

Li, Qiong; Geng, Fangzhi

2018-03-01

Based on the characteristics of complex terrain and different seasons’ weather in Qinghai Tibet Plateau, through statistic the daily rainfall that from 2002 to 2012, nearly 11 years, by Bomi meteorological station, Bomi area rainfall forecast model is established, and which can provide the basis forecasting for dangerous weather warning system on the balloon borne radar in the next step, to protect the balloon borne radar equipment’s safety work and combat effectiveness.

Domestic & International Air Cargo Activity: National and Selected Hub Forecasts.

DTIC Science & Technology

1979-11-01

111371 1991 1887811 2?. 768 :297968 Forecast utilizes 1972 dollar GNP from Wharton’s annual model, December 6, 1978, Post-Meeting Control Solution...mile based on 1973 revenue ton-miles reported in the DOT/CAB, Air Carrier Traffic Statistics. South America - RSA - simple average of American (Latin...9518 F (2/11) = 129.347 DW = 1.41 (b) South America (ESA) 4 = 11.8926 + 18.2908* (GDPSA.C) - 8.94307* ( RSA ) 4 (0.14) (6.08) (-0.46) R2 .8717 F (2/11

Verification of Meteorological and Oceanographic Ensemble Forecasts in the U.S. Navy

NASA Astrophysics Data System (ADS)

Klotz, S.; Hansen, J.; Pauley, P.; Sestak, M.; Wittmann, P.; Skupniewicz, C.; Nelson, G.

2013-12-01

The Navy Ensemble Forecast Verification System (NEFVS) has been promoted recently to operational status at the U.S. Navy's Fleet Numerical Meteorology and Oceanography Center (FNMOC). NEFVS processes FNMOC and National Centers for Environmental Prediction (NCEP) meteorological and ocean wave ensemble forecasts, gridded forecast analyses, and innovation (observational) data output by FNMOC's data assimilation system. The NEFVS framework consists of statistical analysis routines, a variety of pre- and post-processing scripts to manage data and plot verification metrics, and a master script to control application workflow. NEFVS computes metrics that include forecast bias, mean-squared error, conditional error, conditional rank probability score, and Brier score. The system also generates reliability and Receiver Operating Characteristic diagrams. In this presentation we describe the operational framework of NEFVS and show examples of verification products computed from ensemble forecasts, meteorological observations, and forecast analyses. The construction and deployment of NEFVS addresses important operational and scientific requirements within Navy Meteorology and Oceanography. These include computational capabilities for assessing the reliability and accuracy of meteorological and ocean wave forecasts in an operational environment, for quantifying effects of changes and potential improvements to the Navy's forecast models, and for comparing the skill of forecasts from different forecast systems. NEFVS also supports the Navy's collaboration with the U.S. Air Force, NCEP, and Environment Canada in the North American Ensemble Forecast System (NAEFS) project and with the Air Force and the National Oceanic and Atmospheric Administration (NOAA) in the National Unified Operational Prediction Capability (NUOPC) program. This program is tasked with eliminating unnecessary duplication within the three agencies, accelerating the transition of new technology, such as multi-model ensemble forecasting, to U.S. Department of Defense use, and creating a superior U.S. global meteorological and oceanographic prediction capability. Forecast verification is an important component of NAEFS and NUOPC. Distribution Statement A: Approved for Public Release; distribution is unlimited