Search NASA⌕ Search

SEARCH · Search NASA

Results for “Performance Modeling”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,657 records · Page 92

Automatic Waveform Quality Control for Surface Waves Using Machine Learning

Surface-wave seismograms are widely used by researchers to study Earth’s interior and earthquakes. To extract information reliably and robustly from a suite of surface waveforms, the signals require quality control screening to reduce artifacts from signal complexity and noise. This process has usually been completed by human experts labeling each waveform visually, which is time consuming and tedious for large data sets. We explore automated approaches to improve the efficiency of waveform quality control processing by investigating logistic regression, support vector machines, K-nearest neighbors, random forests (RF), and artificial neural networks (ANN) algorithms. To speed up signal quality assessment, we trained these five machine learning (ML) methods using nearly 400,000 human-labeled waveforms. The ANN and RF models outperformed other algorithms and achieved a test accuracy of 92%. We evaluated these two best-performing models using seismic events from geographic regions not used for training. The results show that the two trained models agree with labels from human analysts but required only 0.4% of the time. Although the original (human) quality assignments assessed general waveform signal-to-noise, the ANN or RF labels can help facilitate detailed waveform analysis. Our investigations demonstrate the capability of the automated processing using these two ML models to reduce outliers in surface-wave-related measurements without human quality control screening.

58 GEOSCIENCES↗

Final Report for Cloud-Aerosol Physics in Super-Parameterized Atmospheric Regional Climate Simulations (CAP-SPARCS)(DE-SC0002003) for 8/15/2009 through 8/14/2012

Improving the representation of local and non-local aerosol interactions in state-of-the-science regional climate models is a priority for the coming decade (Zhang, 2008). With this aim in mind, we have combined two new technologies that have a useful synergy: (1) an aerosol-enabled regional climate model (Advanced Weather Research and Forecasting Model with Chemistry WRF-Chem), whose primary weakness is a lack of high quality boundary conditions and (2) an aerosol-enabled multiscale modeling framework (PNNL Multiscale Aerosol Climate Model (MACM)), which is global but captures aerosol-convection-cloud feedbacks, and thus an ideal source of boundary conditions. Combining these two approaches has resulted in an aerosol-enabled modeling framework that not only resolves high resolution details in a particular region, but crucially does so within a global context that is similarly faithful to multi-scale aerosol-climate interactions. We have applied and improved the representation of aerosol interactions by evaluating model performance over multiple domains, with (1) an extensive evaluation of mid-continent precipitation representation by multiscale modeling, (2) two focused comparisons to transport of aerosol plumes to the eastern United States for comparison with observations made as part of the International Consortium for Atmospheric Research on Transport and Transformation (ICARTT), with the first being idealized and the second being linked to an extensive wildfire plume, and (3) the extension of these ideas to the development of a new approach to evaluating aerosol indirect effects with limited-duration model runs by nudging to observations. This research supported the work of one postdoc (Zhan Zhao) for two years and contributed to the training and research of two graduate students. Four peer-reviewed publications have resulted from this work, and ground work for a follow-on project was completed.

54 ENVIRONMENTAL SCIENCES↗

Meeting Global Health Needs via Infectious Disease Forecasting: Development of a Reliable Data-Driven Framework

Infectious diseases (IDs) have a significant detrimental impact on global health. Timely and accurate ID forecasting can result in more informed implementation of control measures and prevention policies. To meet the operational decision-making needs of real-world circumstances, we aimed to build a standardized, reliable, and trustworthy ID forecasting pipeline and visualization dashboard that is generalizable across a wide range of modeling techniques, IDs, and global locations. We forecasted 6 diverse, zoonotic diseases (brucellosis, campylobacteriosis, Middle East respiratory syndrome, Q fever, tick-borne encephalitis, and tularemia) across 4 continents and 8 countries. We included a wide range of statistical, machine learning, and deep learning models (n=9) and trained them on a multitude of features (average n=2326) within the One Health landscape, including demography, landscape, climate, and socioeconomic factors. The pipeline and dashboard were created in consideration of crucial operational metrics—prediction accuracy, computational efficiency, spatiotemporal generalizability, uncertainty quantification, and interpretability—which are essential to strategic data-driven decisions. While no single best model was suitable for all disease, region, and country combinations, our ensemble technique selects the best-performing model for each given scenario to achieve the closest prediction. For new or emerging diseases in a region, the ensemble model can predict how the disease may behave in the new region using a pretrained model from a similar region with a history of that disease. The data visualization dashboard provides a clean interface of important analytical metrics, such as ID temporal patterns, forecasts, prediction uncertainties, and model feature importance across all geographic locations and disease combinations. As the need for real-time, operational ID forecasting capabilities increases, this standardized and automated platform for data collection, analysis, and reporting is a major step forward in enabling evidence-based public health decisions and policies for the prevention and mitigation of future ID outbreaks.

60 APPLIED LIFE SCIENCES↗

Methods for Incorporating Model Uncertainty into Exoplanet Atmospheric Analysis

A key goal of exoplanet spectroscopy is to measure atmospheric properties, such as abundances of chemical species, in order to connect them to our understanding of atmospheric physics and planet formation. In this new era of high-quality JWST data, it is paramount that these measurement methods are robust. When comparing atmospheric models to observations, multiple candidate models may produce reasonable fits to the data. Typically, conclusions are reached by selecting the best-performing model according to some metric. This ignores model uncertainty in favor of specific model assumptions, potentially leading to measured atmospheric properties that are overconfident and/or incorrect. In this paper, we compare three ensemble methods for addressing model uncertainty by combining posterior distributions from multiple analyses: Bayesian model averaging, a variant of Bayesian model averaging using leave-one-out predictive densities, and stacking of predictive distributions. We demonstrate these methods by fitting the Hubble Space Telescope (HST) + Spitzer transmission spectrum of the hot Jupiter HD 209458b using models with different cloud and haze prescriptions. All of our ensemble methods lead to uncertainties on retrieved parameters that are larger but more realistic and consistent with physical and chemical expectations. Since they have not typically accounted for model uncertainty, uncertainties of retrieved parameters from HST spectra have likely been underreported. We recommend stacking as the most robust model combination method. Our methods can be used to combine results from independent retrieval codes and from different models within one code. They are also widely applicable to other exoplanet analysis processes, such as combining results from different data reductions.

79 ASTRONOMY AND ASTROPHYSICS↗

Ultrasonic Transducer Irradiation Test Results

Ultrasonic technologies offer the potential for high accuracy and resolution in-pile measurement of a range of parameters, including geometry changes, temperature, crack initiation and growth, gas pressure and composition, and microstructural changes. Many Department of Energy-Office of Nuclear Energy (DOE-NE) programs are exploring the use of ultrasonic technologies to provide enhanced sensors for in-pile instrumentation during irradiation testing. For example, the ability of small diameter ultrasonic thermometers (UTs) to provide a temperature profile in candidate metallic and oxide fuel would provide much needed data for validating new fuel performance models. These efforts are limited by the lack of identified ultrasonic transducer materials capable of long term performance under irradiation test conditions. To address this need, the Pennsylvania State University (PSU) was awarded an Advanced Test Reactor National Scientific User Facility (ATR NSUF) project to evaluate the performance of promising magnetostrictive and piezoelectric transducers in the Massachusetts Institute of Technology Research Reactor (MITR) up to a fast fluence of at least 1021 n/cm2 . A multi-National Laboratory collaboration funded by the Nuclear Energy Enabling Technologies Advanced Sensors and Instrumentation (NEET ASI) program also provided initial support for this effort. This irradiation, which started in February 2014, is an instrumented lead test and real-time transducer performance data are collected along with temperature and neutron and gamma flux data. The irradiation is ongoing and will continue to approximately mid-2015. To date, very encouraging results have been attained as several transducers continue to operate under irradiation.

Daw, Joshua↗

Display/control requirements for automated VTOL aircraft

A systematic design methodology for pilot displays in advanced commercial VTOL aircraft was developed and refined. The analyst is provided with a step-by-step procedure for conducting conceptual display/control configurations evaluations for simultaneous monitoring and control pilot tasks. The approach consists of three phases: formulation of information requirements, configuration evaluation, and system selection. Both the monitoring and control performance models are based upon the optimal control model of the human operator. Extensions to the conventional optimal control model required in the display design methodology include explicit optimization of control/monitoring attention; simultaneous monitoring and control performance predictions; and indifference threshold effects. The methodology was applied to NASA's experimental CH-47 helicopter in support of the VALT program. The CH-47 application examined the system performance of six flight conditions. Four candidate configurations are suggested for evaluation in pilot-in-the-loop simulations and eventual flight tests.

Hoffman, W. C.↗

Validation for Global Solar Wind Prediction Using Ulysses Comparison: Multiple Coronal and Heliospheric Models Installed at the Community Coordinated Modeling Center

The prediction of the background global solar wind is a necessary part of space weather forecasting. Several coronal and heliospheric models have been installed and/or recently upgraded at the Community Coordinated Modeling Center (CCMC), including the Wang-Sheely-Arge (WSA)-Enlil model, MHD-Around-a-Sphere (MAS)-Enlil model, Space Weather Modeling Framework (SWMF), and Heliospheric tomography using interplanetary scintillation data. Ulysses recorded the last fast latitudinal scan from southern to northern poles in 2007. By comparing the modeling results with Ulysses observations over seven Carrington rotations, we have extended our third-party validation from the previous near-Earth solar wind to middle to high latitudes, in the same late declining phase of solar cycle 23. Besides visual comparison, wehave quantitatively assessed the models capabilities in reproducing the time series, statistics, and latitudinal variations of solar wind parameters for a specific range of model parameter settings, inputs, and grid configurations available at CCMC. The WSA-Enlil model results vary with three different magnetogram inputs.The MAS-Enlil model captures the solar wind parameters well, despite its underestimation of the speed at middle to high latitudes. The new version of SWMF misses many solar wind variations probably because it uses lower grid resolution than other models. The interplanetary scintillation-tomography cannot capture the latitudinal variations of solar wind well yet. Because the model performance varies with parameter settings which are optimized for different epochs or flow states, the performance metric study provided here can serve as a template that researchers can use to validate the models for the time periods and conditions of interest to them.

Jian, L. K.↗

Climate Model Diagnostic Analyzer

The comprehensive and innovative evaluation of climate models with newly available global observations is critically needed for the improvement of climate model current-state representation and future-state predictability. A climate model diagnostic evaluation process requires physics-based multi-variable analyses that typically involve large-volume and heterogeneous datasets, making them both computation- and data-intensive. With an exploratory nature of climate data analyses and an explosive growth of datasets and service tools, scientists are struggling to keep track of their datasets, tools, and execution/study history, let alone sharing them with others. In response, we have developed a cloud-enabled, provenance-supported, web-service system called Climate Model Diagnostic Analyzer (CMDA). CMDA enables the physics-based, multivariable model performance evaluations and diagnoses through the comprehensive and synergistic use of multiple observational data, reanalysis data, and model outputs. At the same time, CMDA provides a crowd-sourcing space where scientists can organize their work efficiently and share their work with others. CMDA is empowered by many current state-of-the-art software packages in web service, provenance, and semantic search.

cloud computing↗

Modeling Quiet Solar Luminosity Variability from TSI Satellite Measurements and Proxy Models During 1980–2018

A continuous record of direct total solar irradiance (TSI) observations began with a series of satellite experiments in 1978. This record requires comparisons of overlapping satellite observations with adequate relative precisions to provide useful long term TSI trend information. Herein we briefly review the active cavity radiometer irradiance monitor physikalisch-meteorologisches observatorium davos (ACRIM-PMOD) TSI composite controversy regarding how the total solar irradiance (TSI) has evolved since 1978 and about whether TSI significantly increased or slightly decreased from 1980 to 2000. The main question is whether TSI increased or decreased during the so-called ACRIM-gap period from 1989 to 1992. There is significant discrepancy between TSI proxy models and observations before and after the gap, which requires a careful revisit of the data analysis and modeling performed during the ACRIM-gap period. In this study, we use three recently proposed TSI proxy models that do not present any TSI increase during the ACRIM-gap, and show that they agree with the TSI data only from 1996 to 2016. However, these same models significantly diverge from the observations from 1981 and 1996. Thus, the scaling errors must be different between the two periods, which suggests errors in these models. By adjusting the TSI proxy models to agree with the data patterns before and after the ACRIM-gap, we found that these models miss a slowly varying TSI component. The adjusted models suggest that the quiet solar luminosity increased from the 1986 to the 1996 TSI minimum by about 0.45 W/sq.m reaching a peak near 2000 and decreased by about 0.15 W/sq.m from the 1996 to the 2008 TSI cycle minimum. This pattern is found to be compatible with the ACRIM TSI composite and confirms the ACRIM TSI increasing trend from 1980 to 2000, followed by a long-term decreasing trend since.

Scafetta, Nicola↗

Modeling Quiet Solar Luminosity Variability from TSI Satellite Measurements and Proxy Models during 1980-2018

A continuous record of direct total solar irradiance (TSI) observations began with a series of satellite experiments in 1978. This record requires comparisons of overlapping satellite observations with adequate relative precisions to provide useful long term TSI trend information. Herein we briefly review the active cavity radiometer irradiance monitor physikalisch-meteorologisches observatorium davos (ACRIM-PMOD) TSI composite controversy regarding how the total solar irradiance (TSI) has evolved since 1978 and about whether TSI significantly increased or slightly decreased from 1980 to 2000. The main question is whether TSI increased or decreased during the so-called ACRIM-gap period from 1989 to 1992. There is significant discrepancy between TSI proxy models and observations before and after the gap, which requires a careful revisit of the data analysis and modeling performed during the ACRIM-gap period. In this study, we use three recently proposed TSI proxy models that do not present any TSI increase during the ACRIM-gap, and show that they agree with the TSI data only from 1996 to 2016. However, these same models significantly diverge from the observations from 1981 and 1996. Thus, the scaling errors must be different between the two periods, which suggests errors in these models. By adjusting the TSI proxy models to agree with the data patterns before and after the ACRIM-gap, we found that these models miss a slowly varying TSI component. The adjusted models suggest that the quiet solar luminosity increased from the 1986 to the 1996 TSI minimum by about 0.45 W/m2 reaching a peak near 2000 and decreased by about 0.15 W/m2 from the 1996 to the 2008 TSI cycle minimum. This pattern is found to be compatible with the ACRIM TSI composite and confirms the ACRIM TSI increasing trend from 1980 to 2000, followed by a long-term decreasing trend since.

total solar irradiance↗

Updated Results of Ultrasonic Transducer Irradiation Test

Ultrasonic technologies offer the potential for high accuracy and resolution in-pile measurement of a range of parameters, including geometry changes, temperature, crack initiation and growth, gas pressure and composition, and microstructural changes. Many Department of Energy-Office of Nuclear Energy (DOE-NE) programs are exploring the use of ultrasonic technologies to provide enhanced sensors for in-pile instrumentation during irradiation testing. For example, the ability of small diameter ultrasonic thermometers (UTs) to provide a temperature profile in candidate metallic and oxide fuel would provide much needed data for validating new fuel performance models. These efforts are limited by the lack of identified ultrasonic transducer materials capable of long term performance under irradiation test conditions. To address this need, the Pennsylvania State University (PSU) was awarded an Advanced Test Reactor National Scientific User Facility (ATR NSUF) project to evaluate the performance of promising magnetostrictive and piezoelectric transducers in the Massachusetts Institute of Technology Research Reactor (MITR) up to a fast fluence of at least 10{sup 21} n/cm{sup 2}. A multi-National Laboratory collaboration funded by the Nuclear Energy Enabling Technologies Advanced Sensors and Instrumentation (NEET-ASI) program also provided initial support for this effort. This irradiation, which started in February 2014, is an instrumented lead test and real-time transducer performance data are collected along with temperature and neutron and gamma flux data. The irradiation is ongoing and will continue to approximately mid-2015. To date, very encouraging results have been attained as several transducers continue to operate under irradiation. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Using Ultrasound Image Augmentation and Ensemble Predictions to Prevent Machine-Learning Model Overfitting

Deep learning predictive models have the potential to simplify and automate medical imaging diagnostics by lowering the skill threshold for image interpretation. However, this requires predictive models that are generalized to handle subject variability as seen clinically. Here, we highlight methods to improve test accuracy of an image classifier model for shrapnel identification using tissue phantom image sets. Using a previously developed image classifier neural network—termed ShrapML—blind test accuracy was less than 70% and was variable depending on the training/test data setup, as determined by a leave one subject out (LOSO) holdout methodology. Introduction of affine transformations for image augmentation or MixUp methodologies to generate additional training sets improved model performance and overall accuracy improved to 75%. Further improvements were made by aggregating predictions across five LOSO holdouts. This was done by bagging confidences or predictions from all LOSOs or the top-3 LOSO confidence models for each image prediction. Top-3 LOSO confidence bagging performed best, with test accuracy improved to greater than 85% accuracy for two different blind tissue phantoms. This was confirmed by gradient-weighted class activation mapping to highlight that the image classifier was tracking shrapnel in the image sets. Overall, data augmentation and ensemble prediction approaches were suitable for creating more generalized predictive models for ultrasound image analysis, a critical step for real-time diagnostic deployment.

60 APPLIED LIFE SCIENCES↗

A performance and reliability model for thermal barrier coatings

A modeling technique for predicting the performance and reliability of TBC's is being developed at Solar Turbines Incorporated. The concept combines experimental coating property data with finite element analyses to predict the thermal and mechanical behavior of coating systems in service. A key feature of Solar's approach is the use of a four point flexure test to estimate coating strength distributions and to predict coating failure probability. This model was used to evaluate the effect of physical variations on coating performance in high heat flux rocket engine applications for NASA. Current work, promoted by Caterpillar Tractor Company for diesel engine applications, is being conducted to measure coating strength as a function of temperature, and future work will document strength degradation with time at temperature. Solar's interest lies in the application of TBCs to gas turbine engine components.

Batakis, A. P.↗

Tokamak divertor plasma emulation with machine learning

Abstract Future tokamak devices that aim to create conditions relevant to power plant operations must consider strategies for mitigating damage to plasma facing components in the divertor. One of the goals of MAST-U tokamak operations is to inform these considerations by researching advanced divertor configurations that aid stable plasma detachment. Machine design, scenario planning and detachment control would all greatly benefit from tools that enable rapid calculation of scenario-relevant quantities given some input parameters. This paper presents a method for generating large, simulated scrape-off layer data sets, which was applied to generate a data set of steady-state Hermes-3 simulations of the MAST-U tokamak. A machine learning model was constructed using a Bayesian approach to hyperparameter optimisation to predict diagnosable output quantities given control-relevant input features. The resulting best-performing model, which is based on a feedforward neural network, achieves high accuracy when predicting electron temperature at the divertor target and carbon impurity radiation front position and runs in around 1 ms in inference mode. Techniques for interpreting the predictions made by the model were applied, and a high-resolution parameter scan of upstream conditions was performed to demonstrate the utility of rapidly generating accurate predictions using the emulator. This work represents a step forward in the design of machine learning-driven emulators of tokamak exhaust simulation codes in operational modes relevant to divertor detachment control and plasma scenario design.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Duct Leakage Modeling in EnergyPlus and Analysis of Energy Savings from Implementing SAV with InCITeTM

This project addressed two significant deficiencies in air-handling systems for large commercial building: duct leakage and duct static pressure reset. Both constitute significant energy reduction opportunities for these buildings. The overall project goal is to bridge the gaps in current duct performance modeling capabilities, and to expand our understanding of air-handling system performance in California large commercial buildings. The purpose of this project is to provide technical support for the implementation of a duct leakage modeling capability in EnergyPlus, to demonstrate the capabilities of the new model, and to carry out analyses of field measurements intended to demonstrate the energy saving potential of the SAV with InCITeTM duct static pressure reset (SPR) technology. A new duct leakage model has been successfully implemented in EnergyPlus, which will enable simulation users to assess the impacts of leakage on whole-building energy use and operation in a coupled manner. This feature also provides a foundation to support code change proposals and compliance analyses related to Title 24 where duct leakage is an issue. Our example simulations continue to show that leaky ducts substantially increase fan power: 10percent upstream and 10percent downstream leakage increases supply fan power 30percent on average compared to a tight duct system (2.5percent upstream and 2.5percent downstream leakage). Much of this increase is related to the upstream leakage rather than to the downstream leakage. This does not mean, however, that downstream leakage is unimportant. Our simulations also demonstrate that ceiling heat transfer is a significant effect that needs to be included when assessing the impacts of duct leakage in large commercial buildings. This is not particularly surprising, given that ?ceiling regain? issues have already been included in residential analyses as long as a decade ago (e.g., ASHRAE Standard 152); mainstream simulation programs that are used for large commercial building energy analyses have not had this capability until now. Our analyses of data that we collected during our 2005 tests of the SAV with InCITeTM duct static pressure reset technology show that this technology can substantially reduce fan power (in this case, by about 25 to 30percent). Tempering this assessment, however, is that cooling and heating coil loads were observed to increase or decrease significantly depending on the time window used. Their impact on cooling and heating plant power needs to be addressed in future studies; without translating the coil loads to plant equipment energy use, it is not possible to judge the net impact of this SPR technology on whole-building energy use. If all of the loads had decreased, such a step would not be as necessary.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Validation of PV-RPM Code in the System Advisor Model.

This paper describes efforts made by Sandia National Laboratories (SNL) and the National Renewable Energy Laboratory (NREL) to validate the SNL developed PV Reliability Performance Model (PV - RPM) algorithm as implemented in the NREL System Advisor Model (SAM). The PV - RPM model is a library of functions that estimates component failure and repair in a photovoltaic system over a desired simulation period. The failure and repair distributions in this paper are probabilistic representations of component failure and repair based on data collected by SNL for a PV power plant operating in Arizona. The validation effort focuses on whether the failure and repair dist ributions used in the SAM implementation result in estimated failures that match the expected failures developed in the proof - of - concept implementation. Results indicate that the SAM implementation of PV - RPM provides the same results as the proof - of - concep t implementation, indicating the algorithms were reproduced successfully.

97 MATHEMATICS AND COMPUTING↗

Supercritical CO 2 Water Displacements and CO 2 Capillary Trapping: Micromodel Experiment and Numerical Simulation

We report that the CO 2 capillary trapping is an important scientific issue in geological carbon sequestration, but few researches focus on the trapping mechanism at pore scale under supercritical CO 2 condition. In this study, based on the high-pressure fluids-microscopy-micromodel experimental system, we performed drainage experiment, i.e. supercritical CO 2 displacing water, and imbibition experiment, i.e. water displacing CO 2 , under the conditions of 45°C and 8.5 MPa. The DSLR camera was used to capture pictures of CO 2 -water two-phase immiscible flow and the microscopy was used to capture the capillary trapping behavior for the supercritical CO 2 at the pore scale. The computational fluid dynamic method was adopted to simulate the two-phase fluid flow processes. The numerical results are generally in agree ment with the experimental observations, and further provide three-dimensional geometries on the interface during the drainage-imbibition processes and the trapped supercritical CO 2 droplet/cluster. Lastly, the capillary trapping curve, i.e. the relationship between the initial CO 2 saturation and the residual saturation, was obtained from the numerical results, and we made an assessment of the three capillary trapping models, i.e. Land's, Jurauld's and Spiteri's trapping models. A comparison of the models performance indicates that Jurauld's model behaves slightly better than Land's model, whereas Spiteri's model behaves poorly. However, given that Land's model only contains one parameter of clear physical meaning, it is recommended for practical use.

58 GEOSCIENCES↗

Effects of error covariance structure on estimation of model averaging weights and predictive performance

When conducting model averaging for assessing groundwater conceptual model uncertainty, the averaging weights are often evaluated using model selection criteria such as AIC, AICc, BIC, and KIC (Akaike Information Criterion, Corrected Akaike Information Criterion, Bayesian Information Criterion, and Kashyap Information Criterion, respectively). However, this method often leads to an unrealistic situation in which the best model receives overwhelmingly large averaging weight (close to 100%), which cannot be justified by available data and knowledge. It was found in this study that this problem was caused by using the covariance matrix, CE, of measurement errors for estimating the negative log likelihood function common to all the model selection criteria. This problem can be resolved by using the covariance matrix, Cek, of total errors (including model errors and measurement errors) to account for the correlation between the total errors. An iterative two-stage method was developed in the context of maximum likelihood inverse modeling to iteratively infer the unknown Cek from the residuals during model calibration. The inferred Cek was then used in the evaluation of model selection criteria and model averaging weights. While this method was limited to serial data using time series techniques in this study, it can be extended to spatial data using geostatistical techniques. The method was first evaluated in a synthetic study and then applied to an experimental study, in which alternative surface complexation models were developed to simulate column experiments of uranium reactive transport. It was found that the total errors of the alternative models were temporally correlated due to the model errors. The iterative two-stage method using Cek resolved the problem that the best model receives 100% model averaging weight, and the resulting model averaging weights were supported by the calibration results and physical understanding of the alternative models. Using Cek obtained from the iterative two-stage method also improved predictive performance of the individual models and model averaging in both synthetic and experimental studies.

error correlation↗