Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Models”

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 397 records · Page 22

Making Steppingstones out of Stumbling Blocks: A Bayesian Model Evidence Estimator with Application to Groundwater Transport Model Selection

Bayesian model evidence (BME) is a measure of the average fit of a model to observation data given all the parameter values that the model can assume. By accounting for the trade-off between goodness-of-fit and model complexity, BME is used for model selection and model averaging purposes. For strict Bayesian computation, the theoretically unbiased Monte Carlo based numerical estimators are preferred over semi-analytical solutions. This study examines five BME numerical estimators and asks how accurate estimation of the BME is important for penalizing model complexity. The limiting cases for numerical BME estimators are the prior sampling arithmetic mean estimator (AM) and the posterior sampling harmonic mean (HM) estimator, which are straightforward to implement, yet they result in underestimation and overestimation, respectively. We also consider the path sampling methods of thermodynamic integration (TI) and steppingstone sampling (SS) that sample multiple intermediate distributions that link the prior and the posterior. Although TI and SS are theoretically unbiased estimators, they could have a bias in practice arising from numerical implementation. For example, sampling errors of some intermediate distributions can introduce bias. We propose a variant of SS, namely the multiple one-steppingstone sampling (MOSS) that is less sensitive to sampling errors. We evaluate these five estimators using a groundwater transport model selection problem. SS and MOSS give the least biased BME estimation at an efficient computational cost. If the estimated BME has a bias that covariates with the true BME, this would not be a problem because we are interested in BME ratios and not their absolute values. On the contrary, the results show that BME estimation bias can be a function of model complexity. Thus, biased BME estimation results in inaccurate penalization of more complex models, which changes the model ranking. This was less observed with SS and MOSS as with the three other methods.

54 ENVIRONMENTAL SCIENCES↗

Evaluation of experimentally measured and model-calculated pH for rock-brine-CO2 systems under geologic CO2 sequestration conditions

pH is an essential parameter for understanding the geochemical reactions that occur in rock-brine-CO2 systems when CO2 is injected into deep geologic formations for long-term storage. Due to a lack of reliable experimental methods, most laboratory studies conducted under geological CO2 sequestration (GCS) conditions have relied on thermodynamic modeling to estimate pH. The accuracy of these model predictions is typically uncertain. In our previous work, we have developed a method for pH determination by in-situ spectrophotometry. In the present work, we expanded the applicable pH range for this method and measured the pH of several rock-brine-CO2 systems at GCS conditions for five rock samples collected from ongoing GCS demonstration projects. Experimental measurements were compared with pH values calculated using several geochemical modeling approaches. The effect of different thermodynamic databases on the accuracy of model prediction was evaluated. Results indicate that the accuracy of model calculations is rock-dependent. For rocks comprised of carbonate and sandstone, model results generally agreed well with experimentally measured pH; however, for basalt, significant differences were observed. These discrepancies may be due to the models’ failure to fully account for certain reaction occurring between the basalt minerals the CO2-saturated brine solutions.

pH, carbon sequestration↗

Application-specific optimal model weighting of global climate models: A red tide example

Global climate models (GCMs) and Earth system models (ESMs) provide many climate services with environmental relevance. The High Resolution Model Inter-comparison Project (HighResMIP) of the Coupled Model Intercomparison Project Phase 6 (CMIP6) provides model runs of GCMs and ESMs to address regional phenomena. Developing a parsimonious ensemble of CMIP6 requires multiple ensemble methods such as independent-model subset selection, prescreening-based subset selection, and model weighting. The work presented here focuses on application-specific optimal model weighting, with prescreening-based subset selection. As such, independent ensemble members are categorized, selected, and weighted based on their ability to reproduce physically-interpretable features of interest that are problem-specific. We discuss the strengths and caveats of optimal model weighting using a case study of red tide prediction in the Gulf of Mexico along the West Florida Shelf. Red tide is a common name of specific harmful algal blooms that occur worldwide, causing adverse socioeconomic and environmental impacts. Our results indicate the importance of prescreening-based subset selection as optimal model weighting can underplay robust ensemble members by optimizing error cancellation. Prescreening-based subset selection also provides insights about the validity of the model weights. By illustrating the caveats of using non-representative models when optimal model weighting is used, the findings and discussion of this study are pertinent to many other climate services.

54 ENVIRONMENTAL SCIENCES↗

North American Carbon Project (NACP) Regional Interim Synthesis: Terrestrial Biospheric Model Intercomparison

Understanding of carbon exchange between terrestrial ecosystems and the atmosphere can be improved through direct observations and experiments, as well as through modeling activities. Terrestrial biosphere models (TBMs) have become an integral tool for extrapolating local observations and understanding to much larger terrestrial regions. Although models vary in their specific goals and approaches, their central role within carbon cycle science is to provide a better understanding of the mechanisms currently controlling carbon exchange. Recently, the North American Carbon Program (NACP) organized several interim-synthesis activities to evaluate and inter-compare models and observations at local to continental scales for the years 2000 to 2005. Here, we compare the results from the TBMs collected as part of the regional and continental interim-synthesis (RCIS) activities. The primary objective of this work is to synthesize and compare the 19 participating TBMs to assess current understanding of the terrestrial carbon cycle in North America.

54 ENVIRONMENTAL SCIENCES↗

Modeling Anti-HIV Activity of HEPT Derivatives Revisited. Multiregression Models Are Not Inferior Ones

Several quantitative structure-activity studies for this data set containing 107 HEPT derivatives have been performed since 1997, using the same set of molecules by (more or less) different classes of molecular descriptors. Multivariate Regression (MR) and Artificial Neural Network (ANN) models were developed and in each study the authors concluded that ANN models are superior to MR ones. We re-calculated multivariate regression models for this set of molecules using the same set of descriptors, and compared our results with the previous ones. Two main reasons for overestimation of the quality of the ANN models in previous studies comparing with MR models are: (1) wrong calculation of leave-one-out (LOO) cross-validated (CV) correlation coefficient for MR models in Luco et al., J. Chem. Inf. Comput. Sci. 37 392-401 (1997), and (2) incorrect estimation/interpretation of leave-one-out (LOO) cross-validated and predictive performance and power of ANN models. More precise and fairer comparison of fit and LOO CV statistical parameters shows that MR models are more stable. In addition, MR models are much simpler than ANN ones. For real testing the predictive performance of both classes of models we need more HEPT derivatives, because all ANN models that presented results for external set of molecules used experimental values in optimization of modeling procedure and model parameters.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Numerical modeling of the solar wind flow with observational boundary conditions

In this paper we describe our group efforts to develop a self-consistent, data-drivenmodel of the solar wind (SW) interaction with the local interstellar medium. The motion of plasma in this model is described with the MHD approach, while the transport of neutral atoms is addressed by either kinetic or multi-fluid equations. The model and its implementation in the Multi-Scale Fluid-Kinetic Simulation Suite (MS-FLUKSS) are continuously tested and validated by comparing our results with other models and spacecraft measurements. In particular, it was successfully applied to explain an unusual SW behavior discovered by the Voyager 1 spacecraft, i.e., the development of a substantial negative radial velocity component, flow turning in the transverse direction, while the latitudinal velocity component goes to very small values. We explain recent SW velocity measurements at Voyager 1 in the context of our 3-D, MHD modeling. We also present a comparison of different turbulence models in their ability to reproduce the SW temperature profile from Voyager 2 measurements. The boundary conditions obtained at 50 solar radii from data-driven numerical simulations are used to model a CME event throughout the heliosphere.

79 ASTRONOMY AND ASTROPHYSICS↗

Phase space effects on fast ion distribution function modeling in tokamaks

Integrated simulations of tokamak discharges typically rely on classical physics to model energetic particle (EP) dynamics. However, there are numerous cases in which energetic particles can suffer additional transport that is not classical in nature. Examples include transport by applied 3D magnetic perturbations and, more notably, by plasma instabilities. Focusing on the effects of instabilities, ad-hoc models can empirically reproduce increased transport, but the choice of transport coefficients is usually somehow arbitrary. New approaches based on physics-based reduced models are being developed to address those issues in a simplified way, while retaining a more correct treatment of resonant wave-particle interactions. The kick model implemented in the tokamak transport code TRANSP is an example of such reduced models. It includes modifications of the EP distribution by instabilities in real and velocity space, retaining correlations between transport in energy and space typical of resonant EP transport. The relevance of EP phase space modifications by instabilities is first discussed in terms of predicted fast ion distribution. Results are compared with those from a simple, ad-hoc diffusive model. It is then shown that the phase-space resolved model can also provide additional insight into important issues such as internal consistency of the simulations and mode stability through the analysis of the power exchanged between energetic particles and the instabilities.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Structure of dark matter halos in warm dark matter models and in models with long-lived charged massive particles

We study the formation of non-linear structures in warm dark matter (WDM) models and in a long-lived charged massive particle (CHAMP) model. CHAMPs with a decay lifetime of about 1 yr induce characteristic suppression in the matter power spectrum at subgalactic scales through acoustic oscillations in the thermal background. We explore structure formation in such a model. We also study three WDM models, where the dark matter particles are produced through the following mechanisms: i) WDM particles are produced in the thermal background and then kinematically decoupled; ii) WDM particles are fermions produced by the decay of thermal heavy bosons; and iii) WDM particles are produced by the decay of non-relativistic heavy particles. We show that the linear matter power spectra for the three models are all characterised by the comoving Jeans scale at the matter-radiation equality. Furthermore, we can also describe the linear matter power spectrum for the long-lived CHAMP model in terms of a suitably defined characteristic cut-off scale k{sub Ch}, similarly to the WDM models. We perform large cosmological N-body simulations to study the non-linear growth of structures in these four models. We compare the halo mass functions, the subhalo mass functions, and the radial distributions of subhalos in simulated Milky Way-size halos. For the characteristic cut-off scale k{sub cut} = 51 h Mpc{sup −1}, the subhalo abundance ( ∼ 10{sup 9}M{sub sun}) is suppressed by a factor of ∼ 10 compared with the standard ΛCDM model. We then study the models with k{sub cut} ≅ 51, 410, 820 h Mpc{sup −1}, and confirm that the halo and the subhalo abundances and the radial distributions of subhalos are indeed similar between the different WDM models and the long-lived CHAMP model. The result suggests that the cut-off scale k{sub cut} not only characterises the linear power spectra but also can be used to predict the non-linear clustering properties. The radial distribution of subhalos in Milky Way-size halos is consistent with the observed distribution for k{sub cut} ∼ 50−800 h Mpc{sup −1}; such models resolve the so-called ''missing satellite problem''.

79 ASTRONOMY AND ASTROPHYSICS↗

Current developments in soil organic matter modeling and the expansion of model applications. A review

It is important to note that Soil organic matter (SOM) is a great natural resource. It is fundamental to soil and ecosystem functions across a wide range of scales, from site-specific soil fertility and water holding capacity to global biogeochemical cycling. It is also a highly complex material that is sensitive to direct and indirect human impacts. In our SOM research, simulation models play an important role by providing a mathematical framework to integrate, examine, and test the understanding of SOM dynamics. Simulation models of SOM are also increasingly used in more ‘applied’ settings to evaluate human impacts on ecosystem function, and to manage SOM for greenhouse gas mitigation, improved soil health, and sustainable use as a natural resource. Within this context, there is a need to maintain a robust connection between scientific developments in SOM modeling approaches and SOM model applications. This need forms the basis of this review. In this review we first provide an overview of SOM modeling, focusing on SOM theory, data-model integration, and model development as evidenced by a quantitative review of SOM literature. Second, we present the landscape of SOM model applications, focusing on examples in climate change policy. Finally, we conclude by discussing five areas of recent developments in SOM modeling including: (1) microbial roles in SOM stabilization; (2) modeling SOM saturation kinetics; (3) temperature controls on decomposition; (4)SOM dynamics in deep soil layers; and (5)SOM representation in earth system models. Our aim is to comprehensively connect SOM model development to its applications, revealing knowledge gaps in need of focused interdisciplinary attention and exposing pitfalls that, if avoided, can lead to best use of SOM models to support policy initiatives and sustainable land management solutions.

54 ENVIRONMENTAL SCIENCES↗

A revised dosimetric characterization of the model S700 electronic brachytherapy source containing an anode-centering plastic insert and other components not included in the 2006 model

Purpose: The model S700 Axxent electronic brachytherapy source by Xoft, Inc., was characterized by Rivard et al. in 2006. Since then, the source design was modified to include a new insert at the source tip. Current study objectives were to establish an accurate source model for simulation purposes, dosimetrically characterize the new source and obtain its TG-43 brachytherapy dosimetry parameters, and determine dose differences between the original simulation model and the current model S700 source design. Methods: Design information from measurements of dissected model S700 sources and from vendor-supplied CAD drawings was used to aid establishment of an updated Monte Carlo source model, which included the complex-shaped plastic source-centering insert intended to promote water flow for cooling the source anode. These data were used to create a model for subsequent radiation transport simulations in a water phantom. Compared to the 2006 simulation geometry, the influence of volume averaging close to the source was substantially reduced. A track-length estimator was used to evaluate collision kerma as a function of radial distance and polar angle for determination of TG-43 dosimetry parameters. Results for the 50 kV source were determined every 0.1 cm from 0.3 to 15 cm and every 1° from 0° to 180°. Photon spectra in water with 0.1 keV resolution were also obtained from 0.5 to 15 cm and polar angles from 0° to 165°. Simulations were run for 10{sup 10} histories, resulting in statistical uncertainties on the transverse plane of 0.04% at r = 1 cm and 0.06% at r = 5 cm. Results: The dose-rate distribution ratio for the model S700 source as compared to the 2006 model exceeded unity by more than 5% for roughly one quarter of the solid angle surrounding the source, i.e., θ ≥ 120°. The radial dose function diminished in a similar manner as for an {sup 125}I seed, with values of 1.434, 0.636, 0.283, and 0.0975 at 0.5, 2, 5, and 10 cm, respectively. The radial dose function ratio between the current and the 2006 model had a minimum of 0.980 at 0.4 cm, close to the source sheath and for large distances approached 1.014. 2D anisotropy function ratios were close to unity for 50° ≤ θ ≤ 110°, but exceeded 5% for θ < 40° at close distances to the sheath and exceeded 15% for θ > 140°, even at large distances. Photon energy fluence of the updated model as compared to the 2006 model showed a decrease in output with increasing distance; this effect was pronounced at the lowest energies. A decrease in photon fluence with increase in polar angle was also observed and was attributed to the silver epoxy component. Conclusions: Changes in source design influenced the overall dose rate and distribution by more than 2% in several regions. This discrepancy is greater than the dose calculation acceptance criteria as recommended in the AAPM TG-56 report. The effect of the design change on the TG-43 parameters would likely not result in dose differences outside of patient applicators. Adoption of this new dataset is suggested for accurate depiction of model S700 source dose distributions.

60 APPLIED LIFE SCIENCES↗

Comparison and Assessment of Three Advanced Land Surface Models in Simulating Terrestrial Water Storage Components over the United States

In preparation for next generation North American Land Data Assimilation System (NLDAS), 3 three advanced land surface models (CLM4.0, Noah-MP, and CLSM-F2.5) were run from 1979 4 to 2014 within the NLDAS-based framework. Monthly total water storage anomaly (TWSA) and 5 its individual water storage components were evaluated against satellite-based and in situ 6 observations, and reference reanalysis products at basin-wide and statewide scales. In general, all 7 three models are able to reasonably capture the monthly and interannual variability and 8 magnitudes for TWSA. However, contributions of the anomalies of individual water 9 components to TWSA are very dependent on the model and basin. A major contributor to the 10 TWSA is the anomaly of total column soil moisture content (SMCA) for CLM4.0 and Noah-MP 11 or groundwater storage anomaly (GWSA) for CLSM-F2.5 although other components such as 12 the anomaly of snow water equivalent (SWEA) also play some role. For each individual water 13 storage component, the models are able to capture broad features such as monthly and 14 interannual variability. However, there are large inter-model differences and quantitative 15 uncertainties in this study. Therefore, it should be thought of as a preliminary synthesis and 16 analysis.

58 GEOSCIENCES↗

Rapid measurement of soluble xylo-oligomers using near-infrared spectroscopy (NIRS) and multivariate statistics: calibration model development and practical approaches to model optimization

Rapid monitoring of biomass conversion processes using techniques such as near-infrared (NIR) spectroscopy can be substantially quicker and less labor-, resource-, and energy-intensive than conventional measurement techniques such as gas or liquid chromatography (GC or LC) due to the lack of solvents and preparation methods, as well as removing the need to transfer samples to an external lab for analytical evaluation. The purpose of this study was to determine the feasibility of rapid monitoring of a biomass conversion process using NIR spectroscopy combined with multivariate statistical modeling, and to examine the impact of (1) subsetting the samples in the original dataset by process location and (2) reducing the spectral range used in the calibration model on model performance. We develop multivariate calibration models for the concentrations of soluble xylo-oligosaccharides (XOS), monomeric xylose, and total solids at multiple points in a biomass conversion process which produces and then purifies XOS compounds from sugar cane bagasse. A single model using samples from multiple locations in the process stream showed acceptable performance as measured by standard statistical measures. However, compared to the single model, we show that separate models built by segregating the calibration samples according to process location show improved performance. We also show that combining an understanding of the sample spectra with simple multivariate analysis tools can result in a calibration model with a substantially smaller spectral range that provides essentially equal performance to the full-range model. We demonstrate that real-time monitoring of soluble xylo-oligosaccharides (XOS), monomeric xylose, and total solids concentration at multiple points in a process stream using NIR spectroscopy coupled with multivariate statistics is feasible. Segregation of sample populations by process location improves model performance. Models using a reduced spectral range containing the most relevant spectral signatures show very similar performance to the full-range model, reinforcing the importance of performing robust exploratory data analysis before beginning multivariate modeling.

09 BIOMASS FUELS↗

Validation and Recalibration of the Solubility Models in Support of the Heater Test in Salt Formations

The US Department of Energy Office of Nuclear Energy is conducting a brine availability heater test to characterize the thermal, mechanical, hydrological and chemical response of salt at elevated temperatures. In the heater test, brines will be collected and analyzed for chemical compositions. In order to support the geochemical modeling of chemical evolutions of the brines during the heater test, we are recalibrating and validating the solubility models for the mineral constituents in salt formations up to 100°C, based on the solubility data in multiple component systems as well as simple systems from literature. In this work, we systematically compare the model-predicted values based on the various solubility models related to the constituents of salt formations, with the experimental data. As halite is the dominant constituent in salt formations, we first test the halite solubility model in the Na-Mg-Cl dominated brines. We find the existing halite solubility model systematically over-predict the solubility of halite. We recalibrate the halite model, which can reproduce halite solubilities in Na-Mg-Cl dominated brines well. As gypsum/anhydrite in salt formations controls the sulfate concentrations in associated brines, we test the gypsum solubility model in NaCl solutions up to 5.87 mol•kg –1 from 25 ° C to 50 ° C. The testing shows that the current gypsum solubility model reproduces the experimental data well when NaCl concentrations are less than 1 mol•kg –1 . However, at NaCl concentrations higher than 1, the model systematically overpredicts the solubility of gypsum. In the Na + —Cl – —SO 4 2– —CO 3 2– system, the validation tests up to 100 ° C demonstrate that the model excellently reproduces the experimental data for the solution compositions equilibrated with one single phase such as halite (NaCl) or thenardite (Na 2 SO 4 ), with deviations equal to, or less than, 1.5 %. The model is much less ideal in reproducing the compositions in equilibrium with the assemblages of halite + thenardite, and of halite + thermonatrite (Na 2 CO 3 •H 2 O), with deviations up to 31 %. The high deviations from the experimental data for the multiple assemblages in this system at elevated temperatures may be attributed to the facts that the database has the Pitzer interaction parameters for Cl – —CO 3 2– and SO 4 2– —CO 3 2– only at 25 ° C. Finally, in the Na + —Ca 2+ —SO 4 2– —HCO 3 – system, the validation tests also demonstrate that the model reproduces the equilibrium compositions for one single phase such as gypsum better than the assemblages of more than one phase.

36 MATERIALS SCIENCE↗

Multiscale Modeling of CO 2 Migration and Trapping in Fractured Reservoirs with Validation by Model Comparison and Real-Site Applications (Final Technical Report)

This report documents the accomplishments achieved during the project titled “Multiscale Modeling of CO 2 Migration and Trapping in Fractured Reservoirs with Validation by Model Comparison and Real-Site Applications” funded by the US Department of Energy, Office of Fossil Energy. The objectives of the project were to develop modeling capabilities for predicting CO 2 and brine migration in fractured reservoirs during geologic carbon storage and to investigate the feasibility of carbon storage in fractured reservoirs using the newly developed modeling capabilities. To achieve these objectives new mass transfer functions were developed to adapt existing dual-continuum approaches – a commonly used approach in hydrocarbon reservoir modeling – to the CO 2 -brine system. In the dual-continuum approach the undisturbed rock matrix and the fractures are modeled as two separate continua, which are coupled through mass exchange of CO 2 and brine between the two continua. Mass transfer functions represent the mass exchange between fractures and the rock matrix, without requiring highly resolved numerical grids of individual fractures, thus allowing large modeling domains relevant to geologic carbon storage questions (tens of meters to hundreds of kilometers). Mass transfer functions for the exchange of both free-phase CO 2 and brine between fractures and rock matrix were developed for both the injection phase and the post-injection phase. During the injection phase the fractures are filled with CO 2 , while the rock matrix is initially filled with brine. Therefore, a mass transfer function was developed that is based on vertical displacement of brine by CO 2 in the rock matrix due to gravitational forces. Once CO 2 injection ceases and CO 2 migrates away from the direct vicinity of the injection site, the fractures will begin to re-fill with brine, while the rock matrix will have high CO 2 saturations. For these conditions a mass transfer functions was developed that is based on spontaneous imbibition, as the pore spaces in the rock matrix are much smaller than in the fractures. For formations with low permeability or high capillary entry pressure of the rock matrix, free-phase CO 2 may not be able to enter the rock matrix. However, free-phase CO 2 at the interface between fractures and rock matrix can dissolve into brine resident in the rock matrix. Therefore, additional mass transfer functions were developed based on aqueous-phase diffusion of CO 2 through the rock matrix. These mass transfer functions are based on existing solutions for diffusion and were expanded to be applicable to a variety of block shapes relevant to dual-continuum modeling. Implementations of the newly developed mass transfer functions showed that they give accurate results when compared to highly resolved models. In addition to developing new transfer functions, a new vertically-integrated dual-continuum modeling framework was developed. This development takes advantage of fast vertical segregation of CO 2 and brine in the fractures due to the density difference between CO 2 and brine and the high permeability of the fractures. Therefore, a vertically-integrated vertical equilibrium model for the fractures is coupled with a vertically-distributed model for the rock matrix. The newly developed approach significantly reduces the computational effort, but is only accurate for fracture permeabilities of 100 mD or higher. Finally, the newly developed models were applied to real and hypothetical geologic carbon storage sites to investigate the feasibility of using fractured reservoirs for carbon storage. The general results were, that storage of CO 2 – both as free-phase and dissolved – is feasible, but that storage capacities are likely lower than in unfractured systems. While the high permeability of the fractures reduces the injection pressure (or permits higher injection rates), fast migration of CO 2 may limit the overall storage volume, because CO 2 may escape laterally with not enough of the rock matrix being swept for effective storage.

20 FOSSIL-FUELED POWER PLANTS↗

Challenge Problem 1: Preliminary Model Development and Assessment of Flexible Heat Transfer Modeling Approaches

This report presents the modeling progress of a university consortium studying flexible modeling for turbulent heat transfer. In the context of advanced nuclear reactor design, where challenges include non-unity Prandtl fluids, free and mixed convection, and transitional flow, computational fluid dynamics methods are underutilized due to low confidence in modeling approaches and high computational cost. This report evaluates the performance of engineering CFD models in diabatic, buoyant turbulent flow. It finds that all turbulence models including several variants of the k-ε and k-ω models struggle to predict accurate turbulent momentum and heat transfer in such flows. The Nusselt numbers have been compared between the models and the DNS data, where calculations have been performed for each case and trends have shown a good agreement between DNS estimated Nusselt numbers and available correlations as well as experimental data. A novel DNS correlation for high Pr cases as they are transitioning from mixed convection to natural convection has been proposed. While Nusselt number errors relative to DNS range from 20% to 50%, the models capture similar trends to DNS with respect to Nusselt suppression and amplification under varying levels of buoyancy effect. This report also highlights model form error as a significant contributor to CFD predictions and proposes a framework for quantifying model error and improving confidence in CFD calculations. To improve the predictive capability of engineering CFD models, data-driven approaches for turbulence models are investigated. Theoretical frameworks based on the invariant tensor / vector basis neural networks for prediction of Reynolds stresses and turbulent heat fluxes are employed. The models are developed using direct numerical simulations data for forced convection flows of different fluids in vertical planar channel domain. The framework is implemented in spectral element solvers Nek5000 / nekRS and has shown a potential for future development and consideration of mixed convection flows.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Efficient surrogate modeling methods for large-scale Earth system models based on machine-learning techniques

Abstract. Improving predictive understanding of Earth system variability and change requires data–model integration. Efficient data–model integration for complex models requires surrogate modeling to reduce model evaluation time. However, building a surrogate of a large-scale Earth system model (ESM) with many output variables is computationally intensive because it involves a large number of expensive ESM simulations. In this effort, we propose an efficient surrogate method capable of using a few ESM runs to build an accurate and fast-to-evaluate surrogate system of model outputs over large spatial and temporal domains. We first use singular value decomposition to reduce the output dimensions and then use Bayesian optimization techniques to generate an accurate neural network surrogate model based on limited ESM simulation samples. Our machine-learning-based surrogate methods can build and evaluate a large surrogate system of many variables quickly. Thus, whenever the quantities of interest change, such as a different objective function, a new site, and a longer simulation time, we can simply extract the information of interest from the surrogate system without rebuilding new surrogates, which significantly reduces computational efforts. We apply the proposed method to a regional ecosystem model to approximate the relationship between eight model parameters and 42 660 carbon flux outputs. Results indicate that using only 20 model simulations, we can build an accurate surrogate system of the 42 660 variables, wherein the consistency between the surrogate prediction and actual model simulation is 0.93 and the mean squared error is 0.02. This highly accurate and fast-to-evaluate surrogate system will greatly enhance the computational efficiency of data–model integration to improve predictions and advance our understanding of the Earth system.

58 GEOSCIENCES↗

A hierarchical analysis of terrestrial ecosystem model Biome-BGC: Equilibrium analysis and model calibration

The increasing complexity of ecosystem models represents a major difficulty in tuning model parameters and analyzing simulated results. To address this problem, this study develops a hierarchical scheme that simplifies the Biome-BGC model into three functionally cascaded tiers and analyzes them sequentially. The first-tier model focuses on leaf-level ecophysiological processes; it simulates evapotranspiration and photosynthesis with prescribed leaf area index (LAI). The restriction on LAI is then lifted in the following two model tiers, which analyze how carbon and nitrogen is cycled at the whole-plant level (the second tier) and in all litter/soil pools (the third tier) to dynamically support the prescribed canopy. In particular, this study analyzes the steady state of these two model tiers by a set of equilibrium equations that are derived from Biome-BGC algorithms and are based on the principle of mass balance. Instead of spinning-up the model for thousands of climate years, these equations are able to estimate carbon/nitrogen stocks and fluxes of the target (steady-state) ecosystem directly from the results obtained by the first-tier model. The model hierarchy is examined with model experiments at four AmeriFlux sites. The results indicate that the proposed scheme can effectively calibrate Biome-BGC to simulate observed fluxes of evapotranspiration and photosynthesis; and the carbon/nitrogen stocks estimated by the equilibrium analysis approach are highly consistent with the results of model simulations. Therefore, the scheme developed in this study may serve as a practical guide to calibrate/analyze Biome-BGC; it also provides an efficient way to solve the problem of model spin-up, especially for applications over large regions. The same methodology may help analyze other similar ecosystem models as well.

54 ENVIRONMENTAL SCIENCES↗

Verification of Compartmental Epidemiological Models using Metamorphic Testing, Model Checking and Visual Analytics

Compartmental models in epidemiology are widely used as a means to model disease spread mechanisms and understand how one can best control the disease in case an outbreak of a widespread epidemic occurs. However, a significant challenge within the community is in the development of approaches that can be used to rigorously verify and validate these models. In this paper, we present an approach to rigorously examine and verify the behavioral properties of compartmen- tal epidemiological models under several common modeling scenarios including birth/death rates and multi-host/pathogen species. Using metamorphic testing, a novel visualization tool and model checking, we build a workflow that provides insights into the functionality of compartmental epidemiological models. Our initial results indicate that metamorphic testing can be used to verify the implementation of these models and provide insights into special conditions where these mathematical models may fail. The visualization front-end allows the end-user to scan through a variety of parameters commonly used in these models to elucidate the conditions under which an epidemic can occur. Further, specifying these models using a process algebra allows one to automatically construct behavioral properties that can be rigorously verified using model checking. Taken together, our approach allows for detecting implementation errors as well as handling conditions under which compartmental epidemiological models may fail to provide insights into disease spread dynamics.

Ramanathan, Arvind↗