Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “predicted”

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 1,081 records · Page 60
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Efficient Probabilistic Prediction and Uncertainty Quantification of Tropical Cyclone–Driven Storm Tides and Inundation

Abstract This study proposes and assesses a methodology to obtain high-quality probabilistic predictions and uncertainty information of near-landfall tropical cyclone–driven (TC-driven) storm tide and inundation with limited time and resources. Forecasts of TC track, intensity, and size are perturbed according to quasi-random Korobov sequences of historical forecast errors with assumed Gaussian and uniform statistical distributions. These perturbations are run in an ensemble of hydrodynamic storm tide model simulations. The resulting set of maximum water surface elevations are dimensionality reduced using Karhunen–Loève expansions and then used as a training set to develop a polynomial chaos (PC) surrogate model from which global sensitivities and probabilistic predictions can be extracted. The maximum water surface elevation is extrapolated over dry points incorporating energy head loss with distance to properly train the surrogate for predicting inundation. We find that the surrogate constructed with third-order PCs using elastic net penalized regression with leave-one-out cross validation provides the most robust fit across training and test sets. Probabilistic predictions of maximum water surface elevation and inundation area by the surrogate model at 48-h lead time for three past U.S. landfalling hurricanes (Irma in 2017, Florence in 2018, and Laura in 2020) are found to be reliable when compared to best track hindcast simulation results, even when trained with as few as 19 samples. The maximum water surface elevation is most sensitive to perpendicular track-offset errors for all three storms. Laura is also highly sensitive to storm size and has the least reliable prediction. Significance Statement The purpose of this study is to develop and evaluate a methodology that can be used to provide high-quality probabilistic predictions of hurricane-induced storm tide and inundation with limited time and resources. This is important for emergency management purposes during or after the landfall of hurricanes. Our results show that sampling forecast errors using quasi-random sequences combined with machine learning techniques that fit polynomial functions to the data are well suited to this task. The polynomial functions also have the benefit of producing exact sensitivity indices of storm tide and inundation to the forecasted hurricane properties such as path, intensity, and size, which can be used for uncertainty estimation. The code implementing the presented methodology is publicly available on GitHub.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2016

Predictability and Decadal Variability of the North Atlantic Ocean State Evaluated from a Realistic Ocean Model

This study investigates the excitation of decadal variability and predictability of the ocean climate state in the North Atlantic. Specifically, initial linear optimal perturbations (LOPs) in temperature and salinity that vary with depth, longitude, and latitude are computed, and the maximum impact on the ocean of these perturbations is evaluated in a realistic ocean general circulation model. The computations of the LOPs involve a maximization procedure based on Lagrange multipliers in a nonautonomous context. To assess the impact of these perturbations four different measures of the North Atlantic Ocean state are used: meridional volume and heat transports (MVT and MHT) and spatially averaged sea surface temperature (SST) and ocean heat content (OHC). It is shown that these metrics are dramatically different with regard to predictability. Whereas OHC and SST can be efficiently modified only by basin-scale anomalies, MVT and MHT are also strongly affected by smaller-scale perturbations. This suggests that instantaneous or even annual-mean values of MVT and MHT are less predictable than SST and OHC. Only when averaged over several decades do the former two metrics have predictability comparable to the latter two, which highlights the need for long-term observations of the Atlantic meridional overturning circulation in order to accumulate climatically relevant data. This study also suggests that initial errors in ocean temperature of a few millikelvins, encompassing both the upper and deep ocean, can lead to ~0.1-K errors in the predictions of North Atlantic sea surface temperature on interannual time scales. This transient error growth peaks for SST and OHC after about 6 and 10 years, respectively, implying a potential predictability barrier.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

G4Boost: a machine learning-based tool for quadruplex identification and stability prediction

Background: G-quadruplexes (G4s), formed within guanine-rich nucleic acids, are secondary structures involved in important biological processes. Although every G4 motif has the potential to form a stable G4 structure, not every G4 motif would, and accurate energy-based methods are needed to assess their structural stability. Here, we present a decision tree-based prediction tool, G4Boost, to identify G4 motifs and predict their secondary structure folding probability and thermodynamic stability based on their sequences, nucleotide compositions, and estimated structural topologies. Results: G4Boost predicted the quadruplex folding state with an accuracy greater then 93% and an F1-score of 0.96, and the folding energy with an RMSE of 4.28 and R 2 of 0.95 only by the means of sequence intrinsic feature. G4Boost was successfully applied and validated to predict the stability of experimentally-determined G4 structures, including for plants and humans. Conclusion: G4Boost outperformed the three machine-learning based prediction tools, DeepG4, Quadron, and G4RNA Screener, in terms of both accuracy and F1-score, and can be highly useful for G4 prediction to understand gene regulation across species including plants and humans.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

Proteome-scale Structure Prediction Data - Pseudodesulfovibrio mercurii

The number of proteins predicted for Pseudodesulfovibrio mercurii is 3,446, each of which have five predicted structures from an AlphaFold run, as well as structural alignment results using the TMscore-based structural alignment method within the APoc program. Specifically, AlphaFold outputs the atoms and coordinates of the protein model in human-readable PDB files and quantitative prediction metrics in Python PICKLE files. The 5 models have been ranked based on the predicted TM-score (pTMS), a quantitative confidence metric output by AlphaFold that reports on protein model quality. The top ranked model has undergone an energy minimization calculation to relax and remove any potential clashes in the atomic coordinates. Structural alignment results are stored in two files for each protein; the top ranked model (as discussed above) is used for all alignment analyses. Both are compressed gzip files that, once unpacked, are human readable. The first file is the TMalign score results and contains the quantitative metrics for the top alignments between the predicted structure and experimental structures from the PDB70, a curated non-redundant database of about 80,000 experimental structures developed by the Soding lab. Each data point in this file is directly associated with one experimental structure; PDB ID and brief meta-data about the protein taken from the PDB70 file are reported alongside the quantitative metrics. The second results file contains the raw results associated with each alignment reported in the score results file. Specifically, the translation and rotation arrays for each alignment are provided so that the structural alignment can be recreated. Additionally, residue-level scores are reported to quantify the closeness of the aligned residues between the predicted and experimental models.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2020

Accuracy of HVAC Load Predictions: Validation of EnergyPlus and DOE-2 using FLEXLAB Measurements

The aim of the project reported here was to better understand the level of accuracy of three building energy simulation (BES) engines (‘engines’) — EnergyPlus™, DOE-2.1e, and DOE-2.2 — by identifying and investigating significant deviations between the performance predicted by these engines and actual performance as measured in the FLEXLAB® test facility at Lawrence Berkeley National Laboratory (LBNL). The specific test conditions included some of those prescribed in ANSI/ASHRAE Standard 140 - Standard Method of Test for the Evaluation of Building Energy Analysis Computer Programs. Detailed measurements of FLEXLAB performance, including indoor temperatures and heat fluxes and air-flow and water flow rates and temperatures in the Heating, Ventilating and Air Conditioning (HVAC) system, together with hourly weather data, were recorded and used in analyzing the simulation results from EnergyPlus v8.8, DOE-2.2 v3.65 and DOE-2.1e v127. These engines are commonly used in the United States for building energy code compliance, federal, state, and utility incentives programs, as well as energy efficient design of new buildings and energy retrofit of existing buildings. Seven conventional overhead mixing ventilation scenarios were tested and each engine was found to have a similar level of agreement with the measurements of space-level heating and sensible cooling loads. These results provide useful information regarding the accuracy of these engines in predicting the cooling and heating load elements of whole building energy performance. This information is intended for practitioners who are concerned about transitioning between simulation tools with different engines and for managers of utility programs leveraging these tools for evaluating and/or projecting measure savings to be incentivized under their programs. The results of the comparisons of simulated and measured performance indicate that the predictions from all three engines are not significantly different. The 24-hour average value of the absolute mean bias indicates the likely magnitude of the error in any particular case. The average mean bias is reduced by cancelation of overprediction in one case by underprediction in another. The daytime absolute mean biases, which may be more important for both energy performance and occupant comfort, are ~6%, presumably because of the greater complexity involved in simulating in the presence of solar radiation. EnergyPlus typically overpredicts the cooling load and/or underpredicts the heating load by ~1.5% and the DOE-2 engines typically underpredict the cooling load by approximately the same amount. The Root Mean Square Error is relatively more sensitive to shorter term variations in the difference between predicted and measured loads; the three engines have similar values, ~10%, suggesting that the uncertainties in their predictions of peak loads may also be similar in magnitude. The implication of these results is that users, both designers and program analysts, can use EnergyPlus, DOE-2.1e, or DOE-2.2 to model conventional commercial buildings equipped with overhead mixing ventilation with a similar level of confidence. Further work is required to better understand the variability in the level of agreement between the engine predictions and FLEXLAB measurements, where a particular engine will agree well with FLEXLAB in some cases and not so well in others and another engine will agree or disagree in different cases. As the sources of this variability are identified and eliminated or reduced significantly, it is recommended that the experimental capabilities and methods developed in the study reported here should be applied to validating heating and cooling load calculations for spaces with different types of furniture and miscellaneous loads. These methods should then be applied to low energy space conditioning systems in EnergyPlus including, in particular, radiant slab and radiant ceiling panel cooling and heating systems and ‘mixed mode’ systems that combine mechanical cooling and natural ventilation systems, focusing on controls, including control of thermal mass. The work reported here addresses the conventional method of heating and cooling occupied spaces; other methods, such as the use of radiant heating and cooling systems have the potential to provide equivalent occupant comfort, or better, with lower energy consumption. These systems are addressed more explicitly in EnergyPlus but there is a need for empirical validation to give users the same level of confidence in modeling these systems that they have, or should have, in modeling conventional systems, based on the results presented here.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2015

Predictive Engineering Tools for Injection-Molded Long-Carbon-Fiber Thermoplastic Composites - Second FY 2015 Quarterly Report

During the second quarter of FY 2015, the following technical progress has been made toward project milestones: 1) Autodesk reviewed 3D fiber orientation distribution (FOD) comparisons and provided support on improving accuracy. 2) Autodesk reviewed fiber length distribution (FLD) data comparisons and provided suggestions, assisted PNNL in FOD and FLD parameter settings optimization, and advised PNNL on appropriate through thickness thermal conductivity for improved frozen layer effect on FOD predictions. Autodesk also participated in project review meetings including preparations and discussions towards passing the go/no-go decision point. 3) Autodesk implemented an improved FOD inlet profile specification method through the part thickness for 3D meshes and provided an updated ASMI research version to PNNL. 4) The University of Illinois (Prof. C.L. Tucker) provided Autodesk with ideas to improve fiber orientation modeling 5) Purdue University re-measured fiber orientation for the fast-fill 50wt% LCF/PA66 edge-gated plaque, and delivered the fiber orientation data for this plaque at the selected locations (named A, B, and C, Figure 1) to PNNL. Purdue also re-measured fiber orientation for locations A on the fast-fill 30wt% LCF/PP and 50wt% LCF/PA66 center-gated plaques, which exhibited anomalous fiber orientation behavior. 6) Purdue University conducted fiber length measurements and delivered the length data to PNNL for the purge materials (slow-fill 30wt% LCF/PP and 30wt% LCF/PA66 purge materials) and PlastiComp plaques selected on the go/no-go list for fiber length model validation (i.e., slow-fill edge-gated 30wt% LCF/PP and 30wt% LCF/PA66 plaques, Locations A, B, and C). 7) PNNL developed a method to recover intact carbon fibers from LCF/PA66 materials. Isolated fibers were shipped to Purdue for length distribution analysis. 8) PNNL completed ASMI mid-plane analyses for all the PlastiComp plaques defined on the go/no-go list for fiber orientation (FO) model validation and compared the predicted fiber orientations with the measured data provided by Purdue at Locations A, B, and C on these plaques. The 15% accuracy criterion based on evaluation of tensile and bending stiffness was used to assess the accuracy in fiber orientation predictions. 9) PNNL completed ASMI mid-plane analyses for all the PlastiComp plaques defined on the go/no-go list for fiber length distribution (FLD) model validation and compared the predicted length distributions with the measured data provided by Purdue at Locations A, B, and C on these plaques. The 15% accuracy criterion based on evaluation of tensile and bending stiffness was used to assess the accuracy in fiber orientation predictions. 10) PNNL tested the new ASMI version received from Autodesk in March 2015, examined and discussed 3D fiber orientation predictions for PlastiComp plaques. 11) PlastiComp, Inc. (PlastiComp), Toyota Research Institute North America (Toyota) and Magna Exteriors and Interiors Corporation (Magna) participated in discussions with team members on the go/no-go plan. Toyota continued the discussion with Magna on tool modification for molding the complex part in order to achieve the target fiber length in the part.

42 ENGINEERING↗
DOE OSTI2016

Uncertainty Quantification and Comparison of Weld Residual Stress Measurements and Predictions.

In pressurized water reactors, the prevention, detection, and repair of cracks within dissimilar metal welds is essential to ensure proper plant functionality and safety. Weld residual stresses, which are difficult to model and cannot be directly measured, contribute to the formation and growth of cracks due to primary water stress corrosion cracking. Additionally, the uncertainty in weld residual stress measurements and modeling predictions is not well understood, further complicating the prediction of crack evolution. The purpose of this document is to develop methodology to quantify the uncertainty associated with weld residual stress that can be applied to modeling predictions and experimental measurements. Ultimately, the results can be used to assess the current state of uncertainty and to build confidence in both modeling and experimental procedures. The methodology consists of statistically modeling the variation in the weld residual stress profiles using functional data analysis techniques. Uncertainty is quantified using statistical bounds (e.g. confidence and tolerance bounds) constructed with a semi-parametric bootstrap procedure. Such bounds describe the range in which quantities of interest, such as means, are expected to lie as evidenced by the data. The methodology is extended to provide direct comparisons between experimental measurements and modeling predictions by constructing statistical confidence bounds for the average difference between the two quantities. The statistical bounds on the average difference can be used to assess the level of agreement between measurements and predictions. The methodology is applied to experimental measurements of residual stress obtained using two strain relief measurement methods and predictions from seven finite element models developed by different organizations during a round robin study.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2017

AGR-5/6/7 Irradiation Test Predictions using PARFUME

PARFUME, (PARticle FUel ModEl) a fuel performance modeling code used for high temperature gas-cooled reactors (HTGRs), was used to model the Advanced Gas Reactor (AGR)-5/6/7 irradiation test using predicted physics and thermal hydraulics data. The AGR-5/6/7 test consists of the combined fifth, sixth, and seventh planned irradiations of the AGR Fuel Development and Qualification Program. The AGR-5/6/7 test train is a multi-capsule, instrumented experiment that is designed for irradiation in the 133.4-mm diameter north east flux trap (NEFT) position of Advanced Test Reactor (ATR). Each capsule contains compacts filled with uranium oxycarbide (UCO) unaltered fuel particles. This report documents the calculations performed to predict the failure probability of tristructural isotropic (TRISO)-coated fuel particles during the AGR-5/6/7 experiment. In addition, this report documents the calculated source term from the driver fuel. The calculations include modeling of the AGR-5/6/7 irradiation that is scheduled to occur from October 2017 to April 2021 over a total of 13 ATR cycles, including nine normal cycles and four Power Axial Locator Mechanism (PALM) cycle for a total between 500 – 550 effective full power days (EFPD). The irradiation conditions and material properties of the AGR-5/6/7 test predicted zero fuel particle failures in Capsules 1, 2, and 4. Fuel particle failures were predicted in Capsule 3 due to internal particle pressure. These failures were predicted in the highest temperature compacts. Capsule 5 fuel particle failures were due to inner pyrolytic carbon (IPyC) cracking causing localized stresses concentrations in the SiC layer. This capsule predicted the highest particle failures due to the lower irradiation temperature. In addition, shrinkage of the buffer and IPyC layer during irradiation resulted in formation of a buffer-IPyC gap. The two capsules at the two ends of the test train, Capsules 1 and 5 experienced the smallest buffer-IPyC gap formation due to the lower irradiation fluences and temperatures. Capsule 3 experienced the largest buffer-IPyC gap formation of just under 24 µm. The release fraction of fission products Ag, Cs, and Sr silver (Ag), cesium (Cs), and strontium (Sr) vary depending on capsule location and irradiation temperature. The maximum release fraction of Ag occurs in Capsule 3, reaching up to 84.8% for the TRISO fuel particles. The release fraction of the other two fission products, Cs and Sr are much smaller and, in most cases, less than 1%. The notable exception is again in Capsule 3, where the release fraction for Cs and Sr reach up to 9.7% and 19.1%, respectively.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2018

Impacts of Process and Prediction Uncertainties on Projected Hanford Waste Glass Amount

Analyses were performed to evaluate the impacts of using the advanced glass models, constraints (Vienna et al. 2016), and uncertainty descriptions on projected Hanford glass mass. The maximum allowable WOL was estimated for waste compositions while simultaneously satisfying all applicable glass property and composition constraints with sufficient confidence. Different components of prediction and composition/process uncertainties were systematically included in the calculations to evaluate their impacts on glass mass. The analyses estimated the production of 23,360 MT of IHLW glass when no uncertainties were taken into accound. Accounting for prediction and composition/process uncertainties resulted in 5.01 relative percent increase in estimated glass mass 24,531 MT. Roughly equal impacts were found for prediction uncertainties (2.58 RPD) and composition/process uncertainties (2.43 RPD). ILAW mass was predicted to be 282,350 MT without uncertainty and with weaste loading “line” rules in place. Accounting for prediction and composition/process uncertainties resulted in only 0.08 relative percent increase in estimated glass mass of 282,562 MTG. Without application of line rules the glass mass decreases by 10.6 relative percent (252,490 MT) for the case with no uncertainties. Addition of prediction uncertainties increases glass mass by 1.32 relative percent and the addition of composition/process uncertainties increase glass mass by an additional 7.73 relative percent (9.06 relative percent increase combined). The glass mass estimate without line rules (275,359 MT) was 2.55 relative percent lower than that with the line rules (282,562 MT), after accounting for all applicable uncertainties.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2018

Impacts of Process and Prediction Uncertainties on Projected Hanford Waste Glass Amount

Analyses were performed to evaluate the impacts of using the advanced glass models, constraints (Vienna et al. 2016), and uncertainty descriptions on projected Hanford glass mass. The maximum allowable waste oxide loading (WOL) was estimated for waste compositions while simultaneously satisfying all applicable glass property and composition constraints with sufficient confidence. Different components of prediction and composition/process uncertainties were systematically included in the calculations to evaluate their impacts on glass mass. The analyses estimated the production of 23,360 MT of immobilized high-level waste (IHLW) glass when no uncertainties were taken into account. Accounting for prediction and composition/process uncertainties resulted in 5.01 relative percent increase in estimated glass mass of 24,531 MT. Roughly equal impacts were found for prediction uncertainties (2.58 RPD) and composition/process uncertainties (2.43 RPD). The immobilized low-activity waste (ILAW) mass was predicted to be 282,350 MT without uncertainty and with waste loading “line” rules in place. Accounting for prediction and composition/process uncertainties resulted in only 0.08 relative percent increase in estimated glass mass of 282,562 MT. Without application of line rules the glass mass decreases by 10.6 relative percent (252,490 MT) for the case with no uncertainties. Addition of prediction uncertainties increases glass mass by 1.32 relative percent and the addition of composition/process uncertainties increase glass mass by an additional 7.73 relative percent (9.06 relative percent increase combined). The glass mass estimate without line rules (275,359 MT) was 2.55 relative percent lower than that with the line rules (282,562 MT), after accounting for all applicable uncertainties

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2018

Impacts of Process and Prediction Uncertainties on Projected Hanford Waste Glass Amount

Analyses were performed to evaluate the impacts of using the advanced glass models, constraints, and uncertainty descriptions on projected Hanford glass mass. The maximum allowable waste oxide loading (WOL) was estimated for waste compositions while simultaneously satisfying all applicable glass property and composition constraints with sufficient confidence. Different components of prediction and composition/process uncertainties were systematically included in the calculations to evaluate their impacts on glass mass. The analyses estimated the production of 23,360 MT of immobilized high-level waste (IHLW) glass when no uncertainties were taken into account. Accounting for prediction and composition/process uncertainties resulted in 5.01 relative percent increase in estimated glass mass of 24,531 MT. Roughly equal impacts were found for prediction uncertainties (2.58 RPD) and composition/process uncertainties (2.43 RPD). The immobilized low-activity waste (ILAW) mass was predicted to be 282,350 MT without uncertainty and with waste loading “line” rules in place. Accounting for prediction and composition/process uncertainties resulted in only 0.08 relative percent increase in estimated glass mass of 282,562 MT. Without application of line rules the glass mass decreases by 10.6 relative percent (252,490 MT) for the case with no uncertainties. Addition of prediction uncertainties increases glass mass by 1.32 relative percent and the addition of composition/process uncertainties increase glass mass by an additional 7.73 relative percent (9.06 relative percent increase combined). The glass mass estimate without line rules (275,359 MT) was 2.55 relative percent lower than that with the line rules (282,562 MT), after accounting for all applicable uncertainties.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2016

Physics-Informed Machine Learning for Predictive Turbulence Modeling: Towards a Complete Framework

Although increased availability of computational resources has enabled high-fidelity simulations (e.g., large eddy simulations) of turbulent flows, the Reynolds-Averaged Navier–Stokes (RANS) models are still the dominant tools in industrial applications. However, the predictive capabilities of RANS models are limited by large model-form discrepancies due to the Reynolds stress closure. Recently, a Physics-Informed Machine Learning (PIML) approach has been proposed to learn the functional form of Reynolds stress discrepancy in RANS simulations based on available data. It has been demonstrated that the learned discrepancy function can be used to improve Reynolds stresses in new flows where data are not available. However, due to a number of challenges, the improvements are only demonstrated in the Reynolds stress prediction but not in corresponding propagated quantities of interest (e.g., velocity field). In this work, we investigate the prediction performance on the velocity field by propagating the corrected Reynolds stresses in PIML approach. To enrich the input features, an integrity basis of invariants is implemented. The fully developed turbulent flow in a square duct is used as the test case. The discrepancy model is trained on flow fields from several Reynolds numbers and evaluated on a duct flow at a higher Reynolds number than any of the training cases. The predicted Reynolds stresses are propagated to velocity field via RANS equations. Numerical results show excellent predictive performances in both Reynolds stresses and their propagated velocities, demonstrating the merits of the PIML approach in predictive turbulence modeling.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2016

Predictive Engineering Tools for Novel Fuels

This project was funded through the Campus Executive Fellowship at University of California (UC) Berkeley, and had two principal aims. First, it sought to explore predictive tools for estimating fuel properties based on molecular structure, with the goal of identifying promising candidates for new fuels to be synthesized. Second, it sought to investigate the possibility of increasing engine efficiency by substituting air for a working fluid with higher efficiency potential employed in a closed loop, namely a mixture of argon and oxygen. In pursuing the predictive tool for novel fuels, a new model was built that proved to be highly predictive of autoignition characteristics for a wide variety of hydrocarbons, esters, ethers and alcohols, and reasonably predictive for furan and tetrahydrofuran compounds, the target class of novel fuels. Obtaining more “training data” for the model improved its predictive capabilities, and further reductions in the uncertainty of the predictions would be possible with more training data. In investigating the concept of a closed-loop engine cycle using an argon-oxygen working fluid, substantial progress was made. Initial engineering models were built showing the feasibility of the concept; numerous collaborations were formed with industry and academic partners; external funding was secured from the California Energy Commission (CEC) to build a dedicated engine platform for research; and this engine platform was designed and constructed. Experimental work and associated modeling studies will take place in late 2016 and early 2017.

10 SYNTHETIC FUELS↗
DOE OSTI2019

Creating Feature Representations of Antibody-Antigen Complexes for Fast Binding Prediction with Machine Learning

Antibody-antigen interactions underlie the function of the immune system. Numerous computational tools exist that aim to predict important quantities associated with antibodyantigen interactions, such as binding energy. Efficient and accurate prediction of such quantities would greatly accelerate vaccine (antigen) and therapeutic antibody design. Current methods and tools, however, rely on varying degrees of computationally expensive steps that limit their use on problems that require many predictions, and moreover, no single method produces consistently accurate estimates. In this project, we determined the feasibility of creating feature representations that concisely describe interacting pairs of antibody and antigen as rapidly computable numerical vectors, given the antibody-antigen co-complex. Such a featurization will allow the use of a statistical machine learning model that jointly models the results of the disparate existing computational tools available. In turn, this will greatly enhance their combined predictive performance. We developed a feature representation consisting of 86 descriptors of the antigen-antibody interface which is then reduced into a lower dimensional representation using a neural network dimensionality reduction step. We have demonstrated that this feature representation serves its intended purpose of rapidly describing antigen-antibody pairs for the task of binding strength prediction. While this feasibility study has proven that such a feature representation is feasible, additional research is likely improve its ability to support various tasks, including binding strength prediction.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

Data-Driven Exploration of Climate Attractor Manifolds For Long-Term Predictability

Focal Area: This white paper responds to Focal Area 3. We seek to gain insight into decadal-scale climate predictability by applying novel manifold-finding probabilistic AI techniques to the complex data produced by Earth System models (ESMs) such as E3SM. The associated portfolio of research activities leverages DOE’s asset mix of HPC platforms, climate expertise, climate simulation codes, and AI expertise. Science Challenge: Climate and climate models are dynamical systems exhibiting properties that are interpretable through chaos theory. The theory contains an important concept that is relevant to multi-decade-scale climate prediction: a chaotic attractor. While the space containing all the possible states of the Earth’s atmosphere and ocean, the possible weather, is large, the realized states tend to stay near the smaller-dimensioned attractor. This behavior is responsible for the “order behind the irregularity” [1] of climate phenomena. Climate change can be thought of as a change in the properties of the attractor, and predicting the climate over years to decades is equivalent to predicting how those properties will change. To date, the attractor has been a useful conceptual tool, but has not been amenable to direct characterization. A new development is the advent of efficient high-dimensional manifold-finding probabilistic AI techniques, which permit a data-driven characterization of the ESM attractor and its probability distribution over weather states. Such a characterization would result in a natural dimensional reduction — a “non-linear Principal Components Analysis (PCA) adapted to climate simulation data” — leading to important advances in scenario-based long-term climate prediction, long-term prediction of water cycle extremes, ESM verification, inter-model comparison, and process model development.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Develop a weather-aware climate model to understand and predict extremes and associated power outages and renewable energy shortages with uncertainty-aware and physics-informed machine learning

Focal Area(s): The focus area is predictive modeling through the use of AI techniques and AI-derived model components with a particular emphasis on extreme weather in Atmospheric Science and power outages and shortages in Energy Science. Science Challenge: Predicting weather extremes (e.g., heavy precipitation, strong wind, and large hailstones), and weather-related power system outages and shortages can mitigate economic losses, save lives, support renewables integration, and improve power system resiliency. However, currently, the poor reliability and large uncertainty associated with the weather extreme prediction in the current climate models make the problem intractable. The key challenges are: (1) physical factors like green-house gases (GHGs), aerosols, and land use and land cover (LULC) can significantly impact extreme storms, but the understanding of these impacts is limited, particularly globally; (2) the convective permitting resolutions needed to model severe convective storms and their impacts are computationally prohibitive with global climate models (GCMs); (3) interactions between weather extremes and power system outages are complex and subject to great uncertainty. Current outage prediction models are short lead (~ 3 days), which do not allow for long-time planning of energy production and distribution. Moreover, we have limited capacity to predict weather events leading to sustained shortages in a renewable-energy-dominated power system. These challenges drive motivation for mechanistic understanding and reliable and efficient predictive modeling of extremes and their impacts from the sub-seasonal to long term projections.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

How AI Predicts the Untrained and Unseen

Focus Area: Model predictability improvements (Primary); Data optimization (secondary); Data complexity insights (secondary). The Scientific Challenge: If we believe that a future under extreme conditions will look very differently from today, we can likely agree that ML/AI models trained on past and present datasets will not be adequate to make reliable predictions into the future. This is true for water cycling, as well as biogeochemistry and other Earth system components and behaviors. Additionally, ML/AI models are inherently non-physical. Despite the flourishing success of ML/AI in many applications, such as computer vision, natural language process, and gaming, even the most sophisticated AI models don’t understand the very basic physical laws. Therefore, a natural question is: Can we trust ML/AI based predictions of Earth system behaviors that are fundamentally driven by physical laws? So, are physics models with meticulous process representation a better choice? Not exactly. Physical models, when firmly rooted in first principles, work great at predicting behaviors of systems with a well-defined set of boundary conditions and variables. However, as a complex system, the number of parameters and the degree of complexity and dynamics in processes, coupling, and scale dependent emergent behaviors make the Earth system behaviors very challenging to predict with physical models composed of deterministic laws. In addition, due to the lack of fundamental understandings, physical models often implement empirical correlations derived from observations with biases from locality of data generation. Because correlation is not necessarily causation or comply with first principles, scaling of model predictions beyond locality is often invalid. Beyond the limitation of models, physics or AI, our knowledge of the Earth system is limited by the lack of observational technologies and resources. Insufficient data density, dimensionality and diversity only offer a sliced (or projected) view of the Earth system, e.g. Plato’s Cave analogy, limiting our capability to better understand and represent fundamental processes in models.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Physics-Informed Learning for Predictive Multi-Scale Modeling of Water Cycle and Extreme Events

Climate and earth system models have limited skill in predicting water cycle and associated extremes such as intense rain- and wind-producing storms and resulting hazards across all timescales. Water cycle predictions are inherently limited by the predictability of the chaotic system, but large potential for improving predictions remains untapped due to limited use of data-model integration to improve predictive modeling. This white paper proposes a hybrid data-physics paradigm based on physics-informed neural networks (PINNs) to improve the Energy Exascale Earth System Model (E3SM). In particular, we aim to predict and quantify uncertainties of water cycle processes and low probability high impact events.

54 ENVIRONMENTAL SCIENCES↗