SEARCH · Search NASA
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.
This dataset contains simulations generated by Sunny.jl of the magnetic excitations on the NiPS3 lattice. The exchange parameters are varied over the system so these simulations can be used for AI model training.
Various machine learning (ML) and deep learning (DL) techniques have been recently applied to the forecasting of laboratory earthquakes from friction experiments. The magnitude and timing of shear failures in stick-slip cycles are predicted using features extracted from the recorded ultrasonic or acoustic emission (AE) signals. In addition, the Rate and State Friction (RSF) constitutive laws are extensively used to model the frictional behavior of faults. In this work, we use data from shear experiments coupled with passive acoustic (variance, kurtosis, and AE rate) interleaved with active source ultrasonic monitoring (transmitted wave amplitude) to develop physics-informed neural network (PINN) models incorporating the RSF law and AE rate generation equation with wave amplitude serving as a proxy for friction state variable. This PINN framework allows learning RSF parameters from stick-slip experiments rather than measuring them through a series of velocity step experiments. We observe that when the stick-slip cycles are irregular, the PINN models outperform the data-driven DL models. Transfer learning (TL) PINN models are also developed by pre-training on data collected at one normal stress level followed by forecasting shear failures and retrieving RSF parameters at other stress levels (i.e., with different recurrence intervals) after retraining on a limited amount of new data. Our findings suggest that TL models perform better compared to standalone models. Both standalone and TL PINN-estimated RSF parameters and their ground truth values show excellent agreements thus demonstrating that RSF parameters can be retrieved from laboratory stick-slip experiments using the corresponding acoustic data and that the transmitted wave amplitude provides a good representation of the evolving frictional state during stick-slips.
The surface energy budget governs Arctic sea-ice growth/melt, yet observations are sparse, and reanalysis data sets suffer from systematic biases. Here, we train a neural network with observational data to bias-correct hourly ERA5 fluxes over Arctic ice-covered regions (≥70°N; sea-ice concentration >80%) for 1994–2024. Training data cover two full seasonal cycles and different sea-ice regimes. The neural network reduces RMSE for net shortwave radiation by ∼40%, downward longwave radiation by ∼16% and the total surface energy budget by ∼55%, eliminating the wintertime warm bias of ∼4 K in ERA5. Wintertime surface cooling is reduced by ∼50%, yielding thermodynamic ice-growth estimates of ∼80–120 cm, consistent with SMOS–CryoSat satellite thickness increases and in contrast to the 150–200 cm growth implied by ERA5. Our bias-corrected data capture the observed clear/cloudy states of the winter boundary layer and can be used to study Arctic climatology, evaluate climate models and drive sea-ice-ocean models.
With the end of Moore’s law and Dennard scaling, efficient training increasingly requires rethinking data volume. Can we train better models with significantly less data via intelligent subsampling? To explore this, we develop SICKLE, a sparse intelligent curation framework for efficient learning, featuring a novel maximum entropy (MaxEnt) sampling approach, scalable training, and energy benchmarking. We compare MaxEnt with random and phase-space sampling on large direct numerical simulation (DNS) datasets of turbulence. Evaluating SICKLE at scale on Frontier, we show that subsampling as a preprocessing step can, in many cases, improve model accuracy and substantially lower energy consumption, with observed reductions of up to 38×.
This data package is associated with the publication “Ecosystem leaf area, gross primary production, and evapotranspiration responses to wildfire in the Columbia River Basin” submitted to Biogeosciences (Shi et al., 2024; doi: 10.22541/au.171053013.30286044/v1). In this research, data products, leaf area index (LAI), gross primary production (GPP), and evapotranspiration (ET), from the Moderate Resolution Imaging Spectroradiometer (MODIS) are used to quantify the resistance and resilience of different ecosystem types in the Columbia River Basin (CRB). A machine learning algorithm, random forest (RF), was used to examine the impacts of precipitation, vapor pressure deficit (VPD), and burn severity from Monitoring Trends in Burn Severity (MTBS) on ecosystem resilience. The data package includes the processed MODIS data products, precipitation, VPD, and burn severity in 138 fire regions in CRB and the input files for RF model training. This data package includes six folders. The MODIS products are included in three MODIS_* folders with shell scripts for data clipping and *ncl files for data processing: (1) “/MODIS_LAI_CRB”; (2) “/MODIS_GPP_CRB”; and (3) “/MODIS_ET_CRB”. All the processed data for each fire event are NetCDF formatted. The MTBS burn severity data and the shell and *ncl scripts used for data processing are in the folder named (4) “MTBS_fire”. The ERA meteorological fields and the data processing scritps are in (5) “ERA_Var_CR”. All the scripts for figure development are in the format of *ncl and in the folder (6) “paper_scripts”. See the file ending in “flmd.csv” for a list of all files contained in this data package and descriptions for each. Tabular column headers and units are described in the data dictionary file ending in “dd.csv”.
A search is presented for a heavy scalar (H) or pseudo-scalar (A) predicted by the two-Higgs-doublet models, where the H/A is produced in association with a top-quark pair $(t\bar{t}H/A),$ and with the H/A decaying into a $t\bar{t}$ pair. The full LHC Run 2 proton–proton collision data collected by the ATLAS experiment is used, corresponding to an integrated luminosity of $139~\text {fb}^{-1}.$ Events are selected requiring exactly one or two opposite-charge electrons or muons. Data-driven corrections are applied to improve the modelling of the $t\bar{t}$ +jets background in the regime with high jet and b-jet multiplicities. These include a novel multi-dimensional kinematic reweighting based on a neural network trained using data and simulations. An H/A-mass parameterised graph neural network is trained to optimise the signal-to-background discrimination. In combination with the previous search performed by the ATLAS Collaboration in the multilepton final state, the observed upper limits on the $t\bar{t}H/A \rightarrow t\bar{t}t\bar{t}$ production cross-section at 95% confidence level range between 14 fb and 5.0 fb for an H/A with mass between 400 GeV and 1000 GeV , respectively. Assuming that both the H and A contribute to the $t\bar{t}t\bar{t}$ cross-section, tan β values below 1.7 or 0.7 are excluded for a mass of 400 GeV or 1000 GeV , respectively. The results are also used to constrain a model predicting the pair production of a colour-octet scalar, with the scalar decaying into a $t\bar{t}$ pair.
This paper presents a convolutional neural network (CNN) developed to identify voltage events in photovoltaic (PV) inverters. The CNN is trained on synthetic data generated using the IEEE 13-bus distribution feeder model and evaluated on field measured data collected from Energy Northwest’s Horn Rapids Solar, Storage, and Training (HRSST) facility. The study focuses on two common voltage events: faults and voltage sags. The CNN is configured to analyze voltage and current waveforms from three-phase PV systems, demonstrating excellent accuracy during training. Field data from the HRSST facility is employed to assess its real-world performance, where the CNN achieves perfect identification of faults and voltage sags in a sample of nine events. This work highlights the potential of the proposed method to enhance PV protection schemes, providing a robust foundation for improved voltage event detection and grid reliability.
Steel Thread is a NA-22 venture that seeks to build trustworthy, reliable AI models that can be used in a wide variety of nonproliferation tasks. A key aspect of building these models is developing appropriate benchmarks and evaluation methods, which will enable the venture to identify and adapt models to provide the most value in the nonproliferation domain. Benchmarks must be relevant to key tasks in this domain, such as question answering, information retrieval, document summarization and classification, consensus analysis, and image and data analysis. This report 1) provides an overview of benchmark design, evaluation, and challenges; 2) reviews a variety of open benchmarks, with a focus on language models and tasks; and 3) identifies benchmarks that are most relevant to Steel Thread. This report is intended to serve as a basis for further efforts to classify and evaluate benchmarks and their correlation with success on nonproliferation-specific tasks. The Steel Thread venture has defined benchmarks to be a particular combination of a dataset (or datasets) and a metric (or metrics) conceptualized as representing one or more specific tasks or sets of abilities for a specific modality. It is adopted by a research community as a shared framework for comparing methods.1 It includes 1) Data: Labeled (a designated subset not used for training, which could be all the data), 2) Metric: A way to quantify performance, 3) Task/Ability: What the benchmark is testing, 4) Protocol: A structured and repeatable evaluation process, 5) Baseline/Reference Model: For comparison; could be statistical, rule-based, SME-derived, or another model, and 6) Maintenance Plan: to update with new information over time; important for long-term utility. For further clarity, the definition includes what a benchmark, in this context, is not. It is not a corpus of training data, specific to a model (it is intended to apply to a range of models), a universal evaluation of performance, a guarantee that the ‘top’ model on the leaderboard will be the best fit for every specific use case, an all-encompassing proof of a model’s universal quality, nor is it a one-size-fits-all measure of success. It does not cover every real-world constraint (like operational, ethical, or cost considerations), a systems integration test, or a unit test. This definition was inspired by and resulted from discussions within the Steel Thread Benchmarking Task Force. This group was formed to define what we would mean as a benchmark within Steel Thread but persisted as the need to develop a thorough understanding of the large and expanding existing benchmarking space. This technical report is a result of the group’s divide and conquer approach to exploring this space. The release of benchmarks might not be progressing as quickly as model development, but it is moving very fast, as many benchmarks quickly become saturated, when state-of-the-art models score so close to the benchmark’s ceiling that their results are virtually indistinguishable. At that point, the test no longer differentiates between new systems, so researchers usually stop reporting scores as the benchmark no longer informs about improvements from the next generation of models. In the OpenAI announcement of GPT-5, they reported results on six flagship public benchmarks (AIME 2025, SWE-bench Verified, Aider Polyglot, MMMU, HealthBench Hard, GPQA) but the full system-card covers roughly thirty-five separate evaluations, comprising hundreds of test task items in total. There have been some efforts to summarize benchmarks in specific fields, like for text-to-image generation, but these surveys have had a narrow methodology scope. Therefore, a comprehensive survey of all benchmarks or even all benchmarks that could be relevant to Steel Thread is outside of the scope of this report. We chose some specific benchmarks to investigate in detail.
Growing vaccine hesitancy is contributing to the decline in immunization rates for highly contagious, vaccine-preventable childhood diseases. Therefore, there has been a significant interest in understanding how hesitancy is spreading at higher spatio-temporal resolutions, enabling more targeted interventions. Motivated by this, we study the problem of prediction of vaccine hesitancy at the ZIP Code level, referred to as the VaxHesitancy problem. A significant challenge for this problem is the lack of high-resolution data that indicates hesitancy. Here, we develop a hybrid VaxHesSTL framework that combines a Graph Neural Network (GNN) and a Recurrent Neural Network (RNN) to address the VaxHesitancy problem. The GNN uses a ZIP Code-level network to capture spatial signals from neighboring areas, while the RNN models the temporal dynamics present in the data. We train and evaluate VaxHesSTL using a large dataset, namely the All-Payer Claims Databases (APCD), for Virginia, consisting of insurance claims from over five million individuals for six years. We find that an aggregated contact network or graph, developed from a detailed activity-based population network, plays an important role in the performance of VaxHesSTL, compared to graph models based solely on spatial proximity. Experiments demonstrate that VaxHesSTL outperforms a range of state-of-the-art baselines, which rely solely on historical time series data without accounting for spatial relationships. Since hesitancy data at higher spatial resolution is often unavailable or hard to get, we incorporate an active learning approach with our VaxHesSTL framework to optimize the training set without compromising the prediction performance. We find that hesitancy data for only 18% of ZIP Codes selected by active learning allows us to forecast hesitancy for all the ZIP Codes in the Virginia.
Individual seismic catalogs can contain multiscale observations from fault level to global scales and associated waveforms from discrete events reflect crustal structure across many different scales and locations. Seismic network aperture, geographic location, and observation distance may not provide informative guidance or intuition on how different catalogs will behave across models trained under different conditions. We rely on uncertainty to provide guardrails for when to trust model decisions, but understanding when our uncertainty is trustworthy is an open challenge. Here, in this work, we explore Bayesian approximation methods for assigning predictive uncertainty in seismic event classification problems. We find that computationally expensive Bayesian approximations do not outperform simple ensemble methods. We also find that when exploiting multiple seismic event catalogs, joint training with data from all the catalogs combined with Bayesian approximations and supervised training for classification can obscure bias and result in less robust uncertainty while also not providing substantial performance benefits compared to training individual models for each catalog.
Growing interest in Artificial Intelligence (AI) has resulted in a surge in demand for faster methods of Machine Learning (ML) model training and inference. This demand for speed has prompted the use of high performance computing (HPC) systems that excel in managing distributed workloads. Because data is the main fuel for AI applications, the performance of the storage and I/O subsystem of HPC systems is critical. In the past, HPC applications accessed large portions of data written by simulations or experiments or ingested data for visualizations or analysis tasks. ML workloads perform small reads spread across a large number of random files. This shift of I/O access patterns poses several challenges to modern parallel storage systems. In this paper, we survey I/O in ML applications on HPC systems, and target literature within a 6-year time window from 2019 to 2024. We define the scope of the survey, provide an overview of the common phases of ML, review available profilers and benchmarks, examine the I/O patterns encountered during offline data preparation, training, and inference, and explore I/O optimizations utilized in modern ML frameworks and proposed in recent literature. Lastly, we seek to expose research gaps that could spawn further R&D.
An important challenge with Machine Learning (ML) is its transferability; that is, whether a ML model trained on one set of data can be applied to a second set of data without requiring a full re-training of the model. Transfer Learning (TL) addresses this challenge by transferring knowledge learned in the source domain (the data it was trained on) to the target domain (a second set of data that is statistically different but related, which the model was not trained on). This study investigates the use of TL for street-scale nuisance flood forecasting by exploring whether a ML model trained on data collected for one set of streets can effectively forecast flooding for another set of streets in the same city using TL. The envisioned use case is a city deploying a new flood depth monitoring sensor on a street and using TL to apply a ML model, trained on sensor data from an existing flood depth sensor network, to this new street. Eventually, the new flood depth sensor will have a sufficient dataset for training its own ML model, but TL can be used to fill the gap in time while this new dataset is being generated. This method is explored using a Long Short-Term Memory (LSTM) model trained on data for the flood-prone streets of Norfolk City, Virginia. The data used for training includes environmental time series (rainfall, tide), topographic features (Digital Elevation Model (DEM), Topographic Wetness Index (TWI), Depth To Water (DTW)), and street-scale flood depth time series obtained from a high-fidelity physics-based model, acting as a synthetic street-scale stream depth sensor dataset since actual stream depth sensor data is generally unavailable for most cities. A set of 180 flood-prone streets was used to train a base model, while another set of 180 flood-prone streets was used to re-train that model using different TL strategies. The results show that full-weight re-training proved most effective and minimal re-training of only the output layer was insufficient. The advantage of TL was most pronounced when target data was limited, meaning data collected at the new water depth sensor location included generally less than 18 flood events. As target data increased beyond 18 flood events, the benefit of TL diminished relative to training a ML model directly on the local flood events. These findings can assist cities as they implement street-scale flood sensing systems to create accurate forecasts for new sensing locations that do not yet have sufficient data records to train a local ML model.
As electric vehicle (EV) adoption rises globally, concerns about the impact on aging electrical grids grow, particularly regarding the charging behavior of EV drivers. This study analyzes real-world driving and charging data from Ford battery electric vehicles (BEVs) collected between 2018 and 2019 to develop interpretable models that characterize charging behavior and quantify influencing factors. Prior research has relied on assumptions regarding driver behavior, often overlooking actual charging patterns. By employing generalized linear mixed models (GLMMs), this work offers insights into how various elements, such as next trip distance and state of charge (SOC), influence charging decisions. The dataset comprises over three million park-trip pairs from 1,997 vehicles, revealing that features related to driving behavior significantly dictate charging behavior, while infrastructure and regional factors have lesser impacts. The findings suggest that existing simulation models may oversimplify EV charging behavior assumptions. This work utilizes real-world EV driving and charging data to train interpretable models that describe charging behavior and quantify the factors most associated with how drivers use charging infrastructure. This research underscores the need for interpretable, data-driven methodologies to inform future EV infrastructure planning and grid management.
Not Available
This study investigates the data requirements of generative artificial intelligence (AI), particularly generative adversarial networks (GANs), for reliable data augmentation in energy applications. Generative AI, though seen as a solution to data limitations, requires substantial data to learn meaningful distributions—a challenge often overlooked. This study addresses the challenge through synthetic data generation for critical heat flux (CHF) and power grid demand, focusing on renewable and nuclear energy. Two variants of GAN employed are conditional GAN (cGAN) and Wasserstein GAN (wGAN). Our findings include the strong dependency of GAN on data size, with performance declining on smaller datasets and varying performance when generalizing to unseen experiments. Mass flux and heated length significantly influence CHF predictions. wGAN is more robust to feature exclusion, making it suitable for constrained synthetic data generation. In energy demand forecasting, wGAN performed well for solar, wind, and load predictions. Longer lookback hours and larger datasets improved predictions, especially for load power. Seasonal variations posed challenges, with wGAN achieving a relatively high error of Root Mean Squared Error (RMSE) of 0.32 for load power prediction, compared to RMSE of 0.07 under same-season conditions. Feature exclusions impacted cGAN the most, while wGAN showed greater robustness. This study concludes that, while generative AI is effective for data augmentation, it requires substantial data and careful training to generate realistic synthetic data and generalize to new experiments in engineering applications.
Early battery life prediction models are most useful for R&D if they help us understand the early changes in battery electrochemical response that correspond with long-term degradation and failure. Linear regression models such as Fused lasso and Partial Least Squares can fit coefficients directly to high-dimensional electrochemical data like capacity-voltage and ΔV–state-of-charge, i.e., Q(V) and ΔV(SOC) curves, learning coefficients that can be physically interpreted. We leverage the ISU-ILCC battery aging data set to learn high-dimensional coefficients for early battery life prediction from traditional slow-rate capacity check data, demonstrating learning on Q(V), d Q· d V −1 , and ΔV(SOC) curves. A thorough study on the dependence of coefficient values on train/test size and data preprocessing methods is made, demonstrating the reliability of high-dimensional regression approaches unless very small amounts of data are used for model training. For this data set, coefficients from Q(V) and d Q· d V −1 models highlight changes in electrode stoichiometry due to lithium loss, while ΔV(SOC) coefficients highlight changes in positive electrode diffusivity due to particle cracking as well as electrode stoichiometry shifts. By directly interpreting the coefficients of a regression model, we make physical insights into battery degradation mechanisms without requiring the assumptions of traditional battery data analysis methods.
Abstract Structural neuroimaging data have been used to compute an estimate of the biological age of the brain (brain‐age) which has been associated with other biologically and behaviorally meaningful measures of brain development and aging. The ongoing research interest in brain‐age has highlighted the need for robust and publicly available brain‐age models pre‐trained on data from large samples of healthy individuals. To address this need we have previously released a developmental brain‐age model. Here we expand this work to develop, empirically validate, and disseminate a pre‐trained brain‐age model to cover most of the human lifespan. To achieve this, we selected the best‐performing model after systematically examining the impact of seven site harmonization strategies, age range, and sample size on brain‐age prediction in a discovery sample of brain morphometric measures from 35,683 healthy individuals (age range: 5–90 years; 53.59% female). The pre‐trained models were tested for cross‐dataset generalizability in an independent sample comprising 2101 healthy individuals (age range: 8–80 years; 55.35% female) and for longitudinal consistency in a further sample comprising 377 healthy individuals (age range: 9–25 years; 49.87% female). This empirical examination yielded the following findings: (1) the accuracy of age prediction from morphometry data was higher when no site harmonization was applied; (2) dividing the discovery sample into two age‐bins (5–40 and 40–90 years) provided a better balance between model accuracy and explained age variance than other alternatives; (3) model accuracy for brain‐age prediction plateaued at a sample size exceeding 1600 participants. These findings have been incorporated into CentileBrain ( https://centilebrain.org/#/brainAGE2 ), an open‐science, web‐based platform for individualized neuroimaging metrics.
Deep learning (DL) has become an indispensable tool in hyperspectral data analysis, automatically extracting valuable features from complex, high-dimensional datasets. Super-resolution reconstruction, an essential aspect of hyperspectral data, involves enhancing spatial resolution, particularly relevant to low-resolution hyperspectral data. Yet, the pursuit of super-resolution in hyperspectral analysis is fraught with challenges, including acquiring ground truth high-resolution data for training, generalization, and scalability. The pressing issue of extended spectral acquisition times, notably for high-resolution scans, is a significant roadblock in hyperspectral imaging. Super-resolution methods offer a promising solution by providing higher spatial resolution data to expedite data collection and yield more efficient outcomes. This paper delves into a practical application of these concepts using Raman imaging, where spectral acquisition times can be prohibitively long. In this context, DL-based super-resolution models demonstrate their efficacy by predicting and reconstructing high-resolution Raman data from low-resolution input, eliminating the need for resource-intensive high-resolution scans. While previous work often relied on substantial high-resolution datasets, this study showcases the ability to achieve similar outcomes even with limited data, presenting a more practical and cost-effective approach. In conclusion, the results offer a glimpse into the transformative potential of this technology to streamline hyperspectral imaging applications by saving valuable time and resources through the successful generation of high-resolution data from low-resolution inputs.