SEARCH · Engineering Papers

Results for “distributed inference”

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.

SourceResearch connectionFrom yearThrough year

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

At least 829 records · Page 46

All sources NASA NTRS DOE OSTI

DOE OSTI2023

Robust sampling for weak lensing and clustering analyses with the Dark Energy Survey

Recent cosmological analyses rely on the ability to accurately sample from high-dimensional posterior distributions. A variety of algorithms have been applied in the field, but justification of the particular sampler choice and settings is often lacking. Here, we investigate three such samplers to motivate and validate the algorithm and settings used for the Dark Energy Survey (DES) analyses of the first 3 yr (Y3) of data from combined measurements of weak lensing and galaxy clustering. We employ the full DES Year 1 likelihood alongside a much faster approximate likelihood, which enables us to assess the outcomes from each sampler choice and demonstrate the robustness of our full results. We find that the ellipsoidal nested sampling algorithm multinest reports inconsistent estimates of the Bayesian evidence and somewhat narrower parameter credible intervals than the sliced nested sampling implemented in polychord. We compare the findings from multinest and polychord with parameter inference from the Metropolis–Hastings algorithm, finding good agreement. We determine that polychord provides a good balance of speed and robustness for posterior and evidence estimation, and recommend different settings for testing purposes and final chains for analyses with DES Y3 data. Our methodology can readily be reproduced to obtain suitable sampler settings for future surveys.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

DOE OSTI2022

Hardware-accelerated inference for real-time gravitational-wave astronomy

The field of transient astronomy has seen a revolution with the first gravitational-wave detections and the arrival of multi-messenger observations they enabled. Transformed by the first detection of binary black hole and binary neutron star mergers, computational demands in gravitational-wave astronomy are expected to grow by at least a factor of two over the next five years as the global network of kilometer-scale interferometers are brought to design sensitivity. With the increase in detector sensitivity, real-time delivery of gravitational-wave alerts will become increasingly important as an enabler of multi-messenger followup. In this work, we report a novel implementation and deployment of deep learning inference for real-time gravitational-wave data denoising and astrophysical source identification. This is accomplished using a generic Inference-as-a-Service model that is capable of adapting to the future needs of gravitational-wave data analysis. Overall, our implementation allows seamless incorporation of hardware accelerators and also enables the use of commercial or private (dedicated) as-a-service computing. Based on our results, we propose a paradigm shift in low-latency and offline computing in gravitational-wave astronomy. Such a shift can address key challenges in peak-usage, scalability and reliability, and provide a data analysis platform particularly optimized for deep learning applications. The achieved sub-millisecond scale latency will also be relevant for any machine learning-based real-time control systems that may be invoked in the operation of near-future and next generation ground-based laser interferometers, as well as the front-end collection, distribution and processing of data from such instruments.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

DOE OSTI2025

Electron density measurements and calculations in a helium capacitively-coupled radio-frequency plasma

We report a comparison of inferred electron density (n e ) in a He capacitively-coupled plasma, deduced from laser-collision induced fluorescence measurements, with values computed using a hybrid simulation framework based on particle-in-cell/Monte Carlo collisions simulations and a fluid model for excited He atoms. The studies were carried out for gas pressures between 50 mTorr and 1000 mTorr and peak-to-peak radio-frequency (13.56 MHz) voltages between 150 V and 350 V, in a highly symmetric source equipped with plane-parallel electrodes. A good agreement is found between the experimental and modeling results for n e except at the lowest operating voltages and gas pressures. The (effective) electron temperature (T e ) values derived by the two methods agree as well reasonably within the plasma bulk. The simulation results are used to compare the density distributions of He + and various He excited levels and their major populating and de-populating channels at 100 mTorr and 1000 mTorr.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2024

Improvement and generalization of ABCD method with Bayesian inference

To find New Physics or to refine our knowledge of the Standard Model at the LHC is an enterprise that involves many factors, such as the capabilities and the performance of the accelerator and detectors, the use and exploitation of the available information, the design of search strategies and observables, as well as the proposal of new models. We focus on the use of the information and pour our effort in re-thinking the usual data-driven ABCD method to improve it and to generalize it using Bayesian Machine Learning techniques and tools. We propose that a dataset consisting of a signal and many backgrounds is well described through a mixture model. Signal, backgrounds and their relative fractions in the sample can be well extracted by exploiting the prior knowledge and the dependence between the different observables at the event-by-event level with Bayesian tools. We show how, in contrast to the ABCD method, one can take advantage of understanding some properties of the different backgrounds and of having more than two independent observables to measure in each event. In addition, instead of regions defined through hard cuts, the Bayesian framework uses the information of continuous distribution to obtain soft-assignments of the events which are statistically more robust. To compare both methods we use a toy problem inspired by pp\to hh\to b\bar b b \bar b p p → h h → b b ‾ b b ‾ , selecting a reduced and simplified number of processes and analysing the flavor of the four jets and the invariant mass of the jet-pairs, modeled with simplified distributions. Taking advantage of all this information, and starting from a combination of biased and agnostic priors, leads us to a very good posterior once we use the Bayesian framework to exploit the data and the mutual information of the observables at the event-by-event level. We show how, in this simplified model, the Bayesian framework outperforms the ABCD method sensitivity in obtaining the signal fraction in scenarios with 1% and 0.5% true signal fractions in the dataset. We also show that the method is robust against the absence of signal. We discuss potential prospects for taking this Bayesian data-driven paradigm into more realistic scenarios.

Alvarez, Ezequiel↗

DOE OSTI2021

A predictive discrete-continuum multiscale model of plasticity with quantified uncertainty

Multiscale models of materials, consisting of upscaling discrete simulations to continuum models, are unique in their capability to simulate complex materials behavior. The fundamental limitation in multiscale models is the presence of uncertainty in the computational predictions delivered by them. In this work, a sequential multiscale model has been developed, incorporating discrete dislocation dynamics (DDD) simulations and a strain gradient plasticity (SGP) model to predict the size effect in plastic deformations of metallic micro-pillars. The DDD simulations include uniaxial compression of micro-pillars with different sizes and over a wide range of initial dislocation densities and spatial distributions of dislocations. An SGP model is employed at the continuum level that accounts for the size-dependency of flow stress and hardening rate. Sequences of uncertainty analyses have been performed to assess the predictive capability of the multiscale model. The variance-based global sensitivity analysis determines the effect of parameter uncertainty on the SGP model prediction. The multiscale model is then constructed by calibrating the continuum model using the data furnished by the DDD simulations. A Bayesian calibration method is implemented to quantify the uncertainty due to microstructural randomness in discrete dislocation simulations (density and spatial distribution of dislocations) on the macroscopic continuum model prediction (size effect in plastic deformation). Here, the outcomes of this study indicate that the discrete-continuum multiscale model can accurately simulate the plastic deformation of micro-pillars, despite the significant uncertainty in the DDD results. Additionally, depending on the macroscopic features represented by the DDD simulations, the SGP model can reliably predict the size effect in plasticity responses of the micropillars with below 10% of error.

36 MATERIALS SCIENCE↗

DOE OSTI2020

Constraining the scatter in the galaxy–halo connection at Milky Way masses

In this work, we develop and implement two new methods for constraining the scatter in the relationship between galaxies and dark matter haloes. These new techniques are sensitive to the scatter at low halo masses, making them complementary to previous constraints that are dependent on clustering amplitudes or rich galaxy groups, both of which are only sensitive to more massive haloes. In both of our methods, we use a galaxy group finder to locate central galaxies in the Sloan Digital Sky Survey main galaxy sample. Our first technique uses the small-scale cross-correlation of central galaxies with all lower mass galaxies. This quantity is sensitive to the satellite fraction of low-mass galaxies, which is in turn driven by the scatter between haloes and galaxies. The second technique uses the kurtosis of the distribution of line-of-sight velocities between central galaxies and neighbouring galaxies. This quantity is sensitive to the distribution of halo masses that contain the central galaxies at fixed stellar mass. Theoretical models are constructed using peak halo circular velocity, Vpeak, as our property to connect galaxies to haloes, and all comparisons between theory and observation are made after first passing the model through the group-finding algorithm. We parametrize scatter as a lognormal distribution in M* at fixed Vpeak, σ[M*|Vpeak]. The cross-correlation technique yields a constraint of σ[M*|Vpeak] = 0.27 ± 0.05 dex at a mean Vpeak of 168 km s-1, corresponding to a scatter in log M* at fixed Mh of σ[M*|Mh] = 0.38 ± 0.06 dex at $M_\text{h}=10^{11.8}\, \text{M}_\odot$. The kurtosis technique yields σ[M*|Vpeak] = 0.30 ± 0.03 at Vpeak = 209 km s-1, corresponding to σ[M*|Mh] = 0.34 ± 0.04 at $M_\text{h}=10^{12.2}\, \text{M}_\odot$. The values of σ[M*|Mh] are significantly larger than the constraints at higher masses, in agreement with the results of hydrodynamic simulations. This increase is only partly due to the scatter between Vpeak and Mh, and it represents an increase of nearly a factor of 2 relative to the values inferred from clustering and group studies at high masses.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2021

Quantitative assessment of Distributed Acoustic Sensing at the Source Physics Experiment, Phase II

In this report, we assess the data recorded by a Distributed Acoustic Sensing (DAS) cable deployed during the Source Physics Experiment, Phase II (DAG) in comparison with the data recorded by nearby 4.5-Hz geophones. DAS is a novel recording method with unprecedented spatial resolution, but there are significant concerns around the data fidelity as the technology is ramped up to more common usage. Here we run a series of tests to quantify the similarity between DAS data and more conventional data and investigate cases where the higher spatial resolution of the DAS can provide new insights into the wavefield. These tests include 1D modeling with seismic refraction and bootstrap uncertainties, assessing the amplitude spectra with distance from the source, measuring the frequency dependent inter-station coherency, estimating time-dependent phase velocity with beamforming and semblance, and measuring the cross-correlation between the geophone and the particle velocity inferred from the DAS. In most cases, we find high similarity between the two datasets, but the higher spatial resolution of the DAS provides increased details and methods of estimating uncertainty.

47 OTHER INSTRUMENTATION↗

DOE OSTI2022

Quantitative assessment of Distributed Acoustic Sensing at the Source Physics Experiment (Phase II)

58 GEOSCIENCES↗

DOE OSTI2020

Automatic and rapid calibration of urban building energy models by learning from energy performance database

Urban building energy modeling (UBEM) is attracting increasing attention in the energy modeling filed. Unlike modeling a single building using detailed building systems information, UBEM generally uses limited high-level building stock data to infer default assumptions about building characteristics and operations. Additionally, this practice inherently brings uncertainty to UBEM. This study introduced a novel method of automatic and rapid calibration of UBEM based on the annual electricity and natural gas energy use data by learning the correlations between crucial model input parameters and the building energy use from the reference building models. A case study was presented to calibrate 72 large office buildings built before 1978 in San Francisco. Seventeen model parameters were selected and Monte Carlo sampling was used to create 1000 samples that reasonably represent the parameter space. Then 1000 simulations were performed for the reference building model to create an energy performance database. The results showed that by learning from the energy performance database, it took less than four simulation runs on average to calibrate a building model. After the calibration, the distributions of each parameter were obtained to replace their single predefined default values. For example, the default lighting power density of 21.39 W/m 2 was calibrated to be 7.50 W/m 2 on average. The case study successfully demonstrated the effectiveness of the novel calibration method for UBEM in the mild climate. The method will be further tested in future for other climate zones and other building types.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

DOE OSTI2021

Determining the jet transport coefficient $\hat{q}$ from inclusive hadron suppression measurements using Bayesian parameter estimation

We report a new determination of $\hat{q}$, the jet transport coefficient of the quark-gluon plasma. We use the JETSCAPE framework, which incorporates a novel multistage theoretical approach to in-medium jet evolution and Bayesian inference for parameter extraction. The calculations, based on the Matter and Lbt jet quenching models, are compared to experimental measurements of inclusive hadron suppression in Au + Au collisions at the BNL Relativistic Heavy Ion Collider (RHIC) and Pb + Pb collisions at the CERN Large Hadron Collider (LHC). The correlation of experimental systematic uncertainties is accounted for in the parameter extraction. The functional dependence of $\hat{q}$ on jet energy or virtuality and medium temperature is based on a perturbative picture of in-medium scattering, with components reflecting the different regimes of applicability of Matter and Lbt. In the multistage approach, the switch between Matter and Lbt is governed by a virtuality scale Q 0 . Comparison of the posterior model predictions to the RHIC and LHC hadron suppression data shows reasonable agreement, with moderate tension in limited regions of phase space. The distribution of $\hat{q}/T^3$ extracted from the posterior distributions exhibits weak dependence on jet momentum and medium temperature T, with 90% credible region (CR) depending on the specific choice of model configuration. The choice of Matter + Lbt, with switching at virtuality Q 0 , has 90% CR of $2<\hat{q}/T^3<4$ for p T jet > 40 GeV=c. The value of Q 0 , determined here for the first time, is in the range 2.0-2.7 GeV.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2020

Flow Channeling in Fracture Networks: Characterizing the Effect of Density on Preferential Flow Path Formation

Flow channelization is a commonly observed phenomenon in fractured subsurface media where the flow of fluids is restricted primarily to highly transmissive fracture networks surrounded by a low-permeability rock matrix. The multiscale structural heterogeneity of these networks results in multiscale flow channelization where preferential flow paths form at length scales ranging from the entire system down to the subfracture size. We present an analysis of how one of the largest scales in fractured media, the network density, influences the degree of flow channeling that occurs using an ensemble of semigeneric three-dimensional discrete fracture network (DFN) simulations. We construct 10 DFNs, whose fracture lengths follow a power law distribution, at four densities for a total of 40 networks. We characterize their structure in terms of the network topology and geometry. Eulerian and Lagrangian observations of the steady-state flow fields obtained within the networks are used to quantify the degree of flow channelization at the network scale. We introduce a measure for the importance-ranking/hierarchy of different flow paths in the network using graph-based analysis of Lagrangian transport by which the degree of flow channeling between networks is compared. These flow observations are then linked to the structural properties of the networks. In general, network-scale flow channeling decreases as the network density increases. However, at low densities, there is more uniform flow within the entire connected network than in high-density networks. We also demonstrate how standard transport observables can be used to infer the degree of flow channelization occurring within a fracture network.

58 GEOSCIENCES↗

DOE OSTI2024

Upgrade of the Lyman-alpha diagnostic system on DIII-D for main chamber edge neutral studies

The LLAMA (Lyman Alpha Measurement Apparatus) pinhole camera diagnostic had previously been deployed on DIII-D to measure radial profiles of the Lyman-α (Ly-α) deuterium neutral line brightness across the plasma boundary in the lower chamber to infer neutral deuterium density and ionization rate profiles. This system has recently been upgraded with a new diagnostic head, named ALPACA, that also encloses two pinhole cameras and duplicates the LLAMA views in the upper chamber. Similar to LLAMA, ALPACA provides two times 20 lines of sight, viewing the plasma edge on the inboard and outboard sides with a radial resolution of ~2.5 cm (FWHM) and an effective time resolution of ~1 ms that allows for the investigation of inter-ELM dynamics. The extended Ly-α system provides better coverage to study neutrals in experiments with various plasma shapes utilizing both the upper and lower divertors. Furthermore, post-campaign calibration of the LLAMA diagnostic has successfully been demonstrated for the first time. This was facilitated by various upgrades to the calibration set-up and detailed measurements of the emissivity distribution of the Ly-α calibration source using a pinhole collimator. It was found that the sensitivity of the inboard LLAMA pinhole camera was reduced by a factor of 2.0 ± 0.2 over the course of six months of plasma operation in 2021. In conclusion, the upgraded Ly-α system, equipped with improved absolute calibration, will provide key input for neutral fueling and pedestal particle transport studies and for 2D edge transport code validation on the DIII-D tokamak.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2024

Definition of fragmentation functions and the violation of sum rules

We point out a problem with the formulation and derivations of sum rules for quark fragmentation functions that impacts their validity in QCD, but which potentially points toward an improved understanding of final states in inclusive hard processes. Fragmentation functions give the distribution of final-state hadrons arising from a parton exiting a hard scattering, and the sum rules for momentum, electric charge, etc. express conservation of these quantities. The problem arises from a mismatch between the quark quantum numbers of the initial quark and the fact that all observed final-state hadrons are confined bound states with color zero. We point that, in a confining theory like QCD, the Wilson line in the operator definition of a fragmentation function entails that the final state in a fragmentation function includes a bound state in the external field generated by the Wilson line. We justify this with the aid of general features of string hadronization. The anomalous bound states are restricted to fractional momentum z = 0 . They tend to invalidate sum rules like the one for charge conservation when applied to the fragmentation functions inferred from experimental data but not the momentum sum rule. We propose to exploit our ideas in future studies as a way to relate the fragmentation functions extracted from inclusive cross sections to more detailed nonperturbative descriptions of final state hadronization. We also describe scenarios wherein the traditional sum rules might remain approximately valid with a reasonably high degree of accuracy. Published by the American Physical Society 2024

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2025

Orphaned oil and gas well methane emission rates quantified using Gaussian plume inversions of ambient observations

Abstract. Annually, ∼ 3.6 million abandoned oil and gas wells in the US emit a combined ∼ 2.6 Tg methane (CH4), adversely affecting climate and regional air quality. However, these estimates depend on emission factors derived from measuring subpopulations of wells that vary by orders of magnitude due to very limited field sampling and poorly characterized distributions. Currently, US protocols to remediate orphaned wells lacks standardized quantification methods needed to both prioritize plugging and account for emission reductions. Therefore, sensitive, reliable, affordable, and scalable CH4 flux quantification methods are needed. We report the use of a simple Gaussian plume method where the dispersion parameters are constrained by in situ ground measurements of CH4 concentration at four locations 7.5–49 m downwind of the orphan well as well as local winds to estimate the leak rate from an orphan well. We derive a flux of 10.53 ± 1.16 kg CH4 h−1 during a venting procedure in April 2023 that agrees with the directly measured volumetric flow rate of 9.00 ± 0.25 kg CH4 h−1. This is 71 % greater than the 5.3 kg CH4 h−1 flux measured 7 months prior. Additionally, we discovered a secondary leak through the surface casing inferred as 0.43–0.67 kg CH4 h−1 both by our ground Gaussian analysis and by transecting the plume with an uncrewed aerial system (UAS). We show that in situ determination of the dispersion parameters used in our Gaussian inversions allows us to measure methane emissions to 15 % accuracy, significantly reducing errors when compared to the standard practice of assuming stability class. Our results help develop simpler methods and protocols for robust orphan well emission quantification that can be used for reporting.

Follansbee, Emily↗

DOE OSTI2026

HydraGNN_Predictive_GFM_2026 - Ensemble of predictive graph foundation models for atomistic materials modeling

This release contains data and parameters of HydraGNN-based graph foundation models trained as a result of the work published in the pre-print "Exascale Multi-Task Graph Foundation Models for Imbalanced, Multi-Fidelity Atomistic Data" by M. Lupo Pasini et al. (https://arxiv.org/abs/2604.15380). We jointly train on 16 open first-principles datasets (544+ million structures covering 85+ elements) using a multi-task architecture with per-dataset heads and a scalable ADIOS2/DDStore data pipeline. On Frontier, we execute six large-scale DeepHyper hyperparameter optimization campaigns in FP64 and promote the top-performing message-passing models to sustained 2,048-node training, yielding a PaiNN-based lead model. The version of HydraGNN used to generate the outputs provided in this release is HydraGNN v5.0 (https://github.com/ORNL/HydraGNN/releases/tag/v5.0) The list of datasets used for the training of the graph foundation model is the following: 1) Alexandria [1] 2) ANI1x [2] 3) MPTrj [3] 4) Open Catalyst 2020 (OC20) [4] 5) Open Catalyst 2022 (OC22) [5] 6) Open Catalyst 2025 (OC25) [6] 7) Open Direct ir Capture 2023 (ODAC23) [7] 8) Open Materials 2024 (OMat24) [8] 9) Open Molecules 2025 (OMol25) [9] 10) OMol25-neutral (subset of OMol25 that contains only molecules with zero total charge) 11) OMol25-non-neutral (subset of OMol25 that contains only molecules with non-zero total charge) 12) Open Polymers 2026 (OPoly2026) [10] 13) Nabla2DFT [11] 14) QCML [12] 15) QM7X [reference 13] 16) transition1x [14] Dataset references: [1] J. Schmidt et al., “A dataset of 175k stable and metastable materials calculated with the PBEsol and SCAN functionals,” Scientific Data, vol. 9, p. 64, 2022. [2] J. S. Smith et al., “The ANI-1ccx and ANI-1x data sets, coupled-cluster and density functional theory properties for molecules,” Scientific Data, vol. 7, p. 134, 2020. [Online]. Available: https: //www.nature.com/articles/s41597-020-0473-z [3] A. Jain et al., “Commentary: The Materials Project: A materials genome approach to accelerating materials innovation,” APL Materials, vol. 1, no. 1, p. 011002, 07 2013. [Online]. Available: https://doi.org/10.1063/1.4812323 [4] L. Chanussot et al., “Open catalyst 2020 (oc20) dataset and community challenges,” ACS Catalysis, vol. 11, no. 10, pp. 6059–6072, 2021. [Online]. Available: https://doi.org/10.1021/acscatal.0c04525 [5] K. Tran et al., “Open catalyst 2022 (oc22) dataset and challenges for oxidation electrocatalysts,” ACS Catalysis, vol. 13, no. 5, pp. 3066–3084, 2023. [Online]. Available: https://doi.org/10.1021/acscatal.2c05426 [6] S. J. Sahoo et al., “The open catalyst 2025 (oc25) dataset and models for solid-liquid interfaces,” arXiv preprint arXiv:2509.17862, 2025. [Online]. Available: https://arxiv.org/abs/2509.17862 [7] A. Sriram et al., “The open DAC 2023 dataset and challenges for sorbent discovery in direct air capture,” ACS Central Science, vol. 10, no. 5, pp. 923–941, 2024. [8] L. Barroso-Luque et al., “Open materials 2024 (omat24) inorganic materials dataset and models,” 2024. [Online]. Available: https://arxiv.org/abs/2410.12771 [9] D. S. Levine et al., “The open molecules 2025 (OMol25) dataset, evaluations, and models,” 2025. [Online]. Available: https://arxiv.org/abs/2505.08762 [10] D. S. Levine et al., The open polymers 2026 (OPoly26) dataset and evaluations,” arXiv preprint arXiv:2512.23117, 2025. [Online]. Available: https://arxiv.org/abs/2512.23117 [11] K. Khrabrov et al., “Nabla2dft: A universal quantum chemistry dataset of drug-like molecules and a benchmark for neural network potentials,” in NeurIPS 2024 Datasets and Benchmarks Track, 2024. [Online]. Available: https://openreview.net/forum?id=ElUrNM9U8c [12] S. Ganscha et al., “The QCML dataset, quantum chemistry reference data from 33.5M DFT and 14.7B semi-empirical calculations,” Scientific Data, vol. 12, p. 406, 2025. [13] J. Hoja et al., “QM7-X, a comprehensive dataset of quantum-mechanical properties spanning the chemical space of small organic molecules,” Scientific Data, vol. 8, p. 43, 2021. [Online]. Available: https://www.nature.com/articles/s41597-021-00812-2 [14] M. Schreiner et al., “Transition1x - a dataset for building generalizable reactive machine learning potentials,” Scientific Data, vol. 9, p. 779, 2022. The folder "datasets_ADIOS2_format" contains the set of pre-processed datasets in Adaptable I/O System (ADIOS) format (https://www.exascaleproject.org/research-project/adios/) that have been used for the development and training of GFMs in this work. The "datasets_ADIOS2_format" directory contains 2 sub-directories, one for the version "v1" of the datasets and one for the version "v2" of the datasets. The version "v1" of the datasets provides values of the total energy as they are extracted from the original data as it was released by the respective institutions. The version "v2" of the datasets provides values of the energy that have been realigned. The realignment was performed by training a linear regression model that predicts the total energy as a function of the chemical composition of the atomistic structure, and then subtract such prediction from the original value of the total energy. Both folders "v1" and "v2" contain 16 sub-directories, each corresponding to an ADIOS2-formatted dataset The folder "DeepHyper-results" contains the configurational files and model's parameters for all the 186 HPO trials that were successfully completed by the scalable hyperparameter optimization (HPO) runs on Frontier. The content of the folder "DeepHyper-results" I structured as follows: 1) task-list.txt: list of mpnn name, jobid, and deephyper task id 2) gfm_${MPNN}_${JOBID}_0.${TASKID}: run directory with checkpoint files 3) gfm_${MPNN}: deephyper summary directory (*.csv) for each specific MPNN type 4) deephyper-experiment-${JOBID}: output and error logs for each job The file "deephyper-sorted.csv" contains the details of each HydraGNN model built and tested by HPO, obtained by merging the (*.csv) filed from each HPO run executed. Out of all the HPO trials, we selected 10 to continue the training of the respective HydraGNN models. Due to limited computational budget available in the LRN070 allocation we could not complete the training till convergence for all these 10 selected models. The folder "models" contains multiple sub-folders, one per each HydraGNN model trained. Each model sub-folder contains the parameters of each HydraGNN model, with multiple checkpoint-restarts. The list of sub-folders are as follows: 1) multidataset_hpo-BEST1-fp64 2) multidataset_hpo-BEST2-fp64 3) multidataset_hpo-BEST3-fp64 4) multidataset_hpo-BEST4-fp64 5) multidataset_hpo-BEST5-fp64 6) multidataset_hpo-BEST6-fp64 7) multidataset_hpo-BEST7-fp64 8) multidataset_hpo-BEST8-fp64 9) multidataset_hpo-BEST9-fp64 10) multidataset_hpo-BEST10-fp64 Within each one of these folders, additional auxiliary log files are provided with descriptions about how the training proceeded. The lead PaiNN-model is contained inside "multidataset_hpo-BEST6-fp64". The file "mlp_branch_weights" contains the parameters of the multi-layer perceptron (MLP) used to reconcile the predictions of the 16 output decoding heads of the HydragNN architectures. The MLP takes in input the chemical composition of the atomistic structure and predicts averaging weights to linearly mix the predictions of each output decoding head toward consolidating them into a single one. The folder "1.1billion-structure-inference" contains 1.1 billion atomistic structures randomly generated. Each structures is associated with energy and forces predicted with the lead-PaiNN model combined with the MLP model for reconciliation of the multi-branch predictions generated by the 16 output decoding heads. The folder "1.1billion-structure-inference" contains 9,300 (*.tar.gz) subdirectories, one per Frontier compute node used to execute the inference at exascale. Once uncompressed, each (*.tar.gz) subdirectory contains an ADIOS2 (*.bp) file container, where each atomistic structure is stored as a PyTorch-Geometric Data object. The file "export_dataset_environment_variables.sh" contains the environment variables that need to be set before running the HydraGNN code to reproduce the results provided in this dataset release. The code that can be used to load the ADIOS2 files, load HydraGNN models, and run inference is available at: https://github.com/ORNL/HydraGNN/releases/tag/v5.0

36 MATERIALS SCIENCE↗

DOE OSTI2024

Using phenology to unravel differential soil water use and productivity in a semiarid savanna

Savannas are water-limited ecosystems characterized by two dominant plant types: trees and an understory primarily made up grass. Different phenology and root structures of these plant types complicate how savanna primary productivity responds to changes in water availability. We tested the hypothesis that productivity in savannas is controlled by the temporal and vertical distribution of soil water content (SWC) and differences in growing season length of understory and tree plant functional types. To quantify the relationship between tree, understory, and savanna-wide phenology and productivity, we used PhenoCam and satellite observations surrounding an eddy covariance tower at a semiarid savanna site in Arizona, USA. We distinguished between SWC across two different depth intervals (shallow, <0–30 cm and deep, >30–100 cm). We found that tree greenness increased with SWC at both depths, while understory greenness was only sensitive to the shallower SWC measurements. Onset of ecosystem dormancy, estimated from satellite observations close to the eddy covariance tower, explained more variability in annual gross primary productivity (GPP) than in other phenometrics. Higher SWC led to an extended growing season, caused by delayed dormancy in trees, but the understory showed no evidence of delayed dormancy in wetter periods. We infer that the timing of ecosystem scale dormancy, driven by trees, is important in understanding changes in a savanna's GPP. These findings highlight the important effects of rainfall during the winter. These findings suggest that savanna GPP is conditional on different responses to moisture availability in each of the dominant vegetation components.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2021

A data-driven global soil heterotrophic respiration dataset and the drivers of its inter-annual variability

Soil heterotrophic respiration (SHR), one of the primary carbon fluxes from terrestrial ecosystems to the atmosphere, is important for carbon-climate feedbacks because of its sensitivity to available litter and soil carbon, climatic conditions, and nutrient availability. However, until recently limited SHR data were available, and most published global SHR estimates have either a short time span, coarse spatial resolution, or reply on overly-simple model formulations. To better understand and quantify the global distribution of SHR and its sensitivity to climate variability, we produced a new global SHR dataset using Random Forest algorithms, up-scaling 455 point data from the Global Soil Respiration Database (SRDB 4.0) with gridded fields of climatic, edaphic and productivity as explanatory variables. We estimated a global total SHR of 46.8 Pg C yr-1 over 1985-2013 (95% confidence interval: 38.6-56.3 Pg C yr-1), with a significant increasing trend of 0.03 Pg C yr-2 during this period. We found that the choice of soil moisture datasets contributes more to the difference among these data-driven SHR members rather than that of productivity, temperature and precipitation data sources. We also analyzed the influence of climatic variables on the inter-annual variability (IAV) of our SHR product. Water availability was the dominant driver of IAV at global scales, although the inferred sensitivity depends on the choice of the soil moisture gridded dataset. At the ecosystem scale, temperature strongly controls the IAV of SHR in tropical forests, while water availability dominates in extra-tropical forest and semi-arid regions. Our machine-learning gridded SHR dataset and outputs from process-based land surface models (TRENDYv6) show agreement for a strong association between water variability and SHR IAV at the global scale, but the two approaches lead to different temporal trend globally and different controlling variables for IAV at the ecosystem scale. Our study provides evidence for the pervasive and important role of water availability in driving SHR, indicating both a direct effect limiting decomposition rates and an indirect effect through the amount of fresh organic matter made available to SHR from productivity. In consideration of potential limitations and uncertainties remaining in our data-driven SHR datasets, we call for a more scientifically designed observation network for SHR, more observation data compilation, and increased use of deep learning methods making maximum use of observation data in hand. This will benefit process-based models, and improve our understanding of SHR response to future anomalous environmental conditions.

Yao, Yitong↗

DOE OSTI2025

WHOLESCALE - Water & Hole Observations Leverage Effective Stress Calculations And Lessen Expenses (Final Technical Report 2020 - 2024)

The WHOLESCALE acronym stands for Water & Hole Observations Leverage Effective Stress Calculations and Lessen Expenses. The goal of the WHOLESCALE project is to simulate the spatial distribution and temporal evolution of stress in the geothermal system at San Emidio in Nevada, United States. To reach this goal, the WHOLESCALE team has developed a methodology to incorporate and interpret data from four methods of measurement into a multi-physics model that couples thermal, hydrological, and mechanical (T H-M) processes. The WHOLESCALE team has applied this methodology at the San Emidio geothermal field, located ~100 km north of Reno, Nevada in the northwestern Basin and Range province. The WHOLESCALE team includes 30 individuals working at two universities, two national laboratories, and one industry partner. Two master-degree students and five post-doctoral researchers have gained professional experience and earned partial financial support via the WHOLESCALE project. The WHOLESCALE team has taken advantage of the perturbations created by changes in pumping operations during planned shutdowns in 2016, 2021, and 2022 to infer temporal changes in the state of stress in the geothermal system at San Emidio, Nevada, U.S. The WHOLESCALE results support the working hypothesis that increasing pore-fluid pressure reduces the effective normal stress acting across fault zones. During normal operations, pumping in deep production wells decreases fluid pressures and thus increases the effective normal stresses on faults, reducing microseismicity. During planned shutdowns, the cessation of production increases pore-fluid pressure and reduces effective normal stress. The WHOLESCALE products generated during the 4-year period between 2020 and 2024 include: three articles published in the open-access, peer-reviewed scientific literature, two master’s theses, 20 presentations or papers at scientific conferences, and 17 data sets available on public repositories. The WHOLESCALE project has been completed in two phases that included three performance periods separated by two Go/No-go Stage Gate Reviews. Tasks were classified by data type (i.e., Geologic Structure, Borehole, Geodesy, Hydrology, Seismology, and Modeling). The first phase of the project started July 31, 2020 and included ongoing project coordination (Task 1), a project kickoff (Task 2), analysis of existing data (Task 3), development of the initial stress model & deployment design (Task 4), and Go/No-go Decision Point #1 (Task 5). Phase II began with implementing the 2022 deployment (Task 6), followed by Go/No-go Decision Point #2 (Task 7) The remainder of Phase II consisted of analyzing data collected during deployment (Task 8), calibration of the stress model on all observations (Task 9), and the Final Review (August 23, 2024) & Reporting (Task 10).

15 GEOTHERMAL ENERGY↗