Search NASA⌕ Search

SEARCH · Search NASA

Results for “mean squared error”

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 127 records · Page 7
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Dual‐Transformer Deep Learning Framework for Seasonal Forecasting of Great Lakes Water Levels

Abstract The Great Lakes of North America form one of the largest freshwater systems on Earth, and their lake‐wide average water levels (lake levels) can fluctuate by more than 0.5 m on a seasonal scale. These fluctuations pose substantial challenges for coastal resilience, flood risk management, and navigation planning. Accurate seasonal forecasting of lake levels using traditional mechanistic models is challenging due to the complex physical mechanisms and coupled hydroclimatic processes involved. Recently, deep learning has gained prominence in geoscience applications for its ability to recognize intricate patterns within multiphysical data sets. Here, we introduce a novel Dual‐Transformer deep learning framework, tested on the Great Lakes. This architecture integrates two modified Transformer models: the Prophet, which predicts underlying trends, and the Critic, which refines the Prophet's predictions. The final lake level prediction is derived by weighting the outputs of both models through a multi‐layer perceptron, jointly trained with the Prophet and Critic to enhance overall accuracy. Our results demonstrate that the innovative learning framework achieves the highest prediction accuracy compared to established deep learning models when using identical input features. It attains a root mean square error of 4–7 cm in predicting lake levels up to 6 months in advance across the lakes. Additionally, the Dual‐Transformer model runs six orders of magnitude faster than conventional mechanistic models, producing results in less than one second on a typical personal computer. These findings suggest that our deep learning framework has strong potential to advance lake level prediction and carries important implications for water management and disaster mitigation, thereby enhancing the quality of life in coastal regions.

Chen, Yi [Great Lakes Research Center Michigan Tec↗
DOE OSTI2025

ELM–Wet: Inclusion of a Wet–Landunit With Sub–Grid Representation of Eco–Hydrological Patches and Hydrological Forcing Improves Methane Emission Estimations in the E3SM Land Model (ELM)

Wetlands are the largest emitters of biogenic methane (CH 4 ) and represent the highest source of uncertainty in global CH 4 budgets. Here, we aim to improve the realism of wetland representation in the U.S. Department of Energy's Exascale Earth System Model land surface model, ELM, thereby reducing uncertainty of CH 4 flux predictions. We develop an updated version, ELM-Wet, where we activate a separate landunit for wetlands that handles multiple wetland-specific eco-hydrological patch functional types. We introduce more realistic hydrological forcing through prescribing site-level constraints on surface water elevation, which allows resolving different sustained inundation depth for different patches, and if data exists, prescribing inundation depth. We modified the calculation of aerenchyma transport diffusivity based on observed conductance per leaf area for different vegetation types. We use Bayesian Optimization to parameterize CO 2 and CH 4 fluxes in the developed wet-landunit. Site-level simulations of a coastal non-tidal freshwater wetland in Louisiana were performed with the updated model. Eddy covariance observations of CO 2 and CH 4 fluxes from 2012 to 2013 were used to train the model and data from 2021 were used for validation. Patch-specific chamber flux observations and observations of CH 4 concentration profiles in the soil porewater from 2021 were used for evaluation of the model performance. Our results show that ELM-Wet reduces the model's CH 4 emission root mean squared error by up to 33% and is able to represent inter-daily CO 2 and CH 4 flux variability across the wetland's eco-hydrological patches, including during periods of extreme dry or wet conditions.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Dissipation Scaled Internal Wave Drag in a Global Heterogeneously Coupled Internal/External Mode Total Water Level Model

This study showcases a global, heterogeneously coupled total water level system wherein salinity and temperature outputs from a coarser-resolution (~12 km) ocean general circulation model are used to calculate density-driven terms within a global, higher-resolution (~2.5 km) depth-averaged total water level model. We demonstrate that the inclusion of baroclinic forcing in the barotropic model requires modification of the internal wave drag term to prevent excess degradation of tidal results compared to the barotropic model. By scaling the internal tide dissipation by an easy to calculate dissipation ratio, the resulting heterogeneously coupled model has complex root mean square errors (RMSE) of 2.27 cm in the deep ocean and 12.16 cm in shallow waters for the M 2 tidal constituent. While this represents a 10%–20% deterioration as compared to the barotropic model, the improvements in total water level prediction more than offset this degradation. Global median RMSE compared to observations of total water levels, 30-day sea levels, and non-tidal residuals improve by 1.86 (18.5%), 2.55 (42.5%), and 0.36 (5.3%) cm respectively. The drastic improvement in model performance highlights the importance of including density-driven effects within global hydrodynamic models and will help to improve the results of both hindcasts and forecasts in modeling extreme and nuisance flooding. With only an 11% increase in model run time compared to the fully barotropic total water level model, this approach paves the way for high resolution coastal water level and flood models to be used alongside climate models, improving operational forecasting of total water levels.

Blakely, Coleman Peter [University of Notre Dame, ↗
DOE OSTI2025

Stable Machine‐Learning Parameterization of Subgrid Processes in a Comprehensive Atmospheric Model Learned From Embedded Convection‐Permitting Simulations

Modern climate projections often suffer from inadequate spatial and temporal resolution due to computational limitations, resulting in inaccurate representations of sub-grid processes. A promising technique to address this is the multiscale modeling framework (MMF), which embeds a kilometer-resolution cloud-resolving model (CRM) within each atmospheric column of a host climate model to replace traditional convection and cloud parameterizations. Machine learning offers a unique opportunity to make MMF more accessible by emulating the embedded CRM and reducing its substantial computational cost. Although many studies have demonstrated proof-of-concept success of achieving stable hybrid simulations, it remains a challenge to achieve near operational-level success with real geography and comprehensive variable emulation that includes, for example, explicit cloud condensate coupling. In this study, we present a stable hybrid model capable of integrating for at least 5 years with near operational-level complexity, including coarse-grid geography, seasonality, explicit cloud condensate and wind predictions, and land coupling. Our model demonstrates skillful online performance, achieving a 5-year zonal mean tropospheric temperature bias within 2 K, water vapor bias within 1 g/kg, and a precipitation root mean square error of 0.96 mm/day. Key factors contributing to our online performance include an expressive U-Net architecture and physical thermodynamic constraints for microphysics. With microphysical constraints mitigating unrealistic cloud formation, our work is the first to demonstrate realistic multi-year cloud condensate climatology under the MMF framework. Despite these advances, online diagnostics reveal persistent biases in certain regions, highlighting the need for innovative strategies to further optimize online performance.

Hu, Zeyuan [NVIDIA Corporation, Santa Clara, CA (U↗
DOE OSTI2025

Deep Learning Advances Arctic River Water Temperature Predictions

The accelerated warming in the Arctic poses serious risks to freshwater ecosystems by altering streamflow and river thermal regimes. However, limited research on Arctic River water temperatures exists due to data scarcity and the absence of robust methodologies, which often focus on large, major river basins. To address this, we leveraged the newly released, extensive AKTEMP data set and advanced machine learning techniques to develop a Long Short-Term Memory (LSTM) model. By incorporating ERA5-Land reanalysis data and integrating physical understanding into data-driven processes, our model advanced river water temperature predictions in ungauged, snow- and permafrost-affected basins in Alaska. Our model outperformed existing approaches in high-latitude regions, achieving a median Nash-Sutcliffe Efficiency of 0.95 and root mean squared error of 1.0°C. The LSTM model learned air temperature, soil temperature, solar radiation, and thermal radiation—factors associated with energy balance—were the most important drivers of river temperature dynamics. Soil moisture and snow water equivalent were highlighted as critical factors representing key processes such as thawing, melting, and groundwater contributions. Glaciers and permafrost were also identified as important covariates, particularly in seasonal river water temperature predictions. Our LSTM model successfully captured the complex relationships between hydrometeorological factors and river water temperatures across varying timescales and hydrological conditions. This scalable and transferable approach can be potentially applied across the Arctic, offering valuable insights for future conservation and management efforts.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Estimating Soil Thermal Inertia Profiles From the Passive Equilibration of a Temperature Probe

Knowledge of the distribution of soil thermal properties is important for understanding subsurface hydrological and biogeochemical processes. This study describes and evaluates quick thermal profiling (QTP), a new measurement technique aimed at providing rapid, depth-resolved measurements of soil thermal inertia at numerous locations across the landscape. A cylindrical probe with temperature sensors at multiple depths is quickly inserted into the ground, and soil thermal inertia is estimated from how quickly the probe temperature equilibrates with the soil. To this end, a finite volume heat transfer model is used to generate temperature equilibration time series across combinations of controlling factors, and a gridded search inversion approach is applied to infer soil thermal inertia. Field tests in the Arctic indicate that QTP measurements have a minimum uncertainty of 0.14 J m −2 K −1 s −1/2 and covary with dual-probe heat pulse thermal analyzer measurements (concordance correlation coefficient = 0.56) with a root-mean-square error of 0.40 J m −2 K −1 s −1/2 . Besides demonstrating the value of QTP for estimating thermal inertia, this study identifies various sources of measurement uncertainty, particularly probe-soil contact resistance and frictional heating. Further, analysis of soil samples indicates that thermal inertia can be used to estimate thermal conductivity and dry bulk density in the studied area, although such inferences are highly site-specific. Overall, the QTP method holds promise to generate thermal inertia data products and to complement other characterization approaches for advancing understanding of soil properties across far more locations than is currently possible.

Lamb, J. R. [Lawrence Berkeley National Laboratory↗
DOE OSTI2026

Crowdsourcing the Frontier: Advancing Hybrid Physics‐ML Climate Simulation via a $\$$50,000 Kaggle Competition

Subgrid machine-learning (machine learning [ML]) parameterizations have the potential to introduce a new generation of climate models that incorporate the effects of higher-resolution physics without incurring the prohibitive computational cost associated with more explicit physics-based simulations. However, important issues, ranging from online instability to inconsistent online performance, have limited their operational use for long-term climate projections. To more rapidly drive progress in solving these issues, domain scientists and ML researchers opened up the offline aspect of this problem to the broader ML and data science community with the release of ClimSim, a NeurIPS Data sets and Benchmarks publication, and an associated Kaggle competition. This paper reports on the downstream results of the Kaggle competition by coupling emulators inspired by the winning teams' architectures to an interactive climate model (including full cloud microphysics, a regime historically prone to online instability) and systematically evaluating their online performance. Our results demonstrate that online stability in the low-resolution real-geography setting is reproducible across multiple diverse architectures, which we consider a key milestone. All tested architectures exhibit strikingly similar offline and online biases, though their responses to architecture-agnostic design choices (e.g., expanding the list of input variables) can differ significantly. Multiple Kaggle-inspired architectures achieve state-of-the-art results on certain metrics such as zonal mean bias patterns and global Root Mean Squared Error, indicating that crowdsourcing the essence of the offline problem is one path to improving online performance in hybrid physics-AI climate simulation.

Environmental sciences↗
DOE OSTI2024

Prediction of the Cu oxidation state from EELS and XAS spectra using supervised machine learning

Abstract Electron energy loss spectroscopy (EELS) and X-ray absorption spectroscopy (XAS) provide detailed information about bonding, distributions and locations of atoms, and their coordination numbers and oxidation states. However, analysis of XAS/EELS data often relies on matching an unknown experimental sample to a series of simulated or experimental standard samples. This limits analysis throughput and the ability to extract quantitative information from a sample. In this work, we have trained a random forest model capable of predicting the oxidation state of copper based on its L-edge spectrum. Our model attains an R 2 score of 0.85 and a root mean square error of 0.24 on simulated data. It has also successfully predicted experimental L-edge EELS spectra taken in this work and XAS spectra extracted from the literature. We further demonstrate the utility of this model by predicting simulated and experimental spectra of mixed valence samples generated by this work. This model can be integrated into a real-time EELS/XAS analysis pipeline on mixtures of copper-containing materials of unknown composition and oxidation state. By expanding the training data, this methodology can be extended to data-driven spectral analysis of a broad range of materials.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Inverse design of photonic surfaces via multi fidelity ensemble framework and femtosecond laser processing

We demonstrate a multi-fidelity (MF) machine learning ensemble framework for the inverse design of photonic surfaces, trained on a dataset of 11,759 samples that we fabricate using high throughput femtosecond laser processing. The MF ensemble combines an initial low fidelity model for generating design solutions, with a high fidelity model that refines these solutions through local optimization. The combined MF ensemble can generate multiple disparate sets of laser-processing parameters that can each produce the same target input spectral emissivity with high accuracy (root mean squared errors < 2%). SHapley Additive exPlanations analysis shows transparent model interpretability of the complex relationship between laser parameters and spectral emissivity. Finally, the MF ensemble is experimentally validated by fabricating and evaluating photonic surface designs that it generates for improved efficiency energy harvesting devices. Our approach provides a powerful tool for advancing the inverse design of photonic surfaces in energy harvesting applications.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Heisenberg-limited Hamiltonian learning for interacting bosons

We develop a protocol for learning a class of interacting bosonic Hamiltonians from dynamics with Heisenberg-limited scaling. For Hamiltonians with an underlying bounded-degree graph structure, we can learn all parameters with root mean square error ϵ using ${\mathcal{O}}(1/\epsilon )$ total evolution time, which is independent of the system size, in a way that is robust against state-preparation and measurement error. In the protocol, we only use bosonic coherent states, beam splitters, phase shifters, and homodyne measurements, which are easy to implement on many experimental platforms. A key technique we develop is to apply random unitaries to enforce symmetry in the effective Hamiltonian, which may be of independent interest.

computer science↗
DOE OSTI2025

Geospatial modeling of near subsurface temperatures of the contiguous United States for assessment of materials degradation

Abstract Understanding subsurface temperature variations is crucial for assessing material degradation in underground structures. This study maps subsurface temperatures across the contiguous United States for depths from 50 to 3500 m, comparing linear interpolation, gradient boosting (LightGBM), neural networks, and a novel hybrid approach combining linear interpolation with LightGBM. Results reveal heterogeneous temperature patterns both horizontally and vertically. The hybrid model performed best achieving a root mean square error of 2.61 °C at shallow depths (50–350 m). Model performance generally decreased with depth, highlighting challenges in deep temperature prediction. State-level analyses emphasized the importance of considering local geological factors. This study provides valuable insights for designing efficient underground facilities and infrastructure, underscoring the need for depth-specific and region-specific modeling approaches in subsurface temperature assessment.

Science & Technology - Other Topics↗
DOE OSTI2025

Real-time elemental analysis of liquids for process monitoring using laser-induced breakdown spectroscopy with a liquid wheel sampling approach

This article presents an engineered sampling system that used a rotating wheel to form a thin liquid layer, permitting the use of laser-induced breakdown spectroscopy (LIBS) for in situ, real-time elemental impurity quantification during liquid processing. The sampling approach was demonstrated on eight elements from across the periodic table (Na, Al, K, Ca, Ti, Sr, Mo, and Yb). Univariate and multivariate calibrations were presented for each element. The average value for percent root mean square errors of cross-validation for the multivariate models was 3.64%, highlighting the method's strong prediction accuracy. Additionally, the limits of detection for each analyte were estimated from their univariate models: Na = 0.0532, Al = 18.5, K = 0.105, Ca = 0.273, Ti = 67.7, Sr = 0.640, Mo = 22.4, and Yb = 22.9 μg mL –1 . Finally, a test in which multivariate models were used to monitor a liquid system for 80 min was performed to investigate the real-time monitoring capabilities of this liquid LIBS sampling approach. Rigorous measurements were performed to effectively predict the absence and concentrations of multiple analytes as they were spiked and diluted. This demonstration showed the feasibility of using LIBS for real-time liquid quantification models with estimated precision ≤ 8.1%. Finally, the limitations of this approach and potential future improvements are discussed.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Computational investigation of the impact of metal–organic framework topology on hydrogen storage capacity

Metal–organic frameworks (MOFs) are promising, tunable materials for hydrogen storage. For application under cryogenic operating conditions, past work has run into a ceiling on performance due to a trade-off in the volumetric deliverable capacity (VDC) versus the gravimetric deliverable capacity (GDC). In this study, we computationally constructed and screened 105 230 MOF structures based on 529 nets to explore the effect of underlying topology on the hydrogen storage performance of the resulting materials. A machine learning model was developed based on simulated hydrogen uptake to facilitate screening of the entire dataset, and it successfully identified the top 10% of materials with a root-mean-square error of approximately 1 g L −1 as validated by subsequent grand canonical Monte Carlo simulations. We identified a promising structure based on the tsx topology that exhibits both VDC and GDC higher than the current benchmark material, MOF-5. Our data-driven analysis indicates that nets with higher net density yield MOFs with enhanced volumetric and gravimetric surface areas, thereby improving maximum VDC while shifting the capacity trade-off toward higher GDC.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Online monitoring of lanthanide species with combined spectroscopy in flowing aqueous aerosol systems

Combined spectroscopic analysis through absorption and laser-induced breakdown spectroscopy (LIBS) was used to monitor Nd and Pr concentrations in a flowing aqueous system. A unique, online sampling approach was employed that allowed for an approximately closed analysis loop of liquids from a reservoir. Absorption spectroscopy was performed with an optical flow cell, and LIBS was performed on an aerosol stream. Multivariate calibrations based on combined absorption and LIBS signals were built for Nd and Pr and then used to monitor concentrations in mixed solutions in a series of spiking tests. In these tests, the concentrations of Nd and Pr in solution were intermittently changed while spectroscopic signals were monitored in real-time. The combined spectroscopic signals and multivariate models were successful in monitoring changing concentrations of lanthanide species with high accuracy and minimal latency. Root-mean squared error of predictions were 0.015 mol L −1 and 0.019 mol L −1 for Nd and Pr respectively, and these lanthanides were able to be monitored to an accuracy of < 0.2 wt%. This work demonstrates both the capabilities of a sampling system for online analysis of liquids and the capabilities of multimodal spectroscopic characterization for real-time, continuous tracking of species in potentially hazardous liquid systems.

absorbance spectroscopy↗
DOE OSTI2025

Remote sensing of Pu in uranyl nitrate crystals using reflectance spectroscopy and chemometrics

Remote quantification of Pu(VI) (0–5 mol%) co-crystallized with U in uranyl nitrate hexahydrate (UNH) crystals was achieved in a glove box using reflectance spectroscopy coupled with chemometric modeling. Reflectance spectra were also acquired for Pu(IV) and Np(VI) (0–5 mol%) crystallized with UNH; revealing spectral features consistent with their solution-phase analogs. Principal component analysis revealed Pu(IV/VI) and Np(VI) concentrations as the primary source of variation in the data, informing the development of a supervised partial least squares regression model for Pu(VI). The resulting calibration demonstrated robust performance, with replicate root mean square errors near 10% and quantifiable limits near 0.2 mol% Pu(VI) relative to U. The Pu(VI) remained stable in the crystalline UNH matrix for at least one week with minimal reduction to Pu(IV). Notably, Pu(VI) and Np(VI) incorporation in UNH quenched U(VI) fluorescence while Pu(IV) did not. This study presents a noninvasive, spectroscopic approach for solid-state Pu quantification, with direct implications for material accountability and nuclear nonproliferation monitoring.

Sadergaski, Luke R. [Oak Ridge National Laboratory↗
DOE OSTI2026

Accurate energies for ππ* excited states via exchange scaling: the XS-CASSCF method

The state-averaged complete-active space self-consistent field method (SA-CASSCF) is a widely employed electronic structure method used for studying photochemistry and dynamics owing to its ability to provide a reliable description even of complicated cases while still retaining computational efficiency. However, SA-CASSCF suffers from one Achilles heel, related to the description of ionic ππ* excited states, whose energy is often overestimated by 1–2 eV. In light of this challenge, we present the XS-CASSCF method, a new approach based on the idea of exchange scaling (XS) that screens the involved energy terms to improve the excitation energies of singlet ionic ππ* states. First, we illustrate the power of the XS-CASSCF method using hexatriene and para-quinodimethane as examples, showing that it corrects the targeted ionic states while leaving the other states largely unaffected, giving root-mean-square errors (RMSE) below 0.2 eV for the four lowest states in both cases. Subsequently, XS-CASSCF vertical excitation energies are tested against theoretical best estimates for a set of 11 molecules and 56 excited states. XS-CASSCF performs exceptionally well for the ππ* states of hydrocarbons, reducing the RMSE over 21 excitation energies from 0.96 to 0.27 eV. In the challenging subset of molecules with heteroatoms and a larger number of ππ* and nπ* states, we find that improvements can also be obtained, albeit not as pronounced. We conclude with an outlook into more realistic molecular materials focusing on their singlet–triplet (S 1 /T 1 ) gaps, finding that significant improvements can be obtained along the whole range of S 1 /T 1 gaps studied, going from 0.1 eV to more than 1.5 eV. Owing to notable improvements across significant classes of molecules combined with its conceptual simplicity, we believe that XS-CASSCF is a promising addition to the electronic structure toolbox, serving both as a standalone electronic structure method and as a starting point for further correlated treatment.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Active Learning of Microgrid Frequency Dynamics Using Neural Ordinary Differential Equations

Accurate frequency modelling of inverter‐based resource (IBR)‐dominated power systems is crucial for ensuring stable, reliable and resilient operations, particularly given their inherent low‐inertia characteristics and fast dynamics that traditional swing equation‐based models inadequately capture. This paper explores neural ordinary differential equations (Neural ODEs) as a computationally efficient, data‐driven framework for modelling power system frequency dynamics, specifically within microgrids integrating high penetrations of distributed energy resources (DERs). The developed neural ODEs framework incorporates a neural network architecture designed to capture input dynamics. By actively perturbing the system with a known signal, the Python‐based neural ODEs framework was trained using measured system states and inputs, without the need for detailed system information. The framework, tested on a model of the Cordova, AK, microgrid, achieved a goodness of fit ranging from 60% to 99% across different state variables and maintained a mean square error in the 10 -6 p.u. range under square and step excitation signals. The proposed approach demonstrated robustness to measurement noise and initial condition variations while maintaining low computational complexity suitable for real‐time power system control applications. Furthermore, transfer learning enabled the neural ODEs model to adapt to the following changes in system topology or generator dispatch, highlighting its effectiveness for dynamic microgrids with frequently evolving configurations and diverse DERs.

Aryal, Tara [South Dakota State Univ., Brookings, ↗
DOE OSTI2025

Prediction of hydration energies of adsorbates at Pt(111) and liquid water interfaces using machine learning

Aqueous phase heterogeneous catalysis is important to various industrial processes, including biomass conversion, Fischer–Tropsch synthesis, and electrocatalysis. Accurate calculation of solvation thermodynamic properties is essential for modeling the performance of catalysts for these processes. Explicit solvation methods employing multiscale modeling, e.g., involving density functional theory and molecular dynamics have emerged for this purpose. Although accurate, these methods are computationally intensive. This study introduces machine learning (ML) models to predict solvation thermodynamics for adsorbates on a Pt(111) surface, aiming to enhance computational efficiency without compromising accuracy. In particular, ML models are developed using a combination of molecular descriptors and fingerprints and trained on previously published water–adsorbate interaction energies, energies of solvation, and free energies of solvation of adsorbates bound to Pt(111). These models achieve root mean square error values of 0.09 eV for interaction energies, 0.04 eV for energies of solvation, and 0.06 eV for free energies of solvation, demonstrating accuracy within the standard error of multiscale modeling. Feature importance analysis reveals that hydrogen bonding, van der Waals interactions, and solvent density, together with the properties of the adsorbate, are critical factors influencing solvation thermodynamics. Furthermore, these findings suggest that ML models can provide rapid and reliable predictions of solvation properties. This approach not only reduces computational costs but also offers insights into the solvation characteristics of adsorbates at Pt(111)–water interfaces.

Adsorption↗