Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “latent feature models”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

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

At least 91 records · Page 5
All sourcesNASA NTRSDOE OSTI
NASA NTRS2001

A Study of the Response of Deep Tropical Clouds to Mesoscale Processes: Modeling Strategies and Simulations of TOGA-COARE Convective Systems - Part 1

Interactions between deep tropical clouds over the western Pacific warm pool and the larger-scale environment are key to understanding climate change. Cloud models are an extremely useful tool in simulating and providing statistical information on heat and moisture transfer processes between cloud systems and the environment, and can therefore be utilized to substantially improve cloud parameterizations in climate models. In this paper, the Goddard Cumulus Ensemble (GCE) cloud-resolving model is used in multi-day simulations of deep tropical convective activity over the Tropical Ocean-Global Atmosphere Coupled Ocean-Atmosphere Response Experiment (TOGA COARE). Large-scale temperature and moisture advective tendencies, and horizontal momentum from the TOGA-COARE Intensive Flux Array (IFA) region, are applied to the GCE version which incorporates cyclical boundary conditions. Sensitivity experiments show that grid domain size produces the largest response to domain-mean temperature and moisture deviations, as well as cloudiness, when compared to grid horizontal or vertical resolution, and advection scheme. It is found that a minimum grid-domain size of 500 km is needed to adequately resolve the convective cloud features. The control experiment shows that the atmospheric heating and moistening is primarily a response to cloud latent processes of condensation/evaporation, and deposition/sublimation, and to a lesser extent, melting of ice particles. Air-sea exchange of heat and moisture is found to be significant, but of secondary importance, while the radiational response is small. The simulated rainfall and atmospheric heating and moistening, agrees well with observations, and performs favorably to other models simulating this case.

Johnson, Daniel E.↗
DOE OSTI2024

Selecting Appropriate Model Complexity: An Example of Tracer Inversion for Thermal Prediction in Enhanced Geothermal Systems

Abstract A major challenge in the inversion of subsurface parameters is the ill‐posedness issue caused by the inherent subsurface complexities and the generally spatially sparse data. Appropriate simplifications of inversion models are thus necessary to make the inversion process tractable and meanwhile preserve the predictive ability of the inversion results. In this study, we investigate the effect of model complexity on fracture aperture inversion and thermal performance prediction in a field‐scale EGS model. Principal component analysis was used to map the aperture field to a low‐dimensional latent space. The complexity of the inversion model was quantitatively represented by the percentage of total variance in the original aperture fields preserved by the latent space. Tracer, pressure and flow rate data were used to invert for fracture aperture through an ensemble‐based inversion method, and the inferred aperture field was used to predict thermal performance. With an over‐simplified aperture model, ensemble collapse occurred. The inverted aperture models failed to resolve necessary flow and transport features, leading to a biased thermal performance prediction. A complex aperture model involved excessive features and was prone to overinterpreting the inversion data. Both the tracer/pressure/flow rate data reproduction and thermal prediction showed significant uncertainties, making it difficult to properly estimate long‐term thermal performance. Fortunately, our results indicate that there exists an appropriate model complexity which can simultaneously match inversion data and predict thermal performance with an acceptable uncertainty. The quality of the fit of tracer data appears to be a useful indicator of such an appropriate model complexity.

15 GEOTHERMAL ENERGY↗
DOE OSTI2026

CTGAN-TVAE

SAND2026-18914O CTGAN-TVAE (Conditional Tabular Generative Adversarial Networks-Tabular Variational Autoencoders) generates extensive sets of variable generation data through a hybrid framework. It enhances latent space representation by combining TVAE's robust feature-embedding with CTGAN's ability to condition categorical variables such as time. CTGAN-TVAE employs a fully connected neural network within a conditional generative adversarial network framework to manage continuous and categorical data effectively, capturing complex feature interactions without needing sequential modeling. This was developed as part of NNSA-MSIPP: Minority Serving Institution Partnership Program, Grant Number DE-NA0004016. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy's National Nuclear Security Administration under contract DE-NA0003525.

Newlun, Cody [Sandia National Lab. (SNL-CA), Liver↗
DOE OSTI2023

Bridging the Gap between Cosmological Simulations with Graph Neural Networks and Domain Adaptation

Deep learning models have been shown to outperform methods that rely on summary statistics, like the power spectrum, in extracting information from complex cosmological data sets. However, due to differences in the subgrid physics implementation and numerical approximations across different simulation suites, models trained on data from one cosmological simulation show a drop in performance when tested on another. Similarly, models trained on any of the simulations would also likely experience a drop in performance when applied to observational data. Training on data from two different suites of the CAMELS hydrodynamic cosmological simulations, we examine the generalization capabilities of Domain Adaptive Graph Neural Networks (DA-GNNs). By utilizing GNNs, we capitalize on their capacity to capture structured scale-free cosmological information from galaxy distributions. Moreover, by including unsupervised domain adaptation via Maximum Mean Discrepancy (MMD), we enable our models to extract domain-invariant features. We demonstrate that DA-GNN achieves higher accuracy and robustness on cross dataset tasks (up to 28% better relative error and up to almost an order of magnitude better χ 2 ). Using data visualizations, we show the effects of domain adaptation on proper latent space data alignment. This shows that DA-GNNs are a promising method for extracting domain-independent cosmological information, a vital step toward robust deep learning for real cosmic survey data.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1988

A four-layer model for the heat budget of homogeneous land surfaces

The present model envisions the heat balance of a homogeneous land surface in terms of available energy, a set of driving potentials, and parameters for the physical state of the soil and vegetation. Two unique features of the model are: (1) the expression of the interaction of evaporation from the soil and from foliage by changes in the value of the saturation vapor pressure deficit of air in the canopy (the conclusions of this interaction being consistent with field observations); and (2) the treatment of sensible and latent heat exchange between the atmosphere and a soil consisting of two discrete layers.

Choudhury, B. J.↗
DOE OSTI2023

Domain Adaptive Graph Neural Networks for Constraining Cosmological Parameters Across Multiple Data Sets

Deep learning models have been shown to outperform methods that rely on summary statistics, like the power spectrum, in extracting information from complex cosmological data sets. However, due to differences in the subgrid physics implementation and numerical approximations across different simulation suites, models trained on data from one cosmological simulation show a drop in performance when tested on another. Similarly, models trained on any of the simulations would also likely experience a drop in performance when applied to observational data. Training on data from two different suites of the CAMELS hydrodynamic cosmological simulations, we examine the generalization capabilities of Domain Adaptive Graph Neural Networks (DA-GNNs). By utilizing GNNs, we capitalize on their capacity to capture structured scale-free cosmological information from galaxy distributions. Moreover, by including unsupervised domain adaptation via Maximum Mean Discrepancy (MMD), we enable our models to extract domain-invariant features. We demonstrate that DA-GNN achieves higher accuracy and robustness on cross-dataset tasks (up to $28\%$ better relative error and up to almost an order of magnitude better $\chi^2$). Using data visualizations, we show the effects of domain adaptation on proper latent space data alignment. This shows that DA-GNNs are a promising method for extracting domain-independent cosmological information, a vital step toward robust deep learning for real cosmic survey data.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2024

Deep kernel methods learn better: from cards to process optimization

Abstract The ability of deep learning methods to perform classification and regression tasks relies heavily on their capacity to uncover manifolds in high-dimensional data spaces and project them into low-dimensional representation spaces. In this study, we investigate the structure and character of the manifolds generated by classical variational autoencoder (VAE) approaches and deep kernel learning (DKL). In the former case, the structure of the latent space is determined by the properties of the input data alone, while in the latter, the latent manifold forms as a result of an active learning process that balances the data distribution and target functionalities. We show that DKL with active learning can produce a more compact and smooth latent space which is more conducive to optimization compared to previously reported methods, such as the VAE. We demonstrate this behavior using a simple cards dataset and extend it to the optimization of domain-generated trajectories in physical systems. Our findings suggest that latent manifolds constructed through active learning have a more beneficial structure for optimization problems, especially in feature-rich target-poor scenarios that are common in domain sciences, such as materials synthesis, energy storage, and molecular discovery. The Jupyter Notebooks that encapsulate the complete analysis accompany the article.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2011

Global Intercomparison of 12 Land Surface Heat Flux Estimates

A global intercomparison of 12 monthly mean land surface heat flux products for the period 1993-1995 is presented. The intercomparison includes some of the first emerging global satellite-based products (developed at Paris Observatory, Max Planck Institute for Biogeochemistry, University of California Berkeley, University of Maryland, and Princeton University) and examples of fluxes produced by reanalyses (ERA-Interim, MERRA, NCEP-DOE) and off-line land surface models (GSWP-2, GLDAS CLM/ Mosaic/Noah). An intercomparison of the global latent heat flux (Q(sub le)) annual means shows a spread of approx 20 W/sq m (all-product global average of approx 45 W/sq m). A similar spread is observed for the sensible (Q(sub h)) and net radiative (R(sub n)) fluxes. In general, the products correlate well with each other, helped by the large seasonal variability and common forcing data for some of the products. Expected spatial distributions related to the major climatic regimes and geographical features are reproduced by all products. Nevertheless, large Q(sub le)and Q(sub h) absolute differences are also observed. The fluxes were spatially averaged for 10 vegetation classes. The larger Q(sub le) differences were observed for the rain forest but, when normalized by mean fluxes, the differences were comparable to other classes. In general, the correlations between Q(sub le) and R(sub n) were higher for the satellite-based products compared with the reanalyses and off-line models. The fluxes were also averaged for 10 selected basins. The seasonality was generally well captured by all products, but large differences in the flux partitioning were observed for some products and basins.

Jimenez, C.↗
DOE OSTI2021

Evolution of Idealized Vortices in Monsoon-Like Shears: Application to Monsoon Depressions

Abstract This study examines processes fundamental to the development of South Asian monsoon depressions using an array of integrations of an idealized convection-permitting numerical model. In each integration, a wave of initially small amplitude is subjected to a different amount of vertical and meridional wind shear, with temperature and moisture fields constructed according to realistic constraints. Based on the evolution of this disturbance into monsoon depression–like vortices, two features of the background environment emerge as important: the low-level gradient of moist static energy (MSE) and the low-level meridional shear. As the low-level MSE gradient steepens, the disturbance becomes stronger and produces more rain. This strengthening results from the interaction of the vortex with latent heat release by convection that is in turn organized by positive MSE advection in the northerly flow west of the vortex. In this region of advection, moister air from the north ascends along upward-sloping isentropes, driving moist convection. The disturbance also becomes stronger with increasing meridional shear, which makes the environment more barotropically unstable. The absence of either of these two features of the background environment prevents substantial growth of the disturbance. Our results suggest that monsoon depression growth in South Asia is fostered by the coexistence of a strong low-level MSE gradient with strong meridional wind shear associated with the monsoon trough.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS2021

Features and Characteristics of the New NASA MicroPuLse NETwork (MPLNET) Automatic Rain Detection Algorithm

The water cycle strongly influences life on Earth. In particular, the precipitation modifies the atmospheric column thermodynamics through the process of evaporation and serves as a proxy for latent heat modulation. For this reason, a correct precipitation parameterization (especially low-intensity precipitation) at global scale, bedsides improving our understanding of the hydrological cycle, it is crucial to reduce the associated uncertainty of the global climate models to correctly forecast future scenarios, i.e. to apply fast mitigation strategies. In this study we developed an algorithm to automatically detect precipitation from lidar measurements obtained by the National and Aeronautics Space Administration (NASA) Micropulse lidar network (MPLNET) permanent observational site in Goddard. The algorithm, once full operational, will deliver in Near Real Time (latency 1.5h) a new rain mask product that will be publicly available on MPLNET website as part of the new Version 3 Level 1.5 data. The methodology, based on an image processing technique, can detect only light precipitation events (defined by intensity and duration) as the morphological filters used through the detection process are applied on the lidar volume depolarization ratio range corrected composite images, i.e. heavy rain events are unusable as the lidar signal is completely extinguished after few meters in the precipitation or no signal detected because of the water accumulated on the receiver optics.

S. Lolli↗
NASA NTRS2001

Quasi-Equilibrium States in the Tropics Simulated by a Cloud-Resolving Model: Specific Features and Budget Analysis - Part 1

A series of long-term integrations using the two-dimensional Goddard Cumulus Ensemble (GCE) model were performed by altering imposed environmental components to produce various quasi-equilibrium thermodynamic states. Model results show that the genesis of a warm/wet quasi-equilibrium state is mainly due to either strong vertical wind shear (from nudging) or large surface fluxes (from strong surface winds), while a cold/dry quasi-equilibrium state is attributed to a remarkably weakened mixed-wind shear (from vertical mixing due to deep convection) along with weak surface winds. In general, latent heat flux and net large-scale temperature forcing, the two dominant physical processes, dominate in the beginning stage of the simulated convective systems, then considerably weaken in the final stage, which leads to quasi-equilibrium states. A higher thermodynamic regime is found to produce a larger rainfall amount, as convective clouds are the leading source of rainfall over stratiform clouds even though the former occupy much less area. Moreover, convective clouds are more likely to occur in the presence of strong surface winds (latent heat flux), while stratiform clouds (especially the well-organized type) are favored in conditions with strong wind shear (large-scale forcing). The convective systems, which consist of distinct cloud types due to the variation in horizontal winds, are also found to propagate differently. Accordingly, convective systems with mixed-wind shear generally propagate in the direction of shear, while the system with strong (multidirectional) wind shear propagates in a more complex way. Based on the results from the temperature (Q1) and moisture (Q2) budgets, cloud-scale eddies are found to act as a hydrodynamic 'vehicle' that cascades the heat and moisture vertically. Several other specific features such as atmospheric stability, CAPE, and mass fluxes are also investigated and found to be significantly different between diverse quasi-equilibrium states. Detailed comparisons between the various states are presented.

Shie, C.-L.↗
NASA NTRSDate not supplied

Use of Design of Experiments and Rule-Based Inference in Determining Neural Network Architectures for Loss of Control Detection

In this work, we describe methods for selecting the neural network architectures and input spaces to implement belief state inference on generic commercial transport aircraft. First, we highlight a case study on the planning, execution, and analysis of a set of experiments to determine the configurations of a conditional variational autoencoder (CVAE). We present a structured method that can be used in a number of aerospace applications, to optimize the structure and training parameters of the CVAE for belief state inference, using Design of Experiments (DOE) statistical methodologies. The motivation for this specific DOE was to identify the appropriate hyperparameters for measuring the CVAE reconstruction probability and latent space, such that the measurements can be used to infer qualitative state changes for the aircraft. We demonstrate that this process yields information about a trained neural network’s utility for this specific application, along with a quantifiable range of certainty. We execute 84 experiments using loss-of-control flight maneuver data from a NASA T-2 aircraft, demonstrating that this empirical process allows us to construct cheap and simple models with specific attributes amenable to belief state inference in aerospace applications. While theoretically, we could create a single CVAE with an input space the size of all measurable flight variables and environmental dynamics, it becomes intractable to use such a neural network in an in-situ intelligent multi-agent system. Using the recommendations from our case study, we introduce a technical approach for feasibly describing the belief space by (1) identifying significant statistical relationships among flight variables using rule induction, (2) using a set of rules that cover all features to define the input space of multiple CVAEs, and (3) forming a belief space based on the joint probability density of their collective latent spaces. This results in a series of relatively small matrix multiplications that can be performed in real time, as opposed to large matrix computations in a single CVAE. We demonstrate the application of this approach on the T-2 flight loss-of control experiments, using the architecture and hyperparameter recommendations from the case study. We compare the utilities of an individual CVAE trained on all flight variables and multiple CVAEs defined on subsets of flight variables for detecting qualitative changes in flight. We demonstrate that the use of multiple CVAEs with smaller input spaces permits the CVAE to capture more granular relationships in the latent space, permitting better state space characterization and loss-of-control detection.

Design of experiments↗
NASA NTRS2000

Shape-dependent control of cell growth, differentiation, and apoptosis: switching between attractors in cell regulatory networks

Development of characteristic tissue patterns requires that individual cells be switched locally between different phenotypes or "fates;" while one cell may proliferate, its neighbors may differentiate or die. Recent studies have revealed that local switching between these different gene programs is controlled through interplay between soluble growth factors, insoluble extracellular matrix molecules, and mechanical forces which produce cell shape distortion. Although the precise molecular basis remains unknown, shape-dependent control of cell growth and function appears to be mediated by tension-dependent changes in the actin cytoskeleton. However, the question remains: how can a generalized physical stimulus, such as cell distortion, activate the same set of genes and signaling proteins that are triggered by molecules which bind to specific cell surface receptors. In this article, we use computer simulations based on dynamic Boolean networks to show that the different cell fates that a particular cell can exhibit may represent a preprogrammed set of common end programs or "attractors" which self-organize within the cell's regulatory networks. In this type of dynamic network model of information processing, generalized stimuli (e.g., mechanical forces) and specific molecular cues elicit signals which follow different trajectories, but eventually converge onto one of a small set of common end programs (growth, quiescence, differentiation, apoptosis, etc.). In other words, if cells use this type of information processing system, then control of cell function would involve selection of preexisting (latent) behavioral modes of the cell, rather than instruction by specific binding molecules. Importantly, the results of the computer simulation closely mimic experimental data obtained with living endothelial cells. The major implication of this finding is that current methods used for analysis of cell function that rely on characterization of linear signaling pathways or clusters of genes with common activity profiles may overlook the most critical features of cellular information processing which normally determine how signal specificity is established and maintained in living cells. Copyright 2000 Academic Press.

Review↗
NASA NTRS1986

A diagnostic study of baroclinic disturbances in polar air streams

Quasi-Lagrangian budgets of mass, vorticity and heat are calculated following disturbances that form within polar air streams. Observed cases are extracted from the European Centre for Medium-range Weather Forecasts analyses during the First GARP Global Experiment. Model-generated cases are extracted from the simulations of extratropical cyclogenesis by Sandgathe. These polar lows grow primarily through basic baroclinic instability processes and exhibit many features of larger maritime extratropical cyclones. Polar lows that originate on the poleward (or Cyclonic - Type C) side of the jet and have considerable midtropospheric positive vorticity advection at formation time are contrasted with lows that form on the equatorward (or Anticyclonic - Type A) side of a nearly straight upper-level jet. The midtropospheric positive vorticity advection must be present to enhance the vertical circulation when the large surface fluxes that are associated with strong outbreaks act to damp the thermal wave amplification. Although latent heat release is an important factor in both types, it is an essential energy source for the Type A low developments on the equatorward side. Although the vorticity balance is initially different for the two types of polar lows, the vorticity budgets during later stages are similar. The heat budget and the thickness tendency equation demonstrated that the self-development process that is present in larger maritime cyclones is also important for polar low intensification. The absence of favorable coupling to a jet stream is the missing factor in a model-generated Type A polar low that failed to develop. Consequently, the mid- and upper-tropospheric wind fields determine which polar lows will intensify to significant amplitudes.

Sinclair, Mark R.↗
NASA NTRS1993

An observational search for CO2 ice clouds on Mars

CO2 ice clouds were first directly identified on Mars by the Mariner 6 and 7 infrared spectrometer limb scans. These observations provided support for early theoretical modeling efforts of CO2 condensation. Mariner 9 IRIS temperature profiles of north polar hood clouds were interpreted as indicating that these clouds were composed of H2O ice at lower latitudes and CO2 ice at higher latitudes. The role of CO2 condensation on Mars has recently received increased attention because (1) Kasting's model results indicated that CO2 cloud condensation limits the magnitude of the proposed early Mars CO2/H2O greenhouse, and (2) Pollack el al.'s GCM results indicated that the formation of CO2 ice clouds is favorable at all polar latitudes during the fall and winter seasons. These latter authors have shown that CO2 clouds play an important role in the polar energy balance, as the amount of CO2 contained in the polar caps is constrained by a balance between latent heat release, heat advected from lower latitudes, and thermal emission to space. The polar hood clouds reduce the amount of CO2 condensation on the polar caps because they reduce the net emission to space. There have been many extensive laboratory spectroscopic studies of H2O and CO2 ices and frosts. In this study, we use results from these and other sources to search for the occurrence of diagnostic CO2 (and H2O) ice and/or frost absorption features in ground based near-infrared imaging spectroscopic data of Mars. Our primary goals are (1) to try to confirm the previous direct observations of CO2 clouds on Mars; (2) to determine the spatial extent, temporal variability, and composition (H2O/CO2 ratio) of any clouds detected; and (3) through radiative transfer modeling, to try to determine the mean particle size and optical depth of polar hood clouds, thus, assessing their role in the polar heat budget.

Bell, James F., III↗
DOE OSTI2022

Dynamic logistic regression and variable selection: Forecasting and contextualizing civil unrest

Civil unrest can range from peaceful protest to violent furor, and researchers are working to monitor, forecast, and assess such events to allocate resources better. Twitter has become a real-time data source for forecasting civil unrest because millions of people use the platform as a social outlet. Additionally, daily word counts are used as model features, and predictive terms contextualize the reasons for the protest. To forecast civil unrest and infer the reasons for the protest, we consider the problem of Bayesian variable selection for the dynamic logistic regression model and propose using penalized credible regions to select parameters of the updated state vector. This method avoids the need for shrinkage priors, is scalable to high-dimensional dynamic data, and allows the importance of variables to vary in time as new information becomes available. A substantial improvement in both precision and F1-score using this approach is demonstrated through simulation. Finally, we apply the proposed model fitting and variable selection methodology to the problem of forecasting civil unrest in Latin America. Our dynamic logistic regression approach shows improved accuracy compared to the static approach currently used in event prediction and feature selection.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Finding simplicity: unsupervised discovery of features, patterns, and order parameters via shift-invariant variational autoencoders *

Abstract Recent advances in scanning tunneling and transmission electron microscopies (STM and STEM) have allowed routine generation of large volumes of imaging data containing information on the structure and functionality of materials. The experimental data sets contain signatures of long-range phenomena such as physical order parameter fields, polarization, and strain gradients in STEM, or standing electronic waves and carrier-mediated exchange interactions in STM, all superimposed onto scanning system distortions and gradual changes of contrast due to drift and/or mis-tilt effects. Correspondingly, while the human eye can readily identify certain patterns in the images such as lattice periodicities, repeating structural elements, or microstructures, their automatic extraction and classification are highly non-trivial and universal pathways to accomplish such analyses are absent. We pose that the most distinctive elements of the patterns observed in STM and (S)TEM images are similarity and (almost-) periodicity, behaviors stemming directly from the parsimony of elementary atomic structures, superimposed on the gradual changes reflective of order parameter distributions. However, the discovery of these elements via global Fourier methods is non-trivial due to variability and lack of ideal discrete translation symmetry. To address this problem, we explore the shift-invariant variational autoencoders (shift-VAEs) that allow disentangling characteristic repeating features in the images, their variations, and shifts that inevitably occur when randomly sampling the image space. Shift-VAEs balance the uncertainty in the position of the object of interest with the uncertainty in shape reconstruction. This approach is illustrated for model 1D data, and further extended to synthetic and experimental STM and STEM 2D data. We further introduce an approach for training shift-VAEs that allows finding the latent variables that comport to known physical behavior. In this specific case, the condition is that the latent variable maps should be smooth on the length scale of the atomic lattice (as expected for physical order parameters), but other conditions can be imposed. The opportunities and limitations of the shift VAE analysis for pattern discovery are elucidated.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Discovering Hidden Geothermal Signatures using Unsupervised Machine Learning

Discovering hidden geothermal resources is a very challenging task. It requires the mining of large datasets, including various diverse data attributes representing subsurface hydrogeological and geothermal conditions. The commonly used Play Fairway Analysis (PFA) typically relies on subject-matter expertise to analyze site or regional data to estimate geothermal conditions and prospectivity. Here, we demonstrate an alternative approach based on machine learning (ML) to process a geothermal dataset of Southwest New Mexico (SWNM). The study region includes low- and medium-temperature hydrothermal systems. However, most of these systems are poorly characterized because of insufficient existing data and limited past explorative studies. This study aims to discover hidden patterns and relationships in the SWNM geothermal dataset to better understand regional hydrothermal conditions. This is achieved by applying an unsupervised machine learning algorithm based on non-negative matrix factorization coupled with customized k-means clustering (NMFk). NMFk can automatically identify (1) hidden (latent) signatures characterizing datasets, (2) the optimal number of these signatures, (3) dominant data attributes associated with each signature, and (4) spatial distribution of the extracted signatures. Here, NMFk is applied to analyze 18 geological, geophysical, hydrogeological, geothermal attributes at 44 locations in SWNM. NMFk successfully finds data patterns and identifies the spatial associations of hydrothermal signatures with the four physiographic provinces in SWNM (Colorado Plateau, Volcanic Field, Basin and Range, and the Rio Grande rift). The algorithm identified up to 5 hydrothermal signatures in the SWNM datasets that differentiate between low- and medium-temperature hydrothermal systems in different provinces. Also, the algorithm identifies two medium-temperature hydrothermal systems in SWNM that require further exploration for geothermal resource development. Based on our analyses, 12 of the attributes are important to identify medium-temperature hydrothermal systems, and the remaining six attributes are critical to characterize low-temperature hydrothermal systems. Based on the obtained results, we identify potential physiographic provinces for further exploration to characterize them as geothermal resources. The resulting NMFk model can be applied to predict geothermal conditions and their uncertainties at new SWNM locations based on limited data from unexplored areas.

58 GEOSCIENCES↗