Search NASA⌕ Search

SEARCH · Search NASA

Results for “Knowledge Discovery”

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 361 records · Page 20
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Automated descriptor selection, volcano curve generation, and active site determination using the DescMAP software

The material space for catalyst discovery is expansive. Volcano curves are traditionally employed to provide physical insights into optimal catalyst characteristics for new material selection. Their generation lies on a single descriptor picked using expert knowledge. Here we present DescMAP, a Python-based software, to automate the selection of descriptors, the generation of volcano maps, and the identification of active sites for structure-sensitive reactions. Here, we consider traditional energy-based and geometric descriptors for structure-sensitive reactions. DescMAP is integrated with the Virtual Kinetic Laboratory (VLab) to provide multiple functionalities. It inputs spreadsheets or template files for flexibility and outputs interactive graphs for post-processing. We demonstrate its features using the non-oxidative dehydrogenation of ethane to ethylene over (111) closed-packed surfaces and the methane total oxidation over various Pt facets. It can be easily applied to other complex chemistries and achieves quick screening of potential catalysts.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
NASA NTRS1997

Gamma-Ray Astrophysics: New Insight Into the Universe

During the 15 years that have passed since the first edition of this book was published, there has been a major increase in our knowledge of gamma-ray astronomy. Much of this advance arises from the extensive results that have been forthcoming from the Compton Gamma-Ray Observatory. There has been the discovery of a new class of gamma-ray objects, namely high-energy gamma- ray-emitting blazars, a special class of Active Galactic Nuclei, whose basic high-energy properties now seem to be understood. A much improved picture of our galaxy now exists in the frequency range of gamma rays. The question of whether cosmic rays are galactic or metagalactic now seems settled with certainty. Significant new information exists on the gamma-ray properties of neutron star pulsars, Seyfert galaxies, and gamma-ray bursts. Substantial new insight has been obtained on solar phenomena through gamma-ray observations. Hence, this seemed to be an appropriate time to write a new edition of this book to add the important scientific implications of these many new findings. The special importance of gamma-ray astrophysics had long been recognized by many physicists and astronomers, and theorists had pursued many aspects of the subject well before the experimental results began to become available. The slower development of the experimental side was not because of a lack of incentive, but due to the substantial experimental difficulties that had to be overcome. Thus, as the gamma-ray results became available in much greater number and detail, it was possible to build upon the theoretical work that already existed and to make substantial progress in the study of many of the phenomena involved. Consequently, a much better understanding of many of the astrophysical phenomena mentioned here and others is now possible. Our principal aims in writing this book are the same as they were for the first edition: to provide a text which describes the significance of gamma-ray astrophysics and to assemble in one place a treatment of gamma rays emitted from bodies in the solar i system, from objects in our galaxy, as well as from interactions between cosmic rays and the interstellar medium, and from beyond our galaxy. Thus, this book is intended for those in astrophysics who wish to have the opportunity to learn more about the evolving field of gamma-ray astronomy and its relationship to the high-energy, evolutionary processes occurring in the universe. The last three chapters of the book provide a general discussion of the experimental aspects of the field that seemed best treated together, separately from the astrophysical aspects of gamma-ray astronomy that are discussed in the first ten chapters.

Fichtel, Carl E.↗
NASA NTRS2018

Exploring Our Solar System with Cubesats and Smallsats: A NASA/JPL Perspective

We are on the threshold of a new era in robotic exploration of our solar system, one in which Cubesats and Smallsats will play an important role. As NASA’s lead center for robotic solar system exploration, JPL has a strategic interest in these new capabilities that enable scientists to expand our knowledge of how our solar system formed, how it works, and even how life originated. JPL’s two MarCO spacecraft are currently being prepared for launch along with the InSight Discovery mission in May of this year. This first pair of interplanetary Cubesats will make their way independently to Mars and provide a relay capability for the InSight lander back to Earth. Lunar Flashlight and Near Earth Asteroid Scout, two Cubesats that are currently under development, will form part of the ‘swarm’ of Cubesats escorted into lunar orbit by NASA’s EM-1 mission – the first launch of its new SLS rocket. This paper will provide an overview of these and other Cubesat and Smallsat mission developments at JPL.

Freeman, Anthony↗
DOE OSTI2023

Hybrid Data‐Driven Discovery of High‐Performance Silver Selenide‐Based Thermoelectric Composites

Optimizing material compositions often enhances thermoelectric performances. However, the large selection of possible base elements and dopants results in a vast composition design space that is too large to systematically search using solely domain knowledge. To address this challenge, a hybrid data-driven strategy that integrates Bayesian optimization (BO) and Gaussian process regression (GPR) is proposed to optimize the composition of five elements (Ag, Se, S, Cu, and Te) in AgSe-based thermoelectric materials. Data is collected from the literature to provide prior knowledge for the initial GPR model, which is updated by actively collected experimental data during the iteration between BO and experiments. Within seven iterations, the optimized AgSe-based materials prepared using a simple high-throughput ink mixing and blade coating method deliver a high power factor of 2100 µW m −1 K −2 , which is a 75% improvement from the baseline composite (nominal composition of Ag 2 Se 1 ). In conclusion, the success of this study provides opportunities to generalize the demonstrated active machine learning technique to accelerate the development and optimization of a wide range of material systems with reduced experimental trials.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Ontologizing health systems data at scale: making translational discovery a reality

Common data models solve many challenges of standardizing electronic health record (EHR) data but are unable to semantically integrate all of the resources needed for deep phenotyping. Open Biological and Biomedical Ontology (OBO) Foundry ontologies provide computable representations of biological knowledge and enable the integration of heterogeneous data. However, mapping EHR data to OBO ontologies requires significant manual curation and domain expertise. We introduce OMOP2OBO, an algorithm for mapping Observational Medical Outcomes Partnership (OMOP) vocabularies to OBO ontologies. Using OMOP2OBO, we produced mappings for 92,367 conditions, 8611 drug ingredients, and 10,673 measurement results, which covered 68–99% of concepts used in clinical practice when examined across 24 hospitals. When used to phenotype rare disease patients, the mappings helped systematically identify undiagnosed patients who might benefit from genetic testing. By aligning OMOP vocabularies to OBO ontologies our algorithm presents new opportunities to advance EHR-based deep phenotyping.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗
DOE OSTI2023

ET-AL: Entropy-targeted active learning for bias mitigation in materials data

Growing materials data and data-driven informatics drastically promote the discovery and design of materials. While there are significant advancements in data-driven models, the quality of data resources is less studied despite its huge impact on model performance. In this work, we focus on data bias arising from uneven coverage of materials families in existing knowledge. Observing different diversities among crystal systems in common materials databases, we propose an information entropy-based metric for measuring this bias. To mitigate the bias, we develop an entropy-targeted active learning (ET-AL) framework, which guides the acquisition of new data to improve the diversity of underrepresented crystal systems. We demonstrate the capability of ET-AL for bias mitigation and the resulting improvement in downstream machine learning models. This approach is broadly applicable to data-driven materials discovery, including autonomous data acquisition and dataset trimming to reduce bias, as well as data-driven informatics in other scientific domains.

36 MATERIALS SCIENCE↗
DOE OSTI2020

Selective Enrichment Coupled with Proteomics to Identify S-Acylated Plasma Membrane Proteins in Arabidopsis

Protein S-acylation, predominately in the form of palmitoylation, is a reversible lipid post-translational modification (PTM) on cysteines that plays important roles in protein localization, trafficking, activity, and complex assembly. The functions and regulatory mechanisms of Sacylation have been extensively studied in mammals, owing to remarkable development of the high-resolution proteomics and the discovery of the S-acylation-related enzymes. However, the advance of S-acylation studies in plants lags farther behind than that in mammals, mainly due to the lack of knowledge about proteins responsible for this process, such as protein acyltransferases (PATs) and their substrates. In this article, a systematic protocol to study global S-acylation in Arabidopsis seedlings is described. The procedures are presented in detail, including preparation of Arabidopsis seedlings, enrichment of plasma membrane (PM) proteins, followed by enrichment of S-acylated proteins/peptides based on the acyl-biotin exchange (ABE) method, and large-scale identification of S-acylated proteins/peptides via mass spectrometry. This approach enables researchers to study S-acylation of PM proteins in plants in a systematic and straightforward way. Basic protocol 1: Preparation of Arabidopsis seedling materials Basic protocol 2: Isolation and enrichment of plasma membrane proteins Support protocol 1: Determination of protein concentration using BCA assay Basic Protocol 3: Enrichment of S-acylated proteins by ABE method Support protocol 2: Protein precipitation by methanol/chloroform method Basic Protocol 4: Trypsin digestion and proteomic analysis Alternate protocol 1: Pre-resin digestion and peptide level enrichment.

S-acylation, palmitoylation, plasma membrane prote↗
DOE OSTI2023

A gated graph transformer for protein complex structure quality assessment and its performance in CASP15

Abstract Motivation Proteins interact to form complexes to carry out essential biological functions. Computational methods such as AlphaFold-multimer have been developed to predict the quaternary structures of protein complexes. An important yet largely unsolved challenge in protein complex structure prediction is to accurately estimate the quality of predicted protein complex structures without any knowledge of the corresponding native structures. Such estimations can then be used to select high-quality predicted complex structures to facilitate biomedical research such as protein function analysis and drug discovery. Results In this work, we introduce a new gated neighborhood-modulating graph transformer to predict the quality of 3D protein complex structures. It incorporates node and edge gates within a graph transformer framework to control information flow during graph message passing. We trained, evaluated and tested the method (called DProQA) on newly-curated protein complex datasets before the 15th Critical Assessment of Techniques for Protein Structure Prediction (CASP15) and then blindly tested it in the 2022 CASP15 experiment. The method was ranked 3rd among the single-model quality assessment methods in CASP15 in terms of the ranking loss of TM-score on 36 complex targets. The rigorous internal and external experiments demonstrate that DProQA is effective in ranking protein complex structures. Availability and implementation The source code, data, and pre-trained models are available at https://github.com/jianlin-cheng/DProQA.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Toward Accelerating Discovery via Physics-Driven and Interactive Multifidelity Bayesian Optimization

Both computational and experimental material discovery bring forth the challenge of exploring multidimensional and often nondifferentiable parameter spaces, such as phase diagrams of Hamiltonians with multiple interactions, composition spaces of combinatorial libraries, processing spaces, and molecular embedding spaces. Often these systems are expensive or time consuming to evaluate a single instance, and hence classical approaches based on exhaustive grid or random search are too data intensive. This resulted in strong interest toward active learning methods such as Bayesian optimization (BO) where the adaptive exploration occurs based on human learning (discovery) objective. However, classical BO is based on a predefined optimization target, and policies balancing exploration and exploitation are purely data driven. In practical settings, the domain expert can pose prior knowledge of the system in the form of partially known physics laws and exploration policies often vary during the experiment. Here, we propose an interactive workflow building on multifidelity BO (MFBO), starting with classical (data-driven) MFBO, then expand to a proposed structured (physics-driven) structured MFBO (sMFBO), and finally extend it to allow human-in-the-loop interactive interactive MFBO (iMFBO) workflows for adaptive and domain expert aligned exploration. These approaches are demonstrated over highly nonsmooth multifidelity simulation data generated from an Ising model, considering spin–spin interaction as parameter space, lattice sizes as fidelity spaces, and the objective as maximizing heat capacity. Detailed analysis and comparison show the impact of physics knowledge injection and real-time human decisions for improved exploration with increased alignment to ground truth. Here, the associated notebooks allow to reproduce the reported analyses and apply them to other systems.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2011

The Atmospheres of Titan and Saturn in the Infrared from Cassini: The Interplay Between Observation and Laboratory Studies

The Composite Infrared Spectrometer (CIRS) aboard the Cassini spacecraft has been recording spectra of Saturn and Titan since its arrival in the Saturn system in 2004. CIRS, a Fourier transform spectrometer, observes the thermal infrared spectrum of both atmospheres from 10 to 1500/cm with resolutions up to 0.5/cm (Flasar et al. 2004). From these data CIRS provides global coverage of the molecular composition of the stratosphere and troposphere, as well as maps of temperature and winds. From such studies CIRS helps reveal the chemistry and evolutionary history of Saturn and Titan and their relationships to other Solar System bodies. The Cassini mission is continuing until 2017, permitting CIRS to search for atmospheric changes during more than a Saturnian season. By combining with results from Voyager (1980, 1981) the baseline becomes more than one Saturnian year (Coustenis et al. 2011). CIRS spectroscopy of the atmospheres of Saturn and Titan has raised a variety of questions that require new laboratory studies. A complete understanding of the CIRS high-resolution atmospheric spectra cannot be fully achieved without new or improved line positions and intensities for some trace molecules (e.g., Nixon et al. 2009). Isotopic variants of some of the more abundant species often need improved line parameters in order to derive isotopic ratios (e.g., Coustenis et al. 2008 and Fletcher et a!. 2009). Isotopic ratios contain information about the history of an atmosphere if experimental fractionation rates are available (Jennings et al. 2009). Some aerosol and haze features continue to defy identification and will not be explained without better knowledge of how these materials are formed and until we obtain their laboratory spectra. The interaction between CIRS investigations and laboratory research has been productive and has already led to new discoveries.

Jennings, D. E.↗
DOE OSTI2025

Phylogenomic discovery and engineering of nitrogen fixation into the bioenergy woody crop poplar

Biological nitrogen fixation (BNF) is a key process enabling plants in specific lineages to convert atmospheric dinitrogen (N₂) into bioavailable ammonia through symbioses with diazotrophic microbes. Expanding this capability beyond native nitrogen-fixing clades into non-nodulating crops would reduce synthetic fertilizer use, lowering energy inputs and environmental impacts in agriculture. Supported by DOE Funding Award DE-SC0018247, the NitFix project advanced foundational knowledge required to engineer root-nodule symbioses in new host species. The team generated the most comprehensive phylogenomic analysis to date of all known nodulating lineages, resolving the evolutionary history of nitrogen-fixing symbiosis and identifying core gene suites retained across nodulating taxa. Through multimodal genomics, transcriptomics, and functional analyses in Medicago truncatula and related species, the project mapped regulatory networks underlying nodule organogenesis, bacterial infection, and nitrogen-fixation efficiency. Key discoveries include the identification of conserved signaling modules for rhizobial recognition, transcription factors controlling nodule differentiation, and metabolic pathways integrating fixed nitrogen into plant growth. The project also developed enabling tools—including optimized transformation pipelines, gene-editing workflows, and imaging-based phenotyping—to accelerate engineering efforts in emerging models. Together, these results refine the mechanistic framework of symbiotic nitrogen fixation and highlight transferable components essential for rewiring these traits into non-nodulating crops.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2026

Ambient-pressure 151-K superconductivity in HgBa 2 Ca 2 Cu 3 O 8+δ via pressure quench

Superconductivity has been a vigorously researched topic since its discovery in 1911. Raising the superconducting transition temperature (T c ) has been the main driving force behind such long-sustained efforts due to its potential for impacting humanity and the fundamental knowledge gained from understanding this macroscopic coherent quantum state at high temperatures. The successful development of high-T c superconductivity will make possible extraordinarily efficient generation, delivery, and utilization of energy and could also enable the development of controlled fusion while impacting other burgeoning fields like quantum computation and quantum electronics. However, progress has been hindered by a longstanding plateau in the record ambient-pressure T c , unchanged since 1993. Subsequent significant advancements in T c have been achieved only under high pressures, preventing the realization of superconductivity’s full potential. To directly address this challenge, we developed a pressure-quench protocol (PQP) to stabilize pressure-induced/-enhanced superconducting states at ambient pressure. Here, we achieve a record ambient-pressure T c of 151 K in the cuprate HgBa 2 Ca 2 Cu 3 O 8+δ via PQP. The experimental results are further supported by synchrotron X-ray diffraction measurements and phonon and electronic structure calculations. This breakthrough opens avenues for stabilizing and exploring ambient-pressure high-T c superconducting states and other quantum states that have been previously only accessible under pressure, paving the way for deeper understanding and practical applications of high-T c superconductivity and beyond.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
NASA NTRS2007

Recovery of Space Shuttle Columbia and Return to Flight of Space Shuttle Discovery

NASA has come a long way in our journey to reduce the risks of operating the Spse Shuttle system. The External Tank bipod Thermal Protection System has been redesigned to eliminate the proximate cause of the Columbia accident. In all areas, we have applied the collective knowledge and capabilities of our Nation to comply with the Columbia Accident Investigation Board recommendations and to raise the bar beyond that. We have taken prudent technical action on potential threats to review and verify the material condition of all critical areas where failure could result in catastrophic loss of the crew and vehicle. We are satisfied that critical systems and elements should operate as intended-safely and reliably. While we will never eliminate all the risks from our human space flight programs, we have eliminated those we can and reduced, controlled, and/or mitigated others. The remaining identified risks will be evaluated for acceptance. Our risk reduction approach has its roots in the system safety engineering hierarchy for hazard abatement long employed in aerospace systems engineering. The components of the hierarchy are, in order of precedence, to: design/redesign; eliminate the hazard/risk; reduce the hazard/risk; and control the hazard/risk and/or mitigate the consequence of the remaining hazard/risk through warning devices, special procedures/capabilities, and/or training. This proven approach to risk reduction has been applied to potential hazards and risks in all critical areas of the Space Shuttle and has guided us through the technical challenges, failures, and successes present in return to flight endeavors. This approach provides the structured deliberation process required to verify and form the foundation for accepting any residual risk across the entire Space Shuttle Program by NASA leadership.

Rudolphi, Michael U.↗
NASA NTRS2012

Comets as Messengers from the Early Solar System - Emerging Insights on Delivery of Water, Nitriles, and Organics to Earth

The question of exogenous delivery of water and organics to Earth and other young planets is of critical importance for understanding the origin of Earth's volatiles, and for assessing the possible existence of exo-planets similar to Earth. Viewed from a cosmic perspective, Earth is a dry planet, yet its oceans are enriched in deuterium by a large factor relative to nebular hydrogen and analogous isotopic enrichments in atmospheric nitrogen and noble gases are also seen. Why is this so? What are the implications for Mars? For icy Worlds in our Planetary System? For the existence of Earth-like exoplanets? An exogenous (vs. outgassed) origin for Earth's atmosphere is implied, and intense debate on the relative contributions of comets and asteroids continues - renewed by fresh models for dynamical transport in the protoplanetary disk, by revelations on the nature and diversity of volatile and rocky material within comets, and by the discovery of ocean-like water in a comet from the Kuiper Belt (cf., Mumma & Charnley 2011). Assessing the creation of conditions favorable to the emergence and sustenance of life depends critically on knowledge of the nature of the impacting bodies. Active comets have long been grouped according to their orbital properties, and this has proven useful for identifying the reservoir from which a given comet emerged (OC, KB) (Levison 1996). However, it is now clear that icy bodies were scattered into each reservoir from a range of nebular distances, and the comet populations in today's reservoirs thus share origins that are (in part) common. Comets from the Oort Cloud and Kuiper Disk reservoirs should have diverse composition, resulting from strong gradients in temperature and chemistry in the proto-planetary disk, coupled with dynamical models of early radial transport and mixing with later dispersion of the final cometary nuclei into the long-term storage reservoirs. The inclusion of material from the natal interstellar cloud is probable, for comets formed in the outer solar system.

Mumma, Michael J.↗
NASA NTRS2021

The Scientific Legacy of NASA’s Operation IceBridge

The National Aeronautics and Space Administration (NASA)’s Operation IceBridge (OIB) was a 13-year (2009–2021) airborne mission to survey land and sea ice across the Arctic, Antarctic, and Alaska. Here, we review OIB’s goals, instruments, campaigns, key scientific results, and implications for future investigations of the cryosphere. OIB’s primary goal was to use airborne laser altimetry to bridge the gap in fine-resolution elevation measurements of ice from space between the conclusion of NASA’s Ice, Cloud, and land Elevation Satellite (ICESat; 2003–2009) and its follow-on, ICESat-2 (launched 2018). Additional scientific requirements were intended to contextualize observed elevation changes using a multisensor suite of radar sounders, gravimeters, magnetometers, and cameras. Using 15 different aircraft, OIB conducted 968 science flights, of which 42% were repeat surveys of land ice, 42% were surveys of previously unmapped terrain across the Greenland and Antarctic ice sheets, Arctic ice caps, and Alaskan glaciers, and 16% were surveys of sea ice. The combination of an expansive instrument suite and breadth of surveys enabled numerous fundamental advances in our understanding of the Earth’s cryosphere. For land ice, OIB dramatically improved knowledge of interannual outlet-glacier variability, ice-sheet, and outlet-glacier thicknesses, snowfall rates on ice sheets, fjord and sub-ice-shelf bathymetry, and ice-sheet hydrology. Unanticipated discoveries included a reliable method for constraining the thickness within difficult-to-sound incised troughs beneath ice sheets, the extent of the firn aquifer within the Greenland Ice Sheet, the vulnerability of many Greenland and Antarctic outlet glaciers to ocean-driven melting at their grounding zones, and the dominance of surface-melt-driven mass loss of Alaskan glaciers. For sea ice, OIB significantly advanced our understanding of spatiotemporal variability in sea ice freeboard and its snow cover, especially through combined analysis of fine-resolution altimetry, visible imagery, and snow radar measurements of the overlying snow thickness. Such analyses led to the unanticipated discovery of an interdecadal decrease in snow thickness on Arctic sea ice and numerous opportunities to validate sea ice freeboards from satellite radar altimetry. While many of its data sets have yet to be fully explored, OIB’s scientific legacy has already demonstrated the value of sustained investment in reliable airborne platforms, airborne instrument development, interagency and international collaboration, and open and rapid data access to advance our understanding of Earth’s remote polar regions and their role in the Earth system.

Joseph A. MacGregor↗
DOE OSTI2023

Predicting the synthesizability of crystalline inorganic materials from the data of known material compositions

Abstract Reliably identifying synthesizable inorganic crystalline materials is an unsolved challenge required for realizing autonomous materials discovery. In this work, we develop a deep learning synthesizability model ( SynthNN ) that leverages the entire space of synthesized inorganic chemical compositions. By reformulating material discovery as a synthesizability classification task, SynthNN identifies synthesizable materials with 7× higher precision than with DFT-calculated formation energies. In a head-to-head material discovery comparison against 20 expert material scientists, SynthNN outperforms all experts, achieves 1.5× higher precision and completes the task five orders of magnitude faster than the best human expert. Remarkably, without any prior chemical knowledge, our experiments indicate that SynthNN learns the chemical principles of charge-balancing, chemical family relationships and ionicity, and utilizes these principles to generate synthesizability predictions. The development of SynthNN will allow for synthesizability constraints to be seamlessly integrated into computational material screening workflows to increase their reliability for identifying synthetically accessible materials.

36 MATERIALS SCIENCE↗
DOE OSTI2020

UC LANL Postdoc Entrepreneur Fellowship Cohort Four

Three Laboratory postdoctoral researchers have been selected for this year's cohort of Entrepreneur Fellows. The six-month fellowship program - a joint initiative of the University of California and Los Alamos, now in its fourth year - supports early career scientists explore applications and evaluate market opportunities for their technologies. The Lab selects Entrepreneur Fellows based on the potential for their projects to have both scientific and commercial impact. The fellows receive six months of facilitated training and mentoring, as well as funding to engage in customer discovery, market exploration and technology advancement

36 MATERIALS SCIENCE↗
DOE OSTI2020

Physically Interpretable Neural Networks for the Geosciences: Applications to Earth System Variability

Abstract Neural networks have become increasingly prevalent within the geosciences, although a common limitation of their usage has been a lack of methods to interpret what the networks learn and how they make decisions. As such, neural networks have often been used within the geosciences to most accurately identify a desired output given a set of inputs, with the interpretation of what the network learns used as a secondary metric to ensure the network is making the right decision for the right reason. Neural network interpretation techniques have become more advanced in recent years, however, and we therefore propose that the ultimate objective of using a neural network can also be the interpretation of what the network has learned rather than the output itself. We show that the interpretation of neural networks can enable the discovery of scientifically meaningful connections within geoscientific data. In particular, we use two methods for neural network interpretation called backward optimization and layerwise relevance propagation, both of which project the decision pathways of a network back onto the original input dimensions. To the best of our knowledge, LRP has not yet been applied to geoscientific research, and we believe it has great potential in this area. We show how these interpretation techniques can be used to reliably infer scientifically meaningful information from neural networks by applying them to common climate patterns. These results suggest that combining interpretable neural networks with novel scientific hypotheses will open the door to many new avenues in neural network‐related geoscience research.

58 GEOSCIENCES↗