Search NASA⌕ Search

SEARCH · Search NASA

Results for “explainable machine learning”

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 181 records · Page 10
All sourcesNASA NTRSDOE OSTI
NASA NTRS2020

A Machine Learning Examination of Hydroxyl Radical Differences Among Model Simulations for CCMI-1

Hydroxyl radical (OH) plays critical roles within the troposphere, such as determining the lifetime of methane (CH4), yet is challenging to model due to its fast cycling and dependence on a multitude of sources and sinks. As a result, the reasons for variations in OH and the resulting CH4 lifetime (TCH4), both between models and in time, are difficult to diagnose. We apply a neural network (NN) approach to address this issue within a group of models that participated in the Chemistry-Climate Model Initiative (CCMI). Analysis of the historical specified dynamics simulations performed for CCMI indicates that the primary drivers of TCH4 differences among ten models are the flux of UV light to the troposphere (indicated by the photolysis frequency JO1D) due mostly to clouds, mixing ratio of tropospheric ozone (O3), the abundance of nitrogen oxides (NOx≡NO+NO2), and details of the various chemical mechanisms that drive OH. Water vapor, carbon monoxide (CO), the ratio of NO:NOx, and formaldehyde (HCHO) explain moderate differences in TCH4, while isoprene, CH4, the photolysis frequency of NO2 by visible light (JNO2), overhead O3 column, and temperature account for little-to-no model variation in CH4. We also apply the NNs to analysis of temporal trends in OH from 1980 to 2015. All models that participated in the specified dynamics historical simulation for CCMI demonstrate a decline in CH4 during the analysed timeframe. The significant contributors to this trend, in order of importance, are tropospheric O3, JO1D, NOx, and H2O, with CO also causing substantial interannual variability in OH burden. Finally, the identified trends in TCH4 are compared to calculated trends in the tropospheric mean OH concentration from previous work, based on analysis of observations. The comparison reveals a robust result for the effect of rising water vapor on OH and CH4, imparting an increasing and decreasing trend of about 0.5% decade(exp -1), respectively. The responses due to NOx, O3 column, and temperature are also in reasonably good agreement between the two studies, though a discrepancy in the CH4 response highlights a need for further examination of the CH4 feedback on the abundance of OH.

Julie M Nicely↗
DOE OSTI2024

On the compatibility of established methods with emerging artificial intelligence and machine learning methods for disaster risk analysis

Abstract There is growing interest in leveraging advanced analytics, including artificial intelligence (AI) and machine learning (ML), for disaster risk analysis (RA) applications. These emerging methods offer unprecedented abilities to assess risk in settings where threats can emerge and transform quickly by relying on “learning” through datasets. There is a need to understand these emerging methods in comparison to the more established set of risk assessment methods commonly used in practice. These existing methods are generally accepted by the risk community and are grounded in use across various risk application areas. The next frontier in RA with emerging methods is to develop insights for evaluating the compatibility of those risk methods with more recent advancements in AI/ML, particularly with consideration of usefulness, trust, explainability, and other factors. This article leverages inputs from RA and AI experts to investigate the compatibility of various risk assessment methods, including both established methods and an example of a commonly used AI‐based method for disaster RA applications. This article utilizes empirical evidence from expert perspectives to support key insights on those methods and the compatibility of those methods. This article will be of interest to researchers and practitioners in risk‐analytics disciplines who leverage AI/ML methods.

Mathematical Methods In Social Sciences↗
DOE OSTI2025

On the Need to Align Intent and Implementation in Uncertainty Quantification for Machine Learning

Quantifying uncertainties for machine learning (ML) models is a foundational challenge in modern data analysis. This challenge is compounded by at least two key aspects of the field: (a) inconsistent terminology surrounding uncertainty and estimation across disciplines, and (b) the varying technical requirements for establishing trustworthy uncertainties in diverse problem contexts. In this position paper, we aim to clarify the depth of these challenges by identifying these inconsistencies and articulating how different contexts impose distinct epistemic demands. We examine the current landscape of estimation targets (e.g., prediction, inference, simulation-based inference), uncertainty constructs (e.g., frequentist, Bayesian, fiducial), and the approaches used to map between them. Drawing on the literature, we highlight and explain examples of problematic mappings. To help address these issues, we advocate for standards that promote alignment between the \textit{intent} and \textit{implementation} of uncertainty quantification (UQ) approaches. We discuss several axes of trustworthiness that are necessary (if not sufficient) for reliable UQ in ML models, and show how these axes can inform the design and evaluation of uncertainty-aware ML systems. Our practical recommendations focus on scientific ML, offering illustrative cases and use scenarios, particularly in the context of simulation-based inference (SBI).

Trivedi, Shubhendu [MIT] (ORCID:0000000312374301)↗
NASA NTRSDate not supplied

Statistical learning framework for safety and failure analysis of a DNN-based autonomous aircraft system

Deep Neural Networks (DNNs) and Machine Learning technology is increasingly used for safety-critical applications in the Aerospace domain. To ensure safe operations, the DNN and the system must undergo rigorous verification and validation, including advanced statistical analyses. Performance and safety of the DNN and system behavior must not only be analyzed for the nominal case, but under numerous off-nominal and failure cases. In this paper we will describe how our statistical learning framework SYSAI can efficiently perform such analyses using the tool’s unique combination of advanced learning modeling and statistical analysis techniques. SYSAI can effectively explore the high-dimensional state and failure space of the system under test; geometrical shape detection of safety regions and boundaries support explainability of the results to the designer. In this paper, we report experiments and results obtained with a vision-based DNN control system (ACT) that is capable of autonomously steering an aircraft down a runway.

Yuning He↗
DOE OSTI2024

Towards verifiable cancer digital twins: tissue level modeling protocol for precision medicine

Cancer exhibits substantial heterogeneity, manifesting as distinct morphological and molecular variations across tumors, which frequently undermines the efficacy of conventional oncological treatments. Developments in multiomics and sequencing technologies have paved the way for unraveling this heterogeneity. Nevertheless, the complexity of the data gathered from these methods cannot be fully interpreted through multimodal data analysis alone. Mathematical modeling plays a crucial role in delineating the underlying mechanisms to explain sources of heterogeneity using patient-specific data. Intra-tumoral diversity necessitates the development of precision oncology therapies utilizing multiphysics, multiscale mathematical models for cancer. This review discusses recent advancements in computational methodologies for precision oncology, highlighting the potential of cancer digital twins to enhance patient-specific decision-making in clinical settings. We review computational efforts in building patient-informed cellular and tissue-level models for cancer and propose a computational framework that utilizes agent-based modeling as an effective conduit to integrate cancer systems models that encode signaling at the cellular scale with digital twin models that predict tissue-level response in a tumor microenvironment customized to patient information. Furthermore, we discuss machine learning approaches to building surrogates for these complex mathematical models. These surrogates can potentially be used to conduct sensitivity analysis, verification, validation, and uncertainty quantification, which is especially important for tumor studies due to their dynamic nature.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2026

Conductivity Spectroscopy for Investigation and Discovery of Photovoltaic Materials

Conductivity spectroscopy is an extremely powerful set of methods for probing the properties of optoelectronic materials, especially photovoltaics, where photoconductivity is one of the best spectroscopic proxies for performance. Despite this power, they are substantially less commonly used than time-resolved photoluminescence (for instance) because they tend to be more expensive to implement (THz) and/or require specialized knowledge (GHz) to construct instruments, which are not widely available. The goal of this review is to illustrate the utility of these experiments in the discovery and study of photovoltaic absorber materials and simultaneously make them more accessible to the community by providing a central tutorial resource. We provide a comprehensive review of how conductivity spectroscopy has developed over the past decade and been applied in the discovery and development of photovoltaic materials, with a primary focus on emerging solution-processable technologies. Along the way we aim to demystify conductivity spectroscopy with focused tutorial sections that explain the physical models used to fit the data and illustrate how to think about “high-frequency conductivity”.

14 SOLAR ENERGY↗
DOE OSTI2025

Uncertainty Quantification for Data-Driven Machine Learning Models in Nuclear Engineering Applications: Where We Are and What Do We Need?

Machine learning (ML) has been leveraged to tackle a diverse range of tasks in almost all branches of nuclear engineering. Many of the successes in ML applications can be attributed to the recent performance breakthroughs in deep learning, the growing availability of computational power, data, and easy-to-use ML libraries. However, these empirical successes have often outpaced our formal understanding of the ML algorithms. An important but under-rated area is uncertainty quantification (UQ) of ML. ML-based models are subject to approximation uncertainty when they are used to make predictions, due to sources including but not limited to, data noise, data coverage, extrapolation, imperfect model architecture and the stochastic training process. The goal of this paper is to clearly explain and illustrate the importance of UQ of ML. We will elucidate the differences in the basic concepts of UQ of physics-based models and data-driven ML models. Various sources of uncertainties in physical modeling and data-driven modeling will be discussed, demonstrated, and compared. We will also present and demonstrate a few techniques to quantify the ML prediction uncertainties, including Monte Carlo dropout, deep ensemble, Bayesian neural networks, Gaussian Processes and conformal prediction. Lastly, we will discuss the need for building a verification, validation and UQ framework to establish ML credibility.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2026

Generative Models for Crystalline Materials

Understanding structure-property relationships in materials is fundamental in condensed matter physics and materials science. Over the past few years, machine learning (ML) has emerged as a powerful tool for advancing this understanding and accelerating materials discovery. Early ML approaches primarily focused on constructing and screening large material spaces to identify promising candidates for various applications. More recently, research efforts have increasingly shifted toward generating crystal structures using end-to-end generative models. This review analyzes the current state of generative modeling for crystal structure prediction and de novo generation. It examines crystal representations, outlines the generative models used to design crystal structures, and evaluates their respective strengths and limitations. Furthermore, the review highlights experimental considerations for evaluating generated structures and provides recommendations for suitable existing software tools. Emerging topics, such as modeling disorder and defects, integration in advanced characterization, incorporating synthetic feasibility constraints, and model explainability are explored. Ultimately, this work aims to inform both experimental scientists looking to adapt suitable ML models to their specific circumstances and ML specialists seeking to understand the unique challenges related to inverse materials design and discovery.

Metni, Houssam [Karlsruhe Inst. of Technology (KIT↗
DOE OSTI2024

Using automated machine learning for the upscaling of gross primary productivity

Estimating gross primary productivity (GPP) over space and time is fundamental for understanding the response of the terrestrial biosphere to climate change. Eddy covariance flux towers provide in situ estimates of GPP at the ecosystem scale, but their sparse geographical distribution limits larger-scale inference. Machine learning (ML) techniques have been used to address this problem by extrapolating local GPP measurements over space using satellite remote sensing data. However, the accuracy of the regression model can be affected by uncertainties introduced by model selection, parameterization, and choice of explanatory features, among others. Recent advances in automated ML (AutoML) provide a novel automated way to select and synthesize different ML models. In this work, we explore the potential of AutoML by training three major AutoML frameworks on eddy covariance measurements of GPP at 243 globally distributed sites. We compared their ability to predict GPP and its spatial and temporal variability based on different sets of remote sensing explanatory variables. Explanatory variables from only Moderate Resolution Imaging Spectroradiometer (MODIS) surface reflectance data and photosynthetically active radiation explained over 70 % of the monthly variability in GPP, while satellite-derived proxies for canopy structure, photosynthetic activity, environmental stressors, and meteorological variables from reanalysis (ERA5-Land) further improved the frameworks' predictive ability. We found that the AutoML framework Auto-sklearn consistently outperformed other AutoML frameworks as well as a classical random forest regressor in predicting GPP but with small performance differences, reaching an r 2 of up to 0.75. We deployed the best-performing framework to generate global wall-to-wall maps highlighting GPP patterns in good agreement with satellite-derived reference data. This research benchmarks the application of AutoML in GPP estimation and assesses its potential and limitations in quantifying global photosynthetic activity.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Observational benchmarks inform representation of soil organic carbon dynamics in land surface models

Abstract. Representing soil organic carbon (SOC) dynamics in Earth system models (ESMs) is a key source of uncertainty in predicting carbon–climate feedbacks. Machine learning models can help identify dominant environmental controllers and establish their functional relationships with SOC stocks. The resulting knowledge can be integrated into ESMs to reduce uncertainty and improve predictions of SOC dynamics over space and time. In this study, we used a large number of SOC field observations (n=54 000), geospatial datasets of environmental factors (n=46), and two machine learning approaches (namely random forest, RF, and generalized additive modeling, GAM) to (1) identify dominant environmental controllers of global and biome-specific SOC stocks, (2) derive functional relationships between environmental controllers and SOC stocks, and (3) compare the identified environmental controllers and predictive relationships with those in models used in Phase 6 of the Coupled Model Intercomparison Project (CMIP6). Our results showed that the diurnal temperature, drought index, cation exchange capacity, and precipitation were important observed environmental predictors of global SOC stocks. While the RF model identified 14 environmental factors that describe climatic, vegetation, and edaphic conditions as important predictors of global SOC stocks (R2=0.61, RMSE = 0.46 kg m−2), current ESMs oversimplify the relationships between environmental factors and SOC, with precipitation, temperature, and net primary productivity explaining > 96 % of the variability in ESM-modeled SOC stocks. Further, our study revealed notable disparities among the functional relationships between environmental factors and SOC stocks simulated by ESMs compared with observed relationships. To improve SOC representations in ESMs, it is imperative to incorporate additional environmental controls, such as the cation exchange capacity, and refine the functional relationships to align more closely with observations.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS2022

Diesel Passenger Vehicle Shares Influenced Covid-19 Changes in Urban Nitrogen Dioxide Pollution

Diesel-powered vehicles emit several times more nitrogen oxides than comparable gasoline-powered vehicles, leading to ambient nitrogen dioxide (NO2) pollution and adverse health impacts. The COVID-19 pandemic and ensuing changes in emissions provide a natural experiment to test whether NO2 reductions have been starker in regions of Europe with larger diesel passenger vehicle shares. Here we use a semi-empirical approach that combines in-situ NO2 observations from urban areas and an atmospheric composition model within a machine learning algorithm to estimate business-as-usual NO2 during the first wave of the COVID-19 pandemic in 2020. These estimates account for the moderating influences of meteorology, chemistry, and traffic. Comparing the observed NO2 concentrations against business-as-usual estimates indicates that diesel passenger vehicle shares played a major role in the magnitude of NO2 reductions. European cities with the five largest shares of diesel passenger vehicles experienced NO2 reductions ∼ 2.5 times larger than cities with the five smallest diesel shares. Extending our methods to a cohort of non-European cities reveals that NO2 reductions in these cities were generally smaller than reductions in European cities, which was expected given their small diesel shares. We identify potential factors such as the deterioration of engine controls associated with older diesel vehicles to explain spread in the relationship between cities’ shares of diesel vehicles and changes in NO2 during the pandemic. Our results provide a glimpse of potential NO2 reductions that could accompany future deliberate efforts to phase out or remove passenger vehicles from cities.

Nitrogen Dioxide↗
DOE OSTI2025

Chromium-doped uranium dioxide fuels: A review

UO 2 doped with parts per million CR 2 O 3 powder is considered a potential near term accident tolerant fuel candidate. Here, the results of decades of industry and academic research into Cr-doped UO 2 are analyzed and their shortcomings are critiqued. Focusing on the incorporation mechanisms of Cr into the fuel matrix, we explore a mechanistic understanding of the characteristic properties of Cr-doped UO 2 , notably, enhanced fission gas retention attributed to enlarged grain sizes following sintering, along with marginal improvements in the thermophysical properties. The findings of recent X-ray Adsorption Near Edge Spectroscopy studies were compared and put into conversation with historic data regarding the incorporation of Cr in UO 2 . On the basis of defect mechanisms, the case is made for the substitutional incorporation of Cr governing the lattice solubility but not the enhanced U diffusivity. Instead, Cr/CR 2 O 3 redox chemistry in a well-defined oxygen potential explains the differences in the U diffusivity and O/M ratio. The primary mechanism of doping enhanced grain growth is found to be liquid assisted sintering due to a CRO (1) eutectic phase at the grain boundaries. The role of inhomogeneities in Cr concentration in UO 2 at various length scales across the materials microstructure is highlighted and connected to promising experimental and modeling work to fill in the gaps in the current understanding of Cr-doped UO 2 . The review considers both the open scientific questions and engineering applications to illustrate the deep connections between the practice and theory in the design of accident tolerant nuclear fuels. In conclusion, the review ends with an outline of future works that combine meticulous irradiation studies and high resolution experiments with next generation modeling and simulations techniques empowered by machine learning advances to accelerate the fabrication and adoption of Cr-doped UO 2 light water reactors.

Cleveland, Mack Wesley [Massachusetts Inst. of Tec↗
DOE OSTI2025

Descriptors for Cu facets for CO2 reduction reaction activity

Computation screening is crucial for designing efficient electrochemical catalysts for carbon dioxide (CO2R) reduction to valuable hydrocarbons and oxygenates. Herein, leveraging density functional theory calculations of the CO adsorption energy ΔE_CO on seventeen Cu terminations, we discover a strong linear correlation between ΔE_CO and the recently experimentally measured CO2R electrochemical currents (ACS Catal. 2022, 12, 11, 6578–6588). Examining the ab initio thermodynamics of the early critical intermediates CO*, COH*, and CHO*, we find that CO* → CHO* is the thermodynamically preferred step, and notably shows a volcano trend with the experimental currents where the maximum CO2R current corresponds to the moderate CHO* formation energy. Importantly, we show that increasing the step and kink density of the Cu termination not only enhances CO adsorption strength but also modulates the CO* → CHO* pathway, as respectively exemplified in the (941) and (741) facets. We also explain why (741) is exceptional with high CO2R activity as measured experimentally due to its relatively low activity toward the hydrogen evolution reaction compared with the other Cu surfaces. Beyond the general CO adsorption energy that only shows a linear trend with CO2R activity, we show that the reaction CO* → CHO* free energy is a descriptor that displays a volcano relationship with the overall CO2R activity on Cu facets.

machine learning↗
DOE OSTI2025

Machine Learning Analysis of Temperature-Strain Relationships for Structural Health Monitoring of Pipes: Self-powered wireless sensor system for health monitoring of liquid-sodium cooled fast reactors

This report presents machine learning (ML) analysis of temperature-strain relationships for structural health monitoring of nuclear reactor stainless steel (SS) pipes with the strain gauge sensor directly printed on the pipe with a 3D conformal aerosol jet printer. We investigate correlations for two sensor pairs installed on the same SS304 pipe: commercial K-type thermocouple with a printed gold strain gauge (TC3-SG3), and commercial K-type thermocouple with commercial Kyowa strain gauge (TC0-SG0). The temperature ranges for the sensor pairs TC0-SG0 and TC3-SG3 are 20.00°C to 266.37°C and 39.95°C to 219.28°C respectively. ML algorithms in this study include Linear Regression (baseline method), Ridge Regression, Lasso Regression, and Gradient Boosting. Performance evaluation metrics include Root Mean Square Error (RMSE), Mean Square Error (MSE), Mean Absolute Error (MAE), R 2 Score, and Explained Variance. Using advanced feature engineering techniques, we extracted 27 temperature-based features and 30 strategic inclusion features. The best performance was obtained with the Gradient Boosting method, which achieves prediction accuracy of R 2 = 0.9999 and RMSE = 7.69 μStrain for TC0-SG0, and R 2 = 0.9998 and RMSE = 18.03 μStrain for TC3-SG3. While the temperature-strain correlations are weaker for the gauge directly printed on the pipe than for the commercial strain gauge, deployment-ready performance exceeding industry standards is achieved for both sensor pairs.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
NASA NTRS2019

Testing a Neural Network Accelerator on a High-Altitude Balloon

The cognitive communications project has been working to re d machine learning approaches to support their deployment and sustained use in space environments. It has historically been difficult to implement such techniques on space platforms, however, due to the computational requirements they levy onto general-purpose avionics hardware. While technologies exist to accelerate the computation of aspects of neural networks, such platforms have not historically been deployed in space environments. Given that testing payloads in such environments can be both cost- and time-prohibitive, high-altitude balloons can be used as a way to approximate a space environment at a much lower cost, thus providing a cost-effective way in which to test newer approaches to hardware acceleration for artificial intelligence which may be deployed onto spacecraft more directly. This paper describes a successful test of a commercial off- the-shelf neural network accelerator on a high-altitude balloon. It begins by explaining our selection criteria when evaluating different commercial neural network acceleration techniques: primary considerations include size, weight, and power (SWaP) as well as ease of integration. Next, the paper describes the development and implementation of an experimental flight test platform: flight and ground components are discussed. Afterward, the paper discusses the experimental payload itself: this includes the experimental procedure as well as the specific image and method used for testing. Finally, the paper concludes with an evaluation of both the experimental device tested at altitude as well as the flight test framework itself, identifying how the existing platform can be used to continue tes g commercial off-the-shelf (COTS) solutions for acceleration.

Clark, Gilbert↗
DOE OSTI2026

Monitoring of Liquid Metal Reactor Heater Zones with Recurrent Neural Network Learning of Temperature Time Series

Advanced high-temperature fluid reactors (ARs), such as sodium fast reactors (SFRs) and molten salt cooled reactors (MSCRs) utilize high-temperature fluids at ambient pressure. To melt the fluid during reactor startup and prevent fluid freezing during cooldown, the thermal–hydraulic systems of such ARs include heater zones consisting of specific heaters with controllers, temperature sensors, and thermal insulation. The failure of heater zones due to insulation material degradation or improper installation, resulting in parasitic heat losses, can lead to fluid freezing. The detection of faults using a heat-transfer model is difficult because of a lack of knowledge of the experimental details. Data-driven machine learning of heater zone temperature time series offers a viable alternative. In this study, we benchmarked the performance of recurrent neural networks (RNNs) in an analysis of heat-up transient temperature time series of heater zones installed on a liquid sodium vessel. The RNN models include long short-term memory (LSTM) and gated recurrent unit (GRU) networks, as well as their bi-directional variants, BiLSTM and BiGRU. Anomalous temperature points were designated using a percentile-based threshold applied to residual fluctuations in the detrended temperature time series. Additionally, the impact of the exponentially weighted moving average (EWMA) method on detection accuracy was examined. The RNN models’ performance was assessed using precision, recall, and F 1 score metrics. Results demonstrated that RNN models effectively detect anomalies in temperature time series with the best models for each heater zone achieving F 1 scores of over 93%. To explain the variations in RNN model performance across different heater zones, we used Kullback–Leibler (KL) divergence to quantify the relative entropy between training and testing data, and the Detrended Fluctuation Analysis (DFA) to assess long-range temporal correlations. For datasets with strong long-range correlations and minimal relative entropy between training and testing data, GRU is the best-performing model. When the data exhibits weaker long-term correlations and a significant relative entropy between training and testing distributions, BiGRU shows the best performance. For the data sets with intermediate values of both KL divergence and DFA, the best performance is obtained with LSTM and BiLSTM, respectively.

gated recurrent unit↗
DOE OSTI2025

Baseflow Identification via Explainable AI With Kolmogorov‐Arnold Networks

Abstract Hydrological models often involve constitutive laws that may not be optimal in every application. We propose to replace such laws with the Kolmogorov‐Arnold networks (KANs), a class of neural networks designed to identify symbolic expressions. We demonstrate KAN's potential on the problem of baseflow identification, a notoriously challenging task plagued by significant uncertainty. KAN‐derived functional dependencies of the baseflow components on the aridity index outperform their original counterparts; they demonstrate that water availability, rather than potential evapotranspiration, drives baseflow by constraining actual evapotranspiration under arid conditions. On a test set, they increase the Nash‐Sutcliffe efficiency (NSE) by 65%, decrease the root mean squared error by 29%, and increase the Kling‐Gupta efficiency by 34%. This superior performance is achieved while reducing the number of fitting parameters from three to two. Next, we use data from 378 catchments across the continental United States to refine the water‐balance equation at the mean‐annual scale. The KAN‐derived equations based on the refined water balance outperform both the current aridity index model, with up to a 105% increase in NSE, and the KAN‐derived equations based on the original water balance. While the performance of our model and tree‐based machine learning methods is similar, KANs offer the advantage of simplicity and transparency and require no specific software or computational tools. This case study focuses on the aridity index formulation, but the approach is flexible and transferable to other hydrological processes. Plain Language Summary Equations used in hydrologic model are often suboptimal, resulting in reduced prediction accuracy and efficiency. We implemented Kolmogorov‐Arnold networks (KAN), a machine learning algorithm for deriving symbolic formulations, to estimate groundwater recharge and showed that it outperforms an existing state‐of‐the‐art semi‐empirical formulation. In hydrology, Nash‐Sutcliffe efficiency (NSE), root mean squared error (RMSE), and Kling‐Gupta efficiency (KGE) are commonly used to evaluate model performance. Higher NSE and KGE values indicate better performance, while lower RMSE values are preferable. Our results show that NSE increased by 71%, RMSE decreased by 32%, and KGE improved by 25%. In addition, KAN identifies an optimal functional form and can be used to derive new analytical formulas using the prior knowledge. The KAN‐inspired equation outperformed the original formulation and reduced the fitting parameters. Furthermore, we refined the water‐balance equation at the mean‐annual scale and showed that, based on the new water‐balance equation, KAN can derive new formulations that are superior to the original aridity index formulations (up to 105% increase in NSE) and KAN‐derived equations based on the original water balance. These findings highlight the significant potential of KAN to advance the scientific understanding of a wide range of hydrologic processes. Key Points Kolmogorov‐Arnold networks (KANs) enhance interpretability of machine‐learned hydrological models KAN‐derived symbolic formulations outperform state‐of‐the‐art semi‐empirical aridity indices KAN‐identified functional form yields an analytical index with fewer fitting parameters and improved performance

baseflow↗