[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-123160-en":3,"doc-seo-123160-105":30,"detail-sidebar-cat-0-en-105":92},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},123160,7971461740909,"Levi","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","A Critical Review of Physics-Informed Machine Learning Applications in Subsurface Energy Systems","Machine learning offers powerful pattern discovery across large datasets, yet many machine and deep learning models suffer from limited interpretability and insufficient incorporation of domain-specific physical knowledge. Physics-informed machine learning (PIML) integrates governing physical principles into data-driven models, improving generalization, enforcing compliance with physical laws, and enhancing interpretability. This review systematically surveys PIML applications in subsurface energy systems, with emphasis on oil and gas tasks such as seismic analysis, reservoir simulation, hydrocarbon production forecasting, and intelligent decision-making, and extends to carbon and hydrogen storage and geothermal systems for more reliable resource management and operational efficiency.","arXiv :2308 .04457v1 [ cs .LG] 6 Aug 2023  \nA Critical Review of Physics-Informed Machine Learning Applications in  \nSubsurface Energy Systems  \nAbdeldjalil Latrach 1 , Mohamed L. Malki 1, 2* , Misael Morales2, 3 , Mohamed Mehana2 , Minou Rabiei 1  \n1 Energy and Petroleum Engineering Department, University of Wyoming  \n2 Environmental and Earth Sciences Group, Los Alamos National Lab  \n3 Petroleum and Geosystems Engineering, University of Texas at Austin  \n* Corresponding author: [mlmalki@lanl.gov](mlmalki@lanl.gov)  \nAbstract Machine learning has emerged as a powerful tool in various fields, including computer vision, natural language processing, and speech recognition. It can unravel hidden patterns within large data sets and reveal unparalleled insights, revolutionizing many industries and disciplines. However, machine and deep learning models lack interpretability and limited domainspecific knowledge, especially in applications such as physics and engineering. Alternatively, physics-informed machine learning (PIML) techniques integrate physics principles into data-driven models. By combining deep learning with domain knowledge, PIML improves the generalization of the model, abidance by the governing physical laws, and interpretability. This paper comprehensively reviews PIML applications related to subsurface energy systems, mainly in the oil and gas industry. The review highlights the successful utilization of PIML for tasks such as seismic applications, reservoir simulation, hydrocarbons production forecasting, and intelligent decision-making in the exploration and production stages. Additionally, it demonstrates PIML’s capabilities to revolutionize the oil and gas industry and other emerging areas of interest, such as carbon and hydrogen storage; and geothermal systems by providing more accurate and reliable predictions for resource management and operational efficiency.  \nKeywords: physics-informed machine learning, deep learning, subsurface, energy, petroleum engineering.  \n1 Introduction  \nArtificial intelligence, particularly machine learning (ML) and deep learning (DL), has remarkably advanced over the past few decades [1] . These breakthroughs have been primarily fueled by the exponential growth of computational power and the abundance of big data, which can be effectively utilized for training and testing these models. Deep learning has revolutionized numerous scientific fields, ranging from computer vision and medical image diagnosis to natural language processing [2, 3 , 4] . Most of these models were developed in the broad deep-learning community to solve generic problems, where the availability of extensive labeled datasets, allowed training models with billions, and even trillions of parameters. On the other hand, the scientific community lacks this luxury of data abundance; additionally, scientists have more rigorous constraints imposed on their procedures and the associated outputs. Hence, we highlight several limitations that hinder the wide adoption of machine learning for serious and rigorous scientific research: 1) Lack of interpretability: Despite the progress in improving neural networks interpretability, such as gradient-weighted class activation mapping (Grad-CAM) [5] and attention maps [6], ML models are still largely considered black boxes (or gray boxes at best) . 2) Data requirements: ML models required large datasets, usually unavailable in several scientific applications. 3) Abidance to physical laws: ML models have no inherent and explicit tendency to respect physical laws, and their predictions can be physically nonsensical. 4) Extrapolation: ML models are notoriously unable to extrapolate outside the distribution of training data.  \nTo address these limitations, there have been several emerging approaches to incorporate domain knowledge and physical constraints into ML models, an approach to which we will refer hereafter by physics-informed machine learning (PIML) . In the context of PIML,","cbCaiomnhkVUP1Gs","https://ap.wps.com/l/cbCaiomnhkVUP1Gs","pdf",3116915,1,33,"English","en",105,"# Introduction\n## Limitations of generic ML/DL in scientific research\n## Physics-informed machine learning (PIML) as an alternative\n## Neural simulation: neural solvers vs neural operators\n## Comparison with numerical simulators","[{\"question\":\"Why is generic machine/deep learning less suitable for rigorous scientific research in subsurface energy contexts?\",\"answer\":\"Generic models often behave as black boxes, require large labeled datasets, do not inherently enforce physical laws, and struggle to extrapolate outside training distributions.\"},{\"question\":\"What distinguishes physics-informed machine learning (PIML) from standard ML approaches?\",\"answer\":\"PIML injects theoretical physical laws as informative priors, creating inductive biases that constrain the solution space toward physically plausible models while improving interpretability and generalization.\"},{\"question\":\"How does the paper categorize neural simulation methods for differential-equation-based problems?\",\"answer\":\"It distinguishes neural solvers, which target ODE/PDE/SDE solution with neural networks, and neural operators, which learn solution maps of parametric differential equations.\"}]","A Critical Review of Physics-Informed Machine Learning Applications in Subsurface Energy Systems | PDF",1785814971,83,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":87,"head_meta":89,"extra_data":91,"updated_unix":28},"a-critical-review-of-physics-informed-machine-learning-applications-in-subsurface-energy-systems","",{"@graph":36,"@context":86},[37,54,69],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/a-critical-review-of-physics-informed-machine-learning-applications-in-subsurface-energy-systems/123160/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-05","2026-08-04",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"Why is generic machine/deep learning less suitable for rigorous scientific research in subsurface energy contexts?","Question",{"text":76,"@type":77},"Generic models often behave as black boxes, require large labeled datasets, do not inherently enforce physical laws, and struggle to extrapolate outside training distributions.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"What distinguishes physics-informed machine learning (PIML) from standard ML approaches?",{"text":81,"@type":77},"PIML injects theoretical physical laws as informative priors, creating inductive biases that constrain the solution space toward physically plausible models while improving interpretability and generalization.",{"name":83,"@type":74,"acceptedAnswer":84},"How does the paper categorize neural simulation methods for differential-equation-based problems?",{"text":85,"@type":77},"It distinguishes neural solvers, which target ODE/PDE/SDE solution with neural networks, and neural operators, which learn solution maps of parametric differential equations.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":93},[94,98,102,106,111,116,121,124,129,132,136],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":46,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":112,"doc_module":4,"doc_module_name":46,"category_name":113,"show_sort_weight":114,"slug":115},6,"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":46,"category_name":118,"show_sort_weight":119,"slug":120},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":122,"slug":123},30,"research-report",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":46,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":133,"doc_module":4,"doc_module_name":46,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":46,"category_name":138,"show_sort_weight":107,"slug":139},19,"General","general"]