[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-81617-en":3,"doc-seo-81617-105":30,"detail-sidebar-cat-0-en-105":91},{"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":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},81617,34359740700684,"Finn","https://ap-avatar.wpscdn.com/avatar/1f400023980c374ae676?_k=1777273430885731487",8,"Research & Report","Bridging the Gap Between Climate Science and Machine Learning in Climate Model Emulation","Physics-based climate models support climate decision-making, but their use is limited by heavy computational and technical barriers. Machine-learning emulators can reduce these costs, yet effective adoption in climate research remains difficult. The document highlights a persistent disconnect: climate scientists often bypass emulators, while ML researchers may present them as methodological demonstrations. It proposes a shared development framework, emphasizing easy-to-adopt emulators, task clarity, and demonstrated reliability to improve real-world usability for applied climate studies.","How Can Machine Learning Emulators Best Support Climate  \nScience?  \nLuca Schmidt ∗  \nCluster of Excellence Machine Learning University of Tübingen [luca.schmidt@uni-tuebingen.de](luca.schmidt@uni-tuebingen.de)  \nNina Effenberger ∗  \nInstitute for Atmospheric and Climate Science ETH Zurich [nina.effenberger@env.ethz.ch](nina.effenberger@env.ethz.ch)  \narXiv :2603 .22320v2 [ cs .LG] 10 Jul 2026  \nVitus Benson  \nMax Planck Institute for Biogeochemistry [vbenson@bgc-jena.mpg.de](vbenson@bgc-jena.mpg.de)  \nPhiline L. Bommer  \nSchool of Geosciences University of Edinburgh [philine.bommer@ed.ac.uk](philine.bommer@ed.ac.uk)  \nRobert Brunstein  \nOtto-von-Guericke-Universität Magdeburg [robert.brunstein@ovgu.de](robert.brunstein@ovgu.de)  \nMikel N. Legasa  \nLaboratoire des Sciences du Climat et de l’Environnement (LSCE) CEA/CNRS/UVSQ, Université Paris Saclay Institut Pierre-Simon Laplace (IPSL) [mikel.legasa@lsce.ipsl.fr](mikel.legasa@lsce.ipsl.fr)  \nMaxim Samarin  \nSwiss Data Science Center EPFL and ETH Zurich [maxim.samarin@sdsc.ethz.ch](maxim.samarin@sdsc.ethz.ch)  \nMaybritt Schillinger  \nSeminar for Statistics  \nETH Zurich  \nmaybritt.schillinger@stat .math .ethz .ch  \nAbstract  \nFor decades, physics-based climate models have been used to provide insights for climate decisionmaking. Their application is, however, constrained by significant computational and technical demands. Machine learning (ML) emulators offer a way to reduce these high computational costs; yet, it remains challenging to use ML emulators effectively in climate research. In practice, climate scientists often bypass emulators altogether, and machine learning researchers frequently develop them as methodological showcases without proving their practical utility. The reasons are diverse, ranging from limited accessibility and a lack of specialized knowledge to broader concerns about the physical grounding of ML methods. Here, we discuss limitations and introduce a framework for guiding emulator development, considering both climate science and machine learning perspectives. We argue that designing easy-to-adopt emulators that address clearly defined tasks and demonstrate their reliability is essential. This offers a promising path towards making machine-learning approaches more relevant and usable for applied climate research.  \n1 Introduction  \nClimate models represent our most advanced understanding of the climate system. Their insights into current and future climate change guide mitigation, adaptation, and other societal and economic decisions [Masson-Delmotte et al., 2021] . Climate model simulations, however, are computationally demanding and technically complex, requiring substantial computational resources and specialized expertise [Balaji et al., 2017] . The increasing demand for climate information, along with the cost and difficulty of running physical climate models, calls for more efficient modeling  \n∗Equal contribution, other authors in alphabetical order  \nPreprint.  \napproaches. Machine learning-based climate model emulators fill this gap. They can approximate specific model parameterizations or act as full statistical and machine learning surrogates (e.g. [Leach et al., 2021, Bouabid et al., 2026, Hickman et al., 2026]) . All such models share the goal of approximating key elements of advanced physical climate models.  \nIn recent years, interest in climate model emulation has increased in both climate science and machine learning, accompanied by rapid methodological development–especially within the methodological community [e.g. Fowler et al., 2025, Rampal et al., 2024b] . For the climate science community, emulators provide an efficient way to augment simulations [Kendon et al., 2025], quantify and reduce uncertainty [Watson-Parris, 2021], and generate climate information that would be infeasible with full climate models. For the machine learning community, particularly in computer vision, climate emulation offers a compelling scientific challenge","cbCaice1hRTjUqtf","https://ap.wps.com/l/cbCaice1hRTjUqtf","pdf",748518,4,1,16,"English","en",105,"# Introduction\n## Motivation for climate model emulation\n## Why communities remain disconnected\n## Need for a shared framework","[{\"question\":\"Why are machine learning emulators considered for climate research?\",\"answer\":\"They reduce the high computational costs and technical demands of running physics-based climate models, enabling more efficient approximation of model components or full statistical surrogates.\"},{\"question\":\"What reasons does the document give for limited practical uptake of ML emulators?\",\"answer\":\"It notes a disconnect between climate scientists and ML developers, including differences in workflows and goals, limited accessibility, and concerns about physical grounding of ML methods.\"},{\"question\":\"What framework does the document recommend for emulator development?\",\"answer\":\"It argues for designing easy-to-adopt emulators that target clearly defined tasks and provide evidence of reliability, integrating both climate science and ML perspectives.\"}]",1784174832,40,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"bridging-the-gap-between-climate-science-and-machine-learning-in-climate-model-emulation","",{"@graph":36,"@context":85},[37,53,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":21},"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":20},"https://docshare.wps.com/document/bridging-the-gap-between-climate-science-and-machine-learning-in-climate-model-emulation/81617/",{"url":52,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":24,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":41,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-25","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why are machine learning emulators considered for climate research?","Question",{"text":75,"@type":76},"They reduce the high computational costs and technical demands of running physics-based climate models, enabling more efficient approximation of model components or full statistical surrogates.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What reasons does the document give for limited practical uptake of ML emulators?",{"text":80,"@type":76},"It notes a disconnect between climate scientists and ML developers, including differences in workflows and goals, limited accessibility, and concerns about physical grounding of ML methods.",{"name":82,"@type":73,"acceptedAnswer":83},"What framework does the document recommend for emulator development?",{"text":84,"@type":76},"It argues for designing easy-to-adopt emulators that target clearly defined tasks and provide evidence of reliability, integrating both climate science and ML perspectives.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,119,122,127,130,134],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":20,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":29,"slug":118},7,"Healthcare","healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]