A machine learning model that outperforms conventional global subseasonal forecast models

A machine learning model that outperforms conventional global subseasonal forecast models | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Physical Sciences - Article A machine learning model that outperforms conventional global subseasonal forecast models Hao Li, Lei Chen, Xiaohui Zhong, Jie Wu, Deliang Chen, Shang-Ping Xie, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3776375/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 30 Jul, 2024 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract Skillful subseasonal forecasts beyond 2 weeks is critical to various sectors of society but pose a grand scientific challenge. Recently, machine learning based weather forecasting models have made remarkable advancements, outperforming the most successful numerical weather predictions (NWP) generated by the European Centre for Medium-Range Weather Forecasts (ECMWF). However, currently, no machine learning based subseasonal forecasting model surpasses conventional models. Here, we introduce FuXi Subseasonal-to-Seasonal (FuXi-S2S), a machine learning based subseasonal forecasting model that provides global daily mean forecasts for up to 42 days, covering 5 upper-air atmospheric variables at 13 pressure levels and 11 surface variables. FuXi-S2S integrates an enhanced FuXi base model with a perturbation module for flow-dependent perturbations in hidden features, which is a crucial for estimating uncertainty and enhancing subseasonal skill. The model is developed using 72 years of daily statistics from ECMWF ERA5 reanalysis data. When compared to the state-of-the-art ECMWF Subseasonal-to-Seasonal (S2S) reforecasts using a conventional model, the FuXi-S2S forecasts demonstrate superior deterministic and ensemble forecasts for total precipitation (TP), outgoing longwave radiation (OLR), and geopotential at 500 hPa (Z500). Moreover, it demonstrates comparable performance in predicting global 2-meter temperature (T2M), with clear advantages in land areas. Regarding forecasts for extreme TP, FuXi-S2S outperforms ECMWF S2S globally. The improved performance of FuXi-S2S is primarily due to its superior capability of predicting the Madden–Julian Oscillation (MJO), a key source of subseasonal predictability and a significant driver of weather patterns around the world. FuXi-S2S successfully extends the skillful MJO prediction from 30 days to 36 days. Physical sciences/Mathematics and computing/Computational science Earth and environmental sciences/Climate sciences/Atmospheric science/Atmospheric dynamics subseasonal forecast machine learning deep learning FuXi MJO Full Text Additional Declarations There is NO Competing Interest. Supplementary Files FuXiSubseasonalSI.pdf Supplementary Information: A machine learning model that outperforms conventional global subseasonal forecast models Cite Share Download PDF Status: Published Journal Publication published 30 Jul, 2024 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-3776375","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Physical Sciences - Article","associatedPublications":[],"authors":[{"id":269430231,"identity":"371dac7e-4b80-4e21-8650-567fa7dfadfc","order_by":0,"name":"Hao Li","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA30lEQVRIiWNgGAWjYBACPmYeEGWRwM98gEECLHSAgBY2iBaJBMm2BGK1MEC1GBwjWgs778EPHxgk8oyP8Rje+LmDQY7vRgLj5wK8DuNLlpzBIFFsdozH2LL3DIOx5I0EZukZ+P1iIM3DIJG47X6PmQRvG0PihhsJUA/i1mL8G6RlcxuPmeTfNoZ6YrSYgW3ZwAZkAG1JMCBGiyXQL4kzjrEVW8u2SRjOPPOwWRqfFn7+M8Y3PjDYJPa3MW+8+bbNRp7vePLBz/i0gAHjPzgTFDWMDYQ0jIJRMApGwSggAABqyD4UF3VXegAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0001-6197-0674","institution":"Fudan University","correspondingAuthor":true,"prefix":"","firstName":"Hao","middleName":"","lastName":"Li","suffix":""},{"id":269430232,"identity":"8195f1ed-a9c6-4600-a2cb-7d322f18943e","order_by":1,"name":"Lei Chen","email":"","orcid":"","institution":"Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Lei","middleName":"","lastName":"Chen","suffix":""},{"id":269430233,"identity":"1b60225f-b0af-4d39-8c00-730d8e65df00","order_by":2,"name":"Xiaohui Zhong","email":"","orcid":"https://orcid.org/0000-0001-6706-2335","institution":"Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Xiaohui","middleName":"","lastName":"Zhong","suffix":""},{"id":269430234,"identity":"bb4f09ed-9f9e-431e-8590-d69baa369800","order_by":3,"name":"Jie Wu","email":"","orcid":"","institution":"National Climate Center","correspondingAuthor":false,"prefix":"","firstName":"Jie","middleName":"","lastName":"Wu","suffix":""},{"id":269430235,"identity":"08d1e056-3c91-449c-90b0-df320cf6431b","order_by":4,"name":"Deliang Chen","email":"","orcid":"https://orcid.org/0000-0003-0288-5618","institution":"University of Gothenburg","correspondingAuthor":false,"prefix":"","firstName":"Deliang","middleName":"","lastName":"Chen","suffix":""},{"id":269430236,"identity":"52ae09c9-b640-47ec-978a-b91d064f3664","order_by":5,"name":"Shang-Ping Xie","email":"","orcid":"https://orcid.org/0000-0002-3676-1325","institution":"Scripps Institution of Oceanography","correspondingAuthor":false,"prefix":"","firstName":"Shang-Ping","middleName":"","lastName":"Xie","suffix":""},{"id":269430237,"identity":"c09ee8ba-53ee-46ed-8f7b-7952ab88e716","order_by":6,"name":"Qingchen Chao","email":"","orcid":"","institution":"China Meteorological Administration","correspondingAuthor":false,"prefix":"","firstName":"Qingchen","middleName":"","lastName":"Chao","suffix":""},{"id":269430238,"identity":"102b10de-f04c-4687-8037-42b749cf7e0b","order_by":7,"name":"Chensen Lin","email":"","orcid":"","institution":"Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Chensen","middleName":"","lastName":"Lin","suffix":""},{"id":269430239,"identity":"81813e64-f64b-4abf-b98c-b3dc926c7dfc","order_by":8,"name":"Zixin Hu","email":"","orcid":"","institution":"Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Zixin","middleName":"","lastName":"Hu","suffix":""},{"id":269430240,"identity":"e4861d11-4571-4673-b744-18a437da4f40","order_by":9,"name":"Bo Lu","email":"","orcid":"","institution":"National Climate Center","correspondingAuthor":false,"prefix":"","firstName":"Bo","middleName":"","lastName":"Lu","suffix":""},{"id":269430241,"identity":"78c8ceaa-7824-4b19-8cf1-aee73f7c33b6","order_by":10,"name":"Yuan Qi","email":"","orcid":"","institution":"Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Yuan","middleName":"","lastName":"Qi","suffix":""}],"badges":[],"createdAt":"2023-12-19 10:35:41","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3776375/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3776375/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-024-50714-1","type":"published","date":"2024-07-30T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":61472831,"identity":"a2bc8b64-92c1-415d-a9cc-5e6acb24f76e","added_by":"auto","created_at":"2024-07-31 07:08:43","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2940909,"visible":true,"origin":"","legend":"Article File","description":"","filename":"FuXiSubseasonalmain.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3776375/v1_covered_cdc66c27-0005-4748-af05-f59e1c28962e.pdf"},{"id":51098287,"identity":"4abe4b7e-355a-4632-8392-ad890dca4b8d","added_by":"auto","created_at":"2024-02-14 05:10:13","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":5053535,"visible":true,"origin":"","legend":"Supplementary Information: A machine learning model that outperforms conventional global subseasonal forecast models","description":"","filename":"FuXiSubseasonalSI.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3776375/v1/15a730a6df432663e6b7cca6.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"A machine learning model that outperforms conventional global subseasonal forecast models","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"subseasonal forecast, machine learning, deep learning, FuXi, MJO","lastPublishedDoi":"10.21203/rs.3.rs-3776375/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3776375/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSkillful subseasonal forecasts beyond 2 weeks is critical to various sectors of society but pose a grand scientific challenge. Recently, machine learning based weather forecasting models have made remarkable advancements, outperforming the most successful numerical weather predictions (NWP) generated by the European Centre for Medium-Range Weather Forecasts (ECMWF). However, currently, no machine learning based subseasonal forecasting model surpasses conventional models. Here, we introduce FuXi Subseasonal-to-Seasonal (FuXi-S2S), a machine learning based subseasonal forecasting model that provides global daily mean forecasts for up to 42 days, covering 5 upper-air atmospheric variables at 13 pressure levels and 11 surface variables. FuXi-S2S integrates an enhanced FuXi base model with a perturbation module for flow-dependent perturbations in hidden features, which is a crucial for estimating uncertainty and enhancing subseasonal skill. The model is developed using 72 years of daily statistics from ECMWF ERA5 reanalysis data. When compared to the state-of-the-art ECMWF Subseasonal-to-Seasonal (S2S) reforecasts using a conventional model, the FuXi-S2S forecasts demonstrate superior deterministic and ensemble forecasts for total precipitation (TP), outgoing longwave radiation (OLR), and geopotential at 500 hPa (Z500). Moreover, it demonstrates comparable performance in predicting global 2-meter temperature (T2M), with clear advantages in land areas. Regarding forecasts for extreme TP, FuXi-S2S outperforms ECMWF S2S globally. The improved performance of FuXi-S2S is primarily due to its superior capability of predicting the Madden–Julian Oscillation (MJO), a key source of subseasonal predictability and a significant driver of weather patterns around the world. FuXi-S2S successfully extends the skillful MJO prediction from 30 days to 36 days.\u003c/p\u003e","manuscriptTitle":"A machine learning model that outperforms conventional global subseasonal forecast models","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-02-14 05:10:08","doi":"10.21203/rs.3.rs-3776375/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"cc3d6c55-8873-4342-8aba-79aa4fb8a74b","owner":[],"postedDate":"February 14th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":28392256,"name":"Physical sciences/Mathematics and computing/Computational science"},{"id":28392257,"name":"Earth and environmental sciences/Climate sciences/Atmospheric science/Atmospheric dynamics"}],"tags":[],"updatedAt":"2024-07-31T07:08:34+00:00","versionOfRecord":{"articleIdentity":"rs-3776375","link":"https://doi.org/10.1038/s41467-024-50714-1","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2024-07-30 04:00:00","publishedOnDateReadable":"July 30th, 2024"},"versionCreatedAt":"2024-02-14 05:10:08","video":"","vorDoi":"10.1038/s41467-024-50714-1","vorDoiUrl":"https://doi.org/10.1038/s41467-024-50714-1","workflowStages":[]},"version":"v1","identity":"rs-3776375","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3776375","identity":"rs-3776375","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}