MedFormer: a data-driven model for forecasting the Mediterranean Sea

MedFormer: a data-driven model for forecasting the Mediterranean Sea | 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 Research Article MedFormer: a data-driven model for forecasting the Mediterranean Sea Italo Epicoco, Davide Donno, Gabriele Accarino, Simone Norberti, and 12 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7899254/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 13 You are reading this latest preprint version Abstract Accurate ocean forecasting is essential for supporting a wide range of marine applications. Recent advances in artificial intelligence have highlighted the potential of data-driven models to outperform traditional numerical approaches, particularly in atmospheric weather forecasting. However, extending these methods to ocean systems remains challenging due to their inherently slower dynamics and complex boundary conditions. In this work, we present MedFormer, a fully data-driven deep learning model specifically designed for medium-range ocean forecasting in the Mediterranean Sea. MedFormer is based on a U-Net architecture augmented with 3D attention mechanisms and operates at a high horizontal resolution of 1/24°. The model is trained on 20 years of daily ocean reanalysis data and fine-tuned with high-resolution operational analyses. It generates 9-day forecasts using an autoregressive strategy. The model leverages both historical ocean states and atmospheric forcings, making it well-suited for operational use. We benchmark MedFormer against the state-of-the-art Mediterranean Forecasting System (MedFS), developed at Euro-Mediterranean Center on Climate Change (CMCC), using both analysis data and independent observations. The forecast skills, evaluated with the Root Mean Squared Difference and the Anomaly Correlation Coefficient, indicate that MedFormer consistently outperforms MedFS across key 3D ocean variables. These findings underscore the potential of data-driven approaches like MedFormer to complement, or even surpass, traditional numerical ocean forecasting systems in both accuracy and computational efficiency. data-driven ocean model machine learning U-Net 3D attention mechanism Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 26 Mar, 2026 Reviews received at journal 20 Mar, 2026 Reviews received at journal 19 Mar, 2026 Reviews received at journal 16 Mar, 2026 Reviewers agreed at journal 26 Feb, 2026 Reviewers agreed at journal 24 Feb, 2026 Reviewers agreed at journal 23 Feb, 2026 Reviews received at journal 27 Nov, 2025 Reviewers agreed at journal 04 Nov, 2025 Reviewers invited by journal 30 Oct, 2025 Editor assigned by journal 30 Oct, 2025 Submission checks completed at journal 20 Oct, 2025 First submitted to journal 19 Oct, 2025 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-7899254","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":542215417,"identity":"15d01505-e335-4be8-b79a-f17d3f5ece53","order_by":0,"name":"Italo Epicoco","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA1UlEQVRIiWNgGAWjYBACNgYGxgNQBhBUQEQP4NXCBlUA0XIGqgW/HmQzGdsYCFvDJ9/84MAHBps8PvbDBz98nGcXzd9+gPHwB7wOYzM4OIMhrZiNJy1Zcua25NwZZxLwOwzoF4PDPAyHE9skeMyYebcdyN0gQcAvbGzsH4Ba/kO1zCFKCw/IlgNQLQ1EackpODjDIBnilxnHQH5JbDhwBo8W+ebjGx98qLDLk28HhtiHGrvc/vbDhz9U4NECAQYMCUg8xgaCGkAggaCKUTAKRsEoGLkAANCGTCnrEhM1AAAAAElFTkSuQmCC","orcid":"","institution":"University of Salento","correspondingAuthor":true,"prefix":"","firstName":"Italo","middleName":"","lastName":"Epicoco","suffix":""},{"id":542215419,"identity":"457969a7-3a68-4172-8806-cdfd06fba196","order_by":1,"name":"Davide Donno","email":"","orcid":"","institution":"University of Salento","correspondingAuthor":false,"prefix":"","firstName":"Davide","middleName":"","lastName":"Donno","suffix":""},{"id":542215421,"identity":"9a58e255-eeb8-4974-97d1-cba1d680fa1a","order_by":2,"name":"Gabriele Accarino","email":"","orcid":"","institution":"Columbia University","correspondingAuthor":false,"prefix":"","firstName":"Gabriele","middleName":"","lastName":"Accarino","suffix":""},{"id":542215422,"identity":"0eb3f776-75e6-42bc-9c1a-3b4d7a570881","order_by":3,"name":"Simone Norberti","email":"","orcid":"","institution":"CMCC Foundation - Euro-Mediterranean Center on Climate Change","correspondingAuthor":false,"prefix":"","firstName":"Simone","middleName":"","lastName":"Norberti","suffix":""},{"id":542215423,"identity":"1eb3f350-27f2-4389-8569-67dc38f9ebf5","order_by":4,"name":"Alessandro Grandi","email":"","orcid":"","institution":"CMCC Foundation - Euro-Mediterranean Center on Climate Change","correspondingAuthor":false,"prefix":"","firstName":"Alessandro","middleName":"","lastName":"Grandi","suffix":""},{"id":542215427,"identity":"aa533074-df15-42e7-8b31-3753c866ea74","order_by":5,"name":"Ronan McAdam","email":"","orcid":"","institution":"CMCC Foundation - Euro-Mediterranean Center on Climate Change","correspondingAuthor":false,"prefix":"","firstName":"Ronan","middleName":"","lastName":"McAdam","suffix":""},{"id":542215428,"identity":"bd2ee193-8b86-44bb-a09a-3a8bb7d8bd5a","order_by":6,"name":"Donatello Elia","email":"","orcid":"","institution":"CMCC Foundation - Euro-Mediterranean Center on Climate Change","correspondingAuthor":false,"prefix":"","firstName":"Donatello","middleName":"","lastName":"Elia","suffix":""},{"id":542215429,"identity":"70b62fd4-439a-44fe-91f8-9990d1d51dee","order_by":7,"name":"Emanuela Clementi","email":"","orcid":"","institution":"CMCC Foundation - Euro-Mediterranean Center on Climate Change","correspondingAuthor":false,"prefix":"","firstName":"Emanuela","middleName":"","lastName":"Clementi","suffix":""},{"id":542215431,"identity":"d6a58e5d-a3ee-452f-9a54-4ed2c9fde205","order_by":8,"name":"Paola Nassisi","email":"","orcid":"","institution":"CMCC Foundation - Euro-Mediterranean Center on Climate Change","correspondingAuthor":false,"prefix":"","firstName":"Paola","middleName":"","lastName":"Nassisi","suffix":""},{"id":542215433,"identity":"50d2e5cf-752b-4fd1-b4ab-d0b81dfbe5f0","order_by":9,"name":"Enrico Scoccimarro","email":"","orcid":"","institution":"CMCC Foundation - Euro-Mediterranean Center on Climate Change","correspondingAuthor":false,"prefix":"","firstName":"Enrico","middleName":"","lastName":"Scoccimarro","suffix":""},{"id":542215434,"identity":"99f7356a-ed80-485c-ba0c-070ad2e204f6","order_by":10,"name":"Giovanni Coppini","email":"","orcid":"","institution":"CMCC Foundation - Euro-Mediterranean Center on Climate Change","correspondingAuthor":false,"prefix":"","firstName":"Giovanni","middleName":"","lastName":"Coppini","suffix":""},{"id":542215436,"identity":"6e9f78e3-4c36-49c2-9a33-a0b7b0d8203a","order_by":11,"name":"Silvio Gualdi","email":"","orcid":"","institution":"CMCC Foundation - Euro-Mediterranean Center on Climate Change","correspondingAuthor":false,"prefix":"","firstName":"Silvio","middleName":"","lastName":"Gualdi","suffix":""},{"id":542215437,"identity":"c93e099a-aad6-40e5-872f-f145a484762d","order_by":12,"name":"Giovanni Aloisio","email":"","orcid":"","institution":"CMCC Foundation - Euro-Mediterranean Center on Climate Change","correspondingAuthor":false,"prefix":"","firstName":"Giovanni","middleName":"","lastName":"Aloisio","suffix":""},{"id":542215438,"identity":"3935ef88-f136-4d85-913a-224d8054de1b","order_by":13,"name":"Simona Masina","email":"","orcid":"","institution":"CMCC Foundation - Euro-Mediterranean Center on Climate Change","correspondingAuthor":false,"prefix":"","firstName":"Simona","middleName":"","lastName":"Masina","suffix":""},{"id":542215439,"identity":"b69e17f1-02be-44a7-930c-4bfc24b50c18","order_by":14,"name":"Giulio Boccaletti","email":"","orcid":"","institution":"CMCC Foundation - Euro-Mediterranean Center on Climate Change","correspondingAuthor":false,"prefix":"","firstName":"Giulio","middleName":"","lastName":"Boccaletti","suffix":""},{"id":542215441,"identity":"fb69a483-7955-4520-be1a-2f103592d25c","order_by":15,"name":"Antonio Navarra","email":"","orcid":"","institution":"CMCC Foundation - Euro-Mediterranean Center on Climate Change","correspondingAuthor":false,"prefix":"","firstName":"Antonio","middleName":"","lastName":"Navarra","suffix":""}],"badges":[],"createdAt":"2025-10-19 14:38:24","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7899254/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7899254/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":95517475,"identity":"0ac4fcbd-98eb-4522-9fa3-e8fd9b054f7b","added_by":"auto","created_at":"2025-11-10 08:45:13","extension":"json","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":15570,"visible":true,"origin":"","legend":"","description":"","filename":"5a8c0f35eca0464aab77efb038736e60.json","url":"https://assets-eu.researchsquare.com/files/rs-7899254/v1/db3e7e8d7fe4658f9c0745b4.json"},{"id":95654889,"identity":"a9ff477a-70de-4a14-bf69-4141f7cb0d68","added_by":"auto","created_at":"2025-11-11 16:13:41","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":21229973,"visible":true,"origin":"","legend":"","description":"","filename":"MedFormer.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7899254/v1_covered_c3b36b77-1aec-4ac2-b9a7-eefa6682e64c.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"MedFormer: a data-driven model for forecasting the Mediterranean Sea","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"earth-science-informatics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"esin","sideBox":"Learn more about [Earth Science Informatics](http://link.springer.com/journal/12145)","snPcode":"12145","submissionUrl":"https://submission.nature.com/new-submission/12145/3","title":"Earth Science Informatics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"data-driven ocean model, machine learning, U-Net, 3D attention mechanism","lastPublishedDoi":"10.21203/rs.3.rs-7899254/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7899254/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Accurate ocean forecasting is essential for supporting a wide range of marine applications. Recent advances in artificial intelligence have highlighted the potential of data-driven models to outperform traditional numerical approaches, particularly in atmospheric weather forecasting. However, extending these methods to ocean systems remains challenging due to their inherently slower dynamics and complex boundary conditions.\nIn this work, we present MedFormer, a fully data-driven deep learning model specifically designed for medium-range ocean forecasting in the Mediterranean Sea. MedFormer is based on a U-Net architecture augmented with 3D attention mechanisms and operates at a high horizontal resolution of 1/24°. The model is trained on 20 years of daily ocean reanalysis data and fine-tuned with high-resolution operational analyses. It generates 9-day forecasts using an autoregressive strategy. The model leverages both historical ocean states and atmospheric forcings, making it well-suited for operational use.\nWe benchmark MedFormer against the state-of-the-art Mediterranean Forecasting System (MedFS), developed at Euro-Mediterranean Center on Climate Change (CMCC), using both analysis data and independent observations. The forecast skills, evaluated with the Root Mean Squared Difference and the Anomaly Correlation Coefficient, indicate that MedFormer consistently outperforms MedFS across key 3D ocean variables. These findings underscore the potential of data-driven approaches like MedFormer to complement, or even surpass, traditional numerical ocean forecasting systems in both accuracy and computational efficiency.","manuscriptTitle":"MedFormer: a data-driven model for forecasting the Mediterranean Sea","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-10 08:45:08","doi":"10.21203/rs.3.rs-7899254/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-03-27T01:22:54+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-20T17:38:19+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-19T17:06:04+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-16T15:21:06+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"228533965772413530646659147019020146911","date":"2026-02-26T19:59:41+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"46310262778166244666258891140338342345","date":"2026-02-24T10:06:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"147348972401220326502213275934599113056","date":"2026-02-23T18:54:29+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-11-27T07:33:15+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"72626426386266413261587201127816471813","date":"2025-11-05T00:44:44+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-10-30T15:27:55+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-10-30T15:25:51+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-10-21T00:59:58+00:00","index":"","fulltext":""},{"type":"submitted","content":"Earth Science Informatics","date":"2025-10-19T14:35:04+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"earth-science-informatics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"esin","sideBox":"Learn more about [Earth Science Informatics](http://link.springer.com/journal/12145)","snPcode":"12145","submissionUrl":"https://submission.nature.com/new-submission/12145/3","title":"Earth Science Informatics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"f8ee2299-d803-42fb-b75f-21f2d21005ac","owner":[],"postedDate":"November 10th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-04T18:23:23+00:00","versionOfRecord":[],"versionCreatedAt":"2025-11-10 08:45:08","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7899254","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7899254","identity":"rs-7899254","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}