Spatiotemporal prediction of Soil Organic Carbon Density in Europe (2000–2022) using Earth Observation and Machine Learning

Spatiotemporal prediction of Soil Organic Carbon Density in Europe (2000–2022) using Earth Observation and Machine Learning | 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 Spatiotemporal prediction of Soil Organic Carbon Density in Europe (2000–2022) using Earth Observation and Machine Learning Xuemeng Tian, Sytze de Bruin, Rolf Simoes, Mustafa Serkan Isik, and 6 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5128244/v2 This work is licensed under a CC BY 4.0 License Status: Posted Version 2 posted You are reading this latest preprint version Show more versions Abstract The paper describes a comprehensive framework for soil organic carbon density (SOCD) (kg/m3) modeling and mapping, based on spatiotemporal Random Forest (RF) and Quantile Regression Forests (QRF). A total of 45,616 SOCD observations and various Earth Observation (EO) feature layers were used to produce 30m SOCD maps for the EU at four-year intervals (2000--2022) and four soil depth intervals (0-20cm, 20-50cm, 50-100cm, and 100-200cm). Per-pixel 95% probability prediction intervals (PIs) and extrapolation risk probabilities are also provided. Model evaluation indicates good overall accuracy (R2 = 0.63 and CCC = 0.76 for hold-out independent tests). Prediction accuracy varies by land cover, depth interval and year of prediction with accuracy the worst for shrubland and deeper soils 100--200cm. PI validation confirmed effective uncertainty estimation, though with reduced accuracy for higher SOCD values. Shapley analysis identified soil depth as the most influential feature, followed by vegetation, long-term bioclimate, and topographic features. While pixel-level uncertainty is substantial, spatial aggregation reduces uncertainty by approximately 66%. Detecting SOCD changes remains challenging but offers a baseline for future improvements. Maps, based primarily on topsoil data from cropland, grassland, and woodland, are best suited for applications related to these land covers and depths. We recommend that users interpret the maps in conjunction with local knowledge and consider the accompanying uncertainty and extrapolation risk layers. All data and code are available under an open license at https://doi.org/10.5281/zenodo.13754343 and https://github.com/AI4SoilHealth/SoilHealthDataCube/ . soil organic carbon machine learning uncertainty trend analysis spatial aggregation model interpretation Full Text Additional Declarations The authors declare potential competing interests as follows: The authors declare that they have no competing interests. Xuemeng Tian, Rolf Simoes, Mustafa Serkan Isik, Robert Minarik, Yu-Feng Ho, Davide Consoli and Tomislav Hengl are employed by OpenGeoHub. Cite Share Download PDF Status: Posted Version 2 posted You are reading this latest preprint version Show more versions 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-5128244","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":358214754,"identity":"8cd44084-d266-451d-9cb1-656031d9ed83","order_by":0,"name":"Xuemeng Tian","email":"data:image/png;base64,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","orcid":"","institution":"OpenGeoHub, Wageningen University \u0026 Research","correspondingAuthor":true,"prefix":"","firstName":"Xuemeng","middleName":"","lastName":"Tian","suffix":""},{"id":358214755,"identity":"36dc36e4-a843-4f88-858f-c77b547bd006","order_by":1,"name":"Sytze de Bruin","email":"","orcid":"","institution":"Wageningen University \u0026 Research","correspondingAuthor":false,"prefix":"","firstName":"Sytze","middleName":"","lastName":"de Bruin","suffix":""},{"id":358214756,"identity":"37170c71-2a9a-4dd9-adab-a93d8a2eec98","order_by":2,"name":"Rolf Simoes","email":"","orcid":"","institution":"OpenGeoHub","correspondingAuthor":false,"prefix":"","firstName":"Rolf","middleName":"","lastName":"Simoes","suffix":""},{"id":358214757,"identity":"a9400715-474b-4455-9e55-fbe589000868","order_by":3,"name":"Mustafa Serkan Isik","email":"","orcid":"","institution":"OpenGeoHub","correspondingAuthor":false,"prefix":"","firstName":"Mustafa","middleName":"Serkan","lastName":"Isik","suffix":""},{"id":358214758,"identity":"686d7a00-3dca-4901-9421-0810a13b8867","order_by":4,"name":"Robert Minarik","email":"","orcid":"","institution":"OpenGeoHub","correspondingAuthor":false,"prefix":"","firstName":"Robert","middleName":"","lastName":"Minarik","suffix":""},{"id":358214759,"identity":"16785f37-3f10-4e03-bbcf-8c47249d7886","order_by":5,"name":"Yu-Feng Ho","email":"","orcid":"","institution":"OpenGeoHub","correspondingAuthor":false,"prefix":"","firstName":"Yu-Feng","middleName":"","lastName":"Ho","suffix":""},{"id":358214760,"identity":"287624e9-1377-484c-b06d-20b9a0d82224","order_by":6,"name":"Murat Şahin","email":"","orcid":"","institution":"Delft University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Murat","middleName":"","lastName":"Şahin","suffix":""},{"id":358214761,"identity":"d79834d1-f99f-4f8d-a199-19f948f42bab","order_by":7,"name":"Martin Herold","email":"","orcid":"","institution":"Helmholtz GFZ German Research Centre for Geosciences","correspondingAuthor":false,"prefix":"","firstName":"Martin","middleName":"","lastName":"Herold","suffix":""},{"id":358214762,"identity":"7b3dde3a-626e-43d4-b7e7-a937016f584e","order_by":8,"name":"Davide Consoli","email":"","orcid":"","institution":"OpenGeoHub","correspondingAuthor":false,"prefix":"","firstName":"Davide","middleName":"","lastName":"Consoli","suffix":""},{"id":358214763,"identity":"98d668c8-b852-457b-9033-398d1c733f84","order_by":9,"name":"Tomislav Hengl","email":"","orcid":"","institution":"OpenGeoHub","correspondingAuthor":false,"prefix":"","firstName":"Tomislav","middleName":"","lastName":"Hengl","suffix":""}],"badges":[],"createdAt":"2024-09-21 10:48:22","currentVersionCode":2,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":true,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-5128244/v2","doiUrl":"https://doi.org/10.21203/rs.3.rs-5128244/v2","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":81773719,"identity":"d3ef7077-57c8-4a8e-b3c6-1bace3da4e19","added_by":"auto","created_at":"2025-05-01 15:17:40","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":22149189,"visible":true,"origin":"","legend":"","description":"","filename":"EUspatialtemporalSOCmodellinganditsuncertainty.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5128244/v2_covered_9608a539-1d5b-4791-95df-d456d8403d02.pdf"}],"financialInterests":"The authors declare potential competing interests as follows: The authors declare that they have no competing interests. Xuemeng Tian, Rolf Simoes, Mustafa Serkan Isik, Robert Minarik, Yu-Feng Ho, Davide Consoli and Tomislav Hengl are employed by OpenGeoHub.","formattedTitle":"\u003cp\u003eSpatiotemporal prediction of Soil Organic Carbon Density in Europe (2000–2022) using Earth Observation and Machine Learning\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"OpenGeoHub","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"soil organic carbon, machine learning, uncertainty, trend analysis, spatial aggregation, model interpretation","lastPublishedDoi":"10.21203/rs.3.rs-5128244/v2","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5128244/v2","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe paper describes a comprehensive framework for soil organic carbon density (SOCD) (kg/m3) modeling and mapping, based on spatiotemporal Random Forest (RF) and Quantile Regression Forests (QRF). A total of 45,616 SOCD observations and various Earth Observation (EO) feature layers were used to produce 30m SOCD maps for the EU at four-year intervals (2000--2022) and four soil depth intervals (0-20cm, 20-50cm, 50-100cm, and 100-200cm). Per-pixel 95% probability prediction intervals (PIs) and extrapolation risk probabilities are also provided. Model evaluation indicates good overall accuracy (R2 = 0.63 and CCC = 0.76 for hold-out independent tests). Prediction accuracy varies by land cover, depth interval and year of prediction with accuracy the worst for shrubland and deeper soils 100--200cm. PI validation confirmed effective uncertainty estimation, though with reduced accuracy for higher SOCD values. Shapley analysis identified soil depth as the most influential feature, followed by vegetation, long-term bioclimate, and topographic features. While pixel-level uncertainty is substantial, spatial aggregation reduces uncertainty by approximately 66%. Detecting SOCD changes remains challenging but offers a baseline for future improvements. Maps, based primarily on topsoil data from cropland, grassland, and woodland, are best suited for applications related to these land covers and depths. We recommend that users interpret the maps in conjunction with local knowledge and consider the accompanying uncertainty and extrapolation risk layers. All data and code are available under an open license at \u003ca href=\"https://doi.org/10.5281/zenodo.13754343\"\u003ehttps://doi.org/10.5281/zenodo.13754343\u003c/a\u003e and \u003ca href=\"https://github.com/AI4SoilHealth/SoilHealthDataCube/\"\u003ehttps://github.com/AI4SoilHealth/SoilHealthDataCube/\u003c/a\u003e.\u003c/p\u003e","manuscriptTitle":"Spatiotemporal prediction of Soil Organic Carbon Density in Europe (2000–2022) using Earth Observation and Machine Learning","msid":"","msnumber":"","nonDraftVersions":[{"code":2,"date":"2025-05-01 15:01:21","doi":"10.21203/rs.3.rs-5128244/v2","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}},{"code":1,"date":"2024-09-24 03:12:37","doi":"10.21203/rs.3.rs-5128244/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"8eb3beb9-55d0-4de1-b1b0-012a2d611921","owner":[],"postedDate":"May 1st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-09-24T03:12:37+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-01 15:01:21","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v2","identity":"rs-5128244","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5128244","identity":"rs-5128244","version":["v2"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}