Zhu, SongyanClement, RobertMcCalmont, JonDavies, Christian A.Hill, Timothy2022-12-272022-12-272022-03-01Zhu, S, Clement, R, McCalmont, J, Davies, C A & Hill, T 2022, 'Stable gap-filling for longer eddy covariance data gaps : A globally validated machine-learning approach for carbon dioxide, water, and energy fluxes', Agricultural and Forest Meteorology, vol. 314, 108777. https://doi.org/10.1016/j.agrformet.2021.1087770168-1923RIS: urn:522F3033020A19B1F03AA9B9E8CC5B31ORCID: /0000-0002-5978-9574/work/107063447https://hdl.handle.net/2164/19784Acknowledgments The authors thank the FLUXNET and the research groups for providing the CC-BY-4.0 (Tier one) open-access eddy covariance data (https://fluxnet.org/login/?redirect_to=/data/download-data/). They also thank the ReddyProc (https://cran.r-project.org/web/packages/REddyProc/index.html) team and scikit-learn (https://scikit-learn.org/stable/install.html) team for the packages that help the implementation and validation for gap-filling approaches. Songyan Zhu would like to acknowledge a Shell funded PhD studentship and Timonthy Hill acknowledge funding from a joint UK NERC-FAPESP grant no. NE/S000011/1 & FAPESP-19/07773-1.101511489engSDG 13 - Climate ActionSDG 15 - Life on LandGlobal land ecosystemsCarbon exchangeEddy covarianceLong gapsRobust gap-fillingQH301 BiologyNatural Environment Research Council (NERC)NE/S000011/119/07773-1.Supplementary InformationQH301Journal article10.1016/j.agrformet.2021.108777314