Dorffer C. auteur aut IRD Jourdin F. auteur aut IRD Nguyen T.T.N. auteur aut IRD Devillers Rodolphe auteur aut IRD Mouillot D. auteur aut IRD Fablet R. auteur aut IRD text journalArticle eng text/pdf reformatted digital access Monitoring optical properties of coastal and open ocean waters is crucial to assessing the health of marine ecosystems. Deep learning offers a promising approach to address these ecosystem dynamics, especially in scenarios where gapfree ground-truth data is lacking, which poses a challenge for designing effective training frameworks. Using an advanced neural variational data assimilation scheme (called 4DVarNet), we introduce a comprehensive training framework designed to effectively train directly on gappy data sets. Using the Mediterranean Sea as a case study, our experiments not only highlight the high performance of the chosen neural network in reconstructing gap-free images from gappy datasets but also demonstrate its superior performance over state-of-the-art algorithms such as DInEOF and end-to-end neural mapping schemes based CNN or UNet architectures. specialized MEDITERRANEE 030 122 126 020 63 4212512 [12 ] 2025 0196-2892 https://www.documentation.ird.fr/hor/fdi:010095766 10.1109/tgrs.2025.3624465 0196-2892 [F B010095766] https://www.documentation.ird.fr/hor/fdi:010095766 https://horizon.documentation.ird.fr/exl-doc/pleins_textes/2025-12/010095766.pdf IRD - Base Horizon / Pleins textes 2025-12-08 2025-12-17 fdi:010095766 fre