Horizon 1 1 17 ACLN : Articles dans des revues avec comité de lecture non répertoriées par l'AERES IEEE Transactions on Geoscience and Remote Sensing 4212512 [12 ] MEDITERRANEE 2025 fdi:010095766 ENG IEEE Transactions on Geoscience and Remote Sensing 0196-2892 ISI:001626459000022 10.1109/tgrs.2025.3624465 https://www.documentation.ird.fr/hor/fdi:010095766 https://horizon.documentation.ird.fr/exl-doc/pleins_textes/2025-12/010095766.pdf 63 Horizon (IRD) 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. 030 ; 122 ; 126 ; 020