Reconstruction of satellite chlorophyll images under heavy cloud coverage using a neural classification method

► This Work describes a method able to fill data gaps in satellite chlorophyll (CHL) images. ► It is based on the principle of Self Organizing Maps (SOM) classification methods. ► It relies on the assumption that an ocean state is locally defined by its values of SST, SSH and CHL. ► A codebook of possible (SST, SSH, CHL) situations, if large enough, allows the reconstruction of incomplete situations of CHL. ► The results proved its efficiency to reproduce CHL patterns in cases of extreme cloud cover (100%).