Indirect assimilation of remote sensing reflectance: case study in the Liguria Sea

The dataset with the most spatial coverage for data assimilation of biogeochemical models in operational systems is the satellite-derived data. Nevertheless, variables derived from Remote Sensing Reflectance (RSR), like the sea surface chlorophyll concentration, for regions like coastal areas, can reach big errors if compared with in situ measurements. For this reason, a suggestion with the aim of improving the assimilated results comes from the direct assimilation of Remote Sensing Reflectance, removing the error derived from inferring the biogeochemical variable before assimilating. In this work, we focus on a case study, using the Biogeochemical Flux Model (BFM) merged with a hydrological model, we study the effects of the direct and indirect assimilation of RSR in a region located in the Ligurian Basin of the northwestern Mediterranean Sea.  For both assimilation experiments, the algorithm used was an Error Subspace Kalman Filter. To assess the results, we compared them with climatologies computed with in situ measurements, highlighting the advantages and disadvantages of both approaches.