The satellite reversion of dissolved organic carbon (DOC) based on the analysis of the mixing behavior of DOC and colored dissolved organic matter: the East China Sea as an example

The retrieval of dissolved organic carbon (DOC) distribution by remote sensing is mainly based on the empirical relationship of DOC concentration and colored dissolved organic matter (CDOM) concentration in many literatures. To investigate the nature of this relationship, the distributions and mixing behaviors of DOC and CDOM are reviewed in the world's major estuaries and bays. It is found that, generally, the CDOM concentration is well correlated with the salinity in most estuaries, while DOC usually shows a nonconservative behavior which leads to a weak correlation between the DOC concentration and the CDOM concentration. To establish a good satellite reversion of the DOC concentration, the East China Sea(ECS) was taken as an example, and the mixing behavior of DOC and CDOM as well as the influence of biogeochemical processes were analyzed except for the physical mixing process with the data from late autumn (November, 2010) and winter (December, 2009) cruises. In the two ECS cruises, the CDOM concentration was found to be tightly correlated with the salinity, influenced little by the photochemical or biological processes. The data from the winter cruise show that DOC followed a conservative mixing along the salinity gradient, while in the late autumn cruise it was significantly affected by the biological activities, resulting in a poor correlation between the DOC and the CDOM. Accordingly, an improvedDOC algorithm(CSDM) was proposed: when the biological influence was significant (Chl a greater than 0.8 A mu g/dm(3)), DOC was retrieved by the conservative and biologicalmodel, and if the conservativemixing was dominant (Chl a less than 0.8 A mu g/dm(3)), the direct DOC concentration and CDOM concentration relationship was used. Based on the proposed algorithm, a reasonable DOC distribution for the ECS from satellite was obtained in this study, and the proposedmethod can be applied to the other large river-dominantmarginal sea.