A species-dependent bio-optical model of case I waters for global ocean color processing

The PHYSAT method, which enables identification of four different phytoplankton groups from their impact on the normalized water-leaving radiance (nLw) spectra, is applied to coincident in situ measurements of both chlorophyll a concentration (Chl a) and nLw. Observations show that measurements acquired in waters dominated by haptophytes, diatoms and Synechococcus-like cyanobacteria have optical properties that deviate significantly from the mean OC4V4 bio-optical model, which is currently used as the standard for global ocean color processing. A specific OC4v4-like relationship, i.e., Chl a as a 4th-order polynomial of the “maximum band ratio”, was fitted for each dominant phytoplankton group identified by PHYSAT. The resulting OC4-SD bio-optical model thus starts from a classification of the dominant phytoplankton group, followed by a species-dependent estimate of Chl a. It has been applied to global daily SEAWIFS data of the year 2001. Monthly mean maps of Chl a derived from OC4-SD or OC4v4 show large regional differences that can reach 50% at high latitudes. The new algorithm leads to lower concentrations in regions where the standard model retrievals are known to be too high, such as the Mediterranean Sea. Conversely, higher concentrations are retrieved in regions dominated by diatom blooms, such as the northern North Atlantic in summer, where previous studies have demonstrated a low bias in standard SEAWIFS Chl a.