A semi-analytic seasonal algorithm to retrieve chlorophyll-a concentration in the Northwest Atlantic Ocean from SeaWiFS data

In a companion paper [Fuentes-Yaco et al., Indian J. Mar. Sci., 34(2005),.341-355], it was demonstrated that the SeaWiFS OC4 algorithm, applied to the Northwest Atlantic, resulted in a systematic bias in the retrieved chlorophyll-a concentration. Their comparison of satellite-derived chlorophyll-a values with matching in situ observations showed that the OC4 algorithm as implemented in the NASA SeaDAS software package, overestimated chlorophyll-a in waters with low pigment concentrations, and underestimated chlorophyll-a for high pigment concentrations. In this paper, we seek an explanation for the observed bias, using a semi-analytic model of ocean colour [Sathyendranath et al., Int. J. Remote Sens, 22(2001), 249-273] modified to account for seasonal and regional variations in the spectral absorption properties of phytoplankton, dissolved matter (yellow substances) and detritus. The model is also extended into the near infrared region to evaluate the possible impact on the atmospheric correction algorithm. The results indicate that much of the bias can be explained by local variations in the inherent optical properties of particulate and dissolved matter present in the region. The algorithm based on the semi-analytical model eliminates practically all the bias (inaccuracy) in the retrieved chlorophyll-a concentrations, but does not improve the precision of retrieval.