Abstract:
The use of ocean colour remote sensing to facilitate the monitoring of phytoplankton biomass in coastal waters is hampered by the high variability in absorption and scattering from substances other than phytoplankton. The eastern Arabian Sea coastal shelf is influenced by river run-off, winter convection and monsoon upwelling. Bio-optical parameters were measured along this coast from March 2009 to June 2011, to characterise the optical water type and validate three Chlorophyll-a (Chl-a) algorithms applied to Moderate Resolution Imaging Spectroradiometer on Aqua (MODIS-Aqua) data against in situ measurements. Ocean Colour 3 band ratio (OC3M), Garver-Siegel-Maritorena Model (GSM) and Generalized Inherent Optical Property (GIOP) Chl-a algorithms were evaluated. OC3M performed better than GSM and GIOP in all regions and overall, was within 11 percent of in situ Chl-a. GSM was within 24 percent of in situ Chl-a and GIOP on average was 55 percent lower. OC3M was less affected by errors in remote sensing reflectance Rrs(lambda) and by spectral variations in absorption coefficient (aCDOM(lambda)) of coloured dissolved organic material (CDOM) and total suspended matter (TSM) compared to the other algorithms. A nine year Chl-a time series from 2002 to 2011 was generated to assess regional differences between OC3M and GSM. This showed that in the north eastern shelf, maximum Chl-a occurred during the winter monsoon from December to February, where GSM consistently gave higher Chl-a compared to OC3M. In the south eastern shelf, maximum Chl-a occurred in June to July during the summer monsoon upwelling, and OC3M yielded higher Chl-a compared to GSM. OC3M currently provides the most accurate Chl-a estimates for the eastern Arabian Sea coastal waters.
