Assessment of bio-optical parameters retrieval in the coastal waters using local to global optimization

Coastal waters are optically complex waters and are dominated by chlorophyll-a (chl-a), inorganic suspended sediments, coloured dissolved organic matter, detritus, etc., which make retrieval of chlorophyll-a difficult in coastal waters. The analyses were carried out using NASA bio-optical marine algorithm (NOMAD) datasets specifically for coastal waters with bathymetry within 50 m. This study emphasizes the local-to-global optimization technique for simultaneous retrieval of chl-a and coloured dissolved organic matter (CDOM). The global optimization, i.e., Monte-Carlo optimization (MCO), is comparable to the local optimization technique, i.e., Levenberg Marquardt (LM), for retrieving chl-a in optically complex coastal waters where the conventional band ratio technique fails. Both the techniques LM and MCO give R2 = 0.7, slope = 0.9 and RMSEat log scale = 0.3 for retrieving chlorophyll-a. The statistical analyses for the combined response of coloured dissolved organic matter and non-algal particle (detritus) absorption coefficient, i.e., ag + ad (1/m) using LM, MCO and comparative studies with empirical approach were also discussed in this manuscript. The results suggest that the technique, whether local or global, converges to a global minimum to get the optimum solution with better accuracy.