MODELING CLOROPHYLL-A CONCENTRATION BASED ON SIMULATED OLCI/SENTINEL-3 AND MSI/SENTINEL-2 BANDS IN A TROPICAL WATER RESERVOIR

Remote sensing is an essential tool for mapping the dy-namics of the optical components in tropical water reservoirs, allowing the identification of eutrophic conditions and their influences. This study is based on in situ data collected at the Itaipu reservoir, located on the Brazil-Paraguay border, to analyze limnological data of chlorophyll-a, employing remote sensing techniques, calibration, and validation of simulated bio-optical algorithms for the OLCI/Sentinel-3 and MSI/Sentinel-2 sensors. The empirical algorithm calibration and validation were based on Monte Carlo’s simulation. The results showed that the second-order derivative effectively revealed the spectral features of chl-a presence in the water. Furthermore, the best model included the slope between the 673 and 681 simulated bands (R² ~0.64 and MAPE of ~35%) with a polynomial adjustment of degree two. The narrower spectral band of the OLCI sensor, compared to that of the MSI sensor, allowed a greater sensitivity to phyto-plankton features, even at the low chl-a concentrations.