An assessment of a semi-analytical model in estimate the Secchi disk depth in Ibitinga Reservoir using OLI/Landsat-8 images

Abstract

Understanding the optical water quality in reservoirs designed in a cascade system is of great importance for water resources management, since it is intrinsically related with environmental forcings, land use and land cover and biogeochemical composition of water column constituents. Overall, in a cascade system, the upstream reservoirs are more eutrophic than those further downstream as a lateral, vertical and longitudinal modifications in the watershed. The water transparency or clarity is an important indicator of water quality and can be measured through the Secchi disk depth Z_SD. The traditional method of water quality mapping using in situ data are expensive and with low spatial and temporal representativity, therefore, it is very important that a more robust system to be implemented. The water quality management based on remote sensing aim to estimate chlorophyll-a (chl-a), suspended solids, as well as Z_SD and nutrients. The chl-a is a proxy to derive biomass production in a waterbody and together with total phosphorus and Z_SD, the determination of trophic state of water is quite possible. Many efforts have been done to map Z_SD using empirical approaches, however, we know that this kind of model is site-specific and depend on biogeochemical condition. Therefore, the semi-analytical models arise to solve this limitation and give more possibilities to monitor water quality systematically in the cascade reservoirs of Tietê river. The model from Lee et al. (2015) built to retrieve Z_SD was chosen because it was developed based on several water bodies including lakes. The model will be applied in a temporal series of OLI/Landsat-8 images. As result, we expect to map water quality based on water transparency so we can raise the main factors leading to the eutrophication process in Tietê river during the studied time period.