Aplicação de sensoriamento remoto como ferramenta de monitoramento de parâmetros de qualidade da água em corpos hídricos.

Abstract

Water resources management aims to guarantee multiple uses of water, even when there is a diversification of uses and always with a better quality as possible. In Brazil, the water monitoring of lakes and rivers usually are not continuous and mostly are not enough to guarantee a good and safe human supply. This study aims to analyze some water quality parameters (turbidity, chlorophyll a and depth of the Secchi Disk) in the "Doce River", using Landsat 8 satellite images available by Google Earth Engine platform. The analysis starts before, during and after the disruption of Fundão dam (Mariana environmental disaster), considered the most environmental disaster of the Brazilian history. Observed data (from government agencies) are compared to combinations of satellite bands. The results showed the effectiveness of remote sensing as a tool for monitoring visible water quality parameters, highlighting drastic changes in the period of the disaster, such as turbidity increase in 445% and 29% in the loss of river transparency. In addition to the analysis for the period, the trajectory of the place (pixel) where each station is located was analyzed using all the images between the years of 2013 and 2018, demonstrating a consistent way of detecting changes in the turbidity pattern of the water body.