Disaster monitoring in rivers using remote sensing: comparative analysis of turbidity and reflectance in Rio Doce

Abstract

The rupture of the Fundão dam in Mariana/MG, was one of the most shocking environmental disasters in the history of Brazil until then. Millions of mining tailings were dumped in the Doce River, causing the death of aquatic biodiversity and terrestrial fauna, as well as causing the loss of several people, the destruction of the municipality of Bento Rodrigues, and the destitution of many families. This event is not an isolated case and the detection and monitoring of the displacement of these wastes is extremely important since in many cases preventive actions are necessary to protect the integrity of populations and economic activities that depends on the water captured in the rivers, but this is costly in (or perform In Locu analysis) to detect the presence of tailings. Thus, this work proposes the development of a methodology for the monitoring of disasters arising from the disruption of mining tailings dams along the watercourses through remote sensing. The Turbidity data of the existing monitoring points along the Doce River, the Institute of Water Management (IGAM), and the Environmental Institute of Espirito Santo (IEMA) were used, as well as scenes from the LandSat 8, Sensor OLI. Atmospheric correction of selected images was performed. From the thirty-one points evaluated, all those with higher turbidity levels presented values ​​of reflectance lower than 1, which led to the conclusion that the higher the turbidity value, the lower the reflectance of the water analyzed after the addition of tailings, from mining of iron, in large volumes contrary to expectations.