Dinâmica da sedimentação através de derivadores e imagens de sensoriamento remoto de alta resolução: implicações para a recuperação do Rio Doce

Abstract

In terms of environmental impact, the worst mining tailings dam collapse accident that has occurred in Brazil in recent decades is known as the Mariana disaster, which significantly affected the main course of the Doce River. The Doce River originates at the confluence of the Carmo River, directly affected by the accident, and the Piranga River, indirectly affected. Studying confluences allows us to understand the role of tributaries in restoring rivers affected by this type of disaster. Monitoring and understanding the spatiotemporal behavior at the confluence is essential, since important processes of mixing, transport and deposition of sediments occur there. In this study, the hydrodynamics and physical parameters of the flow were observed using satellite-tracked drifters to find the most frequent areas for sediment deposition, called lateral and central bars. Geoprocessing tools and spectral indices, NDWI and BSI, were applied to satellite images and ARPs before and after the event, to analyze these bars over time. The central bar was a site of great sedimentation for the Mariana mud and for the flood that occurred in 2022, while the lateral bar showed varied evolution, with less development by the mud residues and greater growth with the 2022 flood. Through the images, analyses of water spectral signatures of the studied rivers were made, which showed the great impact of the disaster on the Carmo River and its recovery in the following years. While the Piranga River showed an increase in the presence of sediments mainly in the flood that occurred in 2022. The results helped to understand the influence of Carmo and Piranga rivers on the hydrodynamic processes that occurred at the confluence, and the location and evolution of important sedimentation zones, producing information that can be useful for recovery projects of this and other rivers and degraded areas.