The Doce River, located in Southeast Brazil, is of great importance in supplying water for urban use, agriculture, fishing and other purposes. It was affected by a disaster that occurred on November 5, 2015 due to the collapse of the Fundão ore tailing dam, in the municipality of Mariana (MG). This disaster generated socio-environmental consequences, as some regions were severely affected due to the waste released in the river, which was considered potentially toxic in the short to long term. This work presents a temporal analysis of images from the Landsat 8 satellite, associating the aspects of rainfall in dry and rainy months, in years before and after the disaster (2014, 2015, 2016, and 2020), together with the Normalized Difference Water Index (NDWI) obtained by image digital processing, to analyze the relation between dry and rainy periods and the spectral response of images. This analysis allowed the understanding of the behavior of the suspended materials, before and after the disaster, to unravel how the dispersion of tailings occurs. Part of the ore tailing that was deposited along the river went into suspension again in the rainy season one year after the disaster and, five years after, the satellite images show no difference compared to pre-disaster conditions. But, seven years after the spill, there are still iron ore tailings deposited and mixed with the river sediments. The fluvial dynamics are incorporating the tailings into the sediments, originating conducive places to provide various types of metals adhered to the particulate matter. Keywords: iron ore tailings, rainfall, remote sensing.
Read full abstract