Sediment source fingerprints of natural processes and anthropogenic pressures: A contribution to manage the Paraopeba River basin impacted by the B1 tailings dam collapse

Abstract

Understanding the origins of sediment transport in river systems is crucial for effective watershed management, especially after catastrophic events. This information is essential for the development of integrated strategies that guarantee water security in river basins. The present study aimed to investigate the rupture of the B1 tailings dam of the Córrego do Feijão mine, which drastically affected the Brumadinho region (Minas Gerais, Brazil). To address this issue, a confluence-based sediment fingerprinting approach was developed through the SedSAT model. Uncertainty was assessed through Monte Carlo simulations and Mean Absolute Error (MAE). Estimates of the overall average contributions of each tributary were quantified for each station and annually during the period 2019–2021. It was observed that the sampling point PT-09, closest to the dam breach, contributed to almost 80% of the Paraopeba River in 2019. Despite the dredging efforts, this percentage increased to 90% in 2020 due to the need to restore the highly degraded area. Additionally, the main tributaries contributing to sediment increase in the river are Manso River “TT-03” (almost 36%), associated with an area with a high percentage of urban land use, and Cedro stream “TT-07” (almost 71%), whose geology promotes erosion, leading to higher sediment concentration. Uncertainties arise from the limited number of available tracers, variations caused by dredging activities, and reduced data in 2020 due to the pandemic. Parameters such as land use, riparian vegetation degradation, downstream basin geology, and increased precipitation are key factors for successfully assessing tributary contributions to the Paraopeba River. The obtained results are promising for a preliminary analysis, allowing the quantification of key areas due to higher erosion and studying how this disaster affected the watershed. This information is crucial for improving decision-making, environmental governance, and the development of mitigating measures to ensure water security. This study is pioneering in evaluating this methodology in watersheds affected by environmental disasters, where restoration efforts are ongoing.