Tecnossolos do trecho afetado por rejeitos de mineração na Bacia do Rio Doce: modelagem, evolução e monitoramento
Loading...
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Universidade Federal de Viçosa
Abstract
O rompimento da barragem de Fundão, em novembro de 2015, representa um marco trágico na história ambiental e social do Brasil. Localizada em Mariana, Minas Gerais, a barragem liberou 43 milhões de m³ de rejeitos de mineração de ferro, afetando vastas áreas ao longo do rio Doce. Este evento desencadeou uma das maiores catástrofes ambientais do país e levantou questões cruciais sobre a gestão de resíduos industriais, os impactos socioeconômicos nas comunidades locais e a sustentabilidade das práticas de mineração. Compreender a magnitude desse desastre evidencia a necessidade de análises aprofundadas sobre suas causas, consequências e lições aprendidas. Esta pesquisa visa explorar a evolução do ambiente, examinar as repercussões imediatas e de longo prazo nos ecossistemas e comunidades afetadas, a partir da análise dos solos presentes nesses locais. Para isso, o estudo foi dividido em três capítulos. O primeiro capítulo teve o objetivo de realizar uma caracterização físico-química dos tecnossolos assentados sobre áreas ao longo do rio Gualaxo do Norte, Carmo e Doce, a partir de uma ampla base de dados de perfis de solos. O segundo capítulo teve o objetivo de espacializar a textura do solo através de técnicas de mapeamento digital de solos, propondo modelos para as condições de pré e pós-rompimento, e situação atual. Foi desenvolvido o primeiro mapeamento digital de solo na área afetada pelo rompimento, gerando mapas texturais com alta resolução espacial. Modelos mais precisos foram obtidos na modelagem pós-rompimento, destacando a importância da homogeneidade do material e da delimitação da mancha de lama. Na modelagem pré-rompimento, as variáveis topográficas foram cruciais, e na situação atual, a distância euclidiana à drenagem ressurgiu como fundamental, indicando o reestabelecimento dos diques aluvionares. O terceiro capítulo investigou as relações entre a textura e concentração de metais em solos na região do alto e médio rio Doce. A distribuição dos elementos químicos analisados em solos de planícies aluviais do rio Doce foi influenciada pela granulometria, evidenciando uma forte correlação entre silte e argila e as concentrações desses elementos. Palavras-chave: modelagem ambiental, aprendizado de máquina, tecnossolos.
The collapse of the Fundão dam in November 2015 marks a tragic milestone in Brazil's environmental and social history. Located in Mariana, Minas Gerais, the dam released 43 million m³ of iron mining waste, affecting vast areas along the Doce River. This event triggered one of the country's largest environmental disasters and raised crucial questions about industrial waste management, the socioeconomic impacts on local communities, and the sustainability of mining practices. Understanding the magnitude of this disaster highlights the need for in-depth analyses of its causes, consequences, and lessons learned. This research aims to explore environmental evolution, examine the immediate and long-term repercussions on ecosystems and affected communities, based on soil analysis in these areas. For this, the study was divided into three chapters. The first chapter aimed to perform a physicochemical characterization of technosols settled over areas along the Gualaxo do Norte, Carmo, and Doce rivers, based on a comprehensive soil profile database. The second chapter aimed to spatialize soil texture through digital soil mapping techniques, proposing models for pre- and post-collapse conditions, and the current situation. The first digital soil mapping in the area affected by the collapse was developed, generating high-resolution spatial texture maps. More accurate models were obtained in post-collapse modeling, highlighting the importance of material homogeneity and the delineation of the mud stain. In pre-collapse modeling, topographic variables were crucial, and in the current situation, the Euclidean distance to drainage re-emerged as fundamental, indicating the re-establishment of alluvial dikes. The third chapter investigated the relationships between texture and metal concentration in soils in the upper and middle Doce River region. The distribution of chemical elements analyzed in floodplain soils of the Doce River was influenced by granulometry, showing a strong correlation between silt and clay and the concentrations of these elements. Keywords: environmental modeling, machine learning, technosols.
The collapse of the Fundão dam in November 2015 marks a tragic milestone in Brazil's environmental and social history. Located in Mariana, Minas Gerais, the dam released 43 million m³ of iron mining waste, affecting vast areas along the Doce River. This event triggered one of the country's largest environmental disasters and raised crucial questions about industrial waste management, the socioeconomic impacts on local communities, and the sustainability of mining practices. Understanding the magnitude of this disaster highlights the need for in-depth analyses of its causes, consequences, and lessons learned. This research aims to explore environmental evolution, examine the immediate and long-term repercussions on ecosystems and affected communities, based on soil analysis in these areas. For this, the study was divided into three chapters. The first chapter aimed to perform a physicochemical characterization of technosols settled over areas along the Gualaxo do Norte, Carmo, and Doce rivers, based on a comprehensive soil profile database. The second chapter aimed to spatialize soil texture through digital soil mapping techniques, proposing models for pre- and post-collapse conditions, and the current situation. The first digital soil mapping in the area affected by the collapse was developed, generating high-resolution spatial texture maps. More accurate models were obtained in post-collapse modeling, highlighting the importance of material homogeneity and the delineation of the mud stain. In pre-collapse modeling, topographic variables were crucial, and in the current situation, the Euclidean distance to drainage re-emerged as fundamental, indicating the re-establishment of alluvial dikes. The third chapter investigated the relationships between texture and metal concentration in soils in the upper and middle Doce River region. The distribution of chemical elements analyzed in floodplain soils of the Doce River was influenced by granulometry, showing a strong correlation between silt and clay and the concentrations of these elements. Keywords: environmental modeling, machine learning, technosols.
Description
Citation
ALMEIDA, Pedro Henrique Araújo. Tecnossolos do trecho afetado por rejeitos de mineração na Bacia do Rio Doce: modelagem, evolução e monitoramento. 2024. 140 f. Tese (Doutorado em Solos e Nutrição de Plantas) - Universidade Federal de Viçosa, Viçosa. 2024.
