Limiares de precipitação para ocorrência de deslizamentos em Belo Horizonte: análise espacial e influência de fatores condicionantes
Loading...
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Universidade Federal de Viçosa
Abstract
Os movimentos de massas em encostas que estão diretamente ligados a um conjunto de fatores condicionantes como as características dos materiais geológicos, os processos geomorfológicos e as alterações antropogênicas. Os eventos associados a movimentos de massas provocados por chuvas aumentaram significativamente em quantidade e magnitude de impacto ao longo do tempo. Nesse contexto, Minas Gerais posiciona-se como o segundo estado brasileiro mais afetado economicamente, além disso, concentra o maior número de ocorrências registradas de movimentos de massa no Brasil. Dessa forma, determinar o limiar de chuva necessário para provocar deslizamentos é um problema de interesse científico, técnico e social. Sobretudo, é importante avaliar como o limiar pode variar de acordo com um conjunto de fatores condicionantes, tais como: geologia, declividade, alterações antropogênicas. Este estudo tem como objetivo definir limites objetivos de duração da chuva (E,D) para a ocorrência de movimentos de massas para a área urbana de Belo Horizonte (MG). Além de avaliar a influência da declividade e da geologia nestes limiares. A precipitação é o principal agente deflagrador de deslizamentos e, portanto, sua medição e observação é uma das ferramentas fundamentais para a compreensão, monitoramento e previsão destes eventos. A falta de uma série histórica de precipitação com dados espacializados em escala adequada é uma grande limitação para desenvolver estudos de gerenciamento de riscos. Este, portanto, deve ser o primeiro passo para viabilizar o cálculo dos limiares de chuva para ocorrência de desastres. Nesse sentido, foram utilizados dados de precipitação diária de 2012 a 2022, extraídos de uma rede de 56 pluviômetros localizados na área urbana de Belo Horizonte. Especificamente, foram avaliados os métodos de interpolação geoestatística: Krigagem Simples e Krigagem de Posto Fixo (FRK). Dentre esses dois preditores, o FRK obteve resultados mais robustos e homogêneos. A variância de krigagem foi utilizada para o cálculo das incertezas enquanto o FRK apresentou menor variabilidade na estimação da incerteza do que a Krigagem. Sequencialmente, os limiares de precipitação para ocorrência de movimentos de massas foram estabelecidos por meio do método estatístico frequentista e considerando um catálogo abrangente de 2.622 eventos. Os limiares para os níveis de probabilidade de excedência foram calculados com incertezas quantificadas. Além disso, uma análise estratificada investigou a influência de fatores condicionantes importantes: complexos geológicos e declividade. Os resultados indicam que a geologia modula significativamente os limites de precipitação. O domínio geológico do Supergrupo Minas (SM) apresentou o limite mais baixo e estatisticamente distinto. Em contrapartida, diferentes classes de inclinação, embora influenciem a densidade dos eventos, não produziram limites de precipitação estatisticamente diferentes. Esses resultados, fornecem uma base técnica crucial para o desenvolvimento de sistemas de alerta precoce mais eficazes e espacialmente diferenciados e para orientar estratégias de gestão de risco e intervenções de planejamento urbano em Belo Horizonte. Palavras-chave: deslizamentos; limiares; interpolação; desastres
Mass movements on slopes are directly linked to a set of conditioning factors such as the characteristics of geological materials, geomorphological processes, and anthropogenic changes. Events associated with mass movements caused by rainfall have increased significantly in number and magnitude of impact over time. In this context, Minas Gerais ranks as the second most economically affected state in Brazil and also has the highest number of recorded mass movements in the country. Thus, determining the rainfall threshold required to cause landslides is a matter of scientific, technical, and social interest. Above all, it is important to assess how the threshold may vary according to a set of conditioning factors, such as geology, slope, and anthropogenic changes. This study aims to define objective limits for rainfall duration (E, D) for the occurrence of mass movements in the urban area of Belo Horizonte (MG). In addition to assessing the influence of slope and geology on these thresholds. Precipitation is the main trigger for landslides and, therefore, its measurement and observation is one of the fundamental tools for understanding, monitoring, and predicting these events. The lack of a historical series of precipitation with spatial data on an adequate scale is a major limitation for developing risk management studies. This, therefore, should be the first step in enabling the calculation of rainfall thresholds for the occurrence of disasters. In this sense, daily precipitation data from 2012 to 2022 were used, extracted from a network of 56 rain gauges located in the urban area of Belo Horizonte. Specifically, the geostatistical interpolation methods Simple Kriging and Fixed-Rank Kriging (FRK) were evaluated. Among these two predictors, FRK obtained more robust and homogeneous results. Kriging variance was used to calculate uncertainties, while FRK showed less variability in uncertainty estimation than Kriging. Sequentially, precipitation thresholds for the occurrence of mass movements were established using the frequentist statistical method and considering a comprehensive catalog of 2,622 events. The thresholds for the probability of exceedance levels were calculated with quantified uncertainties. In addition, a stratified analysis investigated the influence of important conditioning factors: geological complexes and slope. The results indicate that geology significantly modulate precipitation thresholds. The Minas Supergroup (SM) geological domain had the lowest and statistically distinct threshold. In contrast, different slope classes, although influencing event density, did not produce statistically different precipitation thresholds. These results provide a crucial technical basis for the development of more effective and spatially differentiated early warning systems and for guiding risk management strategies and urban planning interventions in Belo Horizonte. Keywords: landslides; thresholds; interpolation; disasters
Mass movements on slopes are directly linked to a set of conditioning factors such as the characteristics of geological materials, geomorphological processes, and anthropogenic changes. Events associated with mass movements caused by rainfall have increased significantly in number and magnitude of impact over time. In this context, Minas Gerais ranks as the second most economically affected state in Brazil and also has the highest number of recorded mass movements in the country. Thus, determining the rainfall threshold required to cause landslides is a matter of scientific, technical, and social interest. Above all, it is important to assess how the threshold may vary according to a set of conditioning factors, such as geology, slope, and anthropogenic changes. This study aims to define objective limits for rainfall duration (E, D) for the occurrence of mass movements in the urban area of Belo Horizonte (MG). In addition to assessing the influence of slope and geology on these thresholds. Precipitation is the main trigger for landslides and, therefore, its measurement and observation is one of the fundamental tools for understanding, monitoring, and predicting these events. The lack of a historical series of precipitation with spatial data on an adequate scale is a major limitation for developing risk management studies. This, therefore, should be the first step in enabling the calculation of rainfall thresholds for the occurrence of disasters. In this sense, daily precipitation data from 2012 to 2022 were used, extracted from a network of 56 rain gauges located in the urban area of Belo Horizonte. Specifically, the geostatistical interpolation methods Simple Kriging and Fixed-Rank Kriging (FRK) were evaluated. Among these two predictors, FRK obtained more robust and homogeneous results. Kriging variance was used to calculate uncertainties, while FRK showed less variability in uncertainty estimation than Kriging. Sequentially, precipitation thresholds for the occurrence of mass movements were established using the frequentist statistical method and considering a comprehensive catalog of 2,622 events. The thresholds for the probability of exceedance levels were calculated with quantified uncertainties. In addition, a stratified analysis investigated the influence of important conditioning factors: geological complexes and slope. The results indicate that geology significantly modulate precipitation thresholds. The Minas Supergroup (SM) geological domain had the lowest and statistically distinct threshold. In contrast, different slope classes, although influencing event density, did not produce statistically different precipitation thresholds. These results provide a crucial technical basis for the development of more effective and spatially differentiated early warning systems and for guiding risk management strategies and urban planning interventions in Belo Horizonte. Keywords: landslides; thresholds; interpolation; disasters
Description
Citation
CORTEZ, Bianca Nespoli. Limiares de precipitação para ocorrência de deslizamentos em Belo Horizonte: análise espacial e influência de fatores condicionantes. 2025. 141 f. Tese (Doutorado em Engenharia Civil) - Universidade Federal de Viçosa, Viçosa. 2025.
