Uso do sensoriamento remoto para análise de eventos de seca em bacias hidrográficas: estudo de caso na sub-bacia do Ribeirão Serra Azul, MG

• Ano de defesa: 2021
• Autor(a) principal: Débora Joana Dutra
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais
• Palavras-chave em Português: Modelagem de dados – Aspectos ambientais; Secas – Minas Gerais; Bacias hidrográficas – Minas Gerais; Sensoriamento remoto
• Link de acesso: https://hdl.handle.net/1843/36692

Land use changes arising mainly from anthropic influence cause several environmental pressures, being associated with several environmental studies, such as monitoring natural resources. In this regard, drought phenomenon has been highlighted in international and national studies, mainly due to world water scarcity several episodes occurred between 2014 and 2015. In Minas Gerais state, Paraopeba River Basin suffered consequences from pressures on environment, mainly associated with water volume decrease, which caused water scarcity State Decree in portion of Serra Azul reservoir during 2015. The objective this work is to analyze possibility of using vegetation condition index (VCI) to drought monitor, as form to subsidize management and monitoring actions in water supply station. In addition, we analyze the occurrence of drought different types over analyze time series. The case study was realized in Serra Azul subbasin, covering Mateus Leme, Igarapé, Juatuba and Itaúna cities, and located between 20º 15 'and 20º 00' south latitude parallels and 44º 15' and 44º 35' west longitude meridians. The methodology consisted of two stages: drought analysis period and vegetation condition index effectiveness for phenomenon monitoring. The event scope analysis in study region, we used statistical analysis to select the best vegetation index to characterize event extent. The analyzed time series (2013-2018) was organized according to season of year for VCI calculation, we classified according to methodology proposed by index creator. To measure index effectiveness for monitoring different drought categories, the stages was divided into four parts: meteorological, hydrological, agricultural and socioenvironmental drought. In addition, we used correlation technique to verify if index explains results found in applied methodologies. The results allowed to identify, within MODIS sensor products, NDVI is the best index to region drought characterize. In meteorological drought, VCI results explain the data reported in official climate reports from INPE and INMET. In addition, external events helped to event aggravate between 2014 and 2015. In relation the hydrological drought, we found moderately strong VCI correlation with data on drought and relative soil severity. Finally, the hydrological and socioenvironmental drought results allowed to analyze, as phenomenon worsens in study region, forest biomass in agricultural areas and region's water supply eservoirdecreases volume. We obtained results can help in future studies on theme, assisting in systems implementation and actions planning to analyzes events consequences. In addition, we suggested for future studies land use and climatological models’ application, for analysis and event prevention next years. For greater process efficiency, we suggested local analysis is through field visits, for greater precision and identification of drought mapping in study region.