Estimativa de recarga natural em aquífero livre e raso com suporte de sensoriamento remoto : aplicação na bacia do ribeirão Serra Azul (MG)

• Ano de defesa: 2021
• Autor(a) principal: Rodrigo de Paula Hamzi
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; Brasil; ENG - DEPARTAMENTO DE ENGENHARIA HIDRÁULICA; Programa de Pós-Graduação em Saneamento, Meio Ambiente e Recursos Hídricos; UFMG
• Palavras-chave em Português: Recarga; Thornthwaite; Sensoriamento remoto; Aquífero livre e raso; VNA; Engenharia sanitária; Recursos hídricos - Desenvolvimento; Aquíferos
• Link de acesso: http://hdl.handle.net/1843/61139

Estimating groundwater recharge is essential both for understanding available water resources and for assessing groundwater vulnerability. However, there is no universal method for its quantification, so it will depend on each limitation or specification of the technique applied. In addition, the techniques are generally punctual and require extensive resources for their operation on a spatial scale. The use of remote sensing makes it possible to obtain climatic variations, which are related to the recharge process, in scales that the punctual methods would not be able to. Thornthwaite methodology (water balance methods in the soil) applied according to Charles et al. (1993) in association with remote sensing would make it possible to simulate recharge spatially through indirect modeling of soil moisture variations and water balance components. Another method that makes it possible to associate remote sensing with spatial recharge is the simplified water balance. Thus, the present work estimated the spatial and temporal distribution of the recharge of a free and shallow aquifer in a sub-basin of Ribeirão Serra Azul, in Minas Gerais, for the hydrological years of 2011/2012, 2012/2013 and 2013/2014 by water balance of the Thornthwaite method by varying rainfall in situ and via remote sensing and simplified water balance using climatic products from remote sensing. From rainfall data from remote sensing, the following performance indices were analyzed: Tropical Rainfall Measuring Mission (TRMM) – Multi-satellite Precipitation Analysis (TMPA) 3B42RT V7; Combined Scheme approach (CoSch); and Precipitation Estimation From Remotely Sensed Information Using Artificial Neural Networks – Cloud Classification System (PERSIANNCCS), having been used in this study the TRMM – TMPA 3B42RT V7 for having presented better performance. The remote sensing evapotranspiration used was from the Operational Simplified Surface Energy Balance (SSEBop) method. Estimates generated from in situ and remote sensing data were compared with recharge estimated by the Water Table Fluctuation (WTF) method by means of evaluation metrics, which showed low correlation of the data on a monthly scale, reflection of the time scale allowed in the model, since according to the recharge estimate by the WTF method there was recharge in the months when the Thornthwaite method considered it null, when the infiltration is lower than the monthly scale evapotranspiration. The water balance methods for recharge estimation used with in situ data and remote sensing did not show good results on the monthly scale when compared to the gross recharge estimated by the WTF method. In general, the simplified water balance demonstrated worse results than the soil water balance, both using in situ precipitation and in sensing. Thus, in the case of the two water balance methods in the soil, the results indicate that it is necessary to improve their variants and specifically in the case of the simplified water balance, the accuracy of the remote sensing products used must be improved.