Padrões e tendências hidroclimáticas de chuva identificados por sensoriamento remoto na bacia do rio Madeira Amazônia

Abstract

Understanding the behavior of precipitation is a major component of the hydrological cycle and hydrological risk management in watersheds, especially in the Amazon. Important by limited monitoring in the Amazon region by conventional rainfall networks is a challenge to be faced. The task of revealing rainfall patterns in the Madeira River basin, which represents the largest Amazon sub-basin, is proposed. The procedure was based on remotely sensed data, rigorously validating them spatially and temporally. CHIRP and CHIRPS data as well as TRMM products 3B42 and 3B42RT were compared with information based on rainfall collected in raingauge stations. Several analyzes were conducted, including the use of wavelets to better understand the occurrence and distribution of extreme events and rainfall trends in the basin, with characterization of cycles and frequencies. It was found that CHIRPS and CHIRP products, when compared to 3B42 and 3B42RT, better represent the monthly accumulated rainfall phenomena of the Madeira river basin. It is also noteworthy that the work allowed to obtain the intensity-duration-frequency spatial curves through satellite data. Spatial variability and spatial probabilistic recurrence of extreme rainfall events at the basin scale were characterized. Trends of rainfall reductions in the wettest months and increased evapotranspiration in the warmer months were identified. Complementarily, the water balance was performed for a small basin in the surroundings of the municipality of Machadinho do Oeste using the SWAT hydrological model, which showed the trends toward reduced rainfall and lower water availability with respect to deforestation scenarios. On the other hand, the equilibrium among the components of the water balance differ accordingly to the spatial scale of analysis, which deserves to be further investigated in timely research in the Amazon region.