Projeções de mudanças na disponibilidade hídrica na bacia hidrográfica do Rio Grande, sudeste do Brasil, sob cenários de mudanças climáticas

Abstract

The crescent debate regarding the climate change impacts on water availability has raised growing concerning, mainly related to the hydroelectricity generation. Future climate projections studies for water availability and hydric safety have been developed using the Global Circulation Models (GCMs). Such models are tools that simulate and help with the understanding of the climatic patterns throughout the time. Thus, they have great importance in the context of water resources management. The objective of this study was to evaluate the water availability in the Grande river basin, Southeastern Brazil, from 1961 to 2099. For that, two GCMs, dynamically regionalized by the Eta model (HadGEM2-ES and Eta- MIROC5), were used considering two scenarios of the climate projections, RCP4.5 and RCP8.5. These predictions were generated in a resolution of 20 km, sliced in four different periods: 1961-2005, 2007-2040, 2041-2070, and 2071-2099. The study was carried out considering the weather seasonality of the region, grouping in the wet (October to March) and the dry (April to September) periods. For the analyzed periods, water balance was calculated by the difference between the daily averages of precipitation and evapotranspiration. Also, the water balance was mapped employing ordinary kriging. The performance of the Eta-HadGEM2-ES and EtaMIROC5 models was evaluated based on the linear correlation between the maps generated using data from meteorological stations of the National Institute of Meteorology (INMET) and the simulated data for the present (1961-2005). The highest correlations were obtained for the dry season, with the Eta-MIROC5 model presenting the best result (0.83). However, for the rainy season, the values were similar, being 0.55 and 0.60 for, respectively, Eta- HadGEM2-ES and Eta-MIROC5. Taking the historical period as a reference, one can note that the northwest region of the Grande river basin has a greater tendency to water surplus. However, the most significant deficits are in the South-Southeast areas of the watershed. Both models also observed the same trend in wet (October to March) and dry (April to September) periods. EtaHadGEM2-ES model showed greater variability of water balance, with greater amplitude and spatial distribution of values. Besides, it presented trends to water availability decrease in both scenarios when compared to the Eta-MIROC5 model.