Impactos das mudanças climáticas e da expansão agrícola na disponibilidade hídrica e geração de energia elétrica na bacia hidrográfica do Alto Rio Grande

Abstract

Changes in climate and in land use land cover can cause important variations in the hydrological regime. The headwater basins in regions with strong orographic effect are highly strategic to satisfy the country's water demand. Considering this, the main objective of this work is to evaluate the impacts on water resources in the Upper Rio Grande watershed from 1985 to 2015 and in future scenarios. Thus, this study intends to understand how changes in climate and land use land cover can affect the water and energy security in the region. Initially, the impacts related to climatic and anthropogenic factors on runoff observed between 1985 and 2015 were separated by the Budyko curve decomposition method. Next, it was analyzed uncertainties related to the parameters of the hydrological model estimated in the Pareto front from a multi- objective optimization method (SPEA2) in the estimation of daily runoff. In addition, uncertainties in the projection of long-term runoff by the following methods were analyzed: Budyko curve and hydrological modeling (MHD-INPE). Thus, from the MHD-INPE calibration, it was evaluated the hydrological impacts in future scenarios of climate change (simulated by a regional climate model, Eta-model, associated with four global circulation models) and a scenario of changes in land use land cover (projected considering the agricultural expansion observed in the historical period using Dinamica EGO model). Finally, it was estimated the variation in electricity generation for Camargos and Itutinga hydroelectric plants from simulated daily discharge in the future scenarios considering the operation strategy in hydroelectric plants to produce firm energy associated with 5% risk in the historical period of each climate model. As a result, during the observed series between 1985 and 2015, variations in aridity index and agricultural areas in the basin were the variables that best represented the variation of the long-term runoff, where the increase of agricultural areas is related to runoff reduction in six of nine sub-basins. In the analysis of uncertainties considering the parameters of the hydrological model (MHD-INPE) in the Pareto front optimization, differences were verified in the extremes of simulated daily flow and few variations were verified in the future projection of the long-term runoff. Budyko's method tends to project higher runoff when compared to MHD-INPE in years with higher precipitation, as well as lower runoff in years with lower precipitation. At last, the climate change scenarios indicate a reduction in future average annual runoff, especially in the magnitude of the minimum discharge with a frequency of exceedance greater than 95%. Consequently, it is estimated a reduction in electricity generation and water storage in the Camargos reservoir, with increase in system failures. In the combined scenario of rainfed agriculture expansion and climate change, the impacts were verified with lower magnitude, mainly in less umid climate conditions due to restrictions on crop transpiration under water stress conditions. Therefore, the estimated impacts for the expansion of rainfed agriculture are milder in more critical climate scenarios.