Avaliação da precipitação simulada pelo modelo global MPI-ESM-LR e pelo modelo regional MPI-ESM-LR-RCA4 na bacia hidrográfica do rio Paraíba

Abstract

Climate prediction, in particular the analysis of rainfall patterns in a watershed, is essential for hydrological studies required for water resources management, especially in areas as the Brazilian Northeast, region influenced by a high spatial and temporal rainfall variability. Accordingly, the prediction of atmospheric variables, as rainfall, is usually performed through Global Climate Models (MCGs) and Regional Climate Models (MCRs) - numerical models able to describe the main physical and dynamic processes of the atmosphere, ocean and surface for the entire world and for a specific area, respectively. This study investigates the performance of a global model (MPI-ESM-LR) and a regional model (RCA4 nested to MPI-ESM-LR) from the Fifth Assessment Report (AR5) of the Intergovernmental Panel on Climate Change (IPCC) in the simulation of rainfall for the current climate (1994 to 2005) on the Paraíba watershed. The rainfall simulated by the models were evaluated by seasonality, intensity and location. The results showed that both models present bias in the rainfall simulation and that, in general, the regional model was not able to improve substantially the quality of the simulation performed by the global model, although some promising results were observed in the quantitative and spatial rainfall evaluation.