Avaliação de simulações de precipitação e vazão por um modelo atmosférico em bacias do semiárido brasileiro.

Abstract

Semiarid regions present high temporal and spatial variability of rainfall as well as the occurrence of extreme events, such as floods and droughts. Forecasting these extreme events is helpful for taking mitigation measures in advance. Atmospheric models can simulate several meteorological variables, among them precipitation, and fluxes of heat and water in the surface, through the Soil‐Vegetation‐Atmosphere Transfer Schemes (SVAT). The SVATs calculate the water balance in the soil and also runoff. Runoff information, however, has not been adequately evaluated with respect to hydrological forecasting. This work evaluates time series of precipitation and runoff simulated by a regional atmospheric model in hydrographic basins in the Brazilian semiarid region during the first six months of the years 2004 to 2009. The simulations were performed by nesting the BRAMS regional model into the CPTEC global model. The global model resolution was 100 km. The regional model was nested in the global model using two grids of 22 km and 5,5 km. For obtaining hydrographs for each basin, runoff simulated by BRAMS at the 5,5 km grid‐points was propagated through the drainage network. The performance was assessed in 13 sub‐basins of the Piranhas‐Açu and Apodi‐Mossoró Basins, located in the semiarid region of the States of Paraíba and Rio Grande do Norte. The results showed: (a) deficiencies in the simulation of daily precipitation, with overestimation of the occurrence in the range of 10 to 20 mm and not representing the frequency distribution of observed precipitation; however, these deficiencies were minimized when aggregating the scale of analysis by areal averages in the basins and 10‐day totals; (b) overestimation of discharge in the majority of simulated hydrographs when compared to the observed ones, due to, in the grid-points, excessive runoff and soil moisture as well as overflow of water storage capacity in the soil.