Efeito da resolução horizontal na simulação de um evento chuvoso extremo.

Abstract

This work presents an evaluation of the performance of a mesoscale atmospheric model applied to a system that caused high precipitation in Northeast Brazil This event occurred between July 30 and August 02, 2000. The model RAMS was used in its version 4.3 with three nested grids in the horizontal: a) grid 1 with 112 km; b) grid 2 with 28 km, and c) grid 3 with 7 km of resolution. The central coordinates for each one of the three grids were those of Recife - PE (8°S e 35°W). The model was run for four days starting at 00:00 UTC of July 30, 2000. The results show that the precipitation simulated through the convection parameterization does not realistically represent the precipitation of the rainy event even though it represents well the large-scale systems. On the other hand, the precipitation resulting from the parameterization of microphysic processes represented well the qualitative aspects of the system. This is more evident in the higher resolution grid. The simulation, as a whole, underestimates the observed precipitation. However, the qualitative analysis of the results shows that the mesoscale models are a powerful tool to forecast and to understand the local atmospheric phenomena, provided they have a proper calibration.