Abstract

A study of a mesoscale convective system using radar data assimilation is presented. Simulations were made using reflectivity and radial velocity data from two radars (Cascavel and Asunción). Different initializations of the WRF-model were performed: without assimilation, with assimilation of conventional data, and with assimilation of radar. Results were compared with CoSch3 precipitation estimates. Reflectivity and radial velocity data were introduced to the model indirectly (by assimilating rain water mixing-ratio). Analysis generated from the data assimilation showed the impact of the radar data assimilation throughout the model vertical structure. We demonstrated that using cycles to initialize the model is fundamental to improve rainfall location forecasts. Assimilating radar data proved to be the best results to forecast intense precipitation cores. The results may contribute to improve early warning systems.

Keywords:
MCS; WRFDA; Atmospheric Modeling; 3DVAR