Rain

Occurrence or no of rain is normally one of the worries of the society relacionadas to the time and weather. In numerical prediction,the occurrence of precipitation is treated by cumulus parameterization. Several schemes of parameterization of cumulus convection come being building with base in data (observacionais) and in the understanding of way how several atmosphere scales interact.The scheme of cumulus parameterization implemented currently by the model Brazilian developments on the regional Atmospheric Modelling System(BRAMS) consist of disturbance around the classical locks of Grell,Arakawa and Schubert, Kain and Fritsch,low-level Omega and (convergencia)of humidity. Like result,it’s have (prognósticos)of precipitation that can be combined of several ways, creating a numerical representation of precipitation and rates of heating and atmospheric moistening .In this version, only one specific lock or a simple average among the members are available. One detailed study of performance of this locks has not been made, that difficult to the chose of what hypotheses must be used in the model. The goal of this work is to verify the performance of the regional model BRAMS to simulate the convective precipitation for the period of 2004 December to 2006 February ,using the differences of hypotheses of locks of convective parameterization available. As this study is preliminary , just the month December were evaluated. Simulations were performed in each 24 hours for each day in this month. The locks used were from Grell, Arakawa and Schubert, Kain and Fritsch,low-level Omega, humidity(convergencia) and the Ensemble lock(given by simple average among members). From de predicted precipitation, the simulations were compared to the estimated data by the satellite Tropcal Rainfall Measuring Mission(TRMM) over all South America. Also, (pontuais)comparisons were made, in some Brazilian capitals , comparing with datas of conventional meteorological stations obtained from