作者: Rubinei Dorneles Machado , Rosmeri Porfírio da Rocha
DOI: 10.1590/S0102-77862011000100011
关键词: Correlation coefficient 、 Horizontal resolution 、 Air temperature 、 Climatology 、 COLA (software architecture) 、 Climate model 、 Geography 、 Global model
摘要: This study presents an assessment of seasonal forecasts from the regional climate model RegCM3 nested to CPTEC/COLA global model. The used version is 60 km horizontal resolution over area that includes large part South America. and were evaluated using rainfall air temperature analyses Climate Prediction Center (CPC) National Centers for Environmental (NCEP), respectively. From May 2005 July 2007, 27 forecast (except with 26 forecasts) three regions Brazil: Northeast (NDE), Southeast (SDE) (SUL). also compared climatologic fields analyses. statistical indexes (bias, correlation coefficient, root mean square error efficiency coefficient) showed (NDE, SDE SUL) predicted closer observations than CPTEC/COLA. Moreover, a best predictor average during period. For temperature, perform better NDE SUL regions, while shows performance ovver SDE. Finally, estimated are observed climatology. These results suggest potential utility be which will explored in future ensemble forecast.