The seasonal climate predictions have effectively contributed to the planning on various sectors of society, such as energy, agriculture and health activities. The need for more detailed forecasts by increasing spatial resolution climate information has concentrated research efforts recently in various regions of the globe. Downscaling technique by regional dynamic modeling, known as dynamic downscaling, has been used for this purpose. Thus, the present study aimed to evaluate and calibrate the regional BRAMS model, unidirectionally nested with AGCM/CPTEC global model, in predicting the extreme weather events of 1983 (El Niño) and 1989 (La Niña) over South America (SA) for March-April-May (MAM). The predictions were produced at spatial resolution of 30 km, having the persisted sea surface temperature and climatological soil moisture as boundary conditions. After the calibration process, the BRAMS was able to capture the rainfall pattern associated with extreme weather events, such as the decrease (increase) in precipitation pattern mainly on the eastern of North region and Northeast Brazil, and the increase (decrease) of rainfall in the Southern region for the year 1983 (1989).
Downscaling; ENSO; seasonal climate predictions
