Indian summer monsoon rainfall simulation by an AGCM ensemble system

Abstract

Ensemble simulation of Indian summer monsoon rainfall (ISMR) and its interannual variation is investigated using Meteorological Research Institute atmospheric general circulation model. Representation of climatological seasonal variation of rainfall is found to be important for a realistic simulation of seasonal mean and interannual variation of ISMR by the ensemble mean. As a result, the ensemble mean captures the extreme ISMR events associated with pronounced large-scale equatorial sea surface temperature (SST) fluctuations. Further, the systematic bias in the simulation of mean seasonal variation of rainfall over the Asia-Pacific region is removed, based on multiple linear regression of daily rainfall from the member simulations against observation. This results in improved simulation of ISMR in the bias-removed ensemble mean which is illustrated by hindcasting the anomalous ISMR of 2002, the most severe drought recorded in recent times. The removal of bias in the mean seasonal variation of rainfall brings out the realistic precipitation response in the model to fluctuations in SST boundary forcing. As a result, the bias-removed ensemble mean captures not only the seasonal anomaly of 2002 ISMR but improves the subseasonal variation of ISMR as well.