VARIABILITY OF THE ASIAN MONSOON IN THE PRISM EXPERIMENTS

This is a comparative study of the influence of SST and land surface conditions on the monsoon circulation and its predictability. To address this issue a special project PRISM has been set up. This project has produced an experimental dataset consisting of nine 10-member ensembles, run with the ECMWF T63L31-cy16r2 model, and integrated for a full year from the beginning of November to the end of the following October. Additionally, the spring/summer part of the integrations has been run with both observed and climatological SST. In the PRISM ensembles the summertime variability of 850-hPa wind and rainfall over south-east Asia is dominated by one pattern (EOF1) corresponding (in positive phase) to an easterly wind anomaly over the north Indian Ocean, a southward shift of the TCZ but with a weak, positive rainfall anomaly over India. In the model, this pattern is not related to El Niño conditions during summer, but rather to positive SST anomalies in the West Pacific and Indian Oceans during late-winter and spring, following an El Niño event in winter. For both 850-hPa wind and rainfall over SE Asia, other modes of summertime variability exist which show a much stronger relationship with El Niño conditions then EOF1, and therefore a higher predictability. In experiments with observed SST these modes account for a negative correlation between Indian rainfall and east-Pacific SST.