Seasonality in the predictability of Indonesian monsoonal climate

The Indonesian archipelago is unquestionably the one region in the world where climate predictability and seasonal climate prediction are synonymous with ENSO (the El Nino-Southern Oscillation phenomenon). Typically, warm ENSO events are associated with a pronounced dry anomaly during the growth and mature stages of an event, which translates into an extended dry season, and into a delayed transition into the wet season. Consequently, the length of the wet season is reduced, and agricultural production suffers (Naylor et al 2001).

However, when one looks up close, one cannot fail but to notice that predictability is lost for the more densely populated region of Indonesia, the island of Java, and for islands to its east, during the core of the monsoon.

The absence of a coherent large-scale signal during the monsoon season has already been noted (e.g. Haylock and McBride, 2001 and McBride et al, 2003), but a full explanation is lacking. We plan to compare observations and ensembles of simulations with an atmospheric General Circulation Model, forced with different configurations of tropical sea surface temperature (SST) fields, to investigate the relative roles of tropical Pacific, Indian and local SSTs in the seasonality of predictability in Indonesia.