Projected mean and extreme precipitation based on bias-corrected simulation outputs of CORDEX Southeast Asia

Abstract

Simulations of seven regional climate models (RCMs) from CORDEX-SEA were bias-corrected using the quantile mapping (QM) method to examine the possible ranges of future mean precipitation and extreme index changes over Southeast Asia (SEA). Eleven rainfall indices including total precipitation (PRCPTOT), maximum one day precipitation (RX1day), the maximum number of consecutive dry days (MCDD), intensity of very wet days (R95pTOT) and the frequency of heavy rainfall days (R20mm) were considered. The QM procedure largely reduced the biases and inter-model variations in the historical period. For future projections, some QM modifications can be found but these were season and location dependent. The projected large changes of mean seasonal precipitation, especially over Indochina, were slightly magnified after the correction. Generally, the projected bias-corrected ensemble mean (ENSMEAN) indicated increased mean rainfall during boreal winters in mainland SEA by as much as 30% depending on the projection periods. During the boreal summer, mean rainfall in the maritime continent was projected to decrease by as much as −30% in some areas in Indonesia. In summary, the possible ranges of projected changes of extreme indices are dependent on the type of indices, scenarios and seasons. For JJA, the possible range over the entire region is between −40% and 60%. For DJF, the possible ranges are between −20% and 60% and −20%–20% over the mainland Indochina and Maritime Continent, respectively.