The Pampanga River Basin in central Luzon Island, Philippines, has been suffered by floods due to typhoons or monsoonal rainfall every year. Flood risk change on global warming is an important issue in this vulnerable area. We studied rainfall distribution change in present (1979-2003) and future (2075-2099) climate scenarios using MRI-AGCM 3.2S (Meteorological Research Institute- Atmospheric General Circulation Model version 3.2 Super high-resolution) experiments, as a part of Japanese Program for Risk Information on Climate Change (SOUSEI). We conducted dynamical downscaling to obtain rainfall distribution fine enough for streamflow simulation in this river basin (catchment size of 11,000 km2). A regional model WRF (Weather Research and Forecasting) was used for downscaling into 5km horizontal resolution.The downscaling overestimated monthly mean basin rainfall by twice as raingauges. Relative frequencies of downscaled station rainfall overestimated in weak to moderate rainfall but underestimated in extreme rainfall compared to raingauges. To correct those biases, we converted cumulative probability of downscaled rainfall by mapping onto cumulative probability of gauged rainfall. The conversion was made in daily basis separately for each month, but those of top 0.5 percentile were converted by cumulative probability of whole year rainfall. The number of raingauge stations were variable depend on the year (8-17 stations in 22 years), and they included many data missing periods. We converted by mapping cumulative probability of basin average rainfall of RCM onto those of raingauges, instead of popular way of mapping RCM grid rainfall onto observed grid. This bias correction slightly underestimated relative frequency of weak rainfall, but it estimated consistent amount of extreme rainfall. Maximum annual 24-hour rainfall in basin average was 186mm in present climate, close to 193mm by raingauge. The extreme rainfall would be changed into 184 to 253mm in future climate on various SST distribution experiments of RCP8.5 scenario. |