69 巻 (2013) 4 号 p. I_133-I_138
Downscaling plays a major role in evaluating future changes of precipitation and setting up appropriate adaptation strategies at regional and local scale. Hence, this study evaluates statistical bias correction and Dynamical Downscaling (DD) approaches by applying them to reproduce present climate of precipitation and discharges in Shikoku Island, Japan. DD captured well spatial and temporal distributions of precipitation owing to influence of the spatial resolution enhancement. Statistical bias correction has limitations in estimating extreme rainfall intensity as well as frequency with seasonal consistency. In particular, bias corrected GCMs failed to reproduce extreme rainfall events and corresponding discharges in July and August. This is a limitation from the statistical bias correction. As DD results were reasonable with observation, the inconsistency resulted in bias correction method can be eliminated by performing Pseudo Global Warming Down-Scaling (PGW-DS) for future climate. The identified systematic biases in DD will be corrected by statistical method and transferred to PGW-DS. This method aims to provide better science-based information of future changes of precipitation and discharges.