Abstract
Global warming projections by global climate models (GCMs) are very helpful to investigate effects of future climate changes. However, the spatial resolutions of GCMs are not enough for detail assessment of future climate conditions and their effects in a river basin. A dynamic downscaling method by a numerical weather prediction model is applied to present and future climate in the Ara River basin in the Kanto region of Japan. Results for present climate show significant overestimation in annual precipitation around the basin, however, patterns of seasonal variation and spatial distribution are similar to the actual conditions. In future, decreasing annual precipitation is shown in one global warming projection, but insignificant variation in another. On the other hand, amplitude of inter-annual variation becomes larger in both results. Minimum annual precipitation in future are smaller than at present and frequencies of small precipitation year increase in future. These results indicate increasing negative effects on future water resources.
