Abstract
In the field of hydrology and water resources, river planning is performed on the basis of observation data. The determination of the long-term effect of climate change for the purpose of risk management strongly depends on the performance of the climate model used. Therefore, it is necessary to consider the reproducibility of Japanese climatology simulated by the climate models. In this study, the reproducibility of the present climate simulation over Japanese land area was tested by analyzing surface meteorological elements (e.g., precipitation) using the Coupled Model Intercomparison Project phase 3 (CMIP3) multi-model dataset for the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report. The result shows that the variability of precipitation and snowfall in each climate model is larger and that some models do not represent the patterns of seasonal change at all. We filtered the climate models by using data from the Japanese 25-year Reanalysis (JRA-25) and weather observation stations as a standard, and reduced the uncertainty of the prediction.
