Abstract
In a flood-control planning in Japan, a high-water level is estimated from selected target rainfall events. To draw up adaptive flood control measures for future climate change, information about possible maximum rainfall are indispensable. Climatological or statistical information about precipitation are not enough for that purpose. In this study, a method to evaluate future variations of a particular extreme weather event is developed by combining ensemble simulation and pseudo global warming method. Perturbations for ensemble members are generated from simulation results by Lagged Average Forecasting method. The perturbations are scaled up/down and added to the original conditions, and ensemble members are prepared. Ensemble technique enables to examine whether differences between the present and future simulations are caused by global warming or chaotic behaviors from perturbation. Probability density curves from ensemble results give clearer view of future variations of a weather event.
