Sensitivity experiments in determining urban effects on precipitation patterns

The present study discusses uncertainties of the standard sensitivity experiment on a convective rainfall event over the Tokyo metropolitan area surrounded by complex terrain. Similar to previous studies, we obtain the positive impact of urbanization from the sensitivity experiment. However, negative or neutral impacts have also been found when the experimental design is marginally changed. Our results indicate that the standard sensitivity experiment is not reliable due to the strong Chaos influence in rainfall simulation. Second, an approach similar to a climate simulation is attempted to decrease the characteristic of the initial value problem and instead strengthen the characteristic of the boundary value problem. Sensitivity experiments based on this method gives some indication of the possibilities of urban impact on the precipitation, although further research should be needed.