| description abstract | This study was undertaken to investigate the effects of a large city on heavy rainfall in Tokyo, Japan, based on an ensemble simulation with a large number of members. An ensemble simulation (24 members) of eight brief heavy rainfall events that occurred from 1999 to 2007 was performed. The ensemble simulation was performed using five objective analysis datasets [Japanese 25-yr Reanalysis (JRA-25), Regional Analysis (RANAL), NCEP Final Operational Global Analysis (NCEP-FNL), NCEP?Department of Energy Global Reanalysis 2 (NCEP/DOE-R2), and Global Analysis (GANAL)]. Land-use distributions of two types were prepared for numerical simulations: actual land use and virtual land use, in which all urban land use was converted to vegetation. Each member was simulated using actual land use and virtual land use. The effects of the urban area were then assessed by comparing the results of these simulations. Results indicate no large differences in the wind systems of the Kanto plain (roughly 100 km ? 100 km), where Tokyo is located, even if cities were converted entirely to vegetation. The influence of cities on wind systems of this scale was negligible. However, changes of wind convergence were found leeward of the urban area, which increased the horizontal vapor flux there. The precipitation also increased there due to the urban effects, despite the decrease in vertical vapor flux from the land surface. It is concluded that urban effects for Tokyo alter the wind characteristics leeward of the urban area and develop wind convergence and rainfall there. | |
