New Prediction of Sediment-related Disaster Critical Rainfall Using Meteorological Model WRF

Toshihide SUGIMOTO, Toshiyuki SAKAI and Hiroshi MAKINO

A large number of sediment-related disasters have recently occurred in Japan due to record heavy rains exceeding 1,000 mm in cumulative rainfall and concentrated heavy rains equivalent to an hourly rainfall of 100 mm. These heavy rainfall events are likely to increase in frequency because of the impact of an increase in water vapor content caused by rising temperatures associated with global warming. Today, sediment disaster alert information is made public to ready people for sediment disasters. However, since calculation is based on the actually measured rainfall, announcement is generally made just before a sediment disaster occurs. There is no sufficient time left before people can leave their homes for shelter. This is one of the major problems related to the current system of sediment disaster alert information announcement. In this research, we conducted rainfall prediction based on rainfall simulation that uses numerical calculation meteorological model Weather Research and Forecasting (WRF) as a new evaluation technique that predicts a rainfall event likely to cause a sediment disaster at an early stage or two to three days in advance and made a comparative review of the simulation results with recent rainfall events that actually caused sediment disasters. Consequently the research results revealed that the technique is precise enough to clarify areas where orographic rainfall likely to cause sediment disasters will occur and suggests it may be able to provide the data qualified as sediment disaster alert information at an early stage before actual rainfall.

2018/1 394-400