2015 年 21 巻 p. 419-424
In this study, we attempt to establish a method that can simultaneously estimate the wind speeds and basin average rainfalls for an enormous number of typhoons. We create by regression analysis an estimation equation to predict the rainfall induced by a typhoon from characteristics of the typhoon. To improve the estimation equation, we divide previous typhoons into four groups depending on the presence or absence of a front and the amount of rainfall. Since several groups don’t have enough samples, we add WRF-simulated typhoons with the tracks perturbed from original tracks of the typhoons with large rainfalls. Owing to this process, we can create estimation equations taking account of the effects of typhoons that would cause catastrophes in the future. These equations well reproduce basin average rainfalls. We apply the equations to stochastically generated typhoons and estimate the rainfall for each basin. In this way, we establish a method that can simultaneously estimate the wind speeds and basin average rainfalls of typhoons for an arbitrary return period.