Research on extreme rainfall caused by climate change

Master(2016) Shinya Takegawa

Recently, it is said that precipitation distribution has been getting more and more localized, concentrated and intense by global warming. In order to avoid and mitigate the damage by these extreme rainfall, it is necessary to conduct detailed climate predictions and effective countermeasures depending on regional characteristics. In Kanae lab, supported from SI-CAT (Social Implementation Program on Climate Change Adaptation Technology), establishment of the method to make extreme rainfall scenarios have been tried for consideration and development of climate change countermeasures conducted by municipalities all over Japan.

【Analysis of the relationship between extreme rainfall and typhoon track patterns in Japan】

Extreme rainfall by typhoons can cause floods and landslide disasters, and bring down serious damage to the society. Typhoon track is one of the important factors which affect the extent of the damage. However, typhoon track data is multivariate data which consists of genesis position, route, termination position and so on, so it is difficult to understand the relationship between extreme rainfall and characteristics of typhoon tracks. In this study, the typhoons which were observed in the Northwest Pacific from 1951 to 2014 were classified into 20 track patterns (nodes) by using self-organizing maps which is suitable for pattern extraction. Then, in the observation points which were selected from each region in Japan, the rate of precipitation caused by typhoons of each node was calculated. In addition, top 3 events of daily precipitation were extracted in each observation point, and the factors of extreme rainfall such as nodes of typhoons and fronts were examined. As a result, it is found that main factors of extreme rainfall are different depending on the observation point like the combination of the particular node and front or only front. Therefore, typhoon track patterns which cause extreme rainfall are also different depending on the observation point. In the future, I will do more detailed analysis in each observation point.