Author(s): Fumihiko Uemura; Shigekazu Masuya; Takatoshi Yoshida; Noriaki Omura; Sho Tomura; Manabu Chiba; Taro Yamamoto; Atsushi Takeda; Tsuyoshi Hoshino; Tomohito Yamada
Linked Author(s):
Keywords: Climate change; Extreme rainfall event; Massive ensemble climate projections; Annual exceedance probability; Uncertainty
Abstract: The IPCC (Intergovernmental Panel on Climate Change) Fifth Assessment Report leaves no doubt that the climate system is warming, and extreme rainfall events might be more severe and more frequent in most mid- latitude land areas. A high likelihood is predicated for the above situation. In August 2016, three typhoons successively made landfall in Hokkaido, Japan, and heavy rains fell in various places. As a result, rivers flooded and sediment disasters occurred, mainly in eastern Hokkaido. Recently, there is the concern that the impact of such climate changes will become apparent. In light of this, flood control measures considering future climate change are required in Japan. Current flood control plans have safety levels that are based on limited observation data from the past several decades to a maximum of about 100 years. It is essential to utilize climate projections data to respond to events occurring under future climate change. In this study, we estimated the annual maximum average rainfall per basin from the output rainfall of a climate model using d4PDF, which is a massive ensemble of climate projections. Moreover, we proposed methods to evaluate the accuracy of this model and to calculate rainfall probability considering the range of uncertainty. The projection data of past climate were found to closely reproduce the observed values. In addition, 1/150 annual exceedance probability (AEP) rainfall has uncertainty: The range will be about 170mm/72hr under the current climate and about 260mm/72hr under the future climate. In the case of a GEV distribution, it became clear that the median of 1/150 AEP rainfall will be about 1.4 times the current rainfall in the future assuming the scenario of RCP8.5.
Year: 2020