Assessment of Regional Flood Risk and Its Uncertainty due to Heavy Rainfall in Low-lying Paddies

Abstract

We developed a method to evaluate regional flood risk in low-lying paddy areas and its uncertainty. First, the uncertainties of heavy rainfall characteristics were extracted from several climate scenarios obtained from GCMs, which were used in the calculation of rainfall probability distributions. Next, the shapes of the distributions were used to estimate river flood risk and the risk of damage to rice paddy fields through a flood analysis. The shapes of these distributions, compared in a time series, defined the uncertainty of the risk assessment of climate change. The results show that the mean value and peak position of the risk distributions increased, indicating that the flood risk would rise in the future. Additionally, the values of the expected shortfall (ES) for rice reduction with more than 50%, 90%, and 99% of each distribution were calculated to characterize each of them. The risk was quantitatively evaluated by comparing the ES values, and the risk increased in the future and in the higher-order RCP scenario. By using this method we can quantitatively evaluate the uncertainty of climate risk assessment. The results can provide useful information for setting specific goals, such as the scale of drainage facilities, to prevent and/or reduce damage from the heavy rainfall disasters that are expected both in the short-, the medium- and long-term future.