Population exposure to extreme precipitation under a changing climate in Eastern China

Under the global warming, the frequency of extreme precipitation has increased, and the return period of it has changed. The original extreme analysis method based on stationary theory will underestimate the risk of extreme precipitation. Based on observed hourly precipitation data during 1960 to 2019, a non-stationary frequency analysis of the annual maximum (AM) precipitation for China was conducted, then estimate the difference between stationary and non-stationary estimated return periods using Bayesian inference. After that, projected the extreme precipitation risk under different SSP-RCPs scenarios by the CMIP6 models. The results shown that the trends of 1-, 2-, 3-, 6-, 12-, 24-, 48-, 96- and 168-hr AM precipitation in China are complex. The shorter the duration, the more stations that show an upward trend. For a 20-yr to 100-yr return period of 1-hr extreme precipitation, the difference between the non-stationarity and stationarity extreme precipitation is large, and at the station with the upward trend that a stationary assumption may lead to underestimation of extreme precipitation about 32%; the average difference over 24-hr is relatively small, and the difference at station with downward trend is about -17%～-23%. The difference between the extreme precipitation return period under non-stationarity and stationarity assumption decreases with the extension of the duration, and the uncertainty increases as the return period increases in all conditions. The ensembled GCMs show that the precipitation in the 21st century show a fluctuating upward trend in China. The 100-yr return period of 24 -, 48 -, 96 - and 168-hr extreme precipitation changed differently under different scenarios in the early period (2021-2040), the middle period (2041-2060) and the later period (2081-20100). The area exposed to extreme precipitation with 1995 to 2014 100-yr return period under different scenarios varies greatly, among which SSP5-8.5 is the largest and SSP1-1.9 is the smallest. In the short, medium and long period, with the increase of extreme precipitation intensity, the exposure area is increasing. Because of the population change, the characteristics of the exposed population and the exposed area are different. In the medium period, the exposed population is also the largest as the population reaches the peak. Under the assumption of a non-stationary climate, the social-economic exposure of extreme precipitation return level and return period providing new methods and scientific information for design, decision-making, and assessing the impacts of climatic events.