Projection of changes in flash flood occurrence under climate change at tourist attractions

Abstract Flash floods cause widespread fatalities and economic losses. Climate change exacerbates the uncertainties of flash flood risk worldwide. Tourist attractions are particularly vulnerable to flash floods. In this study, precipitation-driven changes in flash flood occurrence are projected for the Nanshan Scenic Zone (NSZ), a famous scenic spot with approximately 6 million annual visitors in China. Historical hourly precipitation series (1991-2012) set a baseline, and future hourly precipitation series for 13 selected general circulation models (GCMs) and the multi-model mean (MME) during 2020-2049 and 2070-2099 under the representative concentration pathway (RCP) 4.5 emission scenario are simulated by a stochastic downscaling hourly weather generator. A cumulative event rainfall and rainfall duration (E-D) threshold is used to detect flash flood events in historical and future periods. All 13 GCMs project a unanimous increase in flash flood occurrence for both future periods. MME annual flash flood occurrence is projected to increase by 72% and 78% during 2020-2049 and 2070-2099, respectively. The range projected by 13 GCMs is 22%-111% and 39% -117%, respectively. The MME projections indicate that variations of no less than 50% occur in both future periods in March, April, July, October and November. Variation in extreme heavy precipitation (∼40%) is much greater than variation in mean precipitation (∼10%) in the NSZ. The increase in flash flood occurrence might be attributed to the increase in the total amount of extreme heavy precipitation. Flood managers and stake-holders in the NSZ should adopt adaptation and risk reduction measures since the months with the largest increase in flash flood occurrence coincide with the peak tourist season.