Quantifying the flood coincidence likelihood between Huai River and its tributaries considering the nonstationarity

Abstract

Study region: The Huai River Basin, China. Study focus: This study chooses the daily flow data of Huai River and its tributaries, Shi River and Hong River during 1959–2016. The co-occurrence of floods in different tributaries of the catchment is assessed by analyzing the flood peak and flood timing at three different gauging stations located in the river network. What’s more, the parameters(estimated from the 30-year scale time windows) of the marginal and joint distributions are assumed stationary and nonstationary respectively to explore how the trend of coincidence probability(CP) is effected by them. New hydrological insights for the region: The results show that, over the period analyzed, the most probable time of flood co-occurrence over the gauge stations in the study area tends to move backwards from 10 July to about 20 July. The probability of flood co-occurrence at the Bantai and Jiangjiaji gauging stations, which singularly experience an increase in flood peak, increases from 9.25×10−9(the 1st window) to 1.17×10−5(the 25th window) when the return period of flood is 20 year, while in the same condition, the CP of Bantai and Wangjiaba decreased from 6.16×10−6 to 2.48×10−7.

Keywords

• Flood coincidence
• Copula
• Nonstationary
• Trend