Abstract: Based on temperature data from Tashkurgan meteorological station and flood data from Kulukelangan station, we investigated the probability distribution characteristics of flood peak discharge and 1 d, 3 d, and 7 d total flood volume of Yarkant River by using Copula functions. Additionally, we explored the relationship between climate factor variations and extreme flood events through wavelet coherence analysis. Our findings indicate that the co-occurrence return period of two variables surpasses the single variable return period, which, in turn, exceeds the joint return period of two variables. Both the joint return period and the co-occurrence return period increase with the continuous rises in peak flow and flood volume, leading to a decreasing likelihood of corresponding extreme flood events. We also found a high coherence between summer daily average temperature and flood peak discharge series on an interdecadal scale. The summer daily average temperature tends to change 0.13-0.31 cycles prior to the occurrence of flood peak flow. By using Copula functions, we established a two-dimensional statistical model between summer daily average temperature and flood peak spanning from 1957 to 2010. Notably, as the return period of a single variable increases, the disparity between joint return period and co-occurrence return period of the corresponding two variables widens. As summer daily average temperature rises, the likelihood of floods in different return periods also increases. These research outcomes carry significant scientific value and offer technical support for the flood risk management and adaptive measures in the Yarkant River Basin.

Key words: extreme value of flood, climatic factors, return period, Copula function, Wavelet coherence, probability analysis, source flow area of the Yarkant River