关键词: 洪水遭遇风险, Archimedean Copula函数, 混合von Mises分布, 长江上游, 洞庭湖

Abstract: To better understand the encountering patternof floods from the upper Yangtze River and Dongting Lake, this paper presents a fitting analysis of the probability distributions of annual maximum (AM) flood peaks, 15-day flood volumes, flood occurrence dates, and flood occurrence intervals for the two hydrological stations of Yichang and Chenglingji, based on their daily discharge records of the period 1951-2016. We use the Archimedean Copula functions to examine the encountering risk characteristics of the occurrence dates and discharges (including peak discharges and discharge time variations) of the stations’ floods. To analyze flood peak encountering risks, we formulate a joint distribution probability of the flood peaks and flood intervals of the two stations through considering their flood peak correlations and the correlations between each station’s flood peaks and occurrence dates. Finally, influence of the Three Gorges reservoir on the two stations’ floods encountering risks is analyzed to help the optimal operation of the reservoir and the flood prevention and disaster reduction in the middle and lower Yangtze River.

Key words: floods encountering risk, Archimedean Copula function, mixed von Mises distribution, upper Yangtze River, Dongting Lake