关键词: 丰枯遭遇, 天然径流, 趋势诊断, Copula模型, 洞庭湖, 鄱阳湖, 三峡工程

Abstract: A variety of statistical techniques were used to detect the trend of runoff series from hydrological stations of Three Gorges Project, Dongting Lake and Poyang Lake. According to the results of multiple fitting accuracy estimation methods, the model with higher fitting accuracy was selected from the various marginal distributions and Copula joint distributions, and the wet-dry encounter probability and variation of natural runoff of Three Gorges Project and Dongting Lake and Poyang Lake were calculated and analyzed. The results indicated that the runoff showed an insignificant decreasing trend at Yichang station (Three Gorges Project) and increasing trends at Xiangtan (Dongting Lake) and Waizhou station (Poyang Lake). The wet-dry encounter probability between Three Gorges Project and Poyang Lake is little larger than that of Three Gorges Project and Dongting Lake. The wet-dry encounter probability in the entire year was greater than that in the flood season and the dry season. Synchronous wet probability of the Three Gorges Project with Dongting Lake and Poyang Lake were 42.10% and 44.71%, and synchronous dryness probability were 11.74% and 16.62%, respectively. The study can provide a theoretical reference for water supply dispatching of the Three Gorges Reservoir to alleviate the contradiction between supply and demand in Dongting Lake and Poyang Lake regions.

Key words: synchronous-asynchronous encounter, natural runoff, trend diagnosis, Copula model, Dongting Lake, Poyang Lake, Three Gorges Project