基于Copula函数的赣江流域气象干旱特征分析

赣江流域是我国粮食主要产区之一,也是干旱灾害频发的区域。基于流域内39个站点1960～2018年的逐月降水和气温资料,计算逐月标准化降水蒸散发指数,通过游程理论提取干旱历时和烈度,分析其逐年变化趋势,对干旱历时进行边缘分布拟合前预处理,利用Copula函数计算‘或’和‘且’两种情境下不同等级干旱的重现期,分析干旱历时与烈度的时空特征。研究结果表明,赣江流域干旱历时和烈度两变量均呈减少趋势,跨季节干旱为主要类型;干旱历时的最优边缘分布函数中威布尔分布居多,Frank-Copula函数能很好地描述大部分站点的联合分布状况;重旱及以上两种重现期均小于7年,表明干旱严重且多发。

Analysis of Meteorological Drought Characteristics in Ganjiang River Basin Based on Copula Function

Ganjiang River Basin is one of the main grain-producing areas in China, and it is also a region with frequent drought disasters. Based on the monthly precipitation and temperature data of 39 stations in the basin from 1960 to 2018, the monthly standardized precipitation evapotranspiration index (SPEI) was calculated. The duration and intensity of drought were extracted by run-length theory, and the trend of drought was analyzed. The edge distribution of drought duration was pretreated. The copula function was used to calculate the recurrence periods of different drought degrees under the two scenarios of `or'' and `and''. The temporal and spatial characteristics of drought duration and intensity were analyzed. The results show that the two variables of drought in Ganjiang River Basin show a decreasing trend, and cross-seasonal drought is the main type. Weibull distribution is the most optimal distribution function for drought duration, and Frank-copula function can well describe the joint distribution of most stations. The recurrence periods of severe drought and above are less than 7 years, which indicates that drought is serious and frequent.