基于两变量联合分布的干旱特征分析

以标准降水指数作为干旱指标,通过Copula函数构造干旱历时和干旱程度的联合分布来分析干旱的基本特征,并以汉江上游为例进行了应用研究.结果表明,由于联合分布考虑了两变量之间的各种组合,并且可以求出各干旱历时或干旱程度下的条件概率,因此,能够更全面地反映干旱的特征.

Analysis on Drought Characteristics Based on Bivariate Joint Distribution

Drought is a kind of extreme hydrologic event with multidimensional correlated variables, such as drought duration and drought, and it is difficult to reveal such correlation by analyzing univariate distribution only. Copula function theory, however, is an effective method to describe this correlation with an advantage that the marginal distribution of individual variables can be of any forms. In this paper, a bivariate joint distribution based on copula function using the standardized precipitation index is proposed and developed, a case study about the drought characteristics in the upper reaches of the Hanjiang River is carried out by regarding precipitation as the drought index, and the drought characteristics are analyzed by developing the joint distribution of drought duration and drought severity with Copula functions. Because the various combinations of the proposed bivariate joint distribution are considered, the conditional probability of drought duration and drought severity can be derived, so the practical drought characteristics can be more comprehensively revealed. The headwaters of the central line of South-to-North Water Diversion Project is located in the upper reaches of the Hanjiang River, and drought will directly impact the normal operation of the project, so the study results can be referred in the decision-making of implementing the project.