椭圆型Copulas函数在西安站干旱特征分析中的应用

Elliptical Copulas for Drought Characteristics Analysis of Xi'an Gauging Station

本文研究了干旱发生的联合概率、条件概率和重现期等干旱特征。以陕西省西安站月降水为例,应用Meta-Gaussian Copula和Student t Copula构造了干旱历时、干旱烈度和烈度峰值的联合概率分布,并进行了多变量分布拟合优质评价及拟合检验,在此基础上计算了联合分布的重现期以及2变量和3变量情形下的条件概率与条件重现期。研究表明,Meta-Gaussian Copula可以描述干旱历时、干旱烈度和烈度峰值三者的联合分布。由于多元联合分布可以考虑到多个变量之间的不同组合,能够求得不同干旱历时、干旱烈度或烈度峰值下的条件概率和条件重现期,因而能够更加全面客观地反映干旱的特征。

This paper aims to study drought characteristics such as joint（and conditional） probability distribution and return periods of drought variables.Based on monthly precipitation of Xi＇an gauging station（Shaanxi Province,China）,Meta-Gaussian Copula and Student t Copula were used to construct the joint probability distribution of drought duration,drought severity and severity peak,and goodness-of-fit test as well as evaluation was also completed.On the basis of this,return periods of multivariate joint distribution were calculated,and the bivariate as well as trivariate conditional probability and return periods were explored too.This study revealed that Meta-Gaussian Copula was suitable to build the multivariate joint distribution of drought duration,drought severity and severity peak.Since the multivariate joint distribution could consider the different combinations of multiple variables and obtain the conditional probability and return periods with different drought duration,drought severity or severity peak,it can reflect the drought characteristics more comprehensively and objectively.