Authors: Songbai Song Vijay P Singh
Publish Date: 2009/07/17
Volume: 24, Issue: 3, Pages: 425-444
Abstract
This study aims to model the joint probability distribution of periodic hydrologic data using metaelliptical copulas Monthly precipitation data from a gauging station 410120 in Texas US was used to illustrate parameter estimation and goodnessoffit for univariate drought distributions using chisquare test Kolmogorov–Smirnov test Cramervon Mises statistic AndersonDarling statistic modified weighted Watson statistic and Liao and Shimokawa statistic Pearson’s classical correlation coefficient r n Spearman’s ρ n Kendall’s τ ChiPlots and KPlots were employed to assess the dependence of drought variables Several metaelliptical copulas and GumbelHougaard AliMikhailHaq Frank and Clayton copulas were tested to determine the bestfit copula Based on the root mean square error and the Akaike information criterion metaGaussian and t copulas gave a better fit A bootstrap version based on Rosenblatt’s transformation was employed to test the goodnessoffit for metaGaussian and t copulas It was found that none of metaGaussian and t copulas considered could be rejected at the given significance level The metaGaussian copula was employed to model the dependence and these results were found satisfactory
Keywords: