Copula based assessment of meteorological drought characteristics: Regional investigation of Iran

Abstract

Arid and semi-arid climate of Iran has made it highly vulnerable to droughts. Comprehensive monitoring of drought characteristics is important for better understanding of drought behaviors for mitigation planning. Spatial analysis of multiple characteristics of meteorological droughts in Iran is conducted in this study using Standardized Precipitation Index (SPI) and Copula functions. Monthly rainfall data of 102 synoptic stations are used to estimate three characteristics of droughts namely Severity (S), Peak (P) and Duration (D). Eight Probability Distribution Functions (PDFs) are used to select the best fit marginal distribution of univariate drought characteristics based on Kolmogorov Smirnov and Chi squared statistics. Archimedean Copulas (Clayton, Frank, and Gumbel) are fitted to the joint S-P, S-D and P-D datasets by Maximum Pseudo Likelihood Estimator (MPLE). The cross-validation Copula Information Criterion (CIC) is used to select the best model according to goodness of fit of the Copulas. The best-fit Copula model is used for the generation of 42 return period maps of different classes of drought characteristics in order to investigate the spatial patterns of joint return period of drought characteristics. Further, the distribution of S, P, and D classes is categorized into different return periods (T) for the generation of maps to facilitate drought management. Results revealed that 1-, 2-, and 3-month mild droughts with slight picks occur mostly in the center of Iran (0 < T < 10). Almost the entire country experiences 1-month droughts with a slight peak once in 10–25 years. The 2- and 3-month duration moderate peak droughts have a return period of 25–50 years in the north. The 2- and 3-month droughts with extreme severity is also frequent in the north. The generated maps exhibit a comprehensive view of droughts over Iran which can be used for better drought management.