A detailed study on quantification and modeling of drought characteristics using different copula families

Abstract

This study delves into analyzing drought patterns in Baluchistan by applying copula-based bivariate probabilistic models complemented by Severity Duration Frequency (SDF) curves. The calculation of the Standardized Precipitation Index (SPI) hinges on monthly aggregate precipitation data from ten distinct sites compiled over six-month periods. After evaluating various parametric distributions, the Log-Normal distribution emerges as suitable for modeling drought severity and duration.A range of bivariate copulas is employed to simulate the characteristics of drought severity and duration, which are then compared against observed data. Remarkably, the Gumbel copula classified as an extreme value copula—outperforms its counterparts according to diverse statistical benchmarks. By utilizing the dependence function, we derive the conditional distribution of drought variables: severity and duration. These conditional distributions subsequently inform the calculation of return periods, forming the basis for constructing SDF diagrams at fixed recurrence levels across the study region. The study's finding indicates that a severe drought could occur over the region with higher return periods for a specific duration.The implications of this research are significant, showcasing the potential of copula-based joint modeling techniques to generate frequency curves for drought severity and duration. This development holds promise for effective water resource management and the formulation of strategies to mitigate the impact of drought in vulnerable regions.

Keywords

• Drought
• Copulas
• SPI
• Tail dependence
• SDF curves