Three-dimensional risk analysis of hydro-meteorological drought using multivariate nonlinear index.

Applying combined nonlinear indices is a useful approach in assessing drought condition. This study investigates a combined multivariate index (Joint Deficit Hydro-meteorological Index: JDHMI) in monitoring hydro-meteorological drought to determine whether the index could simultaneously reflect the complicated nonlinear behavior of precipitation and runoff in a watershed. The JDMI (JDHI) index is obtained on the basis of the joint marginal distribution of SPImod1,..., SPImod12 (SRImod1,..., SRImod12) using the empirical copula. The JDHMI then is computed by combining the JDMI and JDHI indices using a suitable theoretical copula. Furthermore, drought characteristics (severity-duration-magnitude) are extracted from the JDHMI and trivariate conditional return periods that are obtained from three-dimensional copulas in four different scenarios. Results of this study indicate that (1) multivariate copulas effectively reflect the complicated and nonlinear relationship between drought variables; (2) comparison between univariate and multivariate drought indices indicated that the JDHMI is slightly more sensitive to the historical events; and (3) spatial distribution of drought hazard is illustrated using conditional return period obtained in four different scenarios in Bandar-Sedij and Kole-Mehran watershed. Additionally, the conditional probabilities provide effective information for forecasting drought conditions. In this regards, on the basis of several specific values of S-D-M, it is estimated that the central and eastern parts of the region will experience frequent hydro-meteorological drought over next 18 years. Finally, due to the climatic changing in the recent years, findings of this study could be useful in reducing drought effects on the natural resources and also help decision-makers in developing water resources. [ABSTRACT FROM AUTHOR]