Spatial analysis of drought severity, duration and frequency using different drought indices (Case study: Fars Province, Iran)

The present drought is a phenomenon that can occur in any climate, hence, due to its creeping and mysterious nature, economic losses, social effects as well as crises in agricultural, natural resources and ecosystems, its study is of great importance. Therefore, in this study, by using 9 drought indices including SPEI, SIAP, DI, SPI, PN, MCZI, CZI, RDI and ZSI, the drought was analyzed using 40 meteorological and synoptic stations in Fars Province, Iran during the last half century. In order to select the best drought index, three methods including minimum amount of precipitation, normal distribution, and correlation were used. Also, the severity, duration and frequency of droughts and their return period were determined using Run Theory (RT) method and SDF curves. Finally, after determining the best index, the drought events of the region were interpolated using ArcGIS techniques along with the simple and conventional kriging methods with spherical, exponential, and Gaussian models as well as the inverse weighted distance (IDW) method. In order to determine the most appropriate interpolation method, Cross-Validation method and MAE and MBE indices were used. The results showed that the SPI index performed as the best indicator to describe the drought. The results of RT method and SDF curves showed that by increasing time scale and return period, drought continuity and magnitude increase and as drought persisted, the severity of drought not increase at a constant rate. According to the results, the most severe and widespread droughts in the province occurred in 1970, 1993, 1999, 2007, 2014 and 2016. Also, Gaussian conventional Kriging method was the best method of drought interpolation in the study area due to its lower error rate. Therefore, by spatial monitoring and distribution of droughts, necessary measures can be taken to better deal with and manage water and natural resources.