Basically, drought is one of the most important and challenging climatic phenomena caused by a long-term water shortage. This phenomenon is known as one of the costly natural disasters whose nature is limited. In fact, a critical issue in water resources management is developing a suitable method for monitoring and predicting droughts; which is, determining the exact time of their onset and duration. Unfortunately, due to the lack of a precise scientific definition for droughts and the complexity of the causes of them, the development of a comprehensive and completely reliable method for eva‎luating droughts through statistical or physical analysis is not easily achievable. In this research, droughts in Fars province were eva‎luated using 6 different drought indices, including SPI, modified SPI, JDI, SPEI, EDDI and CJDI. To calculate drought indices, weather data, including monthly precipitation, temperature, relative humidity, wind speed, radiation, and the number of sunny hours from 9 synoptic stations located in Fars province, in the statistical period of 1991-2020, were used. After analysing these drought indices, it was concluded that the modified SPI index fixes the defects of the SPI index to a large extent, especially its limitation in considering the seasonal variability of rainfall. JDI index can also be used by researchers for more accurate analysis and assesment of droughts, due to considering the joint distribution of 1 to 12 month time scales in an individual timescale. SPEI, EDDI and CJDI drought indices are among the indices that have the ability to monitor droughts based on various weather parameters besides precipitation. Since the climate of Fars province is dry and semi-arid, the air temperature is one of the most important atmospheric parameters that causes drought in these regions. SPEI and CJDI indices simultaneously use precipitation and potential evapotranspiration to eva‎luate droughts. As a consequence, these indices do not highlight the droughts specified by the SPI index which are not significant at the 5% statistical level. Also, these indicators indicate the drought conditions of the region in each month, according to the long-term average of rainfall and temperature of the region. The EDDI drought index is also a single-variable index that uses potential evapotranspiration to eva‎luate droughts. It is notable that this drought index uses a physical method like Penman-Monteith (FAO 56 or ASCE) to calculate potential evapotranspiration; consequently, relative humidity, wind speed and radiation, have a key role in the calculation of evaporation-transpiration, as well as, temperature. That is why the results of the EDDI index are much more accurate than the SPI index. Also, one of the most important results of this study is that the JDI and CJDI indices can accurately determine the onset time of droughts and their continuity, in addition to having the ability to scientifically describe the drought situation. In the next step, two characteristics of drought including intensity and duration of drought were defined based on CJDI index. In this research, seven families of two-dimensional copula functions were compared to create the joint distribution of drought severity and duration. As a result, the Galambos copula function was selected as the best copula function based on error analysis and tail dependence coefficient for 9 stations in the region. After creating the joint distribution based on the Galambos copula function, some probabilistic characteristics related to drought such as joint and conditional probabilities as well as joint, conditional and Kendall return periods were calculated.

سعيد اسلاميان

رسول ميرعباسي

سعيد سلطاني كوپايي، عليرضا گوهري