Investigating the effect of suitable probability distribution of runoff on the characteristics and classes of hydrological drought using ERA5 model data in different climates of the country

The most important method for monitoring hydrological drought is the use of drought indices based on river flow. These indices are generally based on the assumption that the river flow volume data series follows the gamma distribution and the principle of equal probability transfer. Therefore, in this research, the aim is to investigate the suitable distribution of monthly flow data and its effect on the characteristics and classes of drought using the standardized runoff index (SRI) in different climates of the country. For this purpose, monthly data of precipitation, temperature, evaporation and transpiration and runoff were collected from 40 synoptic stations across the country during the period of 1979-2020. Then, from the reanalyzed ERA5 model of the ECMWF datasets with a spatial resolution of 0.5 x 0.5, the mentioned data and runoff during the selected period have been extracted. Statistics such as correlation coefficient (R), standardized mean square error (NRMSE) and mean skew error (MBE) are used to compare model data with observational data. In the following, the SRI index was calculated for two time scales of 3 and 12 months based on the fitted Gamma distribution and the best distribution, and drought characteristics including intensity, duration and frequency were calculated for them. The results of ERA5 model data evaluation indicate more accuracy of temperature variables, evapotranspiration potential and precipitation with average NRMSE equal to 0.09, 0.62 and 1.02 respectively. Considering the change of drought classes by changing the distribution used in the index, the results of this research show the necessity of fitting and choosing the best distribution in the calculation of hydrological drought indices.