The performance of bias-corrected precipitation of the Coupled Model Intercomparison Project Phase 6 (CMIP6) was evaluated using gauge observations from 45 synoptic stations. The spatial distribution and anomaly of two absolute thresholds of heavy (R10mm), very heavy (R20mm), and two relative thresholds of very wet days (R95p) and extremely wet days (R99p) of precipitation were investigated for the historical period (1975–2014) and two future periods (2021–2060 and 2061–2100) in Iran. Five CMIP6 models were evaluated, including GFDL-ESM4, IPSL-CM6A-LR, MPI-ESM1-2-HR, MRI-ESM2-0, and UKESM1-0-LL by using root mean square error (RMSE), mean bias error (MBE), percent bias (PBIAS), correlation coefficient (PCC) statistics, and receiver operating characteristic (ROC) analysis. An ensemble of the models was generated by applying the independent weighted mean (IWM) method. Among the five models, maximum overestimation and minimum underestimation of precipitation amount are seen in IPSL-CM6A-LR and UKESM1-0-LL, respectively. The maximum R10mm and R20mm computed 35.10 and 18.50 days/year for the historical period, which were identified over the south of the Caspian Sea and Zagros Mountain chain. The anomaly of R10mm and R20mm in all river basins of Iran is positive, except for the four basins of Karkheh, Karun, Tashk-Bakhtegan, and Maharloo for the SSP5-8.5. Precipitation thresholds for both indices will be increasing over the coming decades on the coast of the Caspian Sea, while the index for R95p in the western and southwestern regions and R99p in the southern regions of Iran will experience a decreasing trend.

Keywords