A popular method for climate change prediction are General Circulation Models which are at coarse spatial resolution and must be downscaled. In this study, observed data of temperature, precipitation and potential evapotranspiration over a base period under two emission scenarios in three time intervals used to implement SDSM as a downscaling tool for HadCM3 model output. From another standpoint, MPEH5 model predicts data under three emission scenarios for three future periods. Results indicated that all parameters would increase comparing to the base period. Predictions for all periods under all emission scenarios indicated an increasing trend for all parameters, although it is predicted almost constant precipitation trend change for the future. According to predictions by both models, the greatest increase has been estimated for 2080s under A2 scenario. In SDSM model, the greatest increases in mean monthly temperature would be respectively 6.9, 4.5, 6.2 °C for July and for potential evapotranspiration would be in June by 1.08 mm per day, which are predicted in the 2080s under A2 scenario. For precipitation, the greatest reduction under the same conditions, would be in May by 0.9 mm per day. In LARS-WG model, the greatest increase in mean monthly temperature in the studied station predicted respectively by 5.5, 5.5, 5.6 °C for August. The greatest reduction in precipitation, would be in February (by 0.88 mm per day). The future uncertainties result of predicted parameters in both models and various scenarios show that, uncertainty of the predictions increase towards the end of the century.
