Uncertainty Prediction of Seasonal Variations of Rainfall in the Qom-Kahak Using Different Climate Models and Hybrid Developed Model

In this research, the effect of climate change on rainfall is investigated using five AOGCM climatic models (HadCM3, CCSR-NIES, CSIRO-MK2, CGCM2 and GFDL R30) under emission scenarios A2 and B2 and Hybrid developed model resulting from these models based on Bayesian approach, in order to account for the uncertainties in the Qom-Kahak aquifer. Data were downscaled for current (2001-2017) and future (2054-2069) periods. Then, it is found that the HadCM3 and CCSR-NIES compared to other models have better performance using criteria of efficiency. By calculating climate change scenarios and taking into account uncertainties, seasonal variations of future rainfall were compared with observed rainfall. The trend of seasonal variations of rainfall simulated by HadCM3 and CCSR-NIES and Hybrid developed model will be negative under the A2 in the spring and summer. The highest decrease in rainfall was by -45.14% relative to the observed period, which was related to the HadCM3 in spring. Also, the results of prediction of models under the B2 indicate that the trend of rainfall changes in winter will be partly positive for most models and this trend is decreasing in the spring and summer. The highest decrease of rainfall under the B2 is relative to CCSR-NIES in summer (-22.61% compared to observed period). The hybrid model, which is a combination of different climatic models, predicts the negative trend for the rainfall changes of all seasons under the A2, and the highest rainfall reduction in this model is related to summer by -29.44% compared to the observed period.