Uncertainty Analysis in Climate Change Projection Using Bayesian Approach

The evaluation of climate change impact on hydrological cycle includes uncertainty. This study aimed to evaluate the uncertainty of climate change impact on the Zayandeh-Rud Reservoir inflow during the future 2020–2049 period. The outputs of 22 GCM models have been used under the three emission scenarios including RCP2.6, RCP4.5, and RCP8.5. The Bayesian model averaging (BMA) was used as the uncertainty analysis for weighting the outputs of 22 GCM models (precipitation and temperature) based on their ability to simulate baseline 1980–2005 period. Then the time series of equivalent precipitation and temperature were introduced to the hydrological model (i.e. IHACRES) and the output of runoff was estimated under different climate change scenarios. Results showed that different GCM models have different abilities in estimating climatic variables and the application of uncertainty analysis in climate change studies is necessary. The results showed that the winter stream flow will rise by 16 to 55% under different climate change scenarios. However, a reduction of (20–50)% is expected during spring months which complicated the regional water resource management during the following warm seasons. On the annual scale, the Zayandeh-Rud Reservoir’s Inflow will decrease (1–25)% under different emission scenarios which make the region more vulnerable to climate change than before. Therefore, adaptation strategies should be identified and some changes in Zayandeh-Rud Reservoir’s rule curve must be applied by water resource managers in the future.