The prediction of monthly rainfall in Kermanshah Synoptic Station under the social-economic scenarios of the sixth climate change report

Introduction
Climate change has been a crucial environmental issue in recent years, with global warming, water crisis, and ecosystem changes drawing significant attention. Scientists have been developing various scenarios of greenhouse gas emissions to model future climate changes.
Prior research has focused on forecasting rainfall using the fifth climate change report, with little attention to the future socioeconomic impacts. This study addresses monthly rainfall prediction for the Kermanshah synoptic station using data from the sixth climate change report, considering various social and economic scenarios.

Methodology
The Kermanshah province, located in the western part of the country, spans 24.64 square kilometers, in Iran. It stretches between the latitudes of 33° 36' to 35° 15' north and the longitudes of 45° 24' to 48° 30' east. The Kermanshah synoptic station is situated at the coordinates of 34° 21 longitude and 47° 10 latitude.
One challenge with using AOGCM model outputs in regional climate change studies is the mismatch in spatial scale between the model's computational cells and the study area. At the same time, regional studies need data at a resolution of 50 kilometers or less to assess the effects of climate change. The current study utilized the Delta Change Factor (DCF) method to reduce the scale of GCM model data. In CMIP6, models typically feature enhanced resolution and improved dynamic processes. The simulation of future climate changes has utilized the Shared Socioeconomic Pathways (SSP) scenarios. These scenarios, employed in CMIP6, forecast global socioeconomic changes until 2100. The study utilizes the SSP5_8.5, SSP2_4.5, and SSP1_2.6 scenarios.
To assess model accuracy, RMSE (Root Mean Square Error) and NSH (Nash-Sutcliffe Efficiency)are used as indicators.

Results and Discussion
In this study, three climate models - CanESM5, MRI-ESM2-0, and MIROC6 - were utilized to forecast monthly rainfall variability in Kermanshah synoptic station for the sixth climate change report across three future periods: 2026-2050, 2051-2075, and 2076-2100 under SSP126, SSP245, and SSP585 scenarios. The comparison of the trend of monthly rainfall changes in the corrected data indicates that the MIROC6 model predicts the median for the Kermanshah region with the least error compared to the other two models.
In all scenarios, there is a decreasing trend in monthly precipitation. The lowest amount of precipitation is related to the months (JUNE-JULY-AUGUST-SEPTEMBER), while the highest amount of precipitation occurs in the months (APRIL-MARCH-NOVEMBER) across the near future, mid-term, and far future. The average rainfall in the SSP126 scenario is predicted to be higher for the MIROC6 model than for the other two models shortly (2026-2050). In the mid-term future (2051-2075), the MIROC6 model predicts less rainfall than the other two models, and in the far future (2076-2100), the CanESM5 model predicts the highest average rainfall for April. In the SSP245 scenario, shortly (2026-2050), the MRI-ESM2-0 model predicts the highest amount of rainfall for November, and the rest of the values are similar. In the mid-term future (2051-2075), the predictions of the models are almost similar, and in the far future (2076-2100), the MIROC6 model predicts the highest average rainfall for March, the CanESM5 model predicts the highest average rainfall for April, and the MRI-ESM2-0 model predicts the highest average rainfall for November. In the SSP585 scenario, shortly (2026-2050), the MRI-ESM2-0 model predicts more rainfall than the other two models. In the mid-term future (2051-2075), the CanESM5 model predicts the highest amount of rainfall for November, and in the distant future (2076-2100), the MRI-ESM2-0 model predicts the highest amount of rainfall for November, while the CanESM5 model predicts approximately the highest average predicted rainfall.
Conclusion
The most significant changes in monthly rainfall are expected to occur after bias correction in the first period (2026-2050) in the MRI-ESM2-0 model scenarios. The MIROC6 and CanESM5 models show similar predicted rainfall changes. In the second period (2051-2075), the trend of monthly rainfall changes for all three scenarios is more similar. However, the MIROC6 and CanESM5 models under the SSP126 scenario have predicted greater increases in rainfall compared to the SSP245 and SSP585 scenarios of the MRI-ESM2-0 model during this historical period. The MRI-ESM2-0, MIROC6, and CanESM5 models are suitable for use in the study area after bias correction, based on the validation index values for all three scenarios and models. Nonetheless, there are differences in the accuracy of the models for examining various climate change parameters and different climate regions. For instance, the MIROC6 model exhibits the highest accuracy for predicting monthly rainfall, while the MRI-ESM2-0 model has the lowest accuracy. The accuracy of the MRI-ESM2-0 model for predicting monthly rainfall in this study area is lower, but its accuracy for other climate change parameters and different regions may yield different results. The comparison of observational data and historical scenario model data in the study area for predicting future monthly rainfall shows that the best models for the study area are MIROC6 and CanESM5.