Performance analysis of CMIP6 models in projection of temperature and precipitation changes in the Chahnimeh area of Sistan and Baluchistan province

Changes in temperature and precipitation are one of the most important debates in the field of environmental sciences. This phenomenon is very important because of its scientific and practical dimensions, because human systems dependent on climate elements such as water, agriculture, industries and the like are designed and operate on the basis of climate stability. Therefore, predicting and knowing about temperature and precipitation changes in the coming years can be a solution to problems such as drought, sudden floods, high evaporation and environmental destruction. The civilization of several thousand years of Sistan has been completely dependent on the flow of water in the Sistan River. Currently, the only source of water supply in the region is the flow of water in the Hirmand River and the water storage sources of Chahnimeh. Considering the dry climate of the region and the rainfall of about 50 mm per year and the presence of 120-day dry winds in Sistan and the lack of suitable underground water resources in the region, the Hirmand River and the Chahnimeh are the only source of water supply in the region. As a result, investigating climate changes and predicting changes in temperature and precipitation within the limits of these devices is very important to save Sistan. In this research, 8 GCM models from CMIP6 were evaluated according to their high resolution and available meteorological data, and after correcting the bias using the LSBC method, the performance of these models was evaluated in simulating the relevant parameters. The study was conducted in the basic period of these models (1990-2014). After receiving the data, the observational and historical parameter values for each of the studied stations were extracted by preparing a in the MATLAB environment using the closest GCMs grid in the base period (1995-2014). Then the bias correction method was used to correct the data. In the following, the difference between the values of observational and historical parameters was evaluated using different indices. After validating and evaluating the accuracy of different GCM models, using the best model, temperature and precipitation changes in the future periods were predicted under three different scenarios and its changes in the future period compared to the historical period (1995-2014) Was investigated. In order to reveal the trend of changes in temperature and precipitation, non-parametric Mann-Kendall test and Sen,s slope estimator were also investigated.

The results showed that the rainfall in Zabul and Zahak stations had a decreasing trend and this decreasing trend was not significant in any of the mentioned stations. Based on this, the decreasing trend of precipitation in the studied area is the type of short-term fluctuations of water and It is aerial. The highest decrease in precipitation is related to Zabul station with a gradient of -2.4. The trend of temperature also shows an increasing trend in both meteorological stations and this increasing trend is significant at the level of 5% in Zabul station. Examining the slope of temperature changes also shows that the highest slope of temperature changes is related to Zabul station with a slope of 0.07. After bias correction, the performance and accuracy of the models were evaluated in the simulation of the studied parameters using various indicators at the Zabul syndication station as the selected station of the region. The results showed that based on the RMSE index, BCC_CSM2_MR model and then MPI-ESM1-2-LR have higher accuracy than other models for simulating precipitation. The RMSE of the mentioned models with the observed rainfall data in the studied station on a monthly scale is equal to 0.94 and 0.95, respectively. FGOALS-g3 model also has the weakest performance among the investigated models for simulating precipitation in the study area with RMSE equal to 5.6. In general, based on various indicators of accuracy, the BCC_CSM2_MR model is suitable for simulating rainfall in the study area, so that its coefficient of determination is equal to 0.97, and its MSE and MAE values are lower than other models. The results of projection the rainfall in the future period (2021-2050) based on different scenarios show that based on each scenario, the amount of rainfall will decrease in both stations under study, and its amount is on average For Zabul station it is equal to 4.1% and for Zahak station it is equal to 4.2%. Overall, based on the average scenarios studied, the amount of precipitation in the study area will decrease by 1.4% compared to the observation period. The highest and lowest precipitation changes are estimated based on SSP5-8.5 and SSP1-2.6 scenarios, respectively. According to the average scenarios studied in the future period, the temperature in the study area will increase by 1.4 degrees Celsius compared to the observation period. The most changes in terms of stations are related to the drainage station with 5.9 percent. Also, scenarios SSP5-8.5 and SSP1-2.6 show the highest and lowest temperature changes in the studied area with 1.6 and 1.2 degrees Celsius, respectively. In general, according to the results of the CMIP6 model and the straw scale of the LSBC method, the amount of precipitation in the study area will decrease and the temperature will increase. to follow The totality of these conditions can cause a decrease in the storage and supply of water resources in the Chahnimeh, as a result of which the climatic conditions of the region will also change.