| چکیده انگلیسی مقاله |
Abstract
Introduction: The most important problem of the planet in this century is the increase of global temperature and change in climate variables due to industrialization of societies and increase of greenhouse gases. therefore, it is very important to investigate the trend of temperature and change in climate variables on a global and regional scale. So far, several general circulation models (GCMs) have been designed to predict the future climate, but due to the non-optimal use of the output of these models because of limited spatial separation at the local scale, different and new methods have been developed to use the output of these models at the regional and local scale.
Materials and methods: In this study, in order to investigate the changes of climatic variables in Gedarchay Catchment located in the northwest of Iran, the SDSM Downscaling Model has been used. First, the efficiency of this model was evaluated for climatic variables, then the mentioned variables were predicted until 2100. Calibration and validation of SDSM model was performed using observational data of Mahabad synoptic station and NCEP data. Also، to evaluate the model, the statistics of correlation coefficient, mean absolute error and root mean square error were used. And after ensuring the efficiency of the model, the outputs of the CanESM2 and CanESM5 models in the periods of 2031-2050 and 2081-2100 in Gedarchay Catchment under RCP 2.6, 4.5, 8.5 and SSP1-2.6, 2-4.5 and 5-8.5 scenarios were downscaled by SDSM model.
Results: The accuracy of SDSM model based on mean absolute error statistic, using CanESM2 outputs was obtained as 1.645, 0.029 and 0.031 for precipitation, maximum temperature and minimum temperature, respectively and using CanESM5 outputs, was obtained 0.73, 1.10 and 1.89, respectively. The correlation coefficient using CanESM2 model for precipitation, maximum temperature, and minimum temperature is 0.998, 0.999, and 0.999, and using CanESM2 model is 0.999, 0.993 and 0.971, respectively. The root mean square error was also obtained using the CanESM2 model as 2.240, 0.043 and 0.045 and using the CanESM5 model as 0.89, 1.49 and 2.07 respectively. Based on the obtained results, the average maximum temperature in the period of 2031-2050 and under the RCP scenario increased by 0.93℃, but it will remain constant in the period of 2081-2100. While under the SSP scenario, it will increase by 1.24℃ in the period 2031-2050 and 0.35℃ in the period 2081-2100. The increase of average minimum temperature under the RCP scenario will be 0.27 and 0.28℃, and under the SSP scenario, it will be 0.46 and 0.43℃ in the first and second periods, respectively. The increase in precipitation will be 0.59 and 0.38 mm in two periods under the RCP scenario and 2.15 and 1.64 mm in the two periods under the SSP scenario.
Conclusion: The results of evaluating the accuracy of the SDSM model in predicting precipitation, maximum temperature and minimum temperature based on R, MAE and RMSE statistics, indicate a great agreement between the predicted values and the base period. According to the obtained results, an increase in precipitation and minimum temperature for the near and far future, as well as an increase in maximum temperature in the near future and its stability in the far future were observed. |