نوع مقاله : مقاله پژوهشی
موضوعات
عنوان مقاله English
نویسندگان English
The objective of this study is to evaluate the accuracy of CMIP6 models in simulating precipitation in the Lake Urmia Basin (Iran) over the past three decades, based on the Kling-Gupta Efficiency (KGE) statistical index. Data from five synoptic meteorological stations were analyzed using four AOGCM models: AIM-ESM1-2-HR, AIM-CM5-0, AIM-CSM2-MR, and EC-EARTH3-CC. The historical period considered spans from 1985 to 2014. Raw model outputs were downscaled using CMHyd software. Taylor diagrams were generated using both linear scaling and distribution mapping methods to identify the most suitable bias correction technique. Model performance was assessed using the KGE index at each station. According to the results, the highest and lowest KGE correction values were observed at Maragheh and Mahabad stations, respectively, while the maximum and minimum KGE values after calibration occurred at Tabriz and Maragheh stations. The calculations indicate that the MPI model provides the most accurate precipitation simulation across all selected stations in the Lake Urmia Basin, whereas the BCC model performs the weakest. The findings also reveal that raw model outputs contain significant errors and cannot be used directly. The linear scaling method was found to improve GCM outputs effectively. Considering the KGE index values of the MPI model (greater than 0.03 at all five stations after scaling), the model demonstrates reliable capability for assessing precipitation in the Urmia Basin.
کلیدواژهها English
