不同分辨率CWRF模式对中国区域气温模拟的比较研究OA北大核心CSTPCD

Based on the 30 km CWRF regional climate model,a new model was developed with a horizontal resolution further refined to 15 km.ERA-Interim atmospheric reanalysis data from the European Center for Medium Range Forecasts and OISSTv2 sea surface temperature data from the United States were used to drive two CWRF models.Compared with the observed data of China's homogenized air temperature dataset CN05.1,the simulation effect of CWRF models on 2 m-air temperature over China during 1982-2016 and its sensitivity to horizontal resolution were systematically analyzed.The results show that simply modifying the underlying surface information of the model has little effect on the simulation results without improving the model resolution,the 15 km resolution CWRF model has a better simulation effect on the temperature in the complex terrain area,and has a better performance on the spatial distribution,inter-annual variation and extreme temperature simulation.From the perspective of climate mean temperature distribution,the 15 km model reduces the cold deviation to less than 1℃in spring,autumn and winter over Southwest China and Qinghai-Tibet Plateau,and Southwest,South China and Central China in summer,where the cold deviation is more significant simulated by the 30 km CWRF model.From the perspective of inter-annual temperature variation,the simulation results of 15 km model over Central China,North China and southern Northeast China in summer and winter are better than those of 30 km model,and the correlation coefficient increases by 0.4 at the highest.In terms of the simulation of extreme events,compared with the 30 km model,the 15 km model has a great improvement in the simulation of the maximum number of days with the annual daily maximum temperature>25℃from South China to Fujian,Jiangxi,Hunan and eastern Hubei,the root mean square error decreases by 1 d,and the extreme high and low temperature in eastern Xinjiang,Northeast China,the Qinghai-Tibet Plateau and South China are also significantly improved,the root mean square error is reduced by 1℃.Therefore,higher resolution regional climate models are helpful to improve the precision simulation ability of temperature in China.