区域气候模式RegCM3初始和侧边界条件的敏感性分析

Sensitivity experiments on different initial and lateral boundary conditions of Regional Climate Model(RegCM3)

不同初始和侧边界条件的选取对区域气候模式的模拟结果有着直接影响。本文分别采用NNRP1和ERA40两套应用广泛的全球再分析资料作为区域气候模式RegCM3的初始场和侧边值,采用20km水平分辨率和An thes-Kuo积云对流参数化方案对海河流域1998年夏季(6—8月)降雨和温度进行了模拟。结果表明:两套资料驱动下模式都能够模拟出海河流域降雨和温度的大尺度空间分布特征。从降雨中心、高(低)温中心位置和分布范围上,使用ERA40资料要好于NNRP1资料;从不同时间(日、月、季)尺度站点降雨和温度模拟与观测值上,ERA40资料也好于NNRP1资料,但两套资料驱动下对降水极值过程模拟效果均不理想。就本次模拟而言,ERA40再分析资料的可信度更高,更适合海河流域气候变化的数值模拟研究。

Different initial and lateral boundary conditions have direct influence on the simulated results of Regional Climate Model. Using two sets of NNRP1 and ERA40 reanalysis datasets as the initial and boundary driving fields, two simulated schemes by RegCM3 with 20km grid space and the cumulus parameterization scheme of Anthes-Kuo are conducted to analyze the precipitation and temperature in the Hai River Basin in summer （June.-Aug.）, 1998. The simulated results indicate that RegCM3 driven by two datasets of NNRP1 and ERA40 can perform the large scale characteristics of precipitation and temperature in the Hai River Basin. In the aspects of location and distributed scope of precipitation and high （or low） temperature centers, the simulated results of ERA40 are better than those of NNRP1, while the simulated results of precipitation and temperature interpolated to station in daily, monthly and seasonal scales are also better than those of NNRP1. However, the simulation of extreme precipitation process is not good in two schemes. In this case, the ERA40 reanalysis datasets with much reliability is suitable for the numerical simulation of climate change in the Hai River Basin.