摘要
做了３个试验，第一个试验只用大气环流模式（ＧＣＭ），主要考察ＧＣＭ的性能并确定其误差的区域分布。后两个为对比试验，一个试验中，将区域气候模式（ＲＣＭ）（ＮｊＵ－ＲＣＭ）的侧边界放在全球模式（Ｌ９Ｒ１５）中我们感兴趣的区域，未考虑侧边界区ＧＣＭ的误差大小，另一个试验中，ＲＣＭ的侧边界位置根据ＧＣＭ预报误差的空间分布选取，使其落在 ＧＣＭ预报误差较小的区域。 ３个试验都对１９９８年 ５、 ６、 ７月份中国区域的降水过程进行了模拟和比较。结果表明：单独使用ＧＣＭ的效果最差；当用ＧＣＭ－ＲＣＭ嵌套模式对区域气候进行预测时，ＧＣＭ侧边界值的误差对ＲＣＭ的模拟结果有显著的影响，嵌套侧边界若选择在ＧＣＭ系统性误差较小的地区，模拟或预测效果会有明显的改进。
Three experiments are conducted in this paper. The first one uses a GCM only and its purpose is to access the GCM capability of predicting the regional climate change and to determine the model error distribution. The second and the third ones are the comparative nesting experiments. One of them has a model domain that we are interested in and a nesting area without special consideration of the error distribution in the GCM, while the other has a nesting area selected according to the GCM error distribution. In all of the three experiments the precipitation processes in May, June and July of 1998 in China are simulated and comparisons of the models' results are made. It is found that the results of modeling by only use of a GCM are the worst, while the results of the third experiment are the best. Therefore, the errors of the boundary values in a GCM are of important influences on the simulated results when a regional climate model is coupled to the GCM. If the nesting area is selected in the minimum error district of the GCM, the efficiency of regional climate modeling and prediction would be much better than that without such selection of a nesting area.
出处
《大气科学》
CSCD
北大核心
2001年第4期492-502,共11页
Chinese Journal of Atmospheric Sciences
基金
国家自然科学基金资助项目49735170
国家"九五"攻关项目96-908-02-04联合资助
