降尺度方法中的初始资料处理的研究  

THE STUDY OF INITIAL DATA PREPROCESSING IN THE DOWNSCALING METHOD

作　　者：芦新平[1] 陈星[1] 苗曼倩[1] 季劲钧[2] 

机构地区：[1]南京大学大气科学系,南京210093 [2]中国科学院大气物理研究所,北京100029

出　　处：《气象科学》2002年第2期139-148,共10页Journal of the Meteorological Sciences

摘　　要：从 GCM模式预测结果获取区域尺度特征的气候变化的统计方法被称为 Downscaling方法 ,它主要是通过区域气候尺度的预报量与 GCM模式输出或大尺度地面观测资料建立统计模式。在建立模式之初 ,研究各种尺度资料的匹配是至关重要的问题。本文在不同的站点数量、区域网格分辨率和差值影响半径的条件下 ,对资料处理方法的可行性和结果的可靠性作了分析探讨。结果说明 :一个大尺度网格箱内用 2 0个站点资料代替 4 0个站点的资料是可行的 ,并且确定了最佳插值方案 :当次网格分辨率取 1°× 1°时 ,插值影响半径应取 1;当次网格分辨率取 0 .5°× 0 .5°时 ,插值影响半径应取为 2。本文还对温度场、降水场及其对应的 EOF场进行了比较试验 ;并做了温度场、降水场与同步海平面气压场 (SL P)The statistical method used to get the feature of regional scale climatic change from the GCM output,which is called Downscaling method,has been paid much attention to by a lot of researchers.The method is mainly established based on the regional climatic scale forecasting variables and GCM output or macro-scale surface observation data. At the beginning of the model establishing,the most important work is to match the different data in different scales.In this paper,with the different number of stations,local grid resolutions and radii of influence in the interpolation schemes,we analyse and test the feasibility of the data preprocessing method and the reliability of the results.The results show that it is feasible to use 20-station data instead of 40-station data in a macro-scale grid box.We also have got two best interpolation schemes:when the sub-grid resolution is 1°×1°,the radii of influence is 1,and when the sub-grid resolution is 0.5°×0.5°,the radii of influence should be 2. Additionally,the comparative experiments have been performed between temperature field,precipitation field and their correspondence EOF field.The single linear correlation analysis between temperature field,precipitation field and the sea-level pressure (SLP) field have also made to find significant impact factors in the method.

关 键 词：降尺度(Downscaling) 插值影响半径 经验正交函数(EOF) 

分 类 号：P435[天文地球—大气科学及气象学]