我国东南夏季干旱指数的ECC预测方法

Ensemble Canonical Correlation Forecasting of Summer Drought Index over Southeastern China

摘要:

根据中国东南夏季气温和降水显著负相关的特点,构造合适的干旱指数。选取美国NCEP/NCAR再分析月平均海表温度场、北半球500hPa高度场、亚欧大陆表面温度场作为预测因子,取由国家气象中心整理的中国160站月平均降水和温度资料计算出的东南部干旱指数作为预测对象,采用集合典型相关分析方法(ensemble canonical correlation,ECC)预测东南夏季干旱指数。预测与实况之间的空间相关性和时间相关性均表明,该方法优于单因子场典型相关分析(canonical correlationa nalysis,CCA)预测方法,其中采用超级集合平均法又比等权集合平均法具有更高的预测技巧。

Abstract:

Based on the fact that there exists significant negative correlation between summer temperature and precipitation over the southeastern China,a drought index is appropriately constituted.Choosing monthly mean sea-surface temperature,500hPa height in the Northern Hemisphere,surface temperature over the Eurasian continent from the NCEP/NCAR reanalyses as predictors,the summer drought index calculated from the monthly mean precipitation and temperature of 160 stations in southeastern China collected by Beijing Meteorological Center,are predicted with ensemble canonical correlation(ECC) method.The results show that the ECC analysis is superior to the single factor canonical correlation analysis(CCA) in the spatial/temporal correlations between observed and predicted summer drought index fields,and the superensemble mean method has higher overall skill than the equally weighted ensemble mean method.