华东地区月平均气温统计降尺度方法比较

用中国地面气象观测站的逐日气温观测资料和NCEP/NCAR再分析资料,分别使用基于多元线性回归(MLR)和3种主成分分析(PCA)的统计降尺度方法,对1959-2008年的华东地区的月平均气温分两个时段进行统计降尺度分析并加以检验,比较了不同降尺度方法的结果。结果表明:对于华东地区气温的统计降尺度预报,基于MLR的统计降尺度方法相对于3种PCA方法而言,对单站年际变化模拟方面有一定优势。PCA方法应用于统计降尺度时,预报因子的区域选择是影响统计降尺度结果的重要因素之一。对于温度进行统计降尺度分析时,预报因子中包含温度因子是非常必要的;所试验的4种降尺度方法,对各站点多年平均情况的模拟要好于对区域平均的年际变化的模拟。

Comparison of statistical downscaling methods on monthly temperature in East China

Based on the daily temperature observation data from surface meteorological observation sites and NCEP/NCAR reanalysis data, the statistical downscaling methods with Multiple Linear Regression(MLR) and three kinds of Principal Components Analysis (PCA) were used to analyze and validate the monthly temperature in two periods from 1959 to 2008 in East China, furthermore, the results of four kinds of statistical downscaling methods were compared. Results show that the statistical downscaling method based on MLR has certain advantages in the interannual variation simulation of single station in East China compared with the PCA methods. The zone of predictors is one of the main factors that influence the statistical downscaling results when PCA method was applied to statistical downscaling. It is necessary to include temperature factor when statistical downscaling was involved in the temperature. All these four kinds of statistical downscaling methods reveal more skills in the simulation of average annual values than in the zonal mean values.