东亚夏季风模式跨季预测的EOF-相似误差订正  

Empirical Orthogonal Function-analogue Correction of Extra-seasonal Dynamical Prediction of East-Asian Summer Monsoon

作　　者：程娅蓓 任宏利[2] 谭桂容[1] 

机构地区：[1]南京信息工程大学气象灾害教育部重点实验室/气候与环境变化国际合作联合实验室/气象灾害预报预警与评估协同创新中心,南京210044 [2]中国气象局国家气候中心气候研究开放实验室,北京100081

出　　处：《应用气象学报》2016年第3期285-292,共8页Journal of Applied Meteorological Science

基　　金：公益性行业(气象)科研专项(GYHY201406022);国家重点基础研究发展计划(2015CB453203);国家自然科学基金项目(41375062)

摘　　要：利用国家气候中心第2代季节气候预测模式BCC_CSM1.1(m)的1991—2010年每年2月起报的历史回算资料集,考察模式对于5个夏季风指数的预测能力,并通过发展基于经验正交函数分解与相似分析的EOF-相似误差订正方法,对5个夏季风指数的模式预测进行再修正。交叉检验和独立样本检验结果表明:该模式对1991—2010年东亚夏季风指数与西北太平洋夏季风指数预测技巧较高;EOF-相似误差订正方法适用于模式预测技巧较低的指数,这些指数经订正后预测效果均有不同程度改进,而预测技巧较高的夏季风指数经订正后改进效果不明显;在交叉检验中,线性部分订正多优于非线性部分订正效果,而对于独立试报的年份,非线性部分订正多优于线性部分订正效果,显示出良好的应用前景。In terms of the dataset of the second generation climate prediction model of Beijing Climate Center, BCC_CSM1. l（m） hindcasts in February from 1991 to 2010, the model is assessed on its performances in predicting five monsoon indices including the East Asian summer monsoon index（EASMI）, the western North Pacific summer monsoon index（WNPSMI）, the East Asian monsoon index（EAMI）, the Indian summer monsoon index（ISMI） and the East Asian summer monsoon intensity index（EASMII）. A correc- tion method based on the empirical orthogonal function（EOF） analysis and the analogue analysis, called the empirical orthogonal function-analogue correction of errors（EOF-ACE） is used to correct the modes which are poorly predicted by the model. Through EOF analysis, the coordinated variation of the five mon- soon indices can be reflect by the obtained modes. In addition, according to the principle of ensemble fore- cast, three similarity indices are used to select the similarity years. The correction process is divided into two parts, which are linear component correction and non-linear component correction. Assessments indicate that the EASMI and the WNPSMI are predicted well by BCC CSM1. l（m）. However, the prediction result of EAMI, ISMI and EASMII is poor. Through EOF-ACE correction, the model prediction skill of the monsoon indices originally unreasonable are improved, but it＇s unsatisfactory for the rest indices. The linear component correction mostly shows higher skill than non-linear component correction in cross-validation. However, for independent validation, the effect of non-linear component correction is mostly better than linear component correction, which displays application prospects. However, the EOF-ACE has barely effects on indices whose deviations are originally small, and corre- lation coefficients of some summer monsoon indices between observations and predictions are still not sig nificant through correction. It is necessary to make further analysis on the inter-annual var

关 键 词：夏季风指数 动力预测 相似分析 统计订正 

分 类 号：P457.5[天文地球—大气科学及气象学]