重庆市动力气候模式产品降尺度应用检验评估分析  

Analysis and Evaluation on the Application of Downscaling Dynamical Climate Model Products in Chongqing

作　　者：李梗[1] 唐红玉[1] 叶钊[1] 

机构地区：[1]重庆市气候中心,重庆401147

出　　处：《高原山地气象研究》2009年第4期57-62,共6页Plateau and Mountain Meteorology Research

基　　金：重庆市气象局2009年科技计划项目:重庆市气候业务手册相关技术研究

摘　　要：应用多种动力气候模式降尺度解释应用技术，经过2004～2008年的回算检验和2009年以来的业务应用，结果表明其对重庆市区域的月平均气温和降水量有较好的预测能力，MOS和PP两大类方法中，分别用逐步回归、相似离度和最短距离相似等建模方法，用车比雪夫多项式，涡度，动态区域，EOF和R方程等5种不同构建因子方案得出的预测结果PS评分（中国气象局预测业务定量评分标准）显示，重庆市动力气候模式解释应用方法的预测能力超过了中国气象局的目标考核要求（降水≥60％，温度≥65％）。月降水量预测最佳方案是PP-同归-涡度和PP-回归-区域，月气温预测最佳方案是MOS-同归-涡度和MOS-回归-车氏。降水和气温预测效果最好月和最差月平均PS评分差值分别达到14．5％和21．8％，表明预测效果存在明显的月际差异。由2006年3月个例分析可见，降尺度解释应用技术对月尺度气温降水正负距平空间分布形态有较好的预测能力。Operational application of downscaling techniques of dynamical climate model （DDCM） was carried out in 2009 and evaluation was made by means of inverse calculating for 2004 - 2008. The result shows that DDCM has a better capability to predict monthly temperature/precipitation in Chongqing. Stepwise regression （SR） , resemble fitting irrelevance （RFI） and minimum distance resemble（MDR） methods were adopted in modeling for both MOS and PP skills, meanwhile the Ps evaluation of predictions （a national prediction evaluation standard） was made through 5 structure programs, i.e. Chebyshev polynomial （CP）, votex, moving region（MR）, EOF and R function （Rf）. The evaluation has proved that the predicting ability of DDCM has exceeded the National requirement. The best predicting programm of monthly precipitaion were PP - SR - votex and PP - SR - MR, and that of monthly temperature were MOS - SR - votex and MOS - SR - CP. Ps evaluation difference of precipitation/temperature between the best and worst was 14.5% / 21.8% , this means that the predicting has an obvious intermonthly discrepancy. Case analysis for March 2006 also confirmed the better predicting ability of DDCM for monthly temperature/precipitation anomaly space- distribution.

关 键 词：动力气候模式产品 解释应用 检验评估 

分 类 号：P463.6[天文地球—大气科学及气象学]