Downscaling法在贵州冬季气温和降水预测中的应用

基于CGCM模式输出500 hPa位势高度场、NCEP/NCAR再分析500 hPa高度资料、贵州冬季降水和气温历史资料,利用降尺度法,对贵州冬季降水和气温预报的技巧和预测效果进行了预测试验和改进。结果表明,该方法从动力与统计相结合的角度,给出季尺度大气环流与局地降水、气温之间的关系,有明确的动力学背景和天气学意义。20年回算及两年回报试验证明了该关系的合理性;对贵州冬季降水的预测率约70%,而对气温的预测率为65%左右。另外,通过对气温反演方程订正后,其预测率达67%左右;在极端异常年,该方法对降水的预测率变幅不大,而对气温的预测效果影响极大。最后利用该方法对2005年贵州冬季降水和气温趋势进行了展望。

Application of a New Downscaling Model to Winter Temperature and Rainfall Prediction over Guizhou

Based on 500 hPa geo-potential height field from CGCM model output,the observed precipitation data of Guizhou in winter,and NCEP/NCAR reanalysis data(500 hPa),the prediction skills and effects of winter precipitation and temperature were studied by using the downscaling method.The results show that: The first,from the point of statistical-dynamical downscaling,the relationship between season-scale general circulation and regional rainfall or temperature was obtained,which has obvious dynamical background and weather signification.The second,20-year and 2-year reforecastes show that the relationship is reasonable.The third,to the prediction rate of precipitation of Guizhou in winter is about 70 percents,and the temperature is about 65 percent.Moreover,through after temperature inversion equation revising,its forecast rate reaches about 67%.In extremely unusual year,the prediction rate of precipitation waves small,but the temperature is great.According to this method,the precipitation and temperature trends in winter of Guizhou in 2005 was prospected finally.