基于集合预报产品的降尺度降水预报试验

利用降水距平百分率的降尺度预报方法和1951-2008 NCEP资料及我国降水资料,建立了降水距平百分率的预报模型,基于T106L19模式的月动力延伸集合预报结果,进行了2007-2009年3 a的预报试验和效果检验.结果表明,基于集合预报产品的统计降尺度方法对降水距平百分率的预报技巧高于模式降水的预报技巧;500 hPa月平均高度场的预报技巧直接影响到降水距平百分率的预报技巧,平均环流的预报技巧越高,降水距平百分率的预报技巧越高;无论集合成员数为多少,集合预报的结果都明显优于控制预报,随着集合成员数的增多,预报技巧呈增大的趋势;我国降水具有显著的季节性和区域性,以江淮地区的降水距平百分率预报技巧最高,华南地区的预报技巧其次.

Precipitation forecasting experiments based on the products of ensemble forecast by using downscaling method

By using the downscaling forecast method of precipitation anomaly percentage, NECP/NCAR data and the observed precipitation data in China, a forecast model of the precipitation anomaly percentage was set up. Then based on monthly dynamical extended range ensemble forecast products of T106L19 model, forecast experiments for three years from 2007 to 2009 and effects test were completed. The results show that the forecast skill of precipitation anomaly percentage by downscaling method based on the ensemble forecast products is higher than the skill by T106 model. The forecast skill of monthly geopotential height field at 500 hPa would directly influence the forecast skill of precipitation anomaly percentage. The better the forecast skill of mean circulation is, the better the forecast skill of precipitation anomaly percentage is. No matter how many the number of the ensemble members may be, the forecast results of ensemble forecast are obviously better than that of control forecast. As the number of the ensemble members increases, the forecast skill takes its trend of rising. The precipitation has significant seasonal and regional features in China. The forecast skill by downscaling method of precipitation anomaly percentage in Jianghuai region is the highest compared to all the regions, and the South China region followed.