民航京沪穗数值预报系统在广州的本地化试验

Localization Test of the Beijing Shanghai Guangzhou Civil Aviation Numerical Forecast System over Guangdong Area

为了调整优化基于WRF模式的民航京沪穗数值预报系统在广州本地的预报效果，使用3组不同的物理参数化方案和资料同化方案组合，对发生在2011年10月13日～14日广东地区的暴雨过程进行模拟。降水预报结果显示不同物理参数化方案和资料同化方案对降水预报有较大的影响，使用香港城市大学大气研究实验室实时预报系统推荐的参数化方案的降水预报好于加拿大温哥华地区业务运行的参数化方案，使用香港的方案，不同化自动站资料的预报效果好于同化自动站资料。而环流形势场、相对湿度场、水汽通量场和CAPE指数场对不同参数化方案的敏感性要小于降水场，另外还分析了系统连续15天预报结果。最终结果表明，不论是降水场还是形势场，使用香港城市大学的方案并且不同化自动站资料的评分优于其他2种方案，可以作为广州本地业务方案使用。

In order to optimize the forecasting performance of the WRF based Beijing Shanghai Guangzhou civil aviation numerical forecast system in Guangzhou area, tested three different combinations of physics parameterization scheme and data assimilation scheme, conduct numerical simulation to a heavy rainfall event over Guangdong area during October 13 - 14, 2011. The results show that, different combination physics parameterization scheme and data assimilation scheme have great effect to the precipitation field. The result of using the scheme provided by city university of Hong Kong is better than using the scheme provided by Vancouver, Canada（both AWS data are not assimi- lated）. Both using the Hong Kong＇ s scheme, the method of not assimilating AWS data has better result than assimilating AWS data. But for the circulation field, the relative humidity field, the water vapor flux field and the CAPE index field, they are less sensitive to different combinations than the precipitation field. Besides, analyze continuous 15 days forecast results of the system, the results show that, whether for the precipitation field or the situation field, using the Hong Kong＇ s scheme with AWS data are not assimilated gets higher score than the other two combinations, so this scheme is suggested.