多普勒天气雷达风场退模糊方法的研究

STUDY OF DATA PROCESSING OF WIND FIELDS FROM DOPPLER RADAR

文章提出了对多普勒天气雷达的风场信息进行预处理的“K 邻域频数法” ,该方法物理统计意义比较明确 ,能够有效地处理风场信息中的“噪声”问题。在此基础上 ,设计了单径向的速度扩展算法 ,并提出了“双径向 双切向”的方法 ,其效果与美国WSR 88D、NSSL和FSL的算法相当。该文还设计了一个界面十分友好的人机交互方法 ,可以达到约 90 %以上的速度扩展正确率 ,是对传统的人机交互方法的极大改进。

A 'K-Neighborhood Frequency' method for preprocessing the wind fields from Doppler radar is presented. It is relatively clear in physical significance and can effectively process the 'noises' in the wind fields. Based on it, a 'Double Radial' method is used to unfold velocity, and a good result is gained. A new person-computer interactive method for obtaining 90% or higher correction is also designed, which can improve greatly the traditional way of person-computer interaction.