EnKF循环同化相控阵雷达数据与变分雷达反演方法的对比研究:对一次台风外围龙卷个例的分析

针对2018年9月17日发生在广东省佛山市的一次台风龙卷过程,对比分析了EnKF和变分方法同化X波段相控阵雷达得到的风场结构。结果表明,EnKF方法和变分方法在同化了X波段相控阵雷达资料之后,都可以得到龙卷母体风暴的涡旋特征。相比之下,EnKF同化分析的涡旋强度更强,台风龙卷母体风暴及其周边三维风场的结构更加完整,与回波结构匹配更好。变分方法同化得到的流场不够连续,台风龙卷母体风暴的涡旋强度偏弱,弱回波区入流也明显偏弱,并不符合概念模型。而两种方法在没有同化相控阵雷达数据时,都无法产生龙卷母体风暴的流场特征。总体而言,相对于变分方法,EnKF同化系统在同化X波段相控阵雷达数据后可以产生更为合理的涡旋结构,台风龙卷母体风暴及其周边动力场结构更合理,这为以后X波段相控阵雷达的业务应用提供了思路。

Comparison between the EnKF assimilating phased array radar observation and variational-based radar synthesis method: analysis for a tornado case

In this paper, based on the Ensemble Kalman Filter (EnKF) assimilation system and the advanced variational-based technique, the three-dimensional wind of a tornadic storm occurred in Foshan, Guangdong on September 17, 2018 was examined by assimilating X-band phased array radar data. Results show that after assimilating X-band phased array radar data, both EnKF and variational-based technique can produce vortex characteristics of the tornado-parent storm. However, the vortex intensity produced by EnKF system is stronger, and the three-dimensional wind structure of tornadic storm and its surrounding area is more complete, which matches the reflectivity better. In contrast, the flow field obtained by variational-based technique is not continuous enough, and the vortex intensity is weak, which doesn''t conform conceptual model. However, neither EnKF nor variational-based method cannot produce the flow field characteristics of tornado-parent storm when the phased array radar data was not assimilated. In general, compared with the variational method, EnKF assimilation system can produce more reasonable vortex structure after assimilating X-band phased array radar data, and the structure of tornadic storm and its surrounding dynamic field is more reasonable, which provides ideas for the future practical application of X-band phased array radar.