下击暴流自动识别算法研究  

Research on Automatic Detection Algorithm of Downburst

作　　者：陶岚[1] 戴建华[1,2] 

机构地区：[1]上海中心气象台,上海200030 [2]南京大学大气科学学院,江苏南京210093

出　　处：《高原气象》2011年第3期784-797,共14页Plateau Meteorology

基　　金：中国气象局灾害天气短时临近预报预警业务系统建设项目;上海市科委项目(08dz0580900);2009年度公益性行业专项项目(GYHY200906003)共同资助

摘　　要：针对下击暴流反射率因子核心下降和低层强辐散的流场特征,开发了基于多普勒天气雷达资料的下击暴流自动识别算法。首先采用风暴识别和追踪算法对风暴进行分类识别,计算风暴反射率因子核心下降的高度,同时对多普勒雷达最低层径向速度资料进行处理,寻找符合低空辐散特征的一维辐散矢量;其次利用二维识别技术合并成二维辐散特征,并采用一组质量控制方法来去除非下击暴流的辐散场,从而获得下击暴流在多普勒雷达速度场中的重要特征信息;最后利用下击暴流的模拟数据以及三次静止型和移动型实际观测的多普勒雷达数据对该算法进行了测试,结果表明,该算法可以自动识别静止型和移动型的低层辐散现象明显的下击暴流,并提前6～12min警戒,对自动站网可能漏测的一些较小尺度的下击暴流提供了新的监测手段。该算法已经加载到中国气象局灾害天气短时临近预报预警业务系统(SWAN)试运行版本中进行实测。According to the descent of storm core height and strong low-level divergence characteristic in the flow field of a downburst,an Automatic Detection Algorithm of DownBurst（ADADB） using the Doppler radar data has been developed.First,the storm core height has been calculated,according to storm cell identification and tracking results of the SCIT（Storm Cell Identification and Tracking） algorithm,meanwhile the radial velocity data of the lowest elevation has been handled to search the one-dimensional segment with divergent characteristics.Then,two-dimension identification technology has been applied to combine the one-dimension segments into a two-dimension divergent feature,also a kind of quality control method has be used to exclude the divergence field caused by other factors,thus to obtain the significant signal of downburst in the Doppler radar radial velocity field.Some simulated and real radar radial velocity data are used to verify the ADADB.The verifications show that this downburst automatic detection algorithm can detect the downburst with clear low-level divergence characteristic and to make warnings 6~12 min ahead.

关 键 词：下击暴流 低层辐散 多普勒天气雷达 径向速度 

分 类 号：P406[天文地球—大气科学及气象学]