基于L波段探空的云区边界识别改进算法  

Improving Algorithm of Cloud Region Boundary Recognition Based on L-Band Sounding

作　　者：胡树贞[1] 陶法 王志成[1] 杨荣康[1] HU Shuzhen;TAO Fa;WANG Zhicheng;YANG Rongkang(Meteorological Observation Center,China Meteorological Administration,Beijing 100081,China)

机构地区：[1]中国气象局气象探测中心,北京100081

出　　处：《沙漠与绿洲气象》2022年第1期56-61,共6页Desert and Oasis Meteorology

基　　金：国家重点研发计划(2017YFC1501700);国家自然基金重点项目(61531019)。

摘　　要：利用气象业务中使用的L波段探空数据和毫米波云雷达观测资料,分析探空相对湿度在入云和出云时的变化规律,提出一种基于探空相对湿度阈值与梯度相结合的云区边界识别改进算法,并利用云雷达观测数据对算法识别结果进行验证。利用北京市南郊观象台2019年1—6月层状云样本验证分析,结果表明:改进算法相比相对湿度阈值法,对云区边界识别更加合理;改进算法识别结果与云雷达观测的云底高度相对误差为-1.1%,云顶高度相对误差为3.7%,较单纯利用相对湿度阈值法精度均有明显提高。Using the L-band sounding data and millimeter-wave cloud radar observation data,change law of sounding relative humidity is analyzed when it enters or leaves the cloud.An improving algorithm of cloud region boundary recognition which is based on the combination of sounding relative humidity’s threshold and gradient is proposed,and the cloud radar observation data is used to verify the algorithm recognition result.Stratiform cloud samples from January to June in 2019 from Nanjiao station of Beijing are used to verified analysis,and the results show that:compared with the relative threshold method,the improving algorithm is more reasonable in identifying cloud boundary.The relative error between the recognition results of the improving algorithm and the cloud base height observed by cloud radar is-1.1%,and the relative error of cloud top height is 3.7%.Compared with the relative humidity threshold method,the accuracy is significantly improved.

关 键 词：L波段探空 毫米波云雷达 梯度 阈值 

分 类 号：P412[天文地球—大气科学及气象学]