两种面向数值同化的风廓线雷达资料质量控制方法比较分析  

Comparative analysis of two data quality control methods for wind profile radar to numerical assimilation

作　　者：汪学渊 林银杰[4] 刘德强 林立峥[3] WANG Xueyuan;LIN Yinjie;LIU Deqiang;LIN Lizheng(Laboratory of Straits Meteorogy,Xiamen Meteorological Bureau,Fujian Xiamen 361012,China;Fujian Key Laboratory of Severe Weather,Fuzhou 350001,China;Support Center for Atmospheric Observation Technology,Fujian Provincial Meteorological Bureau,Fuzhou,350001,China;Nanping Meteorological Office of Fujian Province,Fujian Nanping 353000,China;Fujian Meteorological Observatory,Fuzhou 350001,China)

机构地区：[1]厦门市气象局海峡气象开放实验室,福建厦门361012 [2]福建省灾害天气重点实验室,福州350001 [3]福建省大气探测技术保障中心,福州350001 [4]福建省南平市气象局,福建南平353000 [5]福建省气象台,福州350001

出　　处：《气象科学》2022年第4期495-505,共11页Journal of the Meteorological Sciences

基　　金：福建省自然科学基金资助项目(2016J05092);海峡气象开放实验室开放课题(2019KX06)。

摘　　要：本文利用2018年2月2—11日福建9部CFL-06风廓线雷达小时风场与ECMWF模式背景再分析资料观测增量数据,首先制定了基于迭代加权最小协方差行列式(IRMCD)和双权重标准差(BSD)方法在处理观测增量数据时获得最优参数的判定指标峰偏值KS,然后对两种方法在处理风场离群值的性能和特性上展开比较分析。结果表明:总体上来说,两种方法在判定风廓线雷达风场数据的离群值都具有良好的效果,但是从波形指标、统计指标、概率密度分布、Q-Q图上看IRMCD比BSD方法能判定更多的离群值,各项指标都更接近于正态分布;从散点分布上看,IRMCD方法获得的非离群值的分布近似“椭圆形”特征,BSD方法获得的非离群值分布近似“方形”特征,同时BSD方法在判定离群值的“方形”四个角中存在一定量的误判,这是由各自机理所决定的,IRMCD方法能够同时处理二维向量并具有防止假阳性误判的机制,而BSD方法在处理二维向量时只能分别进行判定离群值;在风羽图的分析中也能看出IRMCD方法能够有效剔除与实际大气不合理的高低空风。因此,IRMCD方法更适合应用于循环数值同化的风廓线雷达水平风质量控制。Based on the background reanalysis data of nine CFL-06 wind profiler radars hourly wind fields and ECMWF model in Fujian Province from 2 to 11 February, 2018, the Iterated Reweighted Minimum Covariance Determinant(IRMCD) and Biweight Standard Deviation(BSD) were used to obtain the peak bias KS of the optimal parameter when processing the incremental observation data. Then the performance and characteristics of the two methods were compared and analyzed. Results show that: generally speaking, the two methods have good results in determining the outliers of wind field data of wind profiler radar, but from the waveform index, statistical index, probability density distribution and Q-Q diagram, IRMCD can determine more outliers than BSD method, and each index is closer to the normal distribution;from the scatter diagram, the distribution of non-outliers obtained by IRMCD method is similar to the “oval” feature. The non outlier distribution obtained by BSD method is similar to the “square” feature. At the same time, there is a certain amount of misjudgment in the four corners of BSD method, which is determined by their own mechanism.IRMCD method can deal with two-dimensional vector at the same time and has the mechanism of preventing false positive misjudgment, while BSD method can only deal with two-dimensional vector separately. It can also be seen from the analysis of the wind plume that the IRMCD method can effectively eliminate the unreasonable high and low air winds. Therefore, IRMCD method is more suitable for horizontal wind quality control of wind profile radar with cyclic numerical assimilation.

关 键 词：风廓线雷达 IRMCD BSD 质量控制 

分 类 号：P412.25[天文地球—大气科学及气象学]