CINRAD雷达数据质量控制方法初探

Study on CINRAD Radar Data Quality Control Methods

多普勒天气雷达数据质量控制是以提高短临预报产品质量和满足短临预报需要为目的。通过参考国内外雷达数据质量控制的研究成果,分析雷达数据杂波来源,对雷达数据质量控制处理方法进行了试验和探讨,结果表明:在雷达数据质量控制过程中,使用非雷达资料(云图、自记雨量和闪电数据)排除地物杂波,尤其是排除超折射杂波具有一定的优势和局限性。晴空区里使用云图特征来排除大范围超折射杂波效果较好;但有降水存在时,这种方法就受到限制或不适宜使用。较远距离孤立和离散噪声杂波采用统计"突变率"、两层仰角叠加分析和改进型中值滤波算法综合处理,逐步去除数据杂波;较近距离则采用多仰角变换方法(分段混扫)来去除数据杂波。通过分析,得出了地物回波与降水回波的识别特征,引进了包含利用"雷达对"资料、雷达相关产品资料和云图、自动雨量站等非雷达资料的数据互补理念。

Through the analysis of the radar data sources of clutter and the research results of radar data quality control at home and abroad,the processing method of radar data quality control has been tested and investigated.The results show that：（1） Doppler radar data quality control should use the transversal method to deal with different layer of data clutter,with the technical route of close distance transformation, long distance fold,algorithm to filter,and data complementation;（2） the identification feature is obtained from precipitation echo to ground clutter;（3） the isolated and scattering radar clutter in long distance is dealt with comprehensive methods such as mutation rate from statistics,overlying analysis of two-layer elevation angles,and improved median filter algorithm,to eliminate data clutter.However,the close radar clutter is dealt with an alternation method of poly-elevation angles to wipe off;（4） introduce the data complementary conception by using data of＂radar pair＂,radar production data and other ones including satellite image,automatic rainfall station.