风廓线雷达资料的应用:质量评估

Application of wind profiler radar data:quality assessment

摘要:

利用FNL全球再分析资料(Final Operational Global Analysis)、探空资料对2019年6—9月位于中国华北地区20个站点共5种型号(CFL-06、GLC-24、TWP8-L、CFL-03、CLC-11-D)的边界层风廓线雷达资料进行了质量评估。结果表明:各型号雷达均具有较强的探测能力,但不同雷达在水平风资料数据获取率以及有效探测高度上差异极大。不区分天气状况时,所有型号雷达均为V风质量优于U风质量。TWP8-L雷达U风测风质量相对最佳,CFL-03雷达紧随其后,GLC-24雷达U风测风质量最差,V风质量则差异不大,U风数据使用前需进行偏差订正以及质量控制。风廓线雷达观测对于降水较为敏感,降水使各型号雷达数据获取率在底层减小,中高层增加,增幅最大达到53％,但探测能力加强并不代表测风质量增加,统计结果表明降水是造成U风平均误差以及均方根误差较高的重要原因,其中,GLC-24、CLC-11-D雷达对降水最为敏感,降水状态相较于非降水状态均方根误差增幅均达到了5.5 m/s以上,降水情况下的U风及V风资料需进行进一步质量控制才可使用。

Abstract:

A wind profiler radar is a new type of Doppler wind radar designed to measure the horizontal wind vectors, and are able to provide wind profiles at various elevations with high temporal and spatial resolution. It plays an increasingly important role in China's Meteorological business. However, the quality of wind profiler radar data varies widely considering the different types of radars deployed in China. To better utilize these data in numerical weather prediction (NWP), this study assesses the data quality from five different types of wind profiler radars (i.e., CFL-06, GLC-24, TWP8-L, CFL-03 and CLC-11-D) at 20 stations in North China from June to September 2019. Results show that each type of profiler has strong detection ability, but different profilers have great differences in data acquisition rate and effective detection height. Without considering the influence of precipitation, the data quality of V-wind is better than that of U-wind for all types of profilers. In all air conditions, the quality of u-wind data from TWP8-L profiler is relatively the best, followed by CFL-03 radar, the quality of U-wind data from GLC-24 radar is the worst. The differences in V-wind data is not noticeable, while bias correction and quality control are necessary for U-wind data. The data quality of wind profile radar is quite sensitive to precipitation, which reduces the data acquisition rates at low levels and increases the data acquisition rates at the middle and high levels, with the maximum increase of 53％. The statistical results indicate precipitation is responsible for the large mean errors and root mean square errors (RMSEs) of u wind data. GLC-24 and CLC-11-D profilers are the most sensitive to precipitation. The RMSEs are increased by 5.5 m/s in precipitation regions as compared to non-precipitation regions. The U-wind and V-wind data in precipitation regions requires further quality control procedure.