中文摘要：

多普勒测风激光雷达是目前探测晴空风场精细结构的重要遥感手段之一。中国第一部车载多普勒激光雷达已经在中国气象局气象探测中心进行准业务化运行。综述了国内外非相干多普勒激光雷达技术发展概况，介绍了多普勒激光雷达风廓线探测方法，分析了测风数据误差。并以2008年9月24-28日，神舟七号飞船回收气象保障试验期间测风数据为例，对多普勒激光雷达风廓线探测进行评估分析。多普勒激光雷达测风数据与神七飞船主着陆场附近的探空资料和北京数值预报同化资料进行对比的结果表明，多普勒激光雷达的测风数据时空分辨率高，是一种可信的高精度大气风场探测手段。为了消除多普勒激光雷达的测风数据较大的高频随机误差，采用切比雪夫多项式正交展开滤波方法得到的低通分量，更好地揭示了在垂直方向上较大空间尺度波动的大气风场信息。利用这一方法对多普勒测风激光雷达测风数据进行质量控制，用于较大尺度的数值预报模式，可以提高初值场同化的效果。而滤波分析分离出的高通分量，尽管主要是高频随机误差，但其中振幅较大的高频分量也许揭示了大气风场更小尺度的波动，这需要有更多的观测事实和更进一步的分析来判定。

英文摘要：

The Doppler wind lidar is one of the important instruments to detect the structure of the wind field. The first mobile Doppler wind lidar in China has been used for the quasi-operational detection in CMA atmospheric observation center. In this work, the general development of the Doppler lidar technology in the world is summarized, the detecting methods of the wind profile of the Doppler lidar are introduced, and the error of the wind data is analyzed. The wind profiles detected by the Doppler lidar are evaluated and analyzed during its participation of this mobile lidar in a meteorological experiment for the Shenzhou 7 manned spaceship mission in September 24--28, 2008. The wind profile data detected by the Doppler lidar was compared with that of the radio-sounding system near the landing site and calculated with a numerical weather prediction （NWP） model at Beijing Meteorological Bureau. The result shows that the wind profile data detected by the lidar has high spatial and temporal resolution and the Doppler lidar is one of the reliable and highly accurate wind detecting methods. In order to filter the high frequency random error in the wind data detected by the Doppler lidar, the filter of the Chebyshev polynomial is used to get the low-pass components, and large-scale wind information is better revealed and can be used in large-scale NWP models for improving the assimilation effect. Although most of the high-pass components is the high frequency random error, the small-scale fluctuation of the wind is revealed by the large-amplitude high-frequency components, and more observational facts and analysis are needed to confirm this viewpoint.