利用中位数方法对风廓线雷达数据质量控制的研究

A STUDY ON THE QUALITY CONTROL OF WIND PROFILER DATA BY USING THE MEDIAN METHOD

风廓线雷达的测风原理是假定风场是均匀的,而在实际探测中往往会出现不满足均匀风场的假定情况（特别是在有降水系统影响情况下）,进而影响雷达测风精度。采用求取中位数法对风廓线雷达观测资料进行质量控制,并结合气球探空资料,对比分析质量控制效果。结果显示：（1）受风廓线雷达的探测原理以及系统灵敏度的影响,当大气较干时,雷达的探测高度较低,当相对湿度增加时,探测高度随之增大;（2）在晴空条件下,求取中位数法得到的结果以及雷达原始探测结果相对于气球探空观测到的结果随高度变化的趋势一致,误差较小,利用求取中位数法得到的结果对原始结果有一定的改善效果;（3）当有降水发生时,求取中位数法得到的结果以及雷达原始探测结果相对于探空资料误差增大,但利用中位数法得到的结果相对于原始探测资料有明显的改善;（4）相比于低层和高层的误差分布,1~2 km高度上利用中位数方法所得到的风向、风速误差最小,对数据质量的改善效果最好。

Wind field is assumed homogenous when detected by using wind profilers, and the results will be affected when the assumption is not satisfied. In this paper, wind profiler data is quality-controlled by using a median method（MOBS data）, and the new results are compared with the original data（ROBS data）. Firstly, it is indicated that the detection height is higher when the air is wetter. Secondly, under the clear air condition, MOBS data and ROBS data have smaller error with balloon-observed wind data, and the MOBS data has some improvement over the ROBS data. Thirdly, when precipitation happens, MOBS data can get a considerable higher accuracy with fewer errors than the ROBS data. Finally, compared with the error of the low and high layer, the MOBS data on the height of 1~2 km has the best improvement.