一种修正的Sea Winds散射计圆中数滤波法

A modified circle median filter approach for SeaWinds scatterometer

利用最大似然法(MLE)对SeaWinds散射计数据反演得出的风矢量一般存在2—4个模糊解,而且第1模糊解构成的风场在轨道外侧区域容易出现块状模糊,所以不能像SASS和NSCAT等散射计那样,以第1模糊解初始化风场后用圆中数滤波法进行模糊去除。文章根据SeaWinds散射计第1模糊解的空间分布特性,对传统圆中数滤波法进行了修正,使之适合SeaWinds散射计的模糊去除,即在缺乏参考数据进行初始化的前提下,能够有效去除块状模糊问题。

The Maximum Likelihood Estimate （MLE） generally generates 2 4 ambiguities in processing SeaWinds scatterometer data, and block ambiguities are easy to appear in the outside area of the track for the wind field composed by the first ambiguity. So the first ambiguity can＇t be used for ambiguiy removal by the circle median filter after the initialization of the wind field in the same way as SASS and NSCAT. Based on the spatial distributing characteristics of the first ambiguity from SeaWinds, the traditional circle median filter method is modified for SeaWinds to effectively remove the block ambiguities under the circumstance of lack of any reference data to initialize the wind field.