目前的研究已明确了在GPS时序结果中存在着所谓共模误差。对于水平向通过"欧拉-滤波"法可对观测数据水平位置的共模误差进行估计,但对垂向共模误差的估计却无能为力。因此,本文利用"均值-滤波"方法对垂向共模误差进行了识别与估计,结果表明:GPS连续站垂向时序观测资料同样包含着非常明显的共模误差;综合利用"均值-滤波"方法可以对共模误差做出较有效的估计;以天津地区11个测站的计算结果为例,测站的综合中误差、共模中误差、测站自身中误差大小的统计比例为1.6:1.3:1.0。通过对观测值实施共模误差修正可明显地提高信噪比。
Present research has suggested that there exist common-mode errors in GPS time sequence results. As to the horizontal direction, the common-mode errors of observation data can be estimated through the method of ' Euler-Fihering' , however, this method is powerlessness as to the estimation of common-mode errors in the vertical direction. Accordingly, the paper made use of the method of ' Mean-Filtering' to identify and estimate common-mode errors in the vertical direction, and the results suggested : time sequence observation data of the vertical direction of GPS continuous stations also included very obvious common-mode errors; valid estimation of the common-mode errors could be achieved by comprehensive using ' Mean-Filtering' method. Taking results of the 11 stations in Tianjin area as example, the statistical proportion of mean square errors of combined, common-mode and the station was 1.6:1.3:1.0. Signal-to-noise ratio could be obviously improved through the common - mode error correction.