固体潮观测数据的预处理

本文介绍了固体潮观测数据的平滑检验和NAKAI法检验之方法原理和具体应用。平滑检验可以有效地监测並改正观测数据中的随机错误,突跳或坏读数;经平滑改正后,可使观测数据的读数误差小于1毫米。用NAKAI(1975)提出的方法检验观测数据可以监测仪器灵敏度变化情况和数据质量好坏,根据该检验计算获得的有关参数(主要是漂移系数和方差)剔除质量不好的48小时数据组或修正某些质量不好的数据可提高调和分析结果的精度。

THE PRE-PROCESSING METHOD OF THE OBSERVATION DATA OF EARTH TIDES

This paper deals with the theory and application of the pre-processing method in the harmonic analisis of earth tides. Before using the Venedikov filters, we applied two step error detection to observation data. Forst, the data was smoothed by means of several filters. Then the occidental fault data, jump data or bad reading among the observation data were corrected, and thus the reading errors of observation, data were not more than 1 mm. Secondly, a rough least square adjustment was made on each 48 hour interval, that would be covered by a Venedikov filter {Nakai, 1975}. Then the bad groups of 48 hour interval could be detected out and a smoothing scale for each group could be given. By adopting this method the bad groups caused by anomalous drift in a short time (for example, 3-5 hours) might be corrected. Finally, the rest of bad groups, should be discarded in the computation of final results. Through the two step error detection, the accuracy of the computational results was promoted.