Based on variation characters of sea surface height and parameters of ICESat return pulse waveform,a method is proposed for estimating Arctic sea ice freeboard with ICESat observation.This proposed method is compared with the common ‘lowest-level’method.The results show that the lowest-level method is vulnerable to gross errors and present systematic errors in spatial distribution of freeboard. But the proposed method is more rel iable than the lowest-level method.The lowest-level method is improved by the employ of an elevation standard deviation threshold and combined with the proposed method to estimate sea ice freeboard for six ICESat campaigns between 2005 and 2006 to review the seasonal variabi l ity.The results show that the abnormal cl imate in the summer of 2005 have a profound impact on arctic sea ice,which caused the decrease of freeboard and changed the constituent of sea ice as wel l.

利用ICESat数据计算了海冰覆盖的北冰洋海域海面高，获得的北极平均海面高模型ICESat MSS与DTU13MSS模型存在-10.0±9.6 cm的差异。分别根据EGM2008和ArcGP大地水准面模型计算的平均海面动力地形在总体趋势上完全一致，准确地展现了北冰洋的主要洋流特征。根据ICESat计算的海面高分析了北极海平面的季节性变化，与验潮站数据的对比表明不同ICESat任务期之间存在系统差异。

A new filtering integrated with GM( 1, 1) model and traditional Kalman filtering was introduced, which was called gray Kalman filtering. The paper also analyzed the application of this new filtering in settlement monito-ring and the length of the GM( 1, 1) model data. This method was proved better than the traditional Kalman Filte-ring by simulation.

The application of mean spacing method and simple-point displacement component method in the analysis of stability of settlement monitoring network was introduced applying to a project example.The validity,advantages and disadvantages of these two methods were also discussed according to the result of the analysis.