高頻GPS連續站動態定位時間序列之精度分析

Precision analysis of kinematic position time series from high-rate continuous GPS observations

本研究利用GPS連續站觀測資料進行動態定位處理，並利用恆星日濾波法與空間濾波法改進坐標時間序列之重複性，期而促進GPS在短時間監測位移之效能。研究成果發現，1HzGPS觀測資料經過恆星日濾波法以及空間濾波法修正之後，可有效降低標準差至水平方向約0.8-1.0cm，垂直方向約1.8-4.6cm。本文再以2010年3月4日之甲仙地震為例，利用GPS取得地震時間段之同震與地震波波形，並與積分後之強震儀資料進行比對。結果可知，甲仙地震造成測站CISH、MLON與SHWA有東西向1.0cm之同震位移量。相較於強震儀紀錄探測地震波形，GPS與強震儀位移獲得之地震波大致相符，但取樣頻率為每秒1筆之GPS位移仍不足獲得完整正確之波形。 Modified sidereal filtering and Spatial filtering methods are used to improve the precision of kinematic position time series generated from continuous GPS stations. The results show the precision of GPS kinematic position time series after filtering are 0.8-1.0 cm and 1.8-4.6 cm in horizontal and vertical components, respectively. The co-seismic displacements in station CISH、MLON and SHWA from Jiashian earthquake is 1.0 cm in EW direction. The seismic waveform derived from GPS is consistent with integral seismic displacement, but the GPS waveform has aliasing due to lower sampling rate of measurements.

Modified sidereal filtering and Spatial filtering methods are used to improve the precision of kinematic position time series generated from continuous GPS stations. The results show the precision of GPS kinematic position time series after filtering are 0.8-1.0 cm and 1.8-4.6 cm in horizontal and vertical components, respectively. The co-seismic displacements in station CISH、MLON and SHWA from Jiashian earthquake is 1.0 cm in EW direction. The seismic waveform derived from GPS is consistent with integral seismic displacement, but the GPS waveform has aliasing due to lower sampling rate of measurements.