Resolution analysis on high-rate GPS positioning accuracy

• Main Authors: Kuang-ChiHsieh, 謝光紀
• Other Authors: Ruey-Juin Rau
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/26240791835462403767

碩士 === 國立成功大學 === 地球科學系碩博士班 === 100 === My research collects and analyzes 1-Hz GPS data from continuous GPS sites in Taiwan to discuss its feasibility in recording seismic displacements. I use both the relative positioning and the precise point positioning techniques (PPP). Furthermore, I try to obtain co-seismic displacement and seismic wave from Mw 6.4 Jiashian earthquake that occurred on 4 March 2010. For eliminating errors caused by multipath effects and common mode noises, I use modified sidereal filtering and spatial filtering techniques. My results discuss two parts, one is the positioning precision and the other is its temporal behavior during the earthquake period. For positioning precision, I could see good improvement in both horizontal and vertical components after applying modified sidereal filtering in relative positioning. Some stations even achieve a 50 % improvement. In PPP, although the improvement in precision is not as well as the relative positioning, there is still a 5 to 30 % enhancement. Because the modified sidereal filtering focus on time series repeatability in the time domain, both positioning techniques have good improvement depending on the environment around the GPS stations, which is thought to be mainly caused by multipath effects. From this result, I could also infer that multipath effects have higher influence on the vertical component than in the horizontal for the kinematic positioning. After applying modified sidereal filtering, I apply spatial filtering in the processing. In relative positioning, it still has good improvement, a 5 to 60 % in horizontal and a 17 to 64 % in vertical; but does not work well in PPP. The reason which makes this difference is that spatial filtering focuses on spatial-related errors. In relative positioning, the error would propagate from the reference station to the rover station, so the spatial filter could show good result. For the high-rate GPS behavior during the earthquake period, I take 2010 Jiashian earthquake for example. I use the high-rate GPS positioning result of station GS51 which is near Chishan fault to compare with its daily solution. Its co-seismic displacement shows smaller than the result from daily solution – 88% in relative positioning and 70% in PPP. So, I could infer that the co-seismic displacement obtained from daily solution also include the displacement caused by postseismic deformation. I also compare the waveform between strong-motion seismometer and GPS site nearby. After transforming the acceleration to the displacement record, it shows well comparison. With seismometer, high-rate GPS could also be a good source for studying seismology.