GPS台站时间序列分析及其地壳形变应用

发布时间：2018-08-18 13:06

【摘要】：上世纪90年代初IGS（International GPS Service for Geodynamics）的建立标志着GPS（Global Positioning System）应用于地球物理研究的开始。随着卫星定位系统测量定位精度的提高，全球分布的连续运行台站（本论文主要对IGS台站进行分析）积累了数十年的观测资料，得到了许多有价值的认识。 GPS时间序列中不仅包含着构造运动信号，也包含着非构造运动信号、季节性信号等噪声，夹杂的噪声影响GPS解的可靠性，对一些地球物理现象甚至可能做出错误的解释。如何高效、快速的剔除这些噪声，分离非构造运动信号的影响，是GPS坐标时间序列中的关键问题之一。本论文主要对100个IGS连续运行参考站6年的时间序数据进行分析。论文的前半部分主要介绍了时间序列的基本原理与分析方法，同时对空间大地测量数据后处理软件GAMIT和QOCA作了详细介绍。主要从构造信号、共模误差（CME）、地表负荷引起的台站位移、台站周年运动展开，对GPS坐标时间序列中包含的噪声进行分析。通过采用主成分分析法结合Karhunen-Loeve expansion的时空滤波对坐标序列进行滤波处理，结果表明PCA能较好的剔除坐标序列中的非构造信号，提高信号的信噪比；并通过对一个区域台站网进行分析，结果表明GPS坐标序列能较好的反映地震孕育过程（震前形变、同震形变及震后余滑）。论文的后半部分主要采用功率谱分析的方法对GPS时间序列中周期信号进行分析，，发现GPS坐标时间序列存在倍频约为1.04cpy的异常周期信号，并对其物理来源进行探讨。分析得出构造运动、共模误差、地表负荷等引起的台站位移不能对其作出相应的解释。该异常周期信号与GPS星座绕太阳运动的轨道重复周期（GPS年，约315.2天，1.040cpy）相吻合，认为该异常周期信号可能的解释是源自未模型化的GPS系统误差。
[Abstract]:The establishment of IGS (International GPS Service for Geodynamics) in the early 1990s marked the beginning of the application of GPS (Global Positioning System) in geophysical research. With the improvement of the positioning accuracy of the satellite positioning system, the continuous operation of the global distribution stations (this paper mainly analyzes the IGS stations) has accumulated decades of observation data. Many valuable understandings have been obtained. The GPS time series contains not only structural motion signals, but also non-tectonic motion signals, seasonal signals and other noises, which affect the reliability of GPS solutions. Some geophysical phenomena may even be misinterpreted. It is one of the key problems in GPS coordinate time series that how to eliminate these noises efficiently and quickly, and how to separate the influence of non-tectonic motion signals. In this paper, the time order data of 100 IGS continuous operation reference stations are analyzed. In the first half of the paper, the basic principle and analysis method of time series are introduced, and the post-processing software GAMIT and QOCA of spatial geodetic data are introduced in detail. The noise contained in the GPS coordinate time series is analyzed mainly from the construction signal, the station displacement caused by the common mode error (CME), surface load and the annual motion expansion of the station. The principal component analysis (PCA) combined with the spatio-temporal filtering of Karhunen-Loeve expansion is used to filter the coordinate sequence. The results show that PCA can eliminate the non-structural signal in the coordinate sequence and improve the signal-to-noise ratio (SNR) of the signal. Through the analysis of a regional station network, the results show that the GPS coordinate sequence can well reflect the earthquake preparation process (pre-earthquake deformation, co-seismic deformation and aftershock slip). In the second half of this paper, the periodic signal in GPS time series is analyzed by power spectrum analysis method. It is found that there exists abnormal periodic signal in GPS coordinate time series with frequency doubling about 1.04cpy, and its physical source is discussed. It is concluded that the station displacement caused by tectonic movement, common model error and surface load can not be explained. The anomalous periodic signal coincides with the repeated period of orbit around the sun of the GPS constellation (GPS year, about 315.2 days, 1.040cpy). It is considered that the possible explanation of the abnormal periodic signal is derived from the unmodeled GPS system error.
【学位授予单位】：东华理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P228.4

【参考文献】