Characteristic Analysis Of Regional GPS Coordinate Time Series

In the past two decades,the observation accuracy of space geodetic technology represented by GPS has been continuously improved,and a large number of high-precision coordinate time series data have been accumulated,which provides a valuable data foundation for geodesy and geodynamics research.Especially in the field of regional crustal deformation analysis,strong earthquake displacement field detection,global plate division and rebound after the ice period and other fields play an important role.However,since the GPS coordinate time series contains various types of "signal" and "noise" at the same time,and the signal types are various,the noise source is complicated,and the expressions of signals and noise on the horizontal and vertical components are quite different.How to extract reliable "signals" from the original time series containing noise is a difficult and hot topic in current research.Based on the GPS coordinate time series data of 46 IGS stations in Europe from 2000 to 2016,this paper studies the related theories,models and methods of GPS coordinate time series analysis,including the Earth reference frame theory,GPS coordinate time series preprocessing,Seasonal signal analysis and analysis of trend items and noise characteristics.The main research contents of this paper are as follows:1.Analyze the problems of gross error detection and rejection,missing data interpolation and interrupt detection in GPS coordinate time series,and verify the data by using simulated data and measured data respectively.The results show that MSSA-IQR can detect the gross error in the sequence more accurately.When the missing data accounts for 16.6% of the entire time series,MSSA can achieve better precision for the interpolation of missing data.In the aspect of interrupt detection,the sigseg program is used to detect and repair the mutation items in the GPS position time series,and verify the data through the measured data.Effectiveness.2.The Multi-channel Singular Spectrum Analysis is used to analyze the seasonal signals of GPS coordinate time series,and compared with the least squares estimation and singular spectrum analysis methods.The environmental load and common mode error are applied to the regional network GPS coordinate time series.The results show that MSSA is better than SSA in seasonal signal extraction;in terms of environmental load deduction,direct deduction of environmental load will have a certain degree of influence on the velocity estimation of vertical component and RMS.3.The noise model theory is introduced.The expressions of covariance and power spectral density of several common noise models are derived,and the long-term motion trend of the base station is extracted.The motion law of the benchmark station is discussed,and all the research areas are derived.The GPS speed field of the site.