Analysis On Time Series Of GPS Station Coordinates In Taiwan,China

Global Navigation Satellite System(GNSS)technique have played an important role in monitoring geophysical phenomena.Meanwhile,the GNSS observations comprise of signal and noise components.Analysis of the GNSS time series can help to understand the terrain change patterns and the geophysical phenomena.Taiwan has a special geographic location on the earth and complicated topography.Precise point positioning(PPP)technique in GIPSY was used to process the observations to obtain daily position time series.After removing outliers,we marked the contant breaks and offsets and modeled and analyzed the long-term time series based on least squares.The principal component analysis(PCA)spatial filtering method is adopted to perform the filtering the time series,as well as spatio-temporal anlysis of the common-mode error was conducted.The maximum likelihood estimation is used to analyze the noise and analyze the terrain changesThe maximum likelihood estimation method was used to analyze the noise and calculate its spectral index.The results show that the noise spectral index of time series in the north-south(N)and east-west(E)directions of most stations is near-1,indicating that the noise type is about flicker noise.The PCA filtering method is used to process the time series of the GNSS sites network;according to the consistency of the principal component for all sites response,the first principal component is selected as the common mode error,the spatio-temporal distribution of the common mode error is also analyzed,and the spectrum analysis of the common mode error is performed.The results show that the noise spectrum index of the common mode error is also about-1,indicating the common mode error includes flicker noise,too.The spatial responses of the first principal components in the N,E,and U directions of each site showe a good consistency,with average spatial responses of about 66%,77%,and 68%,respectively;The noise analysis with the common mode error removed shows,the amplitude of power law noise decreases by approximately 36%,43%,and 23%in the three directions of N,E,and U respectively,and the average reduction rate of the root mean square error of the time series in N,E,and U is 33%,31%and 19%;the velocity uncertainty is reduced significantly in the U direction,reaching 28%;removing the common mode error can significantly suppress the noise,reduce the amplitude of colored noise,and effectively improve the signal-to-noise ratioWe performed linear and seasonal analysis of the movement patterns of stations in the Taiwan area.In the horizontal direction,the station movement rate at the edge of the Taiwan region is greater than that of the station in the Central Range of Taiwan.The southwestern region of Taiwan shows a clockwise rotation movement.In the vertical direction,the Central Mountain Range of Taiwan shows an obvious uplift pattern.And the maximum uplift rate exceeds 20mm/yr,the average uplift rate is 4mm/yr,meanwhile most of the surrounding areas show a subsidence pattern,especially,the vertical sinking rate is fairly distinct in western and southwestern Taiwan.Tthe maximum settlement rate exceeds 40mm/yr,the average rate of subsidence is 5mm/yr.Combined with the data of groundwater level changes in western Taiwan,it is found that the seasonal changes of groundwater level show a high degree of consistency with that of the coordinates time series in the area.