Characteristics of Coordinate Time Series of Shenzhen Continuously Operating Reference Stations

Based on the coordinate time series from Shenzhen CORS (Continuously Operating Reference Stations) during a period from 2010 to 2015, the algorithm of PCA (Principal Component Analysis) is employed in this paper to remove the CME (Common Mode Error) contained in the original series. After that, the optimal noise models of all stations’ components are determined by means of computing the MLE (Maximum Likelihood Estimation) values of selective noise combinations. Meanwhile, impacts of different surface mass loadings to station displacement are computed, including atmospheric pressure, non-tidal ocean, snow and soil moisture, which partly accounts for the existence of CME. The results of these experiments show that the SNR (Signal Noise Ratio) of the original time series can be improved by 2 or 3 times after the employment of PCA, and the filtered series shows a noise characteristic of flick noise and white noise. When the colored noise is to be considered, the velocity uncertainties is 3–9 times as large as that of white noise. Surface mass loadings as a whole, is one of the sources of CME. After the deducting the impact of CME, construction and depiction of horizontal velocity field of Shenzhen area is presented in the paper.