Noise analysis of common mode error in CMONOC GPS coordinate time series

Abstract

Common mode errors (CMEs) commonly exist in GPS coordinate time series, and it would affect the accuracy and reliability of the position and velocity of a reference station. The reduction of CME decreases the amplitude of seasonal variation and colored noise. This paper analyzes the noise of CME of 27 reference stations in the Crustal Movement Observation Network of China (CMONOC) using maximum likelihood estimation (MLE) method. The mean spectral index of CME is approximately −1 in N, E and U components, indicating that flicker noise may be the main colored noise in CME. Further analysis of optimal noise model shows that the main noise model of CME in CMONOC is white noise plus flicker noise and white noise plus power law noise. In horizontal component, the proportions of these two models are similar. In vertical component, over 60% of stations are behaved as white noise plus power law noise model.