<p>Removal of the Common Mode Error (CME) is very important for the investigation of Global Navigation Satellite Systems (GNSS) technique error and the estimation of accurate GNSS velocity field for geodynamic applications. The commonly used spatiotemporal filtering methods cannot accommodate missing data, or they have high computational complexity when dealing with incomplete data. This research presents the Expectation-Maximization Principal Component Analysis (EMPCA) to estimate and extract CME from the incomplete GNSS position time series. The EMPCA method utilizes an Expectation-Maximization iterative algorithm to search each principal subspace, which allows extracting a few eigenvectors and eigenvalues without covariance matrix and eigenvalue decomposition computation. Moreover, it could straightforwardly handle the missing data by Maximum Likelihood Estimation (MLE) at each iteration. To evaluate the performance of the EMPCA algorithm for extracting CME, 44 continuous GNSS stations located in Southern California have been selected here. Compared to previous approaches, EMPCA could achieve better performance using less computational time and exhibit slightly lower CME relative errors when more missing data exists. Since the first Principal Component (PC) extracted by EMPCA is remarkably larger than the other components, and its corresponding spatial response presents nearly uniform distribution, we only use the first PC and its eigenvector to reconstruct the CME for each station. After filtering out CME, the interstation correlation coefficients are significantly reduced from 0.46, 0.49, 0.42 to 0.18, 0.17, 0.13 for the North, East, and Up (NEU) components, respectively. The Root Mean Square (RMS) values of the residual time series and the colored noise amplitudes for the NEU components are also greatly suppressed, with an average reduction of 25.9%, 27.4%, 23.3% for the former, and 49.7%, 53.9%, and 48.9% for the latter. Moreover, the velocity estimates are more reliable and precise after removing CME, with an average uncertainty reduction of 52.3%, 57.5%, and 50.8% for the NEU components, respectively. All these results indicate that the EMPCA method is an alternative and more efficient way to extract CME from regional GNSS position time series in the presence of missing data. Further work is still required to consider the effect of formal errors on the CME extraction during the EMPCA implementation.</p>
Probabilistic principal component analysis with expectation maximization (PPCA-EM) facilitates volume classification and estimates the missing data
