A Study of Rank Defect and Network Effect in Processing the CMONOC Network on Bernese

Abstract

High-precision GPS data processing on Bernese has been employed to routinely resolve daily position solutions of GPS stations in the Crustal Movement Observation Network of China (CMONOC). The rank-deficient problems of the normal equation (NEQ) system and the network effect on the frame alignment of NEQs in the processing of CMONOC data on Bernese still present difficulties. In this study, we diagnose the rank-deficient problems of the original NEQ, review the efficiency of the controlled datum removal (CDR) method in filtering out the three frame-origin-related datum contents, investigate the reliabilities of the inherited frame orientation and scale information from the fixation of the GPS satellite orbits and the Earth rotation parameters in establishing the NEQ of the CMONOC network on Bernese, and analyze the impact of the network effect on the position time series of GPS stations. Our results confirm the nonsingularity of the original NEQ and the efficiency of the CDR filtering in resolving the rank-deficient problems; show that the frame origin parameters are weakly defined and should be stripped off, while the frame orientation and scale parameters should be retained due to their insufficient redefinition from the minimal constraint (MC) implementation through inhomogeneous and asymmetrical fiducial networks; and reveal the superiority of a globally distributed fiducial network for frame alignment of the reconstructed NEQs via No-Net-Translation (NNT) MC conditions. Finally, we attribute the two apparent discontinuities in the position time series to the terrestrial reference frame (TRF) conversions of the GPS satellite orbits, and identify it as the orbit TRF effect.

Recent Crustal Deformation Based on Interpolation of GNSS Velocity in Continental China

“…The velocity field of the GNSS station is obtained by analyzing the coordinate time series of the station. The results showed that the average error in the horizontal velocity field is about 0.2~0.3 mm/yr, and the average error in the vertical velocity field is about 0.5 mm/yr [39,40].…”

Eliminating TRF-related System Errors Through Recursive TRF Realization Strategy for the CMONOC GPS Stations

“…Firstorder ionospheric delays are eliminated by means of the ionosphere-free linear combination of GPS observations, and higher-order terms are modeled as recommended in Petrie et al (2010). Satellites orbits and the ERPs are introduced from reprocessed (1999-2015) and regular (2015-2019) CODE products, respectively, to assure the consistency of TRF in the daily solution (Wu et al 2018). Meanwhile, they are recommended to be retained fixed in the final least squares estimation of the position parameters, together with phase ambiguities resolved as much as possible and geometrical correlations among GPS stations fully considered (Dach et al 2009).…”

“…Following the suggestion from Wu et al (2018) on the processing of CMONOC GPS data, we sequentially reprocess all daily NEQs. As an example, we present the eigenvalue properties of two coordinate NEQs on the day of 2014, the original one and the filtered one with frame-origin-related TRF contents fully removed, respectively.…”

“…Furthermore, both the complete geometric information of GPS networks and frameorientation/scale-related contents are fully retained in the NEQ matrix (Wu et al 2018), which reflected the superiority of the CDR algorithm on the filtering of redundant TRF contents from NEQs.…”

“…For the high-precision estimation of GPS stations' coordinates, Bernese is one the most widely used software (Dach et al 2015), which provides considerable thorough supports and services, regular updates and has rapid growth of user communities, including the GEONET (Global Navigation Satellite System Earth Observation Network System) (Sagiya 2004), the EPN (Europe Permanent Network) (Kenyeres and Bruyninx 2004) and the CMONOC (Wu et al 2018). As regards to the processing strategies on the Bernese platform, the EPN center and the Center of Orbit Determination in Europe (CODE) have proposed conventional recommendations for its application on regional geodetic network (Dach et al 2009).…”