Impact of Station GNSS Antenna Phase Centre Calibrations on Satellite Orbits and Station Coordinates: Preliminary Results

The electromagnetic phase centre of a Global Navigation Satellite System (GNSS) antenna does not coincide with the physical one and depends on the azimuth and elevation of the incoming signals. To improve processing results, the International GNSS Service has developed antenna phase centre corrections (PCC), which include models for constant and variable biases, i.e. phase centre offsets and phase centre variations. By assuming that the small differences in PCCs between antenna/radome combinations of the same type/model are negligible, so called type-mean PCCs have been generated and, for practical reasons, have been widely applied. The type-mean PCCs are averaged corrections, which are composed from several individual PCCs. Contrary to the general assumption, individual PCCs may sometimes be significantly different for antennas of the same type and it can be argued, that the use of type-mean instead of individual calibrations may degrade GNSS-derived products, such as satellite orbits and station coordinates. Furthermore, through simple geometric considerations it can be shown that inaccuracies in the PCCs may propagate into time series of GNSS solutions and may contribute some power at the orbital frequencies and their harmonics. In this study we assess the impact of the applied antenna PCCs on satellite orbits and station coordinates. As the availability of individual PCCs for GNSS stations is very limited, we analyzed a global network with stations mainly located in Europe. We used 10 years of GPS data in our processing. Despite the limitations imposed by the poor network geometry, we observed improvements in orbit overlaps at day boundaries when individual antenna PCCs were used compared to the respective type-mean solutions. Additionally, we analyzed the impact of the applied PCCs on Precise Point Positioning (PPP) station coordinates using our computed orbits. Our preliminary results showed that out of two PPP runs, the one with individual PCCs and respective orbit set reduced the noise in solutions compared to the other PPP run. We conclude that the use of individual PCCs is advantageous for derived GNSS products, as improvements are observed both in the estimated satellite orbits and station coordinate time series