On Adequate Comparison of Antenna Phase Center Variations

Abstract

One important part for ensuring the high quality of the International GNSS Service's (IGS) products is the collection and publication of receiver - and satellite antenna phase center variations (PCV). The PCV are crucial for global and regional networks, since they introduce a global scale factor of up to 16ppb or changes in the height component with an amount of up to 10cm, respectively. Furthermore, antenna phase center variations are also important for precise orbit determination, navigation and positioning of mobile platforms, like e.g. the GOCE and GRACE gravity missions, or for the accurate Precise Point Positioning (PPP) processing. Using the EUREF Permanent Network (EPN), Baire et al. (2012) showed that individual PCV values have a significant impact on the geodetic positioning. The statements are further supported by studies of Steigenberger et al. (2013) where the impact of PCV for local-ties are analysed. Currently, there are five calibration institutions including the Institut für Erdmessung (IfE) contributing to the IGS PCV file. Different approaches like field calibrations and anechoic chamber measurements are in use. Additionally, the computation and parameterization of the PCV are completely different within the methods. Therefore, every new approach has to pass a benchmark test in order to ensure that variations of PCV values of an identical antenna obtained from different methods are as consistent as possible. Since the number of approaches to obtain these PCV values rises with the number of calibration institutions, there is the necessity for an adequate comparison concept, taking into account not only the numerical values but also stochastic information and computational issues of the determined PCVs. This is of special importance, since the majority of calibrated receiver antennas published by the IGS origin from absolute field calibrations based on the Hannover Concept, Wübbena et al. (2000). In this contribution, a concept for the adequate comparison as well as combination of PCV values will be presented. Since PCV corrections have to be applied in PPP Processing, the effect on the coordinate estimates have to be studied in detail to verify the consistent PCV modeling during the PPP processing and to quantify the impact on the coordinate domain also. Furthermore, due to the high correlation between station height, troposphere and receiver clock estimates, the impact on all estimated parameters has to be quantified. Since PPP is also frequently used for precise frequency comparison, the effect of PCV is of special interest in this field, too.