Estimation of phase center offset corrections for Sentinel satellites

The Copernicus Sentinel Earth observation satellites provide crucial earth observation measurements, e.g., sea surface-height. It is of highest importance that the underlying precise orbit determination (POD) of these low Earth orbiters (LEOs) is of high accuracy. The POD is based on observations from Global Navigational Satellite Systems (GNSS). All Sentinel satellites collect measurements from the Global Positioning System (GPS), whereas Sentinel-6A additionally collects measurements from the Galileo system. To achieve highly accurate POD, it is of crucial importance to have exact knowledge of the phase center position of the LEO receiver antenna for both the GPS and Galileo measurements. The phase center position is composed of the antenna reference point (ARP) and frequency-dependent phase center offsets (PCOs) and phase center variations (PCVs). It is known that the pre-launch characterization of the LEO receiver antennas is difficult and corresponding estimates are therefore less precise than those of the ARP. This makes it necessary to apply in-flight determined corrections to the initial pre-launch values of the PCO.

Previous studies have shown that there are deficits in the PCOs of the Sentinel-1 GPS antennas. For example, different estimates of empirical orbit parameters of similar satellites point to such deficits. The aim of this study is to determine corrections to the currently used PCOs of the Sentinel-1,2,3 and 6A satellites and to investigate their variability and reliability. Initial results show that non-negligible corrections result for the PCOs of the satellites studied.

The estimation of the corrections of the PCOs is performed as part of the POD process, which is performed with the Bernese GNSS software. The application of single receiver ambiguity resolution is necessary because it improves the stability of the estimated PCOs. It is of high importance that the modeling of non-gravitational forces acting on the satellite is as accurate as possible because modeling deficits may degrade the estimation of PCOs. The influence of such modeling deficits on the PCO estimation is investigated in this study. The estimation of PCO corrections can thus serve to not only get a better accuracy of observation modelling, but also to identify potential non-gravitational force modeling deficits.

Since the Sentinel-1,2,3 satellites are identical in construction in pairs (A and B), a direct comparison of the estimated corrections of the PCOs is possible. This can serve as a measure for the plausibility of the PCO correction estimation. Because the Sentinel-3 and Sentinel-6A satellites are altimetry satellites, the radial direction is of particular importance. Therefore, it is important to investigate the possible changes in radial levelling by applying corrections to the PCO. This can be done by analyzing Satellite Laser Ranging (SLR) measurements. The Sentinel-3 and Sentinel-6A satellites are equipped with SLR retroreflectors, which allows for SLR validations, which serves as a reliability test of the PCO correction estimations.