Cross-calibration of the TRIG and PODRIX GNSS receivers onboard Sentinel-6A

Abstract

The Sentinel 6A (S6A) satellite ("Michael Freilich") hosts a unique complement of sensors for precise orbit determination (POD). Aside from a Doppler Orbitography and Radiopositioning Integrated by Satellite (DORIS) and a laser-retroreflector array (LRA) for satellite laser ranging (SLR), the spacecraft is equipped with three Global Navigation Satellite Systems (GNSS) receivers. These include a redundant pair of PODRIX receivers supporting GPS and Galileo tracking as well as a TRIG receiver supporting GPS tracking and radio occultation measurements. All three receivers are connected to high-performance GNSS antennas based on a patch-excited cup (PEC) design with choke rings for enhanced multipath mitigation. Simultaneous operations of individual receivers or receiver pairs of S6A enables determination of the relative antenna phase centers through ambiguity resolved differential carrier phase observations. The results demonstrate a good consistency of the PODRIX-PODRIX antenna baseline with nominal coordinates provided by the manufacturer, but systematic deviations for the TRIG-PODRIX baselines in all axes. While offsets in boresight direction may be expected due to minor differences in the antenna design and a limited availability of phase pattern calibration data, offsets of 9 mm and 14 mm w.r.t. to the design values may be noted in the direction of the longitudinal and lateral s/c axes, respectively. Other than antenna offset calibrations in single-receiver POD, the calibrations of relative antenna positions is unaffected by nongravitational force modeling uncertainties and can be obtained with good confidence from purely kinematic GNSS measurements. Concerning baseline inconsistencies in longitudinal, i.e. flight direction, timing offsets in GNSS receivers have earlier been identified as a cause of along-track position errors in GNSS receivers. By way of example, a 1 micro-second error in the latching of carrier phase measurements will cause a 7 mm along-track position error. To disentangle timing errors from geometric antenna position errors, measurements collected in limited phases of a reversed flight orientation were used. These suggest that the antenna baseline in longitudinal (+x) direction matches the design values, leaving a 1.2 micro-second inconsistency between PODRIX and TRIG receivers as the main cause of the apparent along-track baseline error. Independent validation of TRIG and PODRIX POD results through SLR observations suggest that the timing offset can largely be attributed to the TRIG receiver, while the PODRIX timing appears consistent with the SLR data. Dedicated GNSS signal simulator tests are recommended for the preflight validation of the next Sentinel-6B spacecraft to further consolidate these findings and a thorough review of spacecraft design information is encouraged to resolve the baseline inconsistency in cross-track direction.