Improved Models of the GPS Satellite Antenna Phase- and Group-Delay Variations Using Data from Low-Earth Orbiters (Invited)

Designed to support navigation, the GPS L-band satellite antennas are significantly larger and more complex than the receiver antennas typically used in geodetic applications. The phase- and group-delay variations of the GPS satellite antennas are difficult to model, and remain among the limiting sources of error for the most demanding global geodetic problems such as reference-frame realization. We have developed techniques for estimating the GPS satellite antenna phase variations (APV) using data from low-Earth orbiters. We describe new estimates of the GPS APV based on data from the GRACE (2002-present) and TOPEX/Poseidon (T/P, 1992-2005) missions. These satellites offer a number of substantial advantages for developing APV maps. The scale (mean height) of the orbit solutions are well determined at the cm level from dynamical constraints, and there is no troposphere signal to confound interpretation of the measurements. In both cases, the multipath environment is also very favorable. The GRACE receiver antenna is a choke ring embedded in the surface of a clean spacecraft with a simple profile, while the T/P antenna is mounted on a 4-m boom above the spacecraft bus. Together, the T/P and GRACE missions provide a unique opportunity to observe and compare APV patterns from current as well as legacy GPS satellites. We compare our APV models to the International GNSS standard (based on ground data), and also present early estimates of the antenna group-delay variations for use with pseudorange data. Finally, we apply our new antenna models in realizing the terrestrial reference frame from GPS alone. Current comparisons of our GPS-based TRF (1999-2009) with ITRF2008P show 0.03 ppb/yr agreement for scale rate, and better than 1 mm/yr agreement for origin rate.