GPS Block II/IIA Satellite Antenna Testing using the Automated Absolute Field Calibration with Robot

The characteristics of the GPS satellite antenna has an important impact on precise GPS positioning. The transition from relative phase center offsets and variations (PCV) to absolute PCV for the receiving antennas on the ground requires PCV for the satellite antennas. Currently, the IGS is estimating elevation dependent PCV pattern from global networks and from the ionospheric free linear combination. Absolute PCV field calibration for GPS receiver antennas has been available since 2000. The methodology was developed by Geo++ in cooperation with the Institut fur Erdmessung, Universitat Hannover starting in 1996. The absolute field calibration provides absolute phase variations of GNSS antennas completely independent from any reference antenna or station dependent effects. In 2000 NGS attempted a relative PCV calibration on a BLOCK II/IIA antenna which was used as the qualification antenna. This antenna is identical to the flight antennas. Due to the complex design and small beam width results from this test produced limited results. In 2006 NGS approached the GPS wing who sponsored the shipment of the antenna to Geo++ where the absolute PCV measurement process would be used on the BLOCK II/IIA antenna. The PCV determination proved challenging due to the size of the antenna which caused modifications and redesigns of equipment and procedures use to determine the PCV. Testing was delayed by wet year in the Hanover region of Europe Experiences and results from the testing of the BLOCK II/IIA antenna are presented which cover elevation and azimuth dependent phase variation (including mean offsets). WHY PHASE CENTER MEASUREMENTS ARE IMPORTANT GPS pseudo ranges are measured from transmitting phase center to receiver phase center. As cited by real world experiences the phase centers are not physical points. To eliminate the errors caused by variation of phase center, there is a need to describe how phase centers change with azimuth and elevation. After the launch of the first GPS BLOCK II satellite the orbit was estimated to be smaller than what was predicted. The phase center with respect to the satellite's center of mass is critical e.g. for accurate orbit determination. On the ground, it was realized that especially together with tropospheric scale parameter estimation, height errors were incurred when two different antenna types were used in differential solutions. The need for antenna calibration was served by relative field calibrations (Mader 1999), a very robust and relatively simple technique. The calibration result, however, refer to the used reference antenna model, which was defined to be a AOAD/M_T. The assumption that the reference antenna had a fixed PCV, i.e. the phase center did not change with direction (particularly elevation) was known to be incorrect. However, for shorter baselines the assumption was valid and the effects of different antenna characteristics could be corrected using these relative calibrations. For baselines long enough that the curvature of the earth caused a satellite viewed by two stations to be seen at significantly different elevation angles, the assumption of zero PCV breaks down. Early attempts to measure the absolute PCV of the reference antennas in anechoic chambers were done by Schupler et al. (1994). However, when these calibrations were used in the global solutions, scale errors of about 15 ppb resulted. The development of the absolute antenna field calibration (Wubbena et al. 1997, Menge et al. 1998, Wubbena et al. 2000) and subsequent measurements of absolute PCV produced similar scale errors. A relative field calibration of a BLOCK II/IIA antenna by Mader, Czopek (2001) analyzed the validity of phase offsets for the GPS transmitting antenna, which were based on theoretically computed offsets and was used for all satellites. Satellites passing through the zenith-pointed beam of the antenna provided by Boeing were observed. Several days of observing provided enough multiple satellite occurrences in the beam to obtain a good estimate of the L1 and L2 phase center offsets but insufficient data to compute the PCV. As suspected, the phase centers were about 70 cm closer to the earth than the values being used. This corrected offset removed much of the 15 ppb scale error. This led to further investigations of the satellite antenna characteristics based on globally distributed GPS data. Schmid and Rothacher (2003) demonstrated the estimation of elevation dependent satellite PCV together with other geodetic parameters. The determination of azimuthal variation has been presented in Schmid et al. (2005). Also low earth orbit satellites have been used to determine GPS satellite's PCV (Bar-Sever et al. 2006). With consistent absolute receiver PCV and satellite antenna offsets and PCV a better agreement with other geodetic space techniques was finally achieved. The IGS performed the transition from relative to absolute receiver antenna PCV including satellite PCV in November 2006 (Schmid et al. 2007). The current satellite PCV are limited to the ionospheric linear combination (L0) and the computations are correlated with other parameters, especially station heights and troposphere. Therefore a calibration on the ground is a completely independent approach, which has the advantages to eliminate most of the GPS error components. It is also possible to determine PCV for the L1, L2 observable and to investigate azimuthal variations. The precise knowledge of PCV for receiving and transmitting antenna has shown to be of high importance. The GPS error budget and consequently application accuracy will benefit from further insight or improvement. The current Navstar GPS Constellation Status (07-07-07, http:// gge.unb.ca./ Resources/ GPS Constellation Status. txt ) lists 16 active BLOCK II/IIA satellites. The latest IGS antenna PCV correction file igs05.atx (ftp:// igscb.jpl.nasa.gov/ igscb/ station/ general/ ) contains PCV for 29 BLOCK II (20) and BLOCK IIA (9) satellites derived from GPS data starting back in 1994 (Schmid et al. 2007).