Differences between Galileo and GPS coordinate estimates from separate and combined solutions

Results from the MIT IGS20 trial processing period starting GPS weeks 2222-2230 are analyzed to assess the level of agreement between GPS and Galileo estimates of site coordinated and the phase center locations on the satellite transmission antennas.  The types of analyses presented are (1) Separate GPS and Galileo solutions (separate site coordinates) and a combined solution where coordinates match between the two systems;  (2) Satellite phase-center-offsets (PCO) values fixed to IGS20 values (fix20) and fully estimated (fre20); (3) Empirical radiation parameters (3 constant, 6 once-per-rev DYB axes, twice-per-rev Direct and four-per-rev Direct) all loosely constrained (daily) and with all but constant, B OPR and D 2PR estimated with process noise between days (set by overlaps and scatter of daily estimates) (backs) and (4) Weekly solutions with radiation parameter process noise and selected terms (week with fixed and free PCO values. We conclude (1) There are systematic differences between Galileo and GPS position estimates which are enhanced when satellite phase center offsets are estimated;  (2) Mean difference between GPS and Galileo heights are close to zero when IGS20 PCO values held fixed, but this is expected based on the scaling used to create IGS20 (3)Treatment of radiation parameters has some impact but is probably not the main contributor; (4) Some but not all antenna types with only G01 and G02 calibration have negative bias of Galileo height estimates (5) Fully calibrated antennas also have biases (but not all) and these biases seem to arise from local effects at the antenna installation. This presentation concentrated in this talk on heights, but systematic differences in horizontal components as well. Recommendation: Development of insitu absolute calibration method with repeat surveys to verify environmental change effects.