GNSS antenna calibration tables evaluated by means of large volume metrology

We have developed a novel method in which a pair of GNSS antennas with similar characteristics are used to evaluate hidden systematic errors in existing GNSS calibrations with the help of high-end industrial metrology equipment. We tilt the calibrated antennas out of parallel and sort the observations in individual antenna reference frames rather than epoch time. With the combined and compared measurements, we can sort out the different elevation dependent uncertainties in the GNSS observations and quantify the errors of the calibration methods. We show the extent to which the calibration method error systematically maps as troposphere and height components in the GNSS processing and in the worst case found this to be > 1 cm in the vertical when using the ionosphere-free frequency combination L3. While showing results in the presentation for the full elevation range in 5° elevation cells, we report here the 1σ uncertainties of our method for 30° elevation at ±0.38 mm on L1 and ±0.62 mm on L3 with respect to the antenna phase centers. Once uncertainties have been characterized at this level, the etalon antennas can be deployed as space geodetic anchor points at core sites without compromising existing installations.