Comparison of instrumental and empirical station timing biases for a set of ashtech Z12T GPS receivers

Station timing biases have been determined by two different methods for a set of Ashtech Z12T geodetic GPS receivers located at timing laboratories. The timing biases account for the internal delays within the geodetic GPS receiver and antenna hardware, including the combined effects at both the L1 and L2 frequencies, as well as the intra-laboratory offset to the local source of UTC(k). The direct method uses a standard Z12T unit from the Bureau International des Poids et Mesures (BIPM), which was previously calibrated using absolute methods with an estimated accuracy of ~3.5 ns. It is deployed to differentially calibrate laboratory receivers, each measurement having an estimated uncertainty of ~1.6 ns. The second, empirical method relies upon mutual observations of GPS time, as reported in BIPM´s monthly Circular T and as observed by the International GPS Service (IGS), to extend the previously determined calibration bias of each lab´s main TAI time transfer system differentially to the collocated geodetic receiver. For those stations where the geodetic system uses a frequency standard closely related to the local UTC(k) realization, we find calibration biases can be determined with repeatabilities in the range of one to a few ns. This method allows station timing biases to be monitored continuously, in near real-time, without disruption or need of any special local measurements. Results using both methods have been compared for several timing labs. We find differences of a few ns, except in two cases where there are questions related to local measurement procedures, which is roughly consistent with the overall error budgets for the calibration methods. In an operational time transfer setting, both methods can be usefully applied in complementary ways.