As usual the clock time difference data was produced from USNO IGS-style GIPSY solutions generated at USNO with points output every 7.5 minutes. Two-way satellite time transfer experiments (TWSTT) are currently taken taken between USNO and TUG/GRAZ on a Monday-Wednesday-Friday schedule. Single TWSTT experiments were obtained on each day and were made up of 120 1-pulse-per-second comparisons averaged together to form a single TWSTT experiment value. The TWSTT has been empirically calibrated to the BIPM Circular T. The GIPSY solution time differences have been empirically shifted to the TWSTT data by determination and removal of time biases. A rate of 9.443 (+/- 0.005) nanoseconds/day, standard error estimate = 4.357 was removed from the time data over the interval of available data, MJD 50903 - 50940. The rate was determined by linear least-squares. Figure 1 shows the trend removed time data for both the GPS carrier-phase data and TWSTT. The TWSTT experiments show one-sigma error bars. In all cases where there is GPS CP data allowing for direct comparison of the two methods GPS CP lays within 2 sigma of TWSTT. The RMS difference of the two time transfer methods TWSTT - GPS CP is 1.261.
Figure 1:
Time differences between UTC(USNO(MC2)) - UTC(TUG) via TWSTT and GPS CP
solutions.
The conclusions to be drawn are that these two methods are both very
precise and have good internal consistancy over the space of about 1 month.
TWSTT can be absolutely calibrated and would be useful for daily calibration of
GPS CP. This would require TWSTT experiments to be taken every day. A daily
bias could then be applied to the daily IGS-style GIPSY solutions.
(USNO image - gif - 12Kb)
This page is http://tycho.usno.navy.mil/gpscp_ut.html
(27 December 1998)
Go to GPS Carrier-Phase home page.
Back to Time Service Home Page.