Influence of subdaily model for polar motion on the estimated GPS satellite orbits.

In this contribution, it is shown that GPS orbits are able to absorb some diurnal signals in polar motion. The arising implications for the influence of the subdaily pole model on GPS solutions are discussed. Two signals in polar motion can be absorbed by GPS orbits: a retrograde signal with a period of a sidereal day (23 h 56 min 4 s) and a prograde signal with a period matching the revolution period of the GPS satellites in the terrestrial reference frame (23 h 55 min 56 s). We show that the retrograde signal contributes to the absolute orientation of the orbital planes in space and the prograde signal, due to coincidence of its period with the period of revolution of the GPS satellites, contributes to the position of the geocenter for each individual satellite. It is known from previous studies that there are systematic differences between orbital parameters from GPS solutions computed with different subdaily pole models. We show in this paper that this behavior can be explained by the absorption effects in 1-day GPS orbits. Diurnal signals cannot be spectrally separated over a time interval of 1  day. Adjustment of any diurnal prograde or retrograde signal to a subdaily pole time series given by a subdaily model over 24 h will lead to an estimated signal with a nonzero amplitude. Thus, any subdaily pole model used in the processing of space geodetic observations contains a part which corresponds numerically to the discussed prograde signal and a part which corresponds to the retrograde diurnal signal. Different pole models show different amplitudes of the diurnal signals which will be absorbed by the orbits. As a result, GPS orbits computed with different subdaily pole models have systematically different orientation and position in space. Using 1-day GPS solutions over a time span of 13 years (1994-2007), we show that the systematic variations in orbit position and orientation caused by individual tidal terms in polar motion can be well predicted and explained by the suggested mechanism. [ABSTRACT FROM AUTHOR]