Analysis of sub-daily polar motion derived from different estimation intervals

Abstract

Abstract This study investigates the sub-daily polar motion (PM) derived from different estimation interval solutions ranging from 5-minute to 2-hour. By analyzing a 3-year continuous time series of the PM estimates using Global Positioning System (GPS) observations, we conclude that PM should be parameterized as piecewise constant for intervals no longer than 30 minutes, while piecewise linear parameterization is more appropriate for longer intervals. The inconsistencies between the estimates and the background sub-daily PM model become more pronounced as the estimation intervals become shorter. The results demonstrate that applying continuity constraints enhances the accuracy of PM rate parameter estimation by approximately 20%. However, it is noteworthy that continuous constraints significantly modify and smooth the high-frequency content of the signal in PM. Therefore, when employing piecewise linear estimation, it is not recommended to use continuity constraints. Moreover, we find that sub-daily PM estimates are influenced by artificial non-tidal signals, primarily caused by the resonance between the Earth rotation period and the satellite revolution period. These resonance signals are more obvious as the estimation interval becomes shorter, particularly at 4.8-hour and 8-hour periods in the prograde and retrograde spectra, respectively. Finally, we implemented a sub-daily PM series with a 5-minute temporal resolution and examined the recovery of the tidal coefficients for 38 tides. Overall, the residual signal amplitudes were generally small, with most of the main ocean tides below 5 µas. The largest residual signals were observed for S1 and K1 terms, with amplitudes of 13.1 and 18.0 µas, respectively.