Kinematic and static GPS techniques for estimating tidal displacements with application to Antarctica

For several decades relative gravimetric measurements have allowed the precise observation of harmonic signals related to Earth body and ocean tides. More recently, GPS data have been shown to be precise enough to allow the determination of antenna displacements at tidal frequencies in three dimensions. In this paper I focus on a comparison between 'kinematic' and 'static' tidal displacement estimation techniques using GPS data between similar to 1998.5 and 2003.5 from South Pole (AMUN). The GPS estimates are compared with modelled values using the TPXO.6 and FES99 numerical tide models which themselves are found to be in agreement at the similar to 1/100 mm level except for O-1 and N-2. The kinematic estimates are of lower accuracy to the static estimates and the height time series is dominated by non-tidal errors. The best resolved frequencies in the kinematic analysis are solar-related constituents, suggesting the presence of GPS systematic biases. The static analysis agrees with the model estimates, generally at the sub-mm level, with larger errors evident at S-2, K-1 and K-2 frequencies. A time-variable behaviour of K2 is demonstrated. After combination of all daily data, high correlations (0.7-0.9) are evident between north and east components of each constituent, whilst the remainder of the correlations are less than 0.06. These correlations alter with site latitude and point to the source of the correlations being related to the non-integer ambiguities in the daily GPS estimates which are known to introduce correlations between horizontal and vertical site coordinate components and also change with site latitude. Fixing carrier phase ambiguities to integers may therefore increase the precision of harmonic parameter estimates using GPS. (c) 2005 Elsevier Ltd. All rights reserved.