Improved estimation of ocean tide loading displacements using multi-GNSS kinematic and static precise point positioning

Abstract Multi-GNSS solutions are typically considered to improve GPS-only ocean tide loading (OTL) displacements, as they effectively mitigate the constraints imposed by the single-system GPS orbital configuration on resolving various tidal constituents. Most of such solutions, however, are ambigui...
Ausführliche Beschreibung

Abstract Multi-GNSS solutions are typically considered to improve GPS-only ocean tide loading (OTL) displacements, as they effectively mitigate the constraints imposed by the single-system GPS orbital configuration on resolving various tidal constituents. Most of such solutions, however, are ambiguity-float solutions, particularly affecting the solar tidal constituents (K2, S2, K1, P1) for which the ambiguity resolution (AR) is critical. We comprehensively evaluate the performance of OTL displacements derived from the single-system and multi-GNSS solutions by processing 2.5 years of GPS, GLONASS and Galileo observations from 49 global GNSS stations using kinematic precise point positioning (PPP) with undifferenced AR. Our results show that Galileo improves over GPS by 0.2–0.5 mm for solar tidal constituents (excluding the S2 up component) compared to FES2014b model predictions. GPS + Galileo outperforms GPS, Galileo, GPS + GLONASS and GLONASS + Galileo for most tidal constituents, except for K2 where Galileo and GLONASS + Galileo hold a slight advantage. Ambiguity-float GLONASS limits improvements in results after combining with GPS or Galileo. Despite this, GPS + GLONASS + Galileo with the largest number of visible satellites estimates the best vertical OTL displacements. Since short-term (a few hours) static PPP can provide position time series with sufficient accuracy and temporal resolution while rarely used for OTL estimation, we also compare the results from kinematic and static PPP. Indeed we find that the static PPP with a suitable processing session of 1–2 h can estimate OTL displacements (excluding K2) better than the kinematic PPP. Although the static PPP of 3–4 h is too long to capture M2, N2 and S2 tidal variation without aliasing, the proposed amplitude correction can well correct these tidal constituents. Finally, we report that the residual OTL displacements are generally larger at low latitudes (30°S–30°N) than at mid/high latitudes (30°–90°N/S), possibly associated with S2 atmospheric tide loading effects and higher positioning noise at low latitudes. Ausführliche Beschreibung