Estimation, Validation, and Application of 30-s GNSS Clock Corrections

Abstract

High-rate precise satellite clock corrections are essential for precise point positioning (PPP) with global navigation satellite system (GNSS), especially for kinematic PPP (KPPP) for low earth orbiting satellites or moving vehicles on the ground where positioning precision of a few centimeters is demanded. To estimate high-rate clock corrections in a full network solution using zero-difference observations from a large tracking network (e.g., 100 stations) is quite time-consuming which even goes worse with increasing satellites. An efficient approach with estimation of high-rate epoch-difference clocks and a densification procedure to compute high-rate zero-difference clocks based on lower rate zero-difference clocks is developed at Xi’an Research Institute of Surveying and Mapping (XISM) to routinely produce 30-s “rapid” and “final” satellite clocks for the existing GNSS: GPS, GLONASS, BDS, and Galileo. In this paper, basic principle and data processing procedure are described in detail. The GPS clocks at XISM are compared with IGS final clocks to validate their quality and it is demonstrated that both the XISM 30-s and 5-min clocks are in essence the same in quality as clocks provided by IGS analysis centers. When the XISM clocks are used to assess the frequency stability performance of GPS satellites, a good agreement with IGS final clocks is again demonstrated,which further confirms the good quality of clock products at XISM. With the 30-s clocks, the frequency stability performance of GPS, GLONASS, BDS, and Galileo satellites is assessed for a time interval ranging from 30 s to about 15,000 s, which demonstrate a pretty good stability performance of BDS satellites for short intervals, even superior to GPS Block IIR and GLONASS-M satellites. Finally, experiments for KPPP with individual GPS, GLONASS, or BDS are conducted with the XISM 30-s and 5-min clocks to evaluate the impact of clock sampling rate. The result shows that, compared with the result with 5-min clocks, 3D repeatability with 30-s clocks is improved by about 67, 72, and 24% for GPS-, GLONASS-, and BDS KPPP, respectively, and it is interesting that when using 5-min clocks, KPPP with BDS has better repeatability performance than using GPS or GLONASS, which may benefit from the comparative good stability performance of almost the whole BDS constellation for short time interval.