Real-time precise point positioning using tomographic wet refractivity fields

Abstract

The tropospheric wet delay induced by water vapor is a major error source in precise point positioning (PPP), significantly influencing the convergence time to obtain high-accuracy positioning. Thus, high-quality water vapor information is necessary to support PPP processing. This study presents the use of tomographic wet refractivity (WR) fields in PPP to examine their impacts on the positioning performance. Tests are carried out based on 1-year of 2013 global navigation satellite system (GNSS) observations (30 s sampling rate) from three stations with different altitudes in the Hong Kong GNSS network. Coordinate errors with respect to reference values at a 0.1 m level of convergence is used for the north, east, and up components, whilst an error of 0.2 m is adopted for 3D position convergence. Experimental results demonstrate that, in both static and kinematic modes, the tomography-based PPP approach outperforms empirical tropospheric models in terms of positioning accuracy and convergence time. Compared with the results based on traditional, Saastamoinen, AN (Askne and Nordis), and VMF1 (Vienna Mapping Function 1) models, 23-48% improvements of positioning accuracy, and 5-30% reductions of convergence time are achieved with the application of tomographic WR fields. When using a tomography model, about 35% of the solutions converged within 20 min, whereas only 23%, 25%, 25%, and 30% solutions converged within 20 min for the traditional, Saastamoinen, AN, and VMF1 models, respectively. Our study demonstrates the benefit to real-time PPP processing brought by additional tomographic WR fields as they can significantly improve the PPP solution and reduce the convergence time for the up component.