Estimation and representation of regional atmospheric corrections for augmenting real-time single-frequency PPP

Abstract

Real-time single-frequency precise point positioning (PPP) can be significantly augmented by applying high-quality atmospheric corrections. In previous work, the satellite-and-station-specific slant total electron content (STEC) ionospheric corrections, derived from a regional reference network, are commonly used to augment single-frequency PPP for improving positioning accuracy and faster convergence. However, since the users are required to interpolate STEC ionospheric corrections from nearby reference stations, either duplex communication links should be established or all corrections of the reference network must be retrieved, which makes it inefficient to provide augmentation services to many users. Moreover, the regional tropospheric corrections are generally neglected in augmenting real-time single-frequency PPP. In this study, we present a method to estimate and represent tropospheric and ionospheric corrections from a regional reference network, which can be efficiently disseminated to users through a simplex communication link. First, the uncombined dual-frequency PPP, with external ionospheric constraints derived from international GNSS service predicted global ionospheric map, is used for estimating atmospheric delays with observations from a regional GNSS reference network. Then, the atmospheric delays are properly represented to facilitate real-time transmission by applying a polynomial model for the representation of zenith wet tropospheric corrections, and satellite-specific STEC maps for representing the slant ionospheric corrections. The above results in only simple communication links required to retrieve the regional atmospheric corrections for real-time single-frequency PPP augmentation. Observations from a regional network of 30 GNSS reference stations with inter-station distances of about 70 km during a 1-week-long period, including both quiet and active geomagnetic conditions, are used for generating the regional atmospheric corrections. The results indicate that the average root-mean-square errors of the obtained regional tropospheric and ionospheric corrections are better than 0.01 and 0.05 m when compared with those derived from dual-frequency uncombined PPP, respectively. The positioning accuracy of the single-frequency PPP augmented with regional atmospheric corrections is at 0.141 m horizontally and 0.206 m vertically under a 95% confidence level, a significant improvement compared to single-frequency PPP without atmospheric augmentation. The convergence time is also significantly reduced with 70.4% of the positioning sessions achieving instantaneous 3D convergence.