Abstract
Traditionally, in GPS precise point positioning (PPP), ionosphere-free linear combinations of dual-frequency carrier-phase and pseudorange measurements are used. Unfortunately, with these linear combinations only the first-order ionospheric delay term is removed and higher order ionospheric delay terms are usually not taken into account. Such residual error components may deteriorate the PPP solution and slow down the convergence time. In this paper the second-order ionospheric delay term is modeled and it is shown that its effect on GPS satellite orbit varies from 2.3 mm to 23.8 mm in the radial direction, 3.6 mm to 18.8 mm in the alongtrack direction and 2 mm to 16.3 mm in the crosstrack direction. In addition, GPS satellite clock corrections showed a difference of up to 0.067 ns.
To examine the effect of the second-order ionospheric delay on the PPP solution, new data sets from several IGS stations were processed using a modified version of GPSPace software, which was originally developed by Natural Resources Canada (NRCan). The modified GPSPace accepts the second-order ionospheric corrections as well as the newly developed US National Oceanic and Atmospheric Administration (NOAA) tropospheric signal delay model (NOAATrop). The estimated precise satellite orbit and clock corrections, with secondorder ionospheric delay accounted for, were used in all PPP data processing. It is shown that accounting for the second-order ionospheric delay and applying the NOAATrop model improved the PPP solution convergence time by about 15% and improved the accuracy estimation by 3 mm.