Precise Point Positioning with Ionosphere Estimation and application of Regional Ionospheric Maps

Abstract:

The ionosphere is one of most difficult source of errors to be modelled in the GPS positioning, mainly when applying data collected by single frequency receivers. Considering Precise Point Positioning (PPP) with single frequency data the options available include, for example, the use of Klobuchar model or applying Global Ionosphere Maps (GIM). The GIM contains Vertical Electron Content (VTEC) values that are commonly estimated considering a global network with poor covering in certain regions. For this reason Regional Ionosphere Maps (RIM) have been developed considering local GNSS network, for instance, the La Plata Ionospheric Model (LPIM) developed inside the context of SIRGAS (Geocentric Reference System for Americas). The South American RIM are produced with data from nearly 50 GPS ground receivers and considering these maps are generated for each hour with spatial resolution of one degree it is expected to provide better accuracy in GPS positioning for such region. Another possibility to correct for ionosphere effects in the PPP is to apply the ionosphere estimation technique based on Kalman filter. In this case, the ionosphere can be treated as a stochastic process and a good initial guess is necessary what can be obtained from an ionospheric map. In this paper we present the methodology involved with ionosphere estimation by using Kalman filter and also the application of global and regional ionospheric maps in the PPP as first guess. The ionosphere estimation strategy was implemented in the house software called RT_PPP that is capable of accomplishing PPP either for single or dual frequency data. GPS data from Brazilian station near equatorial region were processed and results with regional maps were compared with those by using global maps. Improvements of the order 15% were observed. In case of ionosphere estimation, the estimated coordinates were compared with ionosphere free solution and after PPP convergence the results reached centimeter accuracy.