Applicability of multi-layered mapping function to STEC/VTEC conversion – validation in PPP positioning and GNSS observation domains

Recently, we have observed remarkable progress in algorithms aimed at precise point positioning (PPP) based on uncombined GNSS observations, also with an integer ambiguity resolution. Despite the high potential of such a model, its performance still depends on several factors, among which is the a priori information on slant total electron content (STEC), which is crucial. The ionospheric corrections may be obtained by converting VTEC from global ionosphere maps (GIM) using the selected ionospheric mapping function (MF). Thus, the accuracy of such STECs depends on the uncertainties introduced by both GIMs and MFs. Considering the latter, the most common approach is using a single-layer model (SLM) with a zenith angle as a parameter. However, it may be less effective for regions with strong TEC gradients that depend on azimuth. Improving such areas seems feasible with the support of ionosphere models and multi-layered mapping functions.

In this study, we evaluate the new multi-layered mapping functions. The new functions were developed, taking the Neustrelitz TEC Model as a basis. In this case, the ionosphere comprises numerous thin shells, and the ratio of aggregated slant and vertical TEC values provides the modeled mapping factors. Such derived MFs were validated based on PPP positioning performance. Also, an agreement analysis using the geometry-free linear combination was performed to assess the MFs in the GNSS observation domain.  The preliminary tests involved GNSS data from several globally distributed stations, corresponding to different daily patterns of the ionosphere. The analysis provides the statistics for the low solar activity period (year 2019).  According to the results, we can report a slight benefit from applying the multi-layered mapping function for PPP performance compared to the standard SLM approach. The advancement is the most noticeable for the positioning initialization period and, therefore, is reflected in the convergence time. The analysis performed with the geometry-free linear combination is consistent with PPP results, and it highlights the highest potential of the multi-layered mapping function for the equatorial region.