A new approach for the generation of real-time GNSS low-latitude ionospheric scintillation maps

Ionospheric scintillation disturbs radio frequency signals affecting GNSS-based navigation, especially in Brazil, due to the large magnetic declination. The generation of real-time scintillation maps is an important way to provide scintillation monitoring. This work considers amplitude scintillation, given by the S4 index. Some existing and some proposed approaches for generating these maps are presented and tested, being each one a combination of an interpolation method with some existing and/or proposed sets of pre-processing options. These approaches are named after the related interpolation method as GRIDDATA, Inverse Distance Weighting, Radial Basis Functions and Gaussian Process Regression. The making of scintillation maps requires the interpolation of Ionospheric Pierce Point (IPP) samples, given by the S4 values for each IPP of each satellite-station link considering the set of GNSS stations of the given area and time interval. Some intervals of time that presented strong scintillation over Brazil were selected and the corresponding sets of IPP samples were used to obtain sequences of maps using all possible combinations of interpolation and pre-processing options. Furthermore, a fifth, more recent, approach was also included in the comparison. The quality of the resulting maps was assessed, concluding that the Gaussian Process Regression approach, with a specific set of pre-processing options, allows to generate the most accurate scintillation maps. The proposed map generation approach is part of a broader proposal being implemented to provide real-time scintillation maps covering the Brazilian territory.