Adaptative ionospheric electron content estimation method

Abstract

In 2011 and 2012, the increase of solar activity has severely disturbed the ionosphere, and caused strong performance degradations of both Safety-of-life Satellite Based Augmentation Systems, WAAS and EGNOS. As the dynamics of the ionosphere remains poorly understood, the algorithms at users level models the global ionosphere as a thin layer as defined by MOPS reference document. The Triangular Interpolation model, also known as TRIN model, approximates this thin layer by a rigid polyhedron with triangular facets in a solar fixed frame. This model assumes that the ionosphere layer behaviour can be considered as spatially linear inside each facet. Since during a strong solar activity period, the variation may be significantly non-linear, the ionosphere dynamic may locally be outside the model validity domain. Based on the analysis of the temporal and spatial distribution of ionosphere pierce points (IPP) – and associated L1 ionosphere delays, we have developed a set of invariants to discriminate different types of ionosphere dynamics around each vertex of TRIN model. These invariants reflect the taxonomy of the IPP distribution around a mesh vertex. On this basis, we propose a method to develop a flexible and adaptive TRIN model by adding new mesh vertex where relevant in order to reduce the process model noise. In the case where a cluster distribution is detected, the process positions new mesh vertex near local optima. The method adapts to the ionosphere’s dynamic seeking periodically local optima. This gives a more realistic mapping of heterogeneities of the ionosphere and its developments. A simple implementation of this method is being developed. In this paper we detail the methodology we are following in order to improve the total electric content computation for the Ionospheric Grid Point using a Flexible TRIN method.