High-solar Activity Ionospheric Modelling Using Machine Learning: A comparison Against Classical Models

Shishir Priyadarshi, Wahyudin P. Syam, Andrés Abelardo García Roqué, Alejandro Pérez Conesa, Guillaume Buscarlet, Raül Orús Pérez, Mickael Dall’ Orso

Abstract:	Ionosphere is a chaotic medium for all kind of trans-ionospheric radio wave communication and ranging. Global Navigation Satellite System (GNSS) signals are very vulnerable to the disturbed ionosphere. With enhanced solar activity electrons and plasma within the ionosphere adopt complicated dynamics and turbulence that change the formation and evolution of the ionospheric irregularities for all the geographic latitude, seasons as well as geomagnetic activity conditions. These ionospheric irregularities widely varies in special scales from a few centimetres to a thousands to kilometres where in timescales their expectancy varies from a fraction of second to days, seasons even years. Using the HARMONY (macHine leARning to MOdel gNss sYstems) ionospheric demonstrator (MLD IONO) we are presenting the comparison result of a few selected cases during the high solar activity at the low-, mid- as well as high-latitude stations. For comparison we use the HARMONY Slant Total Electron Content (STEC) model along with the NeQuick-G and Klobutchar ionospheric model. For Vertical Total Electron Content (VTEC) comparison we use STEC observation close to 90 degree of elevation angle, we also use Ionosphere’s International GNSS Service (IGS) maps to reproduce VTEC for the similar spatial coordinates and time to the HARMONY demonstrator along with the VTEC modelled from NeQuick-G and Klobutchar model. For all the stations, the HARMONY demonstrator performance was within the acceptable key performance indicators (KPIs) correlation coefficient (R2 ) and root mean square error (RMSE) in our case, range and better than the NeQuick as well as Klobuchar ionospheric model.
Published in:	Proceedings of the 36th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2023) September 11 - 15, 2023 Hyatt Regency Denver Denver, Colorado
Pages:	121 - 135
Cite this article:	Priyadarshi, Shishir, Syam, Wahyudin P., Roqué, Andrés Abelardo García, Conesa, Alejandro Pérez, Buscarlet, Guillaume, Pérez, Raül Orús, Orso, Mickael Dall’, "High-solar Activity Ionospheric Modelling Using Machine Learning: A comparison Against Classical Models," Proceedings of the 36th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2023), Denver, Colorado, September 2023, pp. 121-135. https://doi.org/10.33012/2023.19291
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In