Sushant Shekhar (National Centre for Geodesy, IIT Kanpur, India); Rohit Singh Rana (National Centre for Geodesy, IIT Kanpur, India); Onkar Dikshit (National Centre for Geodesy, IIT Kanpur, India)

The accuracy of ionospheric models derived from ground-based multi-Global Navigation Satellite System (GNSS) observations is often limited in regions with sparse tracking station coverage. To address this limitation, previous studies have explored the integration of diverse ionospheric data sources; however, inconsistencies in observational coverage and measurement characteristics remain a challenge. In this study, we evaluate the performance of ionospheric maps generated using two approaches: (i) a ground-based strategy utilizing only GNSS-derived ionospheric data, and (ii) a multi-source strategy that integrates ground-based data with additional geodetic measurements, normalized using the single-layer ionospheric model.
The comparative analysis reveals that the primary discrepancies between the two approaches occur over the Indian Ocean region, with differences typically ranging from −1 to 0 Total Electron Content Units (TECU). Validation using Jason-3 vertical Total Electron Content (VTEC) data indicates that the multi-source strategy achieves a mean root mean square (RMS) error of 4.73 TECU, representing an improvement of approximately 13.8% over the ground-based approach. The maximum improvement reaches up to 19% across different latitude bands. Furthermore, a self-consistency evaluation demonstrates that the multi-source model yields an RMS of 3.1 TECU, corresponding to a 2.60% enhancement, with peak improvements of up to 12% on specific days. These results highlight the effectiveness of multi-source data integration in enhancing ionospheric model accuracy, particularly in data-sparse regions.