Helmert-VCE-Aided Weighted Method for Fusing GNSS Data and IRI-2020 Model Virtual Observations to Model Ionosphere TEC in the Arctic

Using the vertical total electronic content (VTEC) value of the empirical ionospheric model to compensate for the deficiency of ground-based Global Navigation Satellite System (GNSS) observation is an effective method to improve the accuracy and reliability of ionospheric modeling in the ocean and polar regions lacking ground-based GNSS observation stations. In the research of using empirical model virtual observations to enhance GNSS for ionospheric modeling, the empirical model error is usually taken as a fixed value for empirical weight determination when the empirical model and GNSS VTEC are weighted, and the variation of empirical model error with time and space is ignored. In this paper, a GNSS ionospheric modeling method enhanced by IRI-2020 virtual observations based on Helmert variance component estimation is proposed. The weights of GNSS and IRI model VTEC are dynamically determined by Helmert variance component estimation. The models obtained from the GNSS ionospheric modeling enhanced by IRI-2020 model virtual observations using empirical weighting and Helmert variance component weighting methods are defined as GIP and GIH respectively. Taking the Arctic region as an example, the FY-3 VTEC is used as reference value to verify the modeling results from the 79th to 90th days of 2023. The results show that the MAE, RMS and STD of the GIH model are all smaller than those of the GIP, decreasing by 0.20 TECU, 0.30 TECU and 0.24 TECU respectively, which indicates that after weighting with the Helmert variance component estimation, the accuracy of ionospheric modeling in the Arctic can be improved. The analysis of ionosphere TEC accuracy at different longitudes, latitudes and local times shows that GIH has a greater improvement in accuracy compared to GIP in the data blank areas of longitude 60°E-180°E, latitude greater than 84°, and local time 12 -18.