Assessment of spatial and temporal TEC variations derived from ionospheric models over the polar regions: Fid-Bau Portal

Assessment of spatial and temporal TEC variations derived from ionospheric models over the polar regions

Jiang, Hu / Liu, Jingbin / Wang, Zemin / An, Jiachun / Ou, Jikun / Liu, Shulun / Wang, Ningbo

Abstract A comprehensive evaluation of Global Positioning System (GPS) Klobuchar, Galileo broadcast NeQuick2, international reference ionosphere (IRI) and global ionospheric map (GIM) models in estimating ionospheric total electron content (TEC) is performed using GPS-derived, constellation observing system for meteorology, ionosphere, and climate and JASON-2 TECs under various solar conditions in the Arctic and Antarctic. The performances of the four ionospheric models are first analysed for the overall polar regions. In addition, according to the temporal and spatial characteristics of the polar regions, the accuracies of the four models are evaluated for the polar days and nights, the Antarctic ice sheet (AIS) and the Arctic Ocean (AO), the Weddell Sea Anomaly (WSA) as well as for ionospheric storm. The results show that Klobuchar, NeQuick2, IRI2016 and GIM can mitigate the ionospheric delay by 28.69–29.41%, 44.57–56.09%, 43.38–55.99% and 67.17–77.56%, respectively. The performances of the four models during the polar days are obviously worse than those during the polar nights. In the AIS and AO, the Galileo broadcast NeQuick outperforms the GPS broadcast Klobuchar, and the root-mean-square error of IRI2016 performs almost the same as NeQuick2, but IRI2016 has a greater deviation. In addition, the GIM model can only mitigate the ionospheric delay by 46.81–56.72%, which is lower than other regions due to the lack of GPS ground station observations. Under the WSA conditions, the four models underestimate the real TEC to varying degrees, and the night-time deviations of the Klobuchar, NeQuick2 and IRI2016 models are significantly greater than the daytime deviations. The relative accuracy of four ionospheric models is lower during the ionospheric storm period than that in the ionospheric quiet period, especially the NeQuick2 and IRI2016 over the Antarctic.