Klobuchar, NeQuickG, BDGIM, GLONASS, IRI-2016, IRI-2012, IRI-Plas, NeQuick2, and GEMTEC Ionospheric Models: A Comparison in Total Electron Content and Positioning Domains

Global navigation satellite systems (GNSS) provide a great data source about the ionosphere state. These data can be used for testing ionosphere models. We studied the performance of nine ionospheric models (Klobuchar, NeQuickG, BDGIM, GLONASS, IRI-2016, IRI-2012, IRI-Plas, NeQuick2, and GEMTEC) both in the total electron content (TEC) domain-i.e., how precise the models calculate TEC-and in the positioning error domain-i.e., how the models improve single frequency positioning. The whole data set covers 20 years (2000-2020) from 13 GNSS stations, but the main analysis involves data during 2014-2020 when calculations are available from all the models. We used single-frequency positioning without ionospheric correction and with correction via global ionospheric maps (IGSG) data as expected limits for errors. Improvements against noncorrected solution were as follows: GIM IGSG-22.0%, BDGIM-15.3%, NeQuick2-13.8%, GEMTEC, NeQuickG and IRI-2016-13.3%, Klobuchar-13.2%, IRI-2012-11.6%, IRI-Plas-8.0%, GLONASS-7.3%. TEC bias and mean absolute TEC errors for the models are as follows: GEMTEC--0.3 and 2.4 TECU, BDGIM--0.7 and 2.9 TECU, NeQuick2--1.2 and 3.5 TECU, IRI-2012--1.5 and 3.2 TECU, NeQuickG--1.5 and 3.5 TECU, IRI-2016--1.8 and 3.2 TECU, Klobuchar-1.2 and 4.9 TECU, GLONASS--1.9 and 4.8 TECU, and IRI-Plas-3.1 and 4.2 TECU. While TEC and positioning domains differ, new-generation operational models (BDGIM and NeQuickG) could overperform or at least be at the same level as classical empirical models.