Validation of various ionospheric models in the high-latitudinal zone

Abstract

When using ionospheric models, great importance is attached to their testing. The most widespread International Reference Ionosphere model is constantly being tested on a global scale, but the high latitude zone attracts special attention. In recent years, the Empirical Canadian High Arctic Ionospheric Model (E-CHAIM) has been proposed, aimed at high latitudes and having certain advantages. In this paper, both models are tested according to the Russian ionosondes Gorkovskaya (60.27˚ N, 29.38˚ E), Lovozero (67.97° N, 35.02° E), Norilsk (69.4° N, 88.1° E) in addition to the previous results of authors. The study period for Gorkovskaya and Lovozero stations has been increased to one year in conditions of low solar activity (May 2017-October 2018). For the Gorkovskaya station, the average absolute deviations |ΔfoF2| of the critical frequency for all months amounted to 0.33 MHz for the E-CHAIM model and 0.27 MHz for the IRI model. This corresponds to relative errors of 11.56% and 9.51%. For the Lovozero station, the absolute deviations were 0.34 MHz for the E-CHAIM model and 0.31 MHz for the IRI model. The corresponding relative errors were 11.8% and 11%. A comparison for the maximum height hmF2 was carried out for three options of the IRI model (AMTB2013, SHU2015, BSE1979) and the E-CHAIM model with experimental data. For the Norilsk station and the conditions of solar activity of an almost whole cycle, the AMTB2013 and SHU2015 options had no advantages anywhere over the BSE1979 and E-CHAIM options. In summer, the absolute deviations for all models lay in the range of 5–20 km, leading to a standard deviation not exceeding 20%. In winter, absolute deviations can reach 50–70 km, however, with the exception of one case, the standard deviation did not exceed 20%. For disturbed conditions, the results of determining foF2 by the E-CHAIM model were compared with an approach involving the usage of the total electron content TEC together with the equivalent slab thickness of the ionosphere. It was found that both approaches reproduce well the nighttime ionization enhancements associated with bursts of the AE index. The E-CHAIM model may overestimate the values of foF2, however, during positive disturbances both approaches underestimate the values of foF2. Using TEC can have a definite advantage over the E-CHAIM model, since TEC more objectively evaluates specific conditions.