Evaluation of the SSN and F10.7 indices as a cause for performance variation of the IRI 2016 and IRI-Plas 2017 models

Abstract

This paper mainly focuses on the evaluation of efficiency of the Sunspot Number (SSN) and the 10.7 cm Solar Radio Flux (F10.7) indices as a cause for the variation of the performance of the latest versions of the International Reference Ionosphere model (IRI 2016 with NeQuick, IRI01-corr and IRI2001 options for the topside electron density) and the IRI Extended to the Plasmasphere (IRI-Plas 2017) for the modeling of the Total Electron Content (TEC). The Global Positioning System (GPS)-derived TEC data obtained from the dual frequency GPS receivers located at Observation Rock, OBSR (geog 46.90°N, 238.18°W, Geom. 52.46°N) and Husband, HUSB (44.12°N, 238.15°W, Geom. 49.73°N) in the West Pacific region during the recent solar maximum (2012-2014) years have been considered for the validation of the performance of the models. The results show that both the GPS-derived TEC (GPS VTEC) and modelled (IRI 2016 and IRI-Plas 2017 VTEC) seasonal diurnal values tend to peak at 00:00 UT (16:00 LT) and 20:00 (12:00 LT) with the highest being observed mostly at 20:00 (12:00 LT); while, their minima are mostly observed at about 13:00 UT (05:00 LT). In addition, in utilizing the SSN, the best performance is generally observed in the June solstice months, especially by the IRI-Plas 2017 model. However, for the equinoctial and December solstice months, the best performance is generally observed by the IRI 2016 model with NeQuick and IRI01 options. It has also been shown that the root-mean-square deviations between the GPS-derived and modelled VTEC diurnal variation in using the F10.7 index are generally less than those of the SSN option in all months, revealing that both the IRI 2016 and IRI-Plas 2017 models generally show better performance using F10.7 index than the SSN. Hence, the F10.7 option is recommended for better TEC modeling employing the IRI 2016 and IRI-Plas 2017 models during the recent solar maximum years over the West Pacific region. In addition, both models cannot effectively estimate the geomagnetic storm time TEC variation.