Evaluation of the performance of ionospheric models at solar maximum using COSMIC slant TEC measurements

Abstract

AbstractWe report the results of a model validation study that assessed how well several ionospheric models captured the slant total electron content, especially at low latitudes near the equatorial ionization anomaly, where horizontal and vertical density gradients are large. We assessed NeQuick, IRI‐2007, IRI‐2012, SAMI‐3, and the Utah State University version of the Global Assimilation of Ionospheric Measurements (GAIM) model. We used slant total electron content measurements made by the Constellation Observing System for Meteorology, Ionosphere, and Climate (COSMIC) constellation during 5 May to 20 June 2012 to test GAIM, NeQuick, IRI‐2007, and IRI‐2012 and during 1 October 2011 to 31 December 2011 to test SAMI‐3, as the SAMI‐3 model runs were not available for the 2012 time frame. We found that the GAIM data assimilation model showed the lowest biases, although all of the models typically agreed with the COSMIC measurements to ~8% in the worst case. One area of concern with all of the models was that the mean percentage difference between the COSMIC measurements and the calculated total electron content (TEC) showed significant scatter, >15% at the 1 sigma level; this was attributed to all of the models not capturing the density gradients near the equatorial ionization anomaly (EIA). All of the models underestimated the topside electron density and thus also the ionospheric slab thickness. Since ionospheric models are often validated using near‐vertical TEC measurements and the vertical TEC is the product of the electron density at the F region peak and the slab thickness, our results suggest that the peak density values in the models may be too high.