Solar activity variability in the IRI at high latitudes: Comparisons with GPS total electron content

Total electron content (TEC) measurements from 10 dual‐frequency GPS receivers in the Canadian High Arctic Ionospheric Network (CHAIN) are used to evaluate the performance of International Reference Ionosphere (IRI)‐2007 within the Canadian sector. Throughout the region, we see systematic underestimation of daytime TEC, particularly at solar maximum, where summer and equinox root‐mean‐square errors reach as high as 14 total electron content units, 1 TECU = 1016 el m−2 (TECU). It is also shown that the use of a monthly IG index, in place of the IRI's standard IG12 index, leads to an improvement in TEC specification by up to 3 TECU in the polar cap and up to 6 TECU in the subauroral region during periods of short‐term, large amplitude changes in solar activity. On diurnal timescales, variability in TEC is found to be underestimated by the IRI, during equinox periods, by up to 40% at subauroral latitudes and up to 70% in the polar cap region. During the winter, diurnal variations are overestimated by up to 40% in the subauroral region and are underestimated within the polar cap by up to 80%. Using collocated ionosonde data, we find IRI bottomside TEC to be within 1 TECU of observation with errors largest during the equinoxes. For the topside we find good agreement during the winter but significant underestimation of topside TEC by the IRI during summer and equinox periods, exceeding 6 TECU at times. By ingesting measured NmF2 into the IRI, we show that the topside thickness parameterization is the source of the bulk of the observed TEC errors. Key Points The IRI significantly underestimates TEC at high solar activity The IRI is incapable of modeling the ionospheric response to subannual changes in solar activity The majority of IRI TEC errors result from errors in its topside thickness parameterization