Assessment of IRI-2012, NeQuick-2 and IRI-Plas 2015 models with observed equatorial ionization anomaly in Africa during 2009 sudden stratospheric warming event

In Africa, we assessed the performance of all the three options of International Reference Ionosphere 2012, IRI-2012 (i.e. IRI-2001, IRI-2001COR and IRI-NeQuick), NeQuick-2 and IRI-Plas 2015 models prior to and during 2009 sudden stratospheric warming (SSW) event to predict equatorial ionization anomaly (EIA) crest locations and their magnitudes using total electron content (TEC) from experimental records of Global Positioning System (GPS). We confirmed that the IRI-Plas 2015 that appeared as the best compared to all of the models as regard prediction of the EIA crest locations in the northern hemisphere of Africa is due to discontinuities in the GPS data between ∼8° N and 22° N. As regard the predictions of EIA crest magnitudes and the location of EIA crests in the southern hemisphere of Africa, they are not present in all the models. The NeQuick-2 model does not have the capability to predict either the EIA crest location in the northern or southern hemisphere. The SSW effect on the low latitude was able to modify a single EIA crest to pre-noon and post noon EIA crests in the northern hemisphere during the SSW peak phase and significantly reduced the GPS TEC magnitudes over the hemispheres as well. These SSW effects and delays of plasma transportation to higher latitudes in GPS TEC were absent in all the models. For future improvements of IRI-2012, NeQuick-2 and IRI-Plas 2015 models, SSW conditions should be included in order to characterize the effect of lower atmosphere on the ionosphere. The EIA trough modeling is only present in IRI-2001COR and IRI-2001NeQuick options. In the middle latitude, all the model could not predict the location of highest TEC magnitudes found at RBAY (Richardsbay, South Africa).