Global Accuracy Assessment and Analysis of the Ionospheric Model IRI-Plas 2020 and IRI-2020 Based on GNSS Observations

The ionosphere plays a crucial role in aerospace, communication and navigation positioning, and empirical ionospheric models such as the IRI (International Reference Ionosphere) model and IRI-Plas (International Reference Ionosphere Extended to Plasmasphere) model are widely used in estimating ionospheric parameters. Due to the uneven distribution and missing stations in certain regions of GNSS (Global Navigation Satellite System) stations, empirical ionospheric models are often employed in ionospheric research to address data scarcity. Evaluating the latest versions of the IRI-Plas model and the IRI model will assist in understanding the ionospheric response and model performance under different environments, enabling better refinement of existing models. Based on the dual-frequency observations of 135 GNSS stations worldwide, the Vertical Total Electron Content (VTEC) and Slant Total Electron Content (STEC) are extracted in this study to evaluate the IRI-Plas 2020 model with GIM TEC input, the IRI-Plas 2020 model without external TEC input, and the IRI-2020 model. The differences between different ionospheric models and GNSS observations are presented and analyzed in detail, including patterns of ionospheric latitude variation, daily variation and seasonal variation and the accuracy of models under different geomagnetic conditions. The results show that all models can represent the ionospheric characteristics very well, among which the IRI-Plas model value with GIM TEC input has the highest accuracy, while the IRI-Plas model value without external TEC input is high due to the limitation of the model calculation altitude range. Also, the errors of all models exhibit latitudinal variations, usually decreasing with the increase of latitude, and these variations also show seasonal trends.