Evaluation of COSMIC‐2 Satellite Data for foF2 and hmF2 Against Ground‐Based Ionosondes and the PyIRI (2020) Model Across Middle to Low Latitudes

Abstract This study presents a comprehensive evaluation of COSMIC‐2 satellite‐derived ionospheric peak parameters (foF2 and hmF2) against ground‐based ionosonde observations and a comparative analysis with PyIRI (2020) model estimates across middle to low latitudes for the period 2020–2024. COSMIC‐2 data were collocated with ionosonde measurements using an unprecedented 0.5° × 0.5° latitude‐longitude spatial grid, resulting in strong agreement: MAPE values of 5.2%–7.6% and correlations of 0.953–0.985 for foF2, and MAPE of 4.2%–6.3% with correlations above 0.79 for hmF2 across three stations (Austin, Ascension Island, Al Dhafra AFB). Global comparison reveals PyIRI (2020) systematically predicts higher foF2 values than COSMIC‐2, with monthly MAPD ranging from 6.89% (October, F10.7 = 141.91 sfu) to 20.43% (January, F10.7 = 176.66 sfu), demonstrating clear solar activity dependence. Regional analysis over the Pacific (10°–30°N, 180°–165°W) shows even larger seasonal variations, with PyIRI (2020) predicting foF2 values 10.54%–44.35% higher than COSMIC‐2 (maximum in December), while hmF2 differences remain below 5%. Vertical electron density profile comparisons reveal COSMIC‐2 captures well‐defined equatorial ionization anomaly (EIA) structures up to ∼450 km altitude, whereas PyIRI (2020) shows weaker EIA signatures at these altitudes. PyIRI (2020) systematically predicts lower electron density than COSMIC‐2 in both bottomside and topside ionosphere, with distinct seasonal EIA asymmetry patterns: December shows northern crest dominance becoming symmetrical at 300–400 km, while June exhibits the opposite pattern. These findings highlight complementary strengths of satellite observations and empirical models for ionospheric characterization.