Assimilation of GPS TEC and COSMIC foF2/ hmF2 Into SPIM Model Using Unscented Kalman Filter Techniques Over the East African Equatorial Region

Abstract This study aims to improve the estimation of Total Electron Content (TEC), the F2 layer critical frequency (foF2), and its peak height (hmF2) over East Africa by applying the Unscented Kalman Filter (UKF) to assimilate GPS‐derived TEC and foF2/hmF2 from COSMIC‐2 into the Standard Plasmaspheric and Ionospheric Model (SPIM). Assimilation was conducted in two experiments: (a) GPS TEC data were assimilated into the background model during geomagnetic quiet periods (23–26 September 2012; 21–24 May 2015; and 27–30 December 2017) and disturbed periods (7–10 March 2012; 16–19 March 2015; and 6–9 September 2017); and (b) foF2 and hmF2 were assimilated during the disturbed period of 23–27 September 2019. Our results show that the assimilation enhances the SPIM model's estimation of ionospheric parameters. The assimilation demonstrates strong agreement with observations, with correlation coefficient ( R ) compared to the background model ( R = 0.79 at moiu station, , and 0.82 at nege station, ). UKF significantly enhances GPS TEC assimilation, with a maximum percentage error reduction of 49.02% during quiet periods across all stations. During geomagnetic disturbances, TEC assimilation reduced percentage error by 40.0%–59.55% relative to SPIM. Assimilated foF2 and hmF2 values show improved agreement with COSMIC‐2 measurements. The maximum RMSE improvement in foF2 was about 99.35% over moiu station, , and the minimum was 67.56% over mal2 station, . For hmF2, the highest RMSE improvement was 97.31% over adis station, , and the lowest was 60.16% over armi station, , compared to the background SPIM model during the study periods.