Improving the ionospheric state estimate during geomagnetic storm time through assimilation of neutral density data

During geomagnetic storms, communication and navigation instruments can be dramatically affected by the rapid changes that occur in the upper atmosphere. The assimilation of data in physics-based models such as the Coupled Thermosphere Ionosphere Plasmasphere electrodynamics (CTIPe) model through and ensemble Kalman filter, can improve the representation of the thermosphere-ionosphere (TI) system. Due to the coupled nature of the TI system, the ionosphere is affected by, among others, changes in the neutral atmosphere. In this study, we investigate the capability of the CTIPe model to provide better estimates of the ionosphere by improving its specification of the thermosphere via data assimilation. Here, we assimilate thermospheric mass density (TMD) observations from the Swarm mission normalized to 400 km altitude during the 2015 St. Patrick’s Day storm. The changes that occur in the ionosphere due to assimilation of TMD data are measured by means of the difference between the model results with and without assimilation. To measure the improvement gained with data assimilation, we compare with independent measurements of electron density along the orbit of GRACE (Gravity Recovery and Climate Experiment) satellite, that shows a reduction in the root mean square error (RMSE) by a 22% with respect to the non-assimilation run. The impact on the global scale is measured by comparing the CTIPe model results with the corresponding output of the 3D B-Spline electron density model. The results illustrate that the electron density equatorial region is the most affected region by assimilation of TMD, with an average RMSE reduction of 25% at the assimilation altitude of 400 km.