Investigating the Storm-Time Response of the Ionosphere-Thermosphere through Data Assimilation

In the low Earth orbit (LEO) environment, improving space weather estimation and forecasting capabilities requires advancing understanding and physics-based modeling of storm-time dynamics and energetics of the ionosphere-thermosphere (I-T) system. Current I-T models are limited due to model biases, largely stemming from neutral state observation gaps. In contrast to neutral observations, LEO radio occultation (RO) constellations globally produce electron density profiles (EDPs), providing near real-time plasma observations of the three-dimensional structure of the ionosphere. Using strongly coupled data assimilation to combine observations and physics-based models, EDPs are adequate for specifying both plasma and neutral states of the I-T system.

For accomplishing neutral state estimation, we use EDPs to directly update neutral temperatures and neutral winds using cross-covariance state information, exploiting strong I-T coupling. In this work, the Thermosphere Ionosphere Electrodynamics General Circulation Model (TIEGCM) is used as the coupled I-T model and the Ensemble Adjustment Kalman Filter (EAKF) as the DA framework. We demonstrate the capability to use EDPs to constrain neutral states in simulated experiments, improving neutral densities and corresponding orbit errors by 70% over a 2-day period. Additionally, through better specified neutral dynamics, helium compositions can be estimated. With direct state estimates, we aim to produce a storm-time reanalysis using COSMIC-2 EDPs to correct the storm response of global circulation, expansion and heating, and recovery.

Towards plasma state estimation, we assess EDP impact on ionospheric specification and inform future RO constellation design through comprehensive Observing System Simulation Experiments (OSSEs). Synesthetic EDPs are operationally retrieved through Abel inversion from a nature run of the Whole Atmosphere Model-Ionosphere Plasmasphere Electrodynamics (WAM-IPE) and assimilated into the distinct I-T model TIEGCM. With realistically represented observation and forecast model errors, we show greater observation coverage to improve ionospheric specification for altitudes 300 km and above, with 520 km altitude RO constellations performing best with the highest observation counts. Abel inversion error impacts are additionally characterized, highlighting key limitations of EDP observations at altitudes below 200 km and low latitudes with large horizontal gradients.