Storm-Time Atmosphere-Ionosphere-Magnetosphere Coupling with Data Assimilation

Subject to the multi-scale magnetospheric forcing, the storm-time ionosphere-thermosphere (IT) system is highly dynamic and nonlinear. It has been challenging to simulate and understand the localized IT responses to storms due to the limitation of the empirical drivers used in the IT models. Adopting a Lattice Kriging Gaussian model, we use SuperDARN and PFISR ion drifts, and THEMIS and SSUSI aurora data, to produce the assimilation maps of aurora and electric field, respectively. The TIEGCM simulations driven by the assimilation capture the local PFISR observations of secondary peak of electron density induced by the enhancement of auroral precipitation in the E-region, and elevated ion and electron temperatures due to the Joule heating. In addition, the modeled neutral vertical winds reach a magnitude of 50 m/s around 250 km, much stronger than the ones (< 10m/s) from the default run, and closer to the FPI observations. Globally, the model driven by the data assimilation better captures the enhancement of TECs in the auroral region. As the magnetospheric forcings are better constrained, the neutral responses show both richer scales and larger dynamic range than the default run. The temporal and longitudinal variability of the IT response increases by 1.5-3 times. The traveling atmospheric disturbances (TADs) reach larger amplitudes, propagating more extensively to middle and low latitudes, and inducing stronger traveling ionospheric disturbances (TIDs). The small-scale perturbations also lead to significant changes of mean background winds and temperatures, indicating the cross-scale interactions. The zonal and meridional winds show change of ~200 m/s, and zonal mean temperatures show increases of 200-300 K at high latitudes. This study highlights the importance of incorporating realistic magnetospheric forcings in order to accurately assess the real-time storm impacts.