A new approach to data-model comparisons: Using MAGE and Lompe to unravel ionosphere-magnetosphere electrodynamics

Global numerical simulations are a valuable tool for understanding the Sun-Earth interaction as they provide a more complete picture of the system when compared to typically sparse observations. Yet, comparison of global numerical simulations against observations is complicated by the inherent uncertainty that the observed phenomena occurred at the same time and location in the simulation domain. It is therefore common to use aggregate measures of the Sun-Earth interaction, such as the Dst and AL indices and cross-polar cap potential, with the downside of missing details, especially at the mesoscale lengths. Despite being quantitative, these aggregate measures can also hide important physical processes. It is therefore crucial to find metrics or parameters that provide deeper insight into the physics but do not rely precisely on the location of the observations. Recently, data assimilative models to reproduce patterns of high-latitude ionospheric electrodynamics have been improved, and they can be used to reconstruct ionospheric quantities using observations to compare with simulation outputs, providing a new avenue for data-model comparisons. In this paper, we demonstrate this new approach to data-model comparisons by assimilating global MHD simulation (MAGE) data into Lompe for direct comparison with the real-data-assimilation patterns of field-aligned currents and their ionospheric components (the FAC terms associated with the divergence of the electric field, gradient of the Hall conductance, and gradient of the Pedersen conductance) from \citeA{gasparini2024quantifying}. In addition, we calculate reconnection electric fields for the MHD simulations and real-data assimilation, and find that the nightside reconnection rate from the MAGE simulations is 30% higher than in the real data case. We also find that with the MHD simulation the system enters steady-magnetospheric convection, in contrast to the observations which indicated a classic substorm. We conclude by speculating on the possible sources of discrepancies between the model and observations.