Feature‐based validation of the Lyon‐Fedder‐Mobarry magnetohydrodynamical model

Field‐aligned currents (FACs) play an important role in the coupling between the ionosphere and magnetosphere. Numerical simulation of these phenomena is of increasing interest, but validation has been hampered by a lack of a formal framework to compare simulations to satellite‐derived products. We develop a statistical approach to compare FAC simulations from global magnetohydrodynamical models against satellite products. We introduce a robust algorithm that automatically detects and defines regions 1 and 2 FACs. In an example, currents derived from the Iridium satellites are compared against simulated currents from two resolutions of the Lyon‐Fedder‐Mobarry model on one solar event. We assess both average and structured discrepancies, the former being a level shift of the physical model away from the satellite product, while structural discrepancy refers to time‐varying, continuous differences. For this event, the lower resolution version of the Lyon‐Fedder‐Mobarry is shown to be a poor representation of the satellite‐derived FACs, while the higher resolution version substantially reduces discrepancy. Key Points A robust algorithm is developed to detect and extract regions 1 and 2 field‐aligned current features Increased resolution of LFM model improves FAC simulation and reduces bias against AMPERE estimates The introduction of a statistical framework allows for quantitative estimates of model discrepancy