Rapid dynamics empirical modeling of high-latitude three-dimensional ionospheric electric currents during geomagnetic storms

Geomagnetic storms are well-known disturbances of the Earth’s magnetic field associated with extreme solar events. Studies of geomagnetic storms are of interest for numerous scientific and societal reasons as they can strongly alter the Earth’s magnetosphere, the near-Earth geomagnetic field, the ionosphere-thermosphere and the lower atmosphere environments, and the electromagnetic environment below and close to the Earth’s surface due to induction of electric currents in the solid Earth. A key to furthering our understanding of the impact of geomagnetic storms is to better characterize the coupling between the Earth’s magnetic field and the solar wind, which takes place through the electric current system that connects the Earth’s magnetosphere to the high-latitude ionosphere. 

F-region field-aligned and E-region toroidal ionospheric currents play an important part in magnetosphere-ionosphere coupling, and often need to be studied jointly. This can be done using the network of ground vector magnetic measurements, complemented by vector satellite observations at LEO satellite altitudes. Unambiguously interpreting the highly dynamic and spatially complex ionospheric signals in these data, however, is a challenging task, as these measurements include contributions from all other natural sources, and because they only provide incomplete space-time data coverage. One approach to extract the ionospheric signal and synthesize information from several data sources is to construct empirical models. Representing both storm-time E- and F-region ionospheric currents at the appropriate cadence in such models, however, generally requires solving severely underdetermined inverse problems, which can hardly be done robustly given the relatively sparse coverage of modern geomagnetic data. 

We present a new scheme that specifically tackles this issue. It allows to construct fully three-dimensional empirical models of high-latitude E- and F-region ionospheric electric currents and magnetic fields during geomagnetic storms at periodicities down to one minute. The main idea is to reduce the model parameter space by relying on optimized basis functions of space, derived from a set of 5 numerical simulations of the TIEGCM first principle physics model, and optimized basis functions of time, directly derived from ground magnetic observations. We constructed a first model of the high-latitude ionosphere in the Northern hemisphere constrained by ground and magnetic perturbation data from the Iridium satellite constellation for the storm of May 2017. The model shows excellent agreement with an independent TIEGCM numerical simulation of this same storm, as well as with independent data from the Swarm, CryoSat-2, and GRACE  satellites.