Satellite magnetic observations from the Oersted and CHAMP missions have in the past
decade revealed rapid changes in the Earth’s core. The origin of these events is not yet
understood, but they seem to involve a reorganization of flows and magnetic fields within the  outer core. Integration of magnetic field observations from satellites, in particular from the  upcoming ESA Swarm constellation, with specially tailored MHD models of core dynamics is now required in order to make progress in our physical understanding.

The success of this data assimilation scheme for core dynamics depends on a small number of critical issues that we propose to investigate within the framework of an international team. First, there is the need for an improved characterization of the correlated errors inherent in satellite magnetic observations and in derived global  field models. Improved source separation  methods, for example taking into account electrical currents induced in the mantle and oceans,  should also help in better isolating the signals of interest. On the theoretical side, there is a need for continued development of numerical models describing rapid outer core dynamics, such as quasi-geostrophic MHD models, designed for comparisons with satellite observations. Progress on these coupled topics can best be achieved through the concerted effort of a team, comprising both observational and theoretical expertise, working together on algorithms to combine satellite observations and numerical models of core dynamics.