An EAGER project to explore newly developed advanced data assimilation methods and implement them in a global magneto-hydro-dynamic (MHD) simulation of the Earth's magnetosphere. While global MHD magnetospheric models have matured to the point when they can be used for prediction of space weather processes, the fundamental problem that they face is common to all numerical models of any realistic physical system: insufficiently or poorly specified boundary/initial conditions and, possibly, missing physics. To tackle this problem, the atmospheric weather community has been developing methods of "data assimilation" for the past ~50 years. In global magnetospheric modeling, the development and implementation of data assimilation methods is hampered by the enormous spatial scales of the system and inevitably poor spatiotemporal coverage of available measurements. Data assimilation in MHD is particularly challenging because the primary wave modes are non-linear, non-dispersive, and highly anisotropic. In addition, modern global MHD codes utilize very large grids with up to tens of millions of cells leading to huge covariance matrices that cannot be solved by classical data assimilation approaches. The project will utilize major new ionospheric observational assets, especially high-cadence magnetometer data from the >70-satellite Iridium constellation. The low-altitude (~800 km) observations of magnetic perturbations with truly global continuous coverage that are now provided by the NSF AMPERE project, will be used to specify a key inner boundary condition for global MHD models. In this way, the project will attempt the development of the first-ever data assimilation methodology for global magnetospheric MHD simulations. If successful, results from this project will lay the ground for transformative advances in global MHD simulations, both as tools of scientific inquiry and for space weather forecasting.
