Satellite observations of the magnetic field induced by the general ocean circulation could provide new constraints on global oceanic water and heat transports. This opportunity is investigated in a model-based twin experiment by assimilating synthetic satellite observations of the ocean-induced magnetic field into a global ocean model. The general circulation of the world ocean is simulated over the period of 1 month. Idealized daily observations are generated from this simulation by calculating the ocean-induced magnetic field at 450 km altitude and disturbing these global fields with error estimates. Utilizing an ensemble Kalman filter, the observations are assimilated into the same ocean model with a different initial state and different atmospheric forcing. Compared to a reference simulation without data assimilation, the corrected ocean-induced magnetic field is improved throughout the whole simulation period and over large regions. The global RMS differences of the ocean-induced magnetic field are reduced by up to 17%. Local improvements show values up to 54%. RMS differences of the depth-integrated zonal and meridional ocean velocities are improved by up to 7% globally, and up to 50% locally. False corrections of the ocean model state are identified in the South Pacific Ocean and are linked to a deficient estimation of the ocean model error covariance matrices. Most Kalman filter induced changes in the ocean velocities extend from the sea surface down to the deep ocean. Allowing the Kalman filter to correct the wind stress forcing of the ocean model is essential for a successful assimilation.
