An Equivalent Source Method for Modelling the Global Lithospheric Magnetic Field

We present a new technique for modelling the global lithospheric magnetic field at Earth’s surface based on the estimation of equivalent potential field sources. As a demonstration we show an application to magnetic field measurements made by the CHAMP satellite during the period 2009–2010 when it was at its lowest altitude and solar activity was quiet. All three components of the vector field data are utilized at all available latitudes. Estimates of core and large-scale magnetospheric sources are removed from the measurements using the CHAOS-4 model. Quiet-time and night-side data selection criteria are also employed to minimize the influence of the ionospheric field. The model for the remaining lithospheric magnetic field consists of magnetic equivalent potential field sources (monopoles) arranged in an icosahedron grid at a depth of 100 km belowthe surface. The corresponding model parameters are estimated using an iteratively reweighted least-squares algorithm that includes model regularization (either quadratic or maximum entropy) and Huber weighting. Data error covariance matrices are implemented, accounting for the dependence of data variances on quasi-dipole latitude. The resulting equivalent source lithospheric field models show a degree correlation to MF7 greater than 0.7 out to spherical harmonic degree 100. Compared to the quadratic regularization approach, the entropy regularized model possesses notably lower power above degree 70 and a lower number of degrees of freedom despite fitting the observations to a very similar level. Advantages of our equivalent source method include its local nature, the possibility for regional grid refinement and the production of local power spectra, the ability to implement constraints and regularization depending on geographical position, and the ease of transforming the equivalent source values into spherical harmonics.