Unravelling source spatial parameters and magnetisation direction from inversion of TMI, vector component and tensor magnetic field data

Abstract

In this paper I investigate the level of error in magnetisation direction which might pass undetected in TMI inversions, and estimate the corresponding errors that this will cause in estimation of the location and shape of source bodies. I also show that three-component and gradient tensor data can potentially provide increased sensitivity to source magnetisation direction. These studies are undertaken by forward computing magnetic anomalies for bodies with various magnetization directions, and then inverting those anomalies with the assumption of induced-only magnetisation. The misfit between input and output fields after the TMI inversions increases with error in source magnetisation direction. Up to 30° to 40° the misfit is likely to be undetectable. Up to 60° to 70° the misfit can be reduced by invoking a more complex magnetisation distribution, but from 70° to 90° the inability to match an anomaly with any feasible distribution of magnetisation is diagnostic of error in the source magnetisation direction. The associated errors in location and dip of a source body increase with undetected error in its magnetisation direction. Components of the magnetic field behave similarly but not identically to error in source magnetisation direction, and slight differences between component inversions provide additional sensitivity to source magnetisation for errors above 40°. Unfortunately measurement difficulties for component data are likely to restrict the achievement of this improvement. Magnetic field tensor data also increase sensitivity to source magnetisation direction and enable better-constrained inversions of the magnetic field.