Two-dimensional magnetotelluric analysis of three-dimensional bodies: a case study from South Australia

Abstract

Magnetotelluric (MT) tensors have significantly different forms depending on whether the subsurface is one-dimensional (1D), two-dimensional (2D) or three-dimensional (3D). In subsurface geological structures that are not 1D, two-dimensionality is often assumed, as inversion routines for 2D earth models are computationally more tractable than those for full 3D media. In 2D, the MT tensor decouples into two independent modes, the transverse electric (TE) mode and the transverse magnetic (TM) mode. Often only one of these modes is acquired during commercial operations. Field data were collected over an elongate magnetic anomaly of a type that would normally be approximated as 2D but which has a finite strike length and is therefore a 3D body. With this in mind, the applicability of interpreting data defined as TE and TM were assessed using (a) Mohr circles galvanic distortion analyses, (b) determination of strike of local and regional geology, and (c) comparison of 2D inversion techniques. The data were collected with the Mount Isa Mines Distributed Acquisition System (MIMDAS) in the Deep Well prospect of the Curnamona Province in South Australia. We show that the TM mode accurately delineates boundaries and that since boundary-charges are included in the inversion formulation, it also provides accurate values of apparent resistivity. The TE mode provides poor boundary delineation and underestimates the resistivity of the 3D body. Joint inversions provide only a small improvement upon TM-only inversions, but determination of dimensionality, strike and detection of galvanic distortion mean that collection of both data modes is still preferable.