Strategies to invert a suite of magnetic field anomalies due to remanent magnetization: an example from the Georgetown area of Queensland

Abstract

Most magnetic field interpretation is based on the assumption that magnetization is in the same direction as the ambient geomagnetic field, and there is a widely held suspicion that anomalies cannot be inverted without this assumption. In this paper we aim to establish that the problems of uncertain magnetization direction can be largely overcome utilising both staged inversion and magnetic moment analysis (MMA). Induced magnetization has a single, known direction. Not only may a remanent magnetization direction be unknown, but it may also be variable across a wide range of scales. Palaeomagnetic studies address this variability with collection of multiple samples from multiple sites (generally outcrops, quarries or road-cuts) for hierarchical statistical analysis. For rapidly cooled volcanic or sub-volcanic bodies, magnetization directions may vary due to short-term changes in the geomagnetic field. For deeper, more slowly cooled bodies, or for bodies formed by multi-phase intrusions, magnetization direction may vary due to apparent polar wander. Furthermore, remanent magnetization of rocks with a complex thermal history may include over-prints of components of different direction acquired at different times, with variable contributions to the overall magnetization in different samples. Magnetization directions may also vary due to any post-acquisition tectonic rotations. The magnetic field generated by a body depends on the resultant of induced and remanent magnetizations, determined by the Koenigsberger (or ?Q?) ratio, which generally is also highly variable across a wide range of scales. To properly incorporate remanent magnetizations in a magnetic field interpretation we can not expect to just replace the geomagnetic field direction of an induced magnetization with a predetermined remanent magnetization direction. Rather, the interpretation of anomalies caused in substantial part by remanent magnetization requires flexible and adaptive methods. In this paper we firstly review the derivation of magnetization direction by MMA of vector components and the gradient tensor of the magnetic field. Then we investigate the combination of inversion and MMA in a study of the Newcastle Range Volcanics in north-west Queensland. This volcanic province has bodies with a wide range of shapes, sizes, complexities and settings, many with magnetizations dominated by remanence. Each anomaly presents an opportunity to determine magnetization direction with its own particular challenges. A palaeomagnetic study of the same rocks by Anderson et al (2003) provides data with which to validate and classify magnetization directions derived from the magnetic field interpretation.