A comparison of some interpretation techniques for magnetic data

Abstract

Summary Magnetic data are used on a routine basis to help geological mapping. Although geologists often use them in a qualitative fashion to map lithological units, they can be used to determine the depth and dip of these units. Magnetic interpretation techniques can be grouped in three major categories. Firstly, magnetic data can be interpreted by interactively fitting 2D models to profile data. Secondly, inversion techniques can be used where one attempts to fit, in a least squares sense, the calculated response of a model to the measured data. Finally, automated techniques, such as Euler deconvolution (Reid, 1990) and the analytic signal (Roest et al., 1992), allow quick estimation of source location and depth. The latter have become popular in the recent years, mostly because of their apparent ease of use and ease of programming. Phillips (2000) compared horizontal gradient, analytic signal and local wavenumber methods to locate magnetic contacts. This work was based on synthetic data. Here, we use real data to compare magnetic interpretation techniques on both gridded data and line data from the Matheson area, Ontario, Canada. Interpreted source depths are compared to actual depths to basement obtained from the Ontario Drill Hole Database. All techniques, except Werner deconvolution, can be used to interpret profile and gridded data. We first review the theory and concepts underlying these techniques, discuss their applicability and see how they relate to each other. Finally, we present examples of their application on test profiles and compare the results.