Fuzzy c-mean (FCM) integration of geophysical data from an iron-oxide copper gold (IOCG) deposit under thick cover

Abstract

Geophysical methods depend on a range of physical and chemical mechanisms, and each method has different sensitivities and resolution of Earth parameters, and over different scale lengths. Additionally, such geophysical methods will also have their own statistical characteristics. Thus, it is a significant challenge to combine different methods through a single inversion framework. Rather than attempting to find an optimal and single model for different geophysical responses, an alternative approach is to use FCM clustering to identify clusters of parameters from two or more different geophysical data sets or models that have similar statistical properties. In this paper, we apply the FCM approach to integrate data and model sets for an array of 100 broadband magnetotelluric (MT) and 100 passive seismic receivers spaced 1 km apart on a 10 by 10 grid above the Vulcan IOCG prospect in the Olympic Cu–Au Province, southern Australia. The challenge for exploration of the Vulcan prospect is that it lies beneath 750 m of regolith sedimentary cover, no single geophysical method provides a unique characterization of the deposit geometry, and drilling is very expensive. Fuzzy c-mean cluster analyses are undertaken in 2D for gravity data and shear-wave velocity model data for basement depths beneath cover and in 3D for shear-wave velocity and electrical resistivity model data. Fuzzy c-mean clustering is shown to provide a simple and efficient method of integrating different geophysical measurements to produce a geological framework that can be verified with drilling information.