Constraining Airborne Electromagnetic Interpretation with Regolith Stratigraphy and Landscape Evolution Processes

Abstract

A typical product from an airborne electromagnetic (AEM) survey is a conductivity depth image (CDI) along each of the flight lines. These CDIs overcome the problem of non-uniqueness by choosing one model that fits the data, typically a smoothest model. However, in the case of using AEM for describing the stratigraphy of the regolith, an understanding of the landscape evolution processes that formed the regolith gives us knowledge about what the stratigraphic units are that make up the regolith, and also something about their likely geometry. In addition, knowledge of their mineralogy tells us something about their likely ranges of conductivity, and understanding of the processes that formed them tells us about their geostatistical properties. For example, materials which are well mixed, such as channel clays, will typically be homogeneous over large distances, whereas material that has formed by in-situ weathering could be much more heterogeneous. It therefore makes sense to try to invert the AEM data for stratigraphic boundaries and conductivity variations within stratigraphic units rather than smooth models. This immediately gives estimates, with uncertainty bounds, for the depths to various interfaces, which are of more direct interest to a geologist than the conductivity values in the CDI. The work presented here shows how the use of geological contextual information can produce improved inverted models of the regolith vertical and lateral stratigraphy from AEM data. A better understanding of regolith architecture then allows for optimisation of drilling sites for geochemical sampling, focussing on stratigraphic units where the geochemical footprint of an orebody is likely to have been concentrated.