3D Multiple Body Parametric Inversion of Time-domain Airborne EM Data

Abstract

We developed a 3D parametric inversion for time-domain airborne EM data using a skewed ellipsoid representation for multiple conductive or resistive anomalies. The approach aims to simplify the task of imaging thin, potentially highly conductive, anomalies with 3D EM inversion. The algorithm finds the optimal location, shape, size and resistivity of the anomalies in a homogeneous or heterogeneous background by employing a Gauss-Newton style optimization. Our parametric method is tested on a synthetic and field data set. The synthetic model is composed of two narrow dipping conductive anomalies in a resistive background along with a vertical narrow conductor. The survey layout and resistivity structure is based off field data from a greenstone setting. The parametric inversion accurately recovers the spatial extent and dips of the three synthetic anomalies, although the depth extent of the anomalies is exaggerated. In the greenstone field example, the inversion defines the spatial location, extent and dips of three conductive anomalies to provide a new conductivity interpretation of an area where little information is known regarding the true nature of the conductors.