Evaluation of the 3-D Multichannel Deconvolution Method for the Case of Low S/N Inphase Data Collected With Loop–Loop FD-EMI Sensors

Frequency domain electromagnetic induction (FD-EMI) sensors using a double magnetic dipole (loop-loop) geometry are efficient tools to rapidly sense the electrical and magnetic properties of the subsurface. Recent progress including the development of multi-configuration FD-EMI systems, efficient acquisition strategies using total-station tracked platforms, and fast full 3D inversion procedures allows the reconstruction of 3D subsurface electrical and magnetic properties distributions with unprecedented details. In this study, we analyse the capabilities of a recently proposed multi-channel deconvolution (MCD) procedure for the 3D inversion of low S/N FD-EMI inphase data. By using a synthetic experiment, we demonstrate how such a method can unveil subsurface magnetic signatures, which might be hardly detected by a qualitative analysis of the data space without inversion. We further confirm these observations using a field data set acquired at a Gallo-Roman archaeological site in Orcines, France. Here, the 3D MCD of FD-EMI inphase data unveils the presence of archaeological remains that are not visible in the original inphase data maps. In addition, the overall appearance of the image is improved with better resolved signatures. This study confirms the limitations of qualitative analysis of FD-EMI data and thereby demonstrate the potential of 3D acquisition and interpretation procedures, which are able to optimally exploit the information provided by FD-EMI inphase data.