Depth slicing of multi-receiver EMI measurements to enhance the delineation of contrasting subsoil features

Mobile multi-receiver electromagnetic induction sensors provide simultaneous readings of the apparent electrical conductivity (ECa) from overlapping soil volumes. Therefore, small contrasting features can be difficult to identify because they have a limited contribution to the bulk measurement, especially if they are present in the subsoil (i.e. beneath the topsoil). Integrating ECa data from simultaneous measurements with multiple coil configurations has the potential to elucidate the variability within the soil profile as it enables modelling the electrical conductivity (EC) for distinct depth intervals. Therefore, it was our objective to develop a methodology to enhance the delineation of contrasting subsoil features, such as in-filled gullies and archaeological features. We selected a 3.5 ha study site where contrasting features were expected. A three-layered build-up was taken as the initial EC-slicing model. After varying the interface depths, the shallowest and deepest EC-depth slice showed a clear minimum of their combined variances at interface depths of 0.36 m, which corresponded to the ploughing depth, and 0.86 m. This implies that the EC-depth slice in-between these depths, contained a demonstrably higher variability. A sub-area of 0.85 ha was completely excavated to a depth of 0.7 m, revealing the subsoil features and the host material. An automated edge detection algorithm showed that the EC-depth slice was superior to any individual ECa measurement for delineating the contrasting subsoil features. Therefore, we concluded that depth slicing by integrating simultaneous ECa signals from a multi-receiver EMI sensor clearly improved the identification of subsoil features.

► Multi-receiver EMI soil sensing was used to perform EC-depth slicing.
► A three-layered soil model was assumed to model the EC within depth slices.
► Depth intervals for EC-depth slicing were modelled.
► EC-depth slicing clearly improves the discriminating potential of subsoil features.