DC-Electrical Resistivity Imaging for embankment dike investigation: A 3D extended normalisation approach

• The geometry and the reservoir of a dike have an effect on resistivity measurements.
• A 3D forward solver can quantify those 3D external perturbing effects.
• “Conventional” normalisation techniques can cancel part of the previous effects.
• A new extended normalisation technique is developped and tested.
• The new procedure improves the relevance of the result in a cost effective way.

Levee, dike and earth embankment dam structures are difficult to assess because of their length and complexity. Managers often include geophysical investigations in the overall dike condition assessment and the DC-Electrical Resistivity Imaging (ERI) method is particularly applicable owing to its cost-effectiveness and its potential sensitivity to internal erosion. However, due to the truly 3D nature of embankment dikes, implementing inline longitudinal tomographies along with conventional 2D inversion is likely to yield image artefacts. 3D effects from external causes (geometry, water reservoir) can be predicted and therefore we present a new approach based on redefining the normalisation principle to derive apparent resistivities from the measured data. The aim is to provide a set of pre-processed apparent resistivities that are not contaminated by external 3D effects and that yield more reliable results when processed within a 2D conventional inversion scheme. The presented approach is successfully applied to synthetic and real data sets, proving superior to the conventional 2D approach, although data acquisition approach is the same thus keeping the same cost-effectiveness.