Considerations regarding small-scale surface and borehole-to-surface electrical resistivity tomography

Our synthetic study focuses on small-scale Electrical Resistivity Tomography (ERT) used for high resolution imaging of the very near surface region, e.g., the first meter of the subsurface. Corresponding arrays employ an electrode spacing of a few centi- or decimetres, i.e., the electrode spacing is in the same length range as the active electrode length - the part of the electrode in contact with the ground. When simulating such ERT data, electrodes cannot be approximated as point electrodes but have to be considered with their real geometry, i.e., 3D modelling is required. Using 3D modelling, we confirm a change of the configuration factor but observe also a change of the sensitivity and coverage distributions compared to point electrode configurations. Sensitivity and coverage are high at the electrode surfaces but drop rapidly below them. This difference means a reduction in both the depth of investigation of the array and the resolution in the region below the electrodes. The effect increases with increasing active electrode length-to-spacing ratios. One way to compensate for the sensitivity reduction with depth is to add borehole electrodes. We study a borehole-to-surface (b2s) electrode array with the borehole electrodes situated in the middle of the array. As for the surface electrodes, we observe strongly increased sensitivities close to the borehole electrodes. Thus, we obtain a T-shaped region with high coverage and a good model recovery. For our synthetic models, the b2s array outperforms both the pure surface array and a pure borehole measurement in terms of model recovery. Hence, small-scale b2s ERT can improve the limited imaging capabilities of small-scale surface ERT. However, b2s is best suited to a layered subsurface as a good model recovery is limited to a region close to the electrodes. We deduce the following main recommendations for small-scale ERT from our synthetic study. For surface ERT, we found that: (i) Long electrodes (compared to their separation) decrease the model recovery and depth of investigation. For b2s ERT, the following main points can be stated: (ii) b2s ERT gives a high resolution of a layered subsurface, but the model resistivities are only well recovered in the near electrode regions. (iii) High resistivity contrasts between layers lead to artefacts, which can be suppressed by introducing a separate borehole region. Overall, we can confirm: (iv) Small-scale ERT requires 3D modelling with accurate representation and placement of the electrodes. In addition to the synthetic study, we developed a borehole tool with 20 ring electrodes and applied it in a back-filled trench using a b2s array. The inverted resistivity distribution reproduces the artificial subsurface structure and thus confirms our synthetic results. (C) 2019 Elsevier B.V. All rights reserved.