Mapping soil layers using electrical resistivity tomography and validation: Sandbox experiments

Abstract

Knowledge of the geologic structure at a field site is a useful piece of information for hydrologic modeling since it can serve as more site-specific prior information about the hydraulic parameter patterns at the site than generic spatial statistics. Widely accepted electrical resistivity tomography (ERT) survey for mapping subsurface anomalies could be a viable tool for acquiring this information. However, their ability to delineate geologic structures has not been thoroughly investigated. In this study, two-dimensional ERT numerical experiments were first conducted to study the effects of boundary conditions on the dipole-dipole and pole-pole array configurations. An ERT setup subsequently was implemented in a sandbox consisting of complex layers of different sands. A continuous copper wire was installed along the sides of the sandbox to impose potential boundaries. Using data collected with the pole-pole array in the sandbox under different degrees of drainage and the Successive Linear Estimator (SLE) algorithm, we show that ERT yields electrical conductivity estimates of complex layer structures with small uncertainties. In addition, using SLE with physically meaningful correlation scales as prior information can lead to an electrical conductivity field that is consistent with visually observed layer structures. The correlation scale concept also was demonstrated to provide guidance to the design of the electrode spacing in the surveys. Moreover, the estimated field was validated by predicting electrical potential fields from two independent ERT surveys using electrodes at different locations. Results of this study suggest that the combination of ERT and SLE is a viable geophysical survey tool for mapping geologic layer structures. Research SignificanceThis study develops a method to implement prescribed potential boundaries for enhancing the pole-pole ERT survey, illustrates the importance of correlation scales and develops an approach for validating ERT results.