Evaluation of two-dimensional resistivity methods in a fractured crystalline-rock terrane

Abstract

A series of two-dimensional resistivity profiles collected in the Blue Ridge Province of southwest Virginia and results from numerical modeling of synthetic data reveal substantial differences in depth of investigation, resolution, and sensitivity using Wenner, Wenner–Schlumberger, dipole–dipole, and pole–pole data collection techniques. Resistivity profiles were collected using short (2-m electrode spacing, 48-m profile length), intermediate (6-m electrode spacing, 144-m profile length), and long (10-m electrode spacing, 240-m profile length) arrays over shallow unconsolidated soils and regolith overlying crystalline bedrock. Pole–pole data were only collected with the short array. Numerical modeling was used to simulate both vertical and horizontal structures similar to subsurface conditions in the study area. All of the apparent resistivity data were inverted into earth models using a computer program that uses an l 1 norm smoothness constrained inversion technique. Earth models generated from both field data and numerical modeling acquired by the dipole–dipole technique consistently indicated more detail and greater depth of investigation than the other techniques. The dipole–dipole method uniquely imaged thin saturated sands and isolated high resistivity bodies beneath the 48-m-length array, significant horizontal and vertical resistivity variations including a thick transitional resistivity zone in the 144-m-length array, and anomalous low resistivity zones in crystalline bedrock in the 240-m-length array. Earth models created from the Wenner and Wenner–Schlumberger apparent resistivity data had shallower depths of investigation and revealed significantly less geologic detail than the profiles generated from the dipole–dipole surveys. The earth model from the pole–pole data had the greatest depth of investigation but low resolution and limited geologic detail when compared to the dipole–dipole survey.