A Targeted Approach for Imaging Near-Surface Geological Faults using Geophysical Surveying

Fault zones are critical in applications involving subsurface flow, as they can serve as conduits or barriers. Detecting and characterizing faults in the near-surface is challenged by limited surface exposure and sparse borehole data. While electrical resistivity tomography (ERT) and seismic refraction tomography (SRT) are commonly employed to investigate fault systems, traditional smooth inversion methods obscure sharp structural features and provide limited uncertainty information. To overcome these limitations, we introduce an ensemble-based inversion workflow that integrates layer-based parameterization with particle-swarm optimization to generate models with sharp interfaces and realistic fault geometries. Results from a synthetic experiment (resembling a normal fault) highlight the high non-uniqueness of the inverse problem, even for a reduced number of model parameters (< 20). Although significant uncertainties are expected, particularly at depth where sensitivity is low, the inversion results can provide insight into an expected fault zone and guide borehole placement.

Because the proposed parameterization is flexible, additional structural information (e.g., fault dip angle and depth to interfaces) can be easily integrated, leading to more realistic models and reduced parameter uncertainty. Our approach opens new avenues for inverting and interpreting fault systems from geophysical data and could be adapted to more complex geological settings.