Semblance-based topographic migration (SBTM): a method for identifying fracture zones in 3D georadar data

Steep-dipping fracture zones are generally difficult to delineate using traditional ground-penetrating radar (georadar) techniques. Evidence for their presence in standard georadar images may be either completely absent or limited to diffractions and/or chaotic reflection patterns. To address this issue, we present a novel three-dimensional (3D) migration scheme based on computations of semblance. This new approach, which accounts for undulating surface topography, emphasizes diffractors while markedly reducing the effects of specular reflectors. After demonstrating the efficiency of the technique on 3D synthetic data, we apply it to a 3D georadar data set acquired across an unstable mountain slope in the Swiss Alps. This region is characterized by rugged topography and numerous shallow- to steep-dipping fracture zones and faults. Only the shallow- to moderate-dipping structures are imaged as reflectors in conventionally migrated versions of the georadar data. Our semblance-based topographic migration (SBTM) scheme produces a 3D volume containing clouds of high-semblance values. Application of morphology image processing to these clouds reveals the presence of geologically meaningful structures, most of which are very steeply dipping (>75°). Several of the steep-dipping structures project to open fracture zones and associated lineaments at the surface, thus demonstrating the capability of the combined SBTM and morphology procedure for mapping near-vertical fracture zones.