GPR and ERT combined analysis on the basis of advanced wavelet methodology: The Montagnole testing area

Abstract

Ground Penetrating Radar (GPR) and Electric Resistivity Tomography (ERT) are well assessed and accurate geophysical methods for the investigation of subsurface geological sections. In this paper, we present the joint exploitation of these methods al the Montagnole (French Alps) experimental site with the final aim to study and monitor effects of possible catastrophic rockslides in transport infrastructures. It is known that factors as the ambiguity of geophysical field examination, the complexity of geological scenarios, and the low signal-to-noise ratio affect the possibility to build reliable physical-geological models of the investigated subsurface structure. Here, we applied for the GPR and ERT methods at Montagnole site, the recent advances in the wavelet theory and data mining. Wavelet approach was specifically used to achieve enhanced (e.g., coherence portraits) images resulting from the integration of the different geophysical fields. This methodology based on the matching pursuit combined with wavelet packet dictionaries permitted us to extract desired signals in different physical-geological conditions, even in presence of strongly noised data. Such tools as complex wavelets were employed to the coherence portraits, combined GPR-ERT coherency orientation angle, to name a few, enable performing non-conventional operations of integration and correlation in subsurface geophysics. The estimation of the above mentioned parameters proved useful not only for location of buried inhomogeneties but also for a rough estimation of their electromagnetic and related properties. Therefore, the combination of the above approaches has allowed to set-up a novel methodology, which may enhance reliability and confidence of each separate geophysical method and their integration.