Search

Rayleigh‐wave dispersion is observed every time acoustic‐impedance stratification occurs, and its analysis is suitable for vertical shear‐wave profile reconstruction. Accurate dispersion‐curve identification is essential in order to properly determine the shear‐wave velocity distribution of a medium. Data sets characterized by several events generate complex velocity spectra that can lead to possible misinterpretations. We analyse a real data set by taking into account theoretical dispersion curves and synthetic data obtained from numerical simulations in order to avoid possible pitfalls that could arise from the complex trends exhibited in the f−k(frequency–wavenumber) and v−f(velocity–frequency) domains. In the v−fdomain in particular, we show that reflection events and their multiples generate coherences that could be misinterpreted because of their similarity to typical higher‐mode dispersion curves. Another observed signal is interpreted in terms of guided waves and related phenomena. The results of the fundamental‐mode dispersion curve inversion performed via genetic algorithms indicate a sedimentary cover stratification that simple reflection analysis cannot reveal. The present case study highlights the importance of a synergic approach, based on integrated synthetic and field data analysis, for correct interpretation of all the wavefield components in the velocity spectrum.

We focus on the application of Ground Penetrating Radar (GPR) to evaluate limestone characteristics of interest in environmental and engineering studies, and in particular: a) to image joints, bedding planes and cavities; b) to improve accuracy and resolution of the method; c) to evaluate joint/bedding planes characteristics which affect the radar response with particular reference to thickness, sedimentary infilling, water/clay content and spatial frequency. The work is based on experiments carried out close to road cuts and cavities, where the exposed rock face allows calibration and validation of results. The test‐sites are located in NE‐Italy and are part of the Peri‐Adriatic carbonate Platform. The rocks in the area of study date back to the Paleocene and are mainly peritidal regressive limestone sequences consisting of more than 90% carbonates. Joints and bedding planes surfaces, as results from direct inspection at the rock face, are sufficiently smooth to neglect roughness effects. We demonstrate that imaging can be improved by non‐conventional data acquisition paradigms, and in particular by multi‐offset (linear multi‐fold, LMF) methods. Multi‐offset/multi‐azimuth (azimuthal multi‐fold, AMF) techniques were exploited to select optimum grid orientation and offset range for LMF application. Multi‐fold velocity analysis on CMP gathers shows rather constant wave propagation velocities. They slightly vary between different test‐sites (between 10 cm/ns and 12 cm/ns) primarily due to macro/microscopic characteristics of limestone, such as spatial frequency of joints and porosity respectively, which affect water content in the rock mass. Constant velocities allowed application of post‐stack time migration algorithms. Such algorithms attained imaging accuracy below 3% in the reconstruction of the detectable discontinuities at most test sites. A Kirchhoff algorithm proved to be the optimum solution as it effectively handled the steep dips (up to 70°) that characterize most of the examined sets of joints. Enhanced data quality further results from the application of original processing techniques, such as, in particular, Hough Transform based coherent noise/background removal and Wavelet Transform based instantaneous parameters computation and analysis. Maximum penetration depth at the examined test sites ranges between 15 m for 250 MHz (central frequency) bow‐tie shielded antennas and 23 m for 50 MHz unshielded resistively loaded linear dipoles. As for resolution, antennas in the range of 200‐250 MHz apparently provide effective discrimination of joints/bedding planes spaced not less than 40 cm, and seem therefore the adequate choice at the examined test sites. Low frequency antennas, in the range of 50 to 100 MHz (central frequency), provide a maximum 55% increment in penetration depth at the expenses of a substantially diminished resolution (around 20% of that provided by 200‐250 MHz antennas). The comparison of modelling results to multi‐fold data allows discrimination of radar response from joints filled by air and clayey deposits. Such results were validated by geological evidence at the exposed rock face.

An international archaeological and geophysical expedition was carried out to study Iron Age tombs of the Saka culture in southeastern Kazakhstan. Single‐fold and multifold ground‐penetrating radar (GPR) surveys, in two and two and a half dimensions, focused on the identification of burial chambers beneath the mound‐shaped Saka tumuli, named kourgans. Primary objectives of the geophysical study were the testing and optimization of GPR techniques for studying this type of tomb and the reconstruction of subsurface images of use for the design of archaeological excavations. Single‐fold GPR grids with line spacing ranging between 50 and 200 cm were obtained at each of the 13 sites selected. After identification of the sectors of major potential interest from data processing performed directly in the field and interpretation of the single‐fold grids, multifold profiles with average 1200 per cent fold were designed and performed. Processing of the data focused on velocity analyses, pre‐stack coherent noise attenuation by means of FK and KL filters, stacking and analysis of the instantaneous attributes of the complex radar trace. Multifold GPR data processing techniques achieved a substantial increment in the signal to noise ratio at most sites, allowed accurate depth conversions of the stacked profiles (average deviation of estimated target location from the actual target was less than 5% of total depth) and provided indications to identify the characteristic response from the target of interest through the joint interpretation of pre‐stack gathers, velocity analysis and amplitude versus offset analysis. Interpretation of stacked sections and three‐dimensional models obtained from the single‐fold/multifold grid allowed the identification of a burial chamber located at an estimated depth of 140–150 cm, which was excavated successfully by the archaeological team. Further reflectors with similar characteristics are shown from other sites, which will be excavated by the next archaeological expeditions to the area. Copyright © 2001 John Wiley & Sons, Ltd.

An international archaeological and geophysical expedition was carried out to study Iron Age tombs of the Saka culture in southeastern Kazakhstan. Single-fold and multifold ground-penetrating radar (GPR) surveys, in two and two and a half dimensions, focused on the identification of burial chambers beneath the mound-shaped Saka tumuli, named kourgans. Primary objectives of the geophysical study were the testing and optimization of GPR techniques for studying this type of tomb and the reconstruction of subsurface images of use for the design of archaeological excavations. Single-fold GPR grids with line spacing ranging between 50 and 200 cm were obtained at each of the 13 sites selected. After identification of the sectors of major potential interest from data processing performed directly in the field and interpretation of the single-fold grids, multifold profiles with average 1200 per cent fold were designed and performed. Processing of the data focused on velocity analyses, pre-stack coherent noise attenuation by means of FK and KL filters, stacking and analysis of the instantaneous attributes of the complex radar trace. Multifold GPR data processing techniques achieved a substantial increment in the signal to noise ratio at most sites, allowed accurate depth conversions of the stacked profiles (average deviation of estimated target location from the actual target was less than 5% of total depth) and provided indications to identify the characteristic response from the target of interest through the joint interpretation of pre-stack gathers, velocity analysis and amplitude versus offset analysis. Interpretation of stacked sections and three-dimensional models obtained from the single-fold/multifold grid allowed the identification of a burial chamber located at an estimated depth of 140–150 cm, which was excavated successfully by the archaeological team. Further reflectors with similar characteristics are shown from other sites, which will be excavated by the next archaeological expeditions to the area. Copyright © 2001 John Wiley & Sons, Ltd.