The potential for improving near-surface characterization through converted-mode (P-S and S-P) seismic reflection imaging was tested experimentally and assessed numerically. High-resolution nine-component reflection data were acquired over a geologic sequence type (flat-lying unsaturated and saturated overburden materials above consolidated units) commonly encountered during geotechnical investigations. Interface energy partitioning analyses, event predictions, and elastic-wave modeling were conducted using parameters derived from field data. Common-mode (P-P and S-S) reflections from known lithologic boundaries were observed in field data; however, predicted converted-mode events were not. Modeling demonstrated that in theory, potential benefits of converted-mode reflection imaging exist for the tested subsurface conditions, but that such benefits were unable to be practically realized using field measurements, due to the low amplitudes of converted-mode reflections and the detrimental affects of random and coherent noise modes. This study shows and explains why it is not practical to use converted-mode reflection imaging for improving characterization under such subsurface conditions. The developed analysis methods can be applied to investigate converted-wave imaging potential under different conditions, and to identify near-surface media situations that may yield converted-mode reflections with amplitudes comparable (at moderate angles of incidence and over a substantial incident angles range) to those of common-mode reflections.
Get the entire paper (zip)
Go back to the TOC