Elastic-wave identification and extraction through array processing: An experimental investigation at the laboratory scale

The mix of body waves and surface waves is a recurrent problem for deep exploration in geophysical contexts. As surface waves represent up to 70% of the recorded energy, they hide a large part of the information coming from the sub-surface through body waves. Efforts have been made in the past to better filter or remove surface waves: however, their impact is always far from negligible, especially with strong backscattering contributions. In parallel, taking advantage of an always growing number of channels, geophysical explorations face new opportunities to enhance the quality of Earth imaging. For example, better spatial sampling is a way to better use or remove surface waves. There are compromises to find between higher spatial sampling and operational costs, even for on-field tests. In this context, surface-wave studies at the laboratory scale are a flexible way to evaluate new acquisition designs and processing. This study shows how a gel-based phantom can be used successfully to study elastic-wave mixing in the context of geophysics prospection. Small-scale experiments provide the records of thousands of traces. Using projections in the slowness/angle domain, wave separation and identification algorithms are proposed, with the goal of being able to adapt array processing to geophysical-like designs. (C) 2011 Elsevier B.V. All rights reserved.