Parallel-line beamsteering for enhanced anomaly detection

Abstract

Parallel-line beamsteering, a new seismic processing method, was developed to enhance signal from laterally continuous heterogeneities (e.g. tunnel, mine shaft, etc.) and attenuate scatter from localized non-laterally contiguous heterogeneities (e.g. clay lens, boulder, etc.). This method takes advantage of the linear cross line nature of potential targets. Parallel-line beamsteering applies spatial shifting and vertical stacking to parallel-line seismic data in combination with backscatter analysis of surface waves (BASW) and diffraction imaging processing. A processing flow that optimizes the stack and shift process was empirically determined using the standard BASW and diffraction imaging processing flows. This empirically established processing flow was evaluated on parallel-line data acquired at a tunnel test site at the Yuma Proving Grounds in southwestern Arizona. Results show that this method enhances signal from the tunnel by as much 2.8 dB and attenuates other scatter events by as much as 3.9 dB. Beamsteering analysis also reduced the number of possible tunnel locations interpreted from diffraction imaging results using standard interpretation based on spatial matching criteria by up to 66% and from the BASW imaging results by up to 50%. Interpreting results of parallel-line beamsteering requires a less skilled interpretation than conventional tunnel detection methods. This method can also be used to determine the orientation of a tunnel with respect to the two seismic lines. Additionally, this method maintains the small acquisition footprint of current 2D seismic methods required for tunnel detection applications.