An Improved Post-Processing Technique for Array-Based Detection of Underground Tunnels

This paper investigates challenges faced by many geophysical algorithms applied to real-world cases such as the Attenuation Analysis of Rayleigh Waves (AARW). AARW shows great promise in terms of detecting shallow underground tunnels. However, in-situ subsurface anomalies, including those due to anisotropy, and instrument sensitivity to natural conditions can significantly degrade the utility of this technique. To address this problem, this work proposes a data acquisition scheme and develops a new post-processing approach. The first applied measure estimates the confidence level of each detection result. The second processes the data in sub-arrays, and filters out false alarms. The third scans all detections and searches the cluster with the highest cumulative confidence level. This paper provides engineering practitioners with a simple and efficient method to reliably determine tunnel locations. Experimental results derived from data recorded in various testing sites and surface conditions verify the effectiveness of this work.