Buried object detection using cross-correlation and wavelet coherence analysis of MASW datasets

Abstract

This study focuses on buried object detection using the cross-correlation and wavelet coherence of multichannel analysis of surface waves (MASW) datasets. Underground objects (pipes, building foundations, and rocks) create large-scale heterogeneity in the soil medium. The dispersion curves obtained from the multi-configuration roll-along MASW tests can provide approximate locations of buried objects within the corresponding sensor layouts. However, analyzing the responses of pairs of sensors from the MASW wavefields can indicate the more precise location of such objects. Two circular underground pipes and one square-shaped buried object were simulated using finite element modeling, representing different field scenarios. The wave scattering phenomena have been studied by looking at the wave propagation snapshots in these heterogeneous soil mediums. The cross-correlation analysis from the responses of sensors indicated the exact horizontal position of the buried object, further verified by the wavelet coherence of pairs of sensors. Considering the horizontal position of the different buried objects and phase of the high wavelet coherence patch, the values of phase velocities of 3114, 2308, and 2412 m/s, and respective wavelengths of 30.52, 23.08, and 24.12 m were calculated, which were used to interpret the shear wave velocities ranging from 2538 to 3424 m/s and depths ranging from 4 to 6 m of buried objects. These analyses accurately determined the physical horizontal position and embedment depth of square- and circular-shaped buried objects than the conventional 1D MASW approach.