Application of 3D Ambient Noise Tomography for Void Detection

A first-time application of 3D ambient noise tomography (3D ANT) method is presented for detection of deep voids. The method is based on the solution of 3D P-SV elastic wave equations and adjoint-state optimization to directly invert experimental cross-correlation functions (CCF) for extraction of S-wave velocity (Vs) structure. The main advantage of this approach is that it does not rely on assumptions of energy balance and far-field waves as required by methods using characteristics of Green’s function (GF). Instead, the source power-spectrum density is inverted to account for source distribution (location and energy), which allows to exploit full information content of all available CCFs from ambient noise recordings. The 3D ANT capability in detecting deep voids is investigated at a test site in southern Florida. For the field experiment, 72 vertical geophones of 4.5-Hz resonance were deployed in a 4 × 18 grid over 9.0 × 76.5 m area on the ground surface to record noise data for 34 min. The CCFs extracted from the noise recordings have good energy at 5–20 Hz and a consistent wave propagation pattern for the entire test area. The inverted result reveals that the 3D ANT was able to image a large deep void at 28-to 44-m depth, which is generally consistent with results from invasive SPTs.