Application of ambient noise tomography for deep void detection

Abstract

This study aims to take the advantage of ambient noise recordings rich in low-frequency components for deep site characterization. We investigate the capabilities of a recently developed ambient noise tomography (ANT) method for imaging deep buried voids via both synthetic and field experiments. A challenging synthetic model with two deep voids was used to demonstrate the practicality of this ANT approach. To further test the method's capability, we conducted a field experiment at a bridge construction site in Miami, Florida, which contained a large and deep void (28 m to 44 m depth). The cross-correlation functions (CCFs) of the traffic noise recordings were directly inverted to provide subsurface S-wave velocity profiles. The results demonstrate that the method is capable of imaging deep voids. The in-situ standard penetration test (SPT) data was then compared to the inverted S-wave velocity obtained by the ANT approach. It shows that the trend of Vs and SPT values are generally in agreement, including the identification of the void and its depth. The field results suggest that the ANT is a useful geophysical tool for roadway imaging, particularly for detection of deep voids that are difficult to be imaged by active-seismic methods.