Deep void detection with 3D full waveform inversion of surface-based and in-depth source seismic wavefields

Abstract Detection of subsurface voids using nondestructive seismic methods is an ongoing problem in many areas of civil and environmental engineering (e.g., sinkholes and caves), homeland security (e.g., tunnel detection), and mining applications (e.g., abandoned mines). Recent advances in 3D full waveform inversion (FWI) technology have made it possible to scan large volumes of the underlying materials efficiently, providing a glimpse into the state of subsurface conditions. A challenge in applying 3D FWI methods to the detection of voids emerges from their embedment depths. Shallower voids are easier to detect due to their large signature on the surface seismic response, whereas deeper voids have a much smaller signature and are therefore much harder to detect. This is not a limitation of the FWI method, but rather that of the seismic field-testing techniques and data gathering processes. The goal of this study is to investigate ways to overcome these limitations and improve void detection depths. One way to achieve this is through the application of a large surface source, generating more energy at lower frequencies (longer wavelengths), thereby increasing the penetration depth. Another way is by increasing the contribution of body waves and utilizing the diffraction/transmission information embedded in the waveforms. The latter is achieved through the application of a recently developed SPT-seismic method, where the standard penetration test (SPT) device is used to generate wave motion from within the subsurface. Both source methods and a newly developed 3D Gauss-Newton FWI method are utilized here to detect a deep void (25–45 m depth) in limestone, on the southern peninsula of Florida. The results are compared with SPT and Sonar profiles obtained from the test site. Overall, a good image of the deep void is achieved, matching observations from the invasive results. The findings provide useful insight into the application of FWI technology for detecting deep subsurface voids and anomalies that are typically hard to identify.