Abstract Detection of anomalies such as voids in the shallow subsurface using noninvasive geophysical techniques has proved to be challenging at best. Three near-surface seismic methods are introduced, including diffracted body waves, backscattered surface waves, and changes in reflection moveout velocities to detect voids directly or their effects on surrounding material properties using different parts of the wavefield. Examples are presented, including modeled and field data sets to demonstrate each technique. Body-wave diffractions were used to identify and locate man-made tunnels in multiple geologic settings. Variations in shear-wave reflection velocities are shown to correlate to changes in stress over known void locations; backscattered surface waves are shown to correlate with a known void location. Results of the studies show that the field data correlate well with the synthetic, and these methods show promise in furthering the ability to locate subsurface voids and their effects on the surrounding media.

Abstract The utility of two varied approaches to first-arrival time analysis of seismic data acquired at several unique levee sites is demonstrated by solving the inverse refraction-traveltime problem (IRTP). These data were evaluated using conventional refraction tomography and joint analysis of refractions with surface waves (JARS). The JARS approach uses a reference model, derived from surface-wave-calculated shear-wave velocity estimates, as a constraint in reducing refraction nonuniqueness. At those levee sites, conventional refraction-tomography and JARS methods provided different solutions, equally matching the observed data. This observation suggests both approaches are equally possible from a numerical perspective. The JARS images reveal horizontal layering patterns, laterally uniform velocity trends, mild velocity variations, and channel-like features consistent with geologic expectations. In addition, the JARS approach demonstrated the capability for imaging low-velocity layers/zones, something not seen using conventional refraction or refraction-tomography techniques. As a result of these qualitative observations, without ground truth to support an earth model (e.g., from wells), the JARS approach can be viewed as an additional method for finding solutions to the IRTP. However, from all evidence in those studies, the JARS approach represents a possible solution and an example of the potential adverse affect of nonuniqueness. These empirical results support the understanding that for a given refraction data set, significantly different and equally possible velocity-model solutions can exist, resolving which is truly best using invasive ground truth.