Improved cavity detection from coupled seismic and hydrologic models

Seismic methods hold much promise for cavity detection, but the results from field measurements have been frustratingly inconsistent between field sites. The reasons for the inconsistencies are not fully understood, though water saturation in the near-surface may be responsible to some extent. The conventional approach has been to focus on reflections and refractions generated from the impedance contrast of the cavity wall itself, where the dimensions and geometry of the cavity should play key roles. Here, we instead focus on the influence of impedance contrasts that are generated by hydrologic processes in the adjacent porous medium. These contrasts can potentially increase or decrease the reflection/refraction footprint of the cavity itself. Detectable hydrologic anomalies can be created by the simple drainage of groundwater into the cavity (initially saturated conditions) or by the creation of a capillary barrier around the cavity (initially unsaturated conditions). Because both processes ultimately involve unsaturated conditions we use HYDRUS 2D to numerically solve the Richard's equation and simulate flow through the vadose zone. Using the generated soil moisture information and Brutsaert's (1964) saturation-velocity relation, we constructed velocity models. Our simulations suggest several scenarios where changes in saturation due to the cavity may be utilized to enhance cavity detection with seismic waves. One simulation is for unsaturated conditions in the top 10 meters of soil, where capillary forces exert a major influence on velocity. In this case, the impedance contrast is greatest for near-saturated soils. Deeper cavities (100s of meters) in permeable saturated materials are also favorable due to the sharp impedance contrast between saturated and unsaturated material. Our hydrology-determined velocity models are then used in finite-difference wave propagation simulations to determine the effects on seismic waves at various depths and saturations. Saturation features in the seismic data can then be utilized to detect cavities rather than relying on traditional yet inconsistent reflection/refraction approach. In ongoing work, we will ground-truth these models with both laboratory and experimental results. Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy's National Nuclear Security Administration under contract DE-AC04-94AL85000.