Laboratory scale seismic analysis of a spatially variable hydrological surface in unconfined, unconsolidated sand

The ability to measure the depth to the water table can provide information such as reservoir characteristics, soil conditions for agriculture, and actions to be taken with contaminant flow and removal. We attempt to detect this boundary in a larger-than-lab scale experiment. We attempt to simulate a spatially variable water level and attempt to image changes in the depth of this water level. We try to reduce the normally complex natural conditions to those of a nominally homogeneous and isotropic, unconfined sand volume for modeling. These simplified conditions help isolate the effects of remaining complexities such as the variable saturation of the transition zone between the residual saturation zone and the capillary fringe above the completely saturated region. As a part of the experiment, an ultra-high frequency (2-15 kHz) multichannel, multi-component acquisition system was built in-house for a target depth of 30 cm. Complete saturation within the tank is found to be unfeasible. Instead of a water table, a zero-tension surface level is increased in the wavetank (5 m by 5 m by 1 m) from the side. Seismic reflection and refraction arrivals from a linear seismic array are selected for travel time inversion to develop a 1-D velocity model of the system. No seismic arrival is found to directly correspond with the zero-tension interface; however, the increased zero-tension levels in the wavetank appear to decrease the velocity within the wavetank. A velocity-saturation relationship could possibly be established with more information on saturation conditions within the sand body of the wavetank