USING SUBSURFACE SOIL DISPLACEMENT TO ESTIMATE CHANGE IN WATER CONTENT

The water balance at the ground surface is an important component of the behavior of hydrologic systems, but measuring changes at the scale of typical hydrologic problems remains a challenge. We have developed a technique that uses a sensor deployed at depth in a boring to characterize changes in water content in the overlying soil. The area affecting the sensor scales with the depth of the installation, so it is possible to estimate changes over regions that are 10s to 100s m across.

The technique uses a device called a Sand Extensometer, or Sand-X. It is designed for applications in unconsolidated formations where it measures vertical soil displacement in response to changes in load at the ground surface. A change in soil moisture results in a change in load causing the underlying soil to deform. By measuring the displacement, we are able to correlate the deformation to changes in load at the ground surface and ultimately to changes in moisture content.

Several field-scale studies are ongoing at a site near Clemson, SC which is underlain by saprolite weathered from biotite gneiss. Measured displacement correlates to precipitation, ET, and other loads caused by barometric fluctuation and human activity. A typical rain event abruptly compresses the soil by more than 1 µm and is followed by a gradual expansion at rates of 150 to 800 nm/da, presumably due to ET. Diurnal fluctuations in displacement of up to 1 µm/da correlate to barometric pressure changes and can mask short-term, low-magnitude, hydrologic processes. To address this problem, we have developed a poroelastic model that predicts displacements due to barometric loading and which can be used to reduce these effects.

The soil response is calibrated to equivalent water load by correlating measured rainfall to the corresponding soil displacement. For example, one instrument responds at a rate of about 185 µm/mm equivalent water load. Estimates of ET track within 25% of monthly averages of daily ET based on historical pan evaporation data. In contrast, daily averages reach double monthly averages of daily ET during wet periods as a result of increased availability of soil moisture for evaporation. Current results suggest that this technique is a viable tool for estimating processes affecting the water balance.