| Abstract: | Upon desiccation, the matrix of Vertisols and other expansive soils shrinks. Matrix shrinkage results in the formation of cracks that can alter the hydrology of the soil. Despite the importance of cracks, many hydrologic models do not account for cracking due in part to a lack of reliable information on the development and morphology of cracks. Electrical resistivity tomography (ERT) has shown promise as a new, non-destructive method of monitoring cracking in the field. We investigated the use and limitation of ERT for monitoring the spatial degree and extent of cracking in a Texas Vertisol. First, we examined the relationship between soil water content and ERT derived bulk soil electrical resistivity. Results showed that when the soil was cracked, ERT is insensitive to changes in water content with the electrical resistivity of the soil much greater than would be predicted from changes in water content alone. For a direct measurement of the degree and extent of cracking, we filled cracks with cement, excavated the soil, and photographed the exposed cracks. Comparing direct crack measurements with ERT images of the electrical resistivity of the subsoil, we found that a simple linear model could describe the relationship between crack volume and bulk electrical resistivity. Unfortunately, the fit of this model was poor (R² from 0.4-0.6) and it showed little promise for accurately estimating crack volume. As a tool for monitoring cracks, it appears that ERT is best suited for identifying probable locations of cracks rather than quantitative evaluation of crack morphology. The electronic version of this dissertation is accessible from http://hdl.handle.net/1969.1/151861 |
|---|