CONSTRAINING ELECTRICAL RESISTIVITY TOMOGRAPHY DATA WITH SOIL THICKNESS AND WATER TABLE DEPTHS IN A STEEPLAND CATCHMENT, ELLIOTT STATE FOREST, OREGON

Accurate determination of soil thickness and water table location is crucial for assessing hillslope stability. Point measurements can be time-consuming to collect and inadequate for characterizing the spatial variability in soil thickness and moisture content in landslide-prone topography. We integrated 2-D spatially continuous data from a DC resistivity survey using a Wenner-alpha type array with point measurements of soil and weathered bedrock thickness, soil water content and soil suction measurements, and water-table position to constrain a laterally-contiguous profile of these parameters at depth. Three resistivity profiles perpendicular to the fall line (electrode spacing 0.7 m) and one profile along the axis of the catchment (electrode spacing 1.0 m) were surveyed to reach subsurface depths consistent with measured depths of the soil-bedrock contact. The study was conducted in a steep (35-40^o) zero-order catchment in the Oregon Coast Range. The site was chosen for its potential susceptibility to landsliding owing to the steep slopes and recent (within 4 years) clearing of forest cover. The surficial mantle of the catchment is a sandy colluvial soil with interlayered organics and bedrock clasts between 5 and 40 mm in diameter. The soil is underlain by weathered yellow, medium- to coarse-grained sandstone with interbedded gray mudstone layers 1-5 cm thick.

The contact between the overlying soil and weathered bedrock was observed in soil pits and auger holes to range from 0.5 m below the surface on catchment side slopes to as deep as 3.3 m along the catchment axis. A saturated zone was observed at 3.1 m depth in the catchment axis near the lowest profile. Least-squares inversions of the geophysical data show an overall decreasing trend in resistivity with depth below the ground surface consistent with measured soil suction and water content profiles, and changes in resistivity at depths consistent with observations of the soil-weathered bedrock interface and water table position. Results suggest that seasonal water table fluctuations could be monitored using a similar geophysical survey, and the resulting tomographic information when combined with point measurements of pressure heads and moisture contents could improve assessments of conditions leading to landslide failure.