THREE-DIMENSIONAL GEOPHYSICAL IMAGING OF THE NEW RIVER FLOODPLAIN, BOONE, NC

We present results from 69 Ground-Penetrating Radar (GPR) and direct current electrical resistivity surveys that constrain a three-dimensional model of a floodplain along the New River in Boone, NC. Existing geologic maps of the region suggest crystalline bedrock (Cranberry Gneiss) overlain by fluvial sediments. Because fluvial systems are heterogeneous and the study area is a city park, the primary goal of these surveys is to image the 3D properties of the site using non-invasive methods. GPR and resistivity data complement one another, in that GPR provides a detailed image of subsurface structure (i.e. stratigraphic patterns), while resistivity data can be used to infer the composition and/or water content of subsurface materials. For example, if a subsurface sediment layer has a low resistivity (e.g. a clay layer), the electromagnetic waves of GPR will rapidly attenuate, whereas resistivity data can readily image the material. Furthermore, the depth scale in GPR data requires knowledge of radar velocity, which can be spatially variable and is not known. Thus, by iteratively adjusting GPR velocities and comparing identical features to those imaged with resistivity data, we can effectively calibrate the GPR velocity to achieve accurate depth scales. Once calibrated, the two data types are overlaid to facilitate subsurface interpretation. From the combined geophysical dataset, we interpret the uppermost layers of the site to consist of silty to sandy floodplain deposits that vary from 500-4000 Ohm-m (depending on grain size) but are readily identifiable by the presence of laterally-continuous horizontal reflectors in the GPR data. We identify several buried meandering channels within the site that are identified by asymmetric channel shaped GPR reflectors at depth. Identification of bedrock was complicated by the presence of two geophysically-distinct varieties of bedrock. The first variety has high resistivity values of > 5000 Ohm-m (as is expected for the local Cranberry Gneiss), while the second has low resistivity values of < 200 Ohm-m. Throughout the site, depth to bedrock varies from ~2.0-6.5 m with the deepest bedrock depths occurring in buried channels. Results suggest that the combination of GPR and resistivity data offers a powerful tool for characterizing shallow and complex fluvial systems.