Complex Structures in Sediments Overlying Sinkholes: 3D-GPR and Azimuthal Resistivity Imaging

Tuesday, 16 December 2014
Sarah Kruse1, Henok G Kiflu1, Abdallah Ibrahim Ammar Sr1,2, Peter Karashay III1, Anita Marie Marshall1 and Christine M McNiff1, (1)University of South Florida, Tampa, FL, United States, (2)National Water Research Center (NWRC), Research Institute for Groundwater (RIG),, Cairo, Egypt
3D GPR surveys in the covered karst terrain of west-central Florida, USA, reveal surprising geometries of surficial sediments. Several meters of surficial sands overlie progressively more clay-rich sediments, which in turn overlie weathered limestone. The top of a clay-rich horizon produces an exceptionally clear GPR reflector visible from depths between 0.5 and ~8 meters. On length scales of 10-20 meters, the geometry of this horizon as it drapes over underlying weathered limestone suggests that depressions are not conical, but instead more complex troughs that surround domed stratigraphic highs. Azimuthal semi-variograms of the clay horizon depth show greatest correlation in directions that are aligned with the direction of elevated resistivities at depths to 10-14 meters. One possible interpretation is that dissolution in underlying limestone is concentrated in elongated zones rather than in columnar or spherical voids. Elongated sand-filled depressions in the clay layer produce azimuthal resistivity highs in the direction of the elongation. This direction in turn corresponds to the major axis of depressions in the clay-rich GPR reflecting horizon. Groundwater recharge in this area is concentrated into conduits that breach the clay-rich units that overlie the limestone aquifer. This study suggests that the conduits themselves may be elongated features rather than cylindrical in form. Recharge flow paths may be more complex than previously recognized. The high-resolution GPR images require 3D surveys with 250 MHz and 500 MHz antennas, with 10-cm line spacings, careful corrections for antenna positions and 3D migrations of the data.