SHALLOW SUBSURFACE ARCHITECTURE STUDY OF CHANNEL BELT DEPOSITS INTEGRATING FIELD DATA WITH GROUND PENETRATING RADAR (GPR) IN THE FERRON SANDSTONE, HANKSVILLE, UTAH

The evolution and subsurface architecture of an ancient channel belt deposit was studied by integrating Ground Penetrating Radar GPR and field data in the Cretaceous Ferron Sandstone, Hanksville, Utah. Due to lack of plan view exposures, most ancient fluvial deposits are assessed from vertical cliff faces solely. This kind of two dimensional studies might lead to misinterpretation of changes in channel orientation and erroneous estimation of channel architecture. Proper assessment of channel architecture in three dimensions is required for precise estimation of fluvial reservoir volume calculation. The studied channel belt deposit is exposed extensively in plan-view. Large-scale braid bar deposits were previously identified based on paleo flow and grain size variation within this channel belt. The belt was built by migration and amalgamation of smaller unit bars. In this study a GPR survey was conducted on this channel belt deposit in order to document the shallow subsurface architecture. Six GPR profiles, four flow parallel and two flow transvers, were collected using a 200MHz antenna. A high amplitude radar reflection was encountered in all GPR profiles at 3-3.5 m which indicates the water table. Four Radar reflection patterns have been identified based on the shape, extent and inclination pattern of the reflection: 1) continuous, inclined, >15m wide and ~1.5m deep reflections; 2) >10m wide and ~2m deep steeply dipping continuous reflections; 3) Long continuous horizontal reflections of 20m length; 4) wavy small reflections of 0.2-0.5m length. In the flow traverse profiles, steeply inclined more than 10m wide and 1.5-2m deep reflections were correlated with the unit bar boundaries as seen in the plan view exposures. Flow parallel profiles are characterized by inclined stratification dipping parallel to the flow direction. Horizontal continuous reflections in these profiles are interpreted as bar boundaries. This radar stratigraphy can be correlated with the surface stratigraphy recorded in previous study. Future work will involve extracting dune height and cross-set thickness for paleohydrological estimations.