MODELING SEDIMENT TRANSPORT IN FULLERS CAVE, CULVERSON CREEK CAVE SYSTEM, WEST VIRGINIA, USA

Sediment transport affects how cave passages grow and behave during floods. For sediment to begin moving, a critical shear stress and minimum discharge must be reached. Often, assumptions about these variables are necessary to model sediment transport and a variety of methods are utilized to calculate shear stresses when predicting where sediment transport is likely to occur. We are studying sediment transport in a narrow, steep vadose cave passage in Fullers Cave, located in Greenbrier County, West Virginia, USA. Flow depth measurements by five in-cave probes are used with known discharges to reconstruct stream processes and sediment transport along a 93 meter reach. A sediment trap was installed and trapped coarse bedload transported during successive floods and entrapped sediments are used to calculate the minimum shear stress generated by the flood. The largest trapped cobble had a critical shear stress for d₈₄ of 214 N m^-2, which is similar to shear stresses previously determined in the cave. This data is combined with sediment tracing: painted rocks were moved by the floods that affected the sediment trap. We report sediment transport distances and other results based on these tracer rocks. The mobilized tracer grains had a median grain size of 74mm and the maximum b axis length was 268mm. The median grain size is comparable to the 77mm median grain size obtained from Wolman counts in the affected reaches prior to the floods. Floodwater velocities are known to be between 1.5 and 1.7 m s^-1 in the study reach for moderate floods of similar size to those that moved the tracer rocks. The critical shear stress for d₈₄ among the tracer rocks is 96 N m^-2, which is substantially less than shear stresses of 278-628 N m^-2 generated by moderate floods with discharges of ~1.5 m³ s^-1. The preliminary results indicate that the cave is capable of transporting larger grains than are currently found in the study reach, which is consistent with intermittent exposures of bedrock in the channel bed. Using the probes and field data, HEC-RAS models are being used to model unsteady flow and sediment transport in the study reach. The model is being constrained by sediment input and output observations to determine sediment influx, efflux, and transport rates when possible. The results of the modeling will be reported.