3D Sediment Transport Modeling of a Hyper-Eutrophic Lake

Abstract

The dynamics of fine sediment resuspension in Newnans Lake (mean depth of 1.6m and area 27 km 2 ) in north-central Florida was examined using the 3D EFDC model as part of a study to identify the mechanism by which sediment is released from a pliant, 2-m thick muck-laden bottom. Wind not only generates the waves, but it creates a variable wave field, and results in a weak correlation of the total suspended solids (TSS) with wind and waves. At an instrument platform towards the south end of the lake, physical and water quality parameters were measured from December 2003 to September 2004. There are two tributaries in the north end of the lake, which carry sediment into the lake. Part of the sediment load deposits in the lake and the remainder is transported out of the south end of the lake in Prairie Creek. The EFDC model was used to simulate the hydrodynamics and cohesive sediment transport in the lake. The hydrodynamic model was able to predict the measured water surface elevation and both the inflow and outflow discharges were very close to the measured values. Sediment properties (bulk density, organic content, settling velocity, bed shear strength, etc.) used in the model was determined using laboratory experiments on core and water column samples collected in the lake. The highly organic lake sediment has very low bed shear strength and a low settling velocity, this resulted in a fairly constant background concentration of 50 mg/L in the water column. The model was calibrated to achieve an optimal agreement between the measured and predicted vertical suspended sediment concentration profiles at the tower. The sediment transport model was adequately validated by comparing the measured and predicted sediment loads in Prairie Creek. The results from the calibrated and validated sediment transport model, that predicted both current and wave-induced resuspension of the sediment, showed more fluctuations in the simulated TSS time series than that seen in the measured time series. The major spikes in the suspended solids (TSS) data (recorded at the tower) were correlated with sediments transported by the tributaries into the lake during a runoff event. The magnitude and timing of these runoff-induced TSS spikes were adequately predicted by the model.