Localized Density Instabilities Driven By Interface Shear and Their Influence on Removal of Sediment from Buoyant Plumes

Abstract:

Sediment removal rates from buoyant river discharges are typically scaled with particle settling velocity. However, some field and laboratory data suggest that removal can take place at rates higher than those predicted by individual particle settling. It is possible that these enhanced removal rates could potentially be due, at least in part, to mass settling of fluid and sediment near the fresh and saltwater interface. Fluid shear at the interface can mix the freshwater and sediment with the underlying saltwater and lead to pockets or bands of saltwater and sediment that are more dense than the underlying clear saltwater; resulting in an unstable configuration that can lead to rapid vertical transport of sediment. In this study, we perform laboratory experiments to study the enhancement of sediment removal from buoyant plumes under the effect of shear-driven gravitational instabilities. To do this, we ran a 5 cm deep layer of freshwater with flocculated kaolinite over a 50 cm deep basin of clear saltwater in a 1 m long glass flume under a range of upper layer velocities, concentrations, and initial stratification ratios. A Vectrino profiler is placed such that it constantly records velocity profiles from 2 cm above to 2 cm below the original interface. The velocity of the buoyant layer is controlled using a vertical head pipe at the flume inlet, and the Richardson number is varied from 0.05 to 0.25. The interface is monitored with a digital camera and a laser sheet, and Rhodamine B is added to the buoyant layer for better visualization. Snapshots from the video are used to observe the overall dynamics and developments of instabilities at the interface. The sediment concentration of the inflow and outflow of the system are continuously measured with a pair of OBS sensors, and floc size is measured with a floc imaging system and converted to a floc settling velocity. The difference in sediment concentration between the two OBS sensors is used along with a mass conservation box model to back calculate an effective settling velocity between the two sensors. This data is then used to test the hypothesis that the shear-driven effective settling velocities are higher than individual particle settling velocities and that greater removal rates are observed at lower Richardson numbers.