EP13A-3504:
Testing the limits of the dune phase-stability diagram: the influence of velocity profile shape on dune morphology

Monday, 15 December 2014
Christopher Adam Unsworth1, Daniel R Parsons2, Arnold Jan Herman Reesink3 and Stuart McLelland1, (1)University of Hull, Hull, United Kingdom, (2)University of Hull, Hull, HU6, United Kingdom, (3)University of Southampton, Geography and Environment, Southampton, United Kingdom
Abstract:
The past 100 years of research on fluvial dunes and their deposits has produced bedform scaling laws based on flow depth, grain size and flow velocity. Such flow-form-deposit scaling is used ubiquitously for a wide range of paleo-environmental interpretations and in predictions of river bed roughness in floods. Recent research from marine environments, density currents, and fluvial flows with strong secondary circulation shows that these laws are often extrapolated beyond the limits of the original research. In submarine density currents, for example, paleo-hydraulic reconstructions commonly predict dune forming flow conditions, but preserved dune cross strata are rarely found. One particular difference between these geophysical flows is the velocity profile shape and bed shear stress that results.

In a series of novel laboratory experiments the shape of the mean downstream velocity profile was systematically altered so the velocity maximum was lowered toward the bed through the addition of roughness elements at the water surface; whilst maintaining flow depth and depth-averaged velocities. This produced velocity profile shapes closer to those in density currents, and open-channel flows with strong secondary circulation. The initial lowering of the velocity maximum position increased dune height and length by 250%. The lowest velocity maximum position produced a stable upper-stage plane bed, whilst predictions based on flow depth and mean velocity remained within the dune regime phase-space. The results therefore demonstrate that the vertical position of the downstream velocity maximum can be a better predictor of equilibrium bedform geometries than flow depth or depth averaged velocity and also highlight that paleo-hydraulic reconstructions need to account for the possible variation in profile shape between geophysical flows.

This research improves fundamental understanding of fluvial bedform stability and flow-form-deposit scaling laws for extrapolation into in a broader variety of depositional environments beyond open-channel flows.