Dynamics and Geometry of Equilibrium Laboratory Channels Well Above Threshold

The mechanisms governing the equilibrium channel geometry of alluvial rivers which transport sediment at well above the threshold of motion remain poorly understood. In nature, bed-load dominated gravel rivers are organized such that the channel-forming Shields stress is just at the threshold of motion on the banks. Most equilibrium channels created in laboratory experiments conform to this threshold geometry. However, sandy rivers in nature typically exhibit a Shields stress well above critical, and transport grains in suspension; this condition has not been replicated in the laboratory. The equilibrium geometry of these channels allow for sediment transport on the banks, leading to the stable channel paradox: bank stability seems incompatible with sediment transport. There must be a mechanism for delivering sediment back to the banks to counteract erosion. It is unclear whether bed-load and suspended-load rivers represent two unique stable equilibrium configurations, or whether there exists a continuum of possible geometries ranging from threshold to many times above threshold. Here we present laboratory experiments that produce a single-thread channel that transports grains at a Shields stress well above critical, using low-density acrylic sediment driven by a turbulent flow. Water and sediment are recirculated, and experiments are run at several different water discharge values to explore the transition from bed-load to suspension-dominated channels. Channel cross-section topographic scans and overhead images are used to quantify channel geometry and dynamics, and ensure that experiments reach a statistical steady state. Although statistical fluctuations are apparent, definite equilibrium channel geometries arise for each discharge. We relate channel geometry to dominant transport conditions using measured sediment flux and Shields stress values associated with steady-state dynamics, and characterize the nature of the transition from bed load to suspended load. Results are compared to empirical channel geometry relations for gravel and sand streams.This experimental model is a novel approach to understanding the equilibrium channel geometry of sandy rivers.