Steep Bedload-laden Flows: Near-Critical?

Abstract

Steep gravel-bed rivers sometimes experience flood events that dramatically rework river bed structure and topography. Hazard assessments and paleo-event reconstructions require better knowledge of such phenomena. This paper explores a data set measured on a generic Froude- scaled model. Bedload-laden flood events were introduced in a narrow flume and spread into a five-times wider unconfined area with a 0.1-steep slope (m/m). Image analysis enabled measurements taken at an unprecedented level of accuracy on unconfined flows laden with bedload. A flow reconstruction procedure was used to compute depth, Froude Fr and Shields T* numbers on millions of pixels based on a friction law and measurements of surface velocity using LSPIV, and slope and roughness using SfM. Despite the steep slope, Froude numbers proved to be mostly subcritical in all but the regions experiencing the most active sediment transport. Competent flows, identified by the transport stage higher than unity (T*/Tcr*>1), were observed to be near-critical and seldom had Fr>1.3-1.5. This demonstrates that, providing that bed width and structure can adjust, hydraulic features such as standing waves, hydraulic jumps and lateral shock waves dissipate energy very efficiently in addition to adjusting channel features. These competent flows also tend to rework channels to approach the energy minimum of near-critical flows. This hypothesis was postulated by Gordon Grant (1997, Wat. Resour. Res 33(2), 349-358), regularly confirmed by post-disaster flow computations on a few river sections but, to the best of our knowledge, not yet demonstrated on such a large sample of very steep, unconfined flows. Considering near-critical flows during discharges that are high enough to be clearly competent in laterally unconfined reaches therefore seems reasonable as a first approximation in steep gravel-bed channels. This result may ease discharge evaluation for paleo-flow reconstruction and can help hazard assessments by constraining flow stage - discharge relationships.