A document by Jonathan Malarkey. Click on the document to view its contents.

Geometric characteristics of subaqueous bedforms, such as height, length and leeside angle, are crucial for determining hydraulic form roughness and interpreting sedimentary records. Traditionally, bedform existence and geometry predictors are primarily based on uniform, cohesionless sediments. However, mixtures of sand, silt and clay are common in deltaic, estuarine, and lowland river environments, where bedforms are ubiquitous. Therefore, we investigate the impact of fine sand and silt in sand-silt mixtures on bedform geometry, based on laboratory experiments conducted in a recirculating flume. We systematically varied the content of sand and silt for different discharges, and utilized a UB-Lab 2C (a type of acoustic Doppler velocimeter) to measure flow velocity profiles. The final bed geometry was captured using a line laser scanner. Our findings reveal that the response of bedforms to an altered fine sediment percentage is ambiguous, and depends on, among others, bimodality-driven bed mobility and sediment cohesiveness. When fine, non-cohesive material (fine sand or coarse silt) is mixed with the base material (medium sand), the hiding-exposure effect comes into play, resulting in enhanced mobility of the coarser material and leading to an increase in dune height and length. However, the addition of weakly-cohesive fine silt reduces the mobility, suppressing dune height and length. Finally, in the transition from dunes to upper stage plane bed, the bed becomes unstable and bedform heights vary over time. The composition of the bed material does not significantly impact the hydraulic roughness, but mainly affects roughness via the bed morphology, especially the leeside angle.

Sediments composed of mixed cohesive clay and non-cohesive sand are widespread in a range of aquatic environments. The dynamics of ripples in mixed sand–clay substrates have been studied under pure current and pure wave conditions. However, the effect of cohesive clay on ripple development under combined currents and waves has not been examined, even though combined flows are common in estuaries, particularly during storms. Based on a series of large flume experiments, we identified robust inverse relationships between initial bed clay content, C0, and wave–current ripple growth rates. The experimental results also revealed two distinct types of equilibrium combined–flow ripples on mixed sand–clay beds: (a) large asymmetrical ripples with dimensions and plan geometries comparable to clean-sand counterparts for C0 ≤ 10.6%; and (b) small, flat ripples for C0 > 11%. The increase in bed cohesion contributed to this discontinuity, expressed most clearly in a sharp reduction in equilibrium ripple height, and thus a significant reduction in bed roughness, which implies that the performance of existing ripple predictors can be improved by the incorporation of this physical cohesive effect. For C0 ≤ 10.6%, strong clay winnowing efficiency under combined flows resulted in the formation of equilibrium clean-sand ripples and clay loss at depths far below the ripple base. In natural environments, this ‘deep cleaning’ of bed clay may cause a concurrent sudden release of a large amount of pollutants during storms, leading to a sudden reduction in post-storm resistance to erosion of mixed sand–clay substrates.

Cohesive properties of clay promote the formation of clay flocs and gels and relatively small suspended clay concentrations can enhance or suppress turbulence in a flow. Flows are naturally non-uniform, varying in space and time, yet the dynamics of non-uniform open-channel clay suspension flows are poorly understood. To research the influence of suspended cohesive clay on changing flow dynamics under non-uniform flow conditions, new experiments were conducted using decelerating and accelerating clay suspension open-channel flows in a recirculating flume. The flows transition between clay flow types, with different degrees of turbulence enhancement and attenuation as the flow adapts to the change in velocity. The experimental results show that decelerating clay suspension flows have a longer adaptation time than accelerating clay suspension flows. The formation of bonds between cohesive sediment particles is a time-dependent process and establishing clay bonds, as in the decelerating flows, requires more time than breaking them, as in the accelerating flows. This hysteresis is more pronounced for higher concentration decelerating flows that pass through a larger variety of flow phases of turbulence enhancement and attenuation. These different adaptation time scales and associated clay flow type transitions are likely to affect erosional and depositional processes in a variety of fluvial and submarine settings.