Bedform development in mixed sand–mud: The contrasting role of cohesive forces in flow and bed

Abstract

The majority of subaqueous sediment on Earth consists of mixtures of cohesive clay and cohesionless sand and silt, but the role of cohesion on the development and stability of sedimentary bedforms is poorly understood. The results of new laboratory flume experiments on bedform development in cohesive, mixed sand–mud beds are compared with the results of previous experiments in which cohesive forces in high concentration clay flows dominated bedform development. Even though both series of mixed sand–mud experiments were conducted at similar flow velocities, the textural and structural properties of the bedforms were sufficiently different to permit the designation of key criteria for identifying bedform generation under cohesive flows against bedform generation on cohesive substrates. These criteria are essential for improving bedform size predictions in sediment transport modelling in modern sedimentary environments and for the reconstruction of depositional processes in the geological record. The current ripples developing on the cohesive, mixed sand–mud beds, with bed mud fractions of up to 18%, were significantly smaller than equivalent bedforms in noncohesive sand. Moreover, the bedform height showed a stronger inversely proportional relationship with initial bed mud fraction than the bedform wavelength. This is in contrast with the bedforms developing under the cohesive clay flows, which tend to increase in size with increasing suspended clay concentration until the flow turbulence is fully suppressed. Selective removal of clay from the mixed beds, i.e., clay winnowing, was found to be an important process, with 82–100% clay entrained into suspension after 2h of bedform development. This winnowing process led to the development of a sand-rich armouring layer. This armouring layer is inferred to have protected the underlying mixed sand–mud from prolonged erosion, and in conjunction with strong cohesive forces in the bed may have caused the smaller size of the bedforms. Winnowing was less efficient for the bedforms developing under the cohesive clay flows, where bedforms consisting of muddy sand were more characteristic. The winnowed sand was also found to heal irregularly scoured topography, thus reestablishing classic quasitriangular bedform shapes. In cohesive flows, the bedforms had more variable shapes, and the healing process was confined to lower transitional plug flows in which strong turbulence is only present close to the sediment bed. Furthermore, the bedforms on the cohesive beds tended to form angle-of-repose cross lamination, whereas low angle cross lamination was more common in bedforms under cohesive flows. In general terms, erosional bedforms prevail when cohesive forces in the bed dominate bedform dynamics, whereas depositional bedforms prevail when cohesive forces in the flow dominate bedform dynamics. Empirical relationships between the proportion of cohesive mud in the mixed sand–mud bed and the development rate and size of the bedforms are defined for future use in field and laboratory studies.