Bedforms are omnipresent in many environments (e.g., fluvial, eolian, glaciofluvial, deltaic and deep sea), although there is still some debate on how they develop. There are two separate, though not mutually exclusive, models of bedform initiation: defect initiation and instantaneous initiation.
Defect Initiation The defect theory proposes that the turbulent sweeps that are generated in turbulent flows entrain sediment that upon deposition generates defects in a non-cohesive material. These deposits then propagate downstream via a flow separation process, thus developing bedform fields. The origin of the defects is thought to be linked to packets of hairpin vortex structures. These
coherent turbulent structures give rise to entrainment corridors on the mobile bed, forming grain lineations that interact with the low-speed streaks generating an agglomeration of grains. Once a critical height of grains is reached, flow separation occurs over the new structure. Sediment will be eroded close from the reattachment point and deposited downstream creating a new defect. This new defect will thus induce formation of another defect and the process will continue, propagating downstream while the accumulations of grains quickly evolve into small bedforms.
Instantaneous Initiation In general, the defect propagation theory plays a bigger role at low
sediment transport rates since for high rates defects maybe washed away and bedforms generally initiated across the entire bed spontaneously. Venditti et al. (2005) report that instantaneous initiation begins with the formation of a cross-hatch pattern, which leads to chevron-shaped forms that migrate independently of the pattern structure. This chevron-like structure reorganizes to form the future crest lines of the bedforms. Venditti et al. (2006), based on the earlier model by Liu (1957), proposed that instantaneous initiation is a manifestation of an interfacial hydrodynamic instability of
Kelvin-Helmholtz type between a highly active pseudofluid sediment layer and the fluid above it. In addition, Venditti et al. (2005) imply that there is no linkage between the instantaneous initiation and coherent turbulent flow structures, since spatially- and temporally-random events should lock in place to generate the cross-hatch pattern. Moreover, there is no clear explanation of the effect of turbulence in the formation of bedforms since bedforms may also occur under
laminar flows. Laminar-generated bedform studies used the temporally-averaged flow conditions to determine the degree of
turbulence, indicating
Reynolds number in the laminar regime. However, instantaneous process, such as burst and sweeps, which are infrequent at low Reynolds number but still present, can be the driving mechanisms to generate the bedforms. The generation of bedforms in laminar flows is still a topic of debate within the scientific community, since if true, it suggests that there should be other processes for defect development other than the one suggested by Best (1992). This alternative model for bedform development at low
sediment transport rates should explain the generation of defects and bedforms for cases where the flow is not turbulent. ==Bedform phase diagrams==