Experiments with saline currents show that most flow phenomena and sedimentary structures characteristic of supercritical flow in open channels can be duplicated with density underflows. Antidunes, breaking antidunes, and chutes-and-pools formed spontaneously on an erodible bed at densiometric Froude numbers greater than 1.0. Runs using fixed model antidunes show that equations derived for antidunes and hydraulic jumps in open channels are equally valid for density currents, provided a term is introduced to take account of the reduced density contrast between fluid layers. Antidune wavelength (L) is related to flow velocity (U) by the equation for internal waves:  
\[\ U\ =\ \sqrt{gL/2{\pi}\ ({\rho}\ -\ {\rho}{^\prime})/{\rho}\ +\ {\rho}{^\prime})}\ \]
where ρ and ρ′ are densities of the current and overlying fluid. Critical slope for large scale density flows should be about 0.001. If so, most natural turbidity currents are supercritical. Wavelength of antidunes formed by density currents must be > or =12.6 times the effective flow thickness. For turbidity currents, implied wavelengths are on the order of tens or hundreds of meters. Antidunes of this scale might be manifest as long period pinch-and-swell of individual turbidites, but internal lamination should be essentially parallel to bedding. Under some circumstances, it may be that many successive turbidites comprise the "laminae" of large, composite antidunes. Density currents provide a range of flow conditions where "upper flow regime" bedforms can coexist with "lower flow regime" forms, pointing up a defect in such classification. Ripple lamination does not in itself imply subcritical conditions.
You do not currently have access to this article.