Bedforms that develop at the interface between a fluid flow and a loose sediment bed are among some of the most fundamental morphodynamic processes, perhaps among the greatest examples of canonical autogenic adjustments between flow and sediments. Because different types of bedforms develop under specific combinations of flow and sediment properties, these sedimentary features have commonly been used to aid interpretations of flow conditions and infer the nature of depositional environments. While subaerial (river) bedforms are relatively well understood, their counterparts in deep water (i.e., related to gravity underflows, namely, density or turbidity currents) remain somewhat elusive, largely due to the difficulty of direct observation in their natural setting, due to the limited number of experimental studies, and due to their inherent process complexity. Although widely practiced, extrapolation of equilibrium regime diagrams developed for subaerial bedforms to the deep-water realm remains questionable, particularly in light of recent experimental and field observations that suggest some departures from the subaerial counterpart. Herewe present results from an experimental program aimed at investigating equilibrium bedforms resulting from saline density currents under bypass conditions. Saline density currents have been typically treated as the surrogate of muddy turbidity currents for which sediments never settle. More than 500 separate experiments were run, comprising currents that spanned a wide range of the densimetric Froude number including all flow regimes (supercritical, critical, subcritical: Fr_d = 0.6 to 2.8). Results confirm some similarities between subaerial and gravity flow bedforms both in process and product but also reveal some interesting differences. For example, ripples form under both subcritical and supercritical density currents, while supercritical currents yield dunes and both small-wavelength, downstream-migrating, and long-wavelength, upstream-migrating antidunes, where the latter may transition to cyclic steps. Supercriticality of the flow, the proportion of bedload to suspended load (when looking at the sediment composing the bed), and the bed characteristic sediment size are the major controls on the prevailing bedform observed. To investigate the flow and morphodynamic mechanisms related to some of the observed bedforms (e.g., supercritical dunes), detailed analyses of flow structure over the bed features were performed using particle image velocimetry techniques.

Outcrop examples are presented to demonstrate that the gravity flow bedforms we observed experimentally might have counterparts at the field scale. Our findings underscore the rich spectrum of potential bed states produced by dense underflows and their deviation from bed behavior in open-channel flows. As a result, we argue that inversion of gravity flow bed features based on known subaerial bedform regimes might be potentially misleading.