Evidence for Bi-Stable Transport States in Alluvial Rivers

Theory and observations have demonstrated that coarse-grained rivers organize their channel geometry and grain size such that transport is close to the threshold of motion. Empirically, such "bedload rivers" correspond to natural river channels with median grain size of order 10 mm and larger. Sandy rivers, on the other hand, typically correspond to flow stresses far in excess of threshold. Although there is no widely accepted explanation for these "suspension rivers", it has been proposed that they represent a state that is distinct from bedload rivers. With the acquisition of new data, researchers have recently argued that there exists a continuum of transport behaviors from near threshold to pure suspension. In the absence of a guiding theory, we re-visit this issue through the utilization of: large, global data sets; longitudinal river profiles that transition from bed- to suspended-load dominated; and laboratory experiments. One difference from previous studies is that we explicitly account for the dependence of the threshold fluid stress on channel slope, and examine transport state as the ratio of fluid stress to (slope-corrected) threshold stress. We identify compelling evidence for two stable states of alluvial river systems, i.e., a tendency for rivers to organize to near-threshold or far-above-threshold transport regimes. In the global dataset, this is manifest as a strongly bimodal distribution of (slope-corrected) transport states, where gravel rivers are near threshold and sandy rivers are far above it. For longitudinal profiles of individual rivers, we find that transport state flickers between the two states identified in the global dataset as the channel transits the gravel-sand transition. There is nothing special about sand and the suspension state, however; experiments, and rivers that originate in sand, exhibit the same near-threshold behavior that is observed for gravel rivers. Based upon these findings, we speculate that: there are two stable states (bi-stability) of river channel geometry for a given parameter set of grain size, discharge and slope; and that upstream inheritance of channel width from the gravel portion of a river forces the downstream sandy channel into a suspension state. The development of a mechanistic understanding, however, remains an open challenge.