Sediment Mobility in Steep Channels and the Transition to Landsliding

Abstract

The mobility of sediment in steep mountain rivers controls the denudation rate and height of mountain ranges worldwide. Sediment movement within the steepest terrain often occurs as catastrophic shallow landsliding, posing significant hazards to those living downstream. Despite the importance of steep channels, our observations of sediment transport are mostly limited to rivers with slopes of less than 2°. This prevents us from predicting the runoff required to transport sediment throughout most of the drainage network and from knowing the mode of transport that should dominate (dilute river transport vs. landsliding). I performed a series of laboratory experiments in an artificial river with an adjustable slope to test the flow depths required to transport sediment on slopes up to the dry angle of repose. Counterintuitively, sediment becomes harder to move on steeper slopes by dilute river processes. Laboratory observations of flow hydraulics and field observations of cobble stability reveal that this reduced mobility is a hydraulic effect resulting from the shallow flows that are inherent to steep channels. In experiments that were conducted at slopes steeper than half of the dry angle of repose, sediment was more easily transported by shallow landsliding than dilute river processes. Within this landsliding regime, sediment was again observed to be more stable than predicted by traditional theory. Documentation of these experimental failures with high-speed video revealed that failures occur with a characteristic length scale that is shorter than predicted, and that these short failures experience a strong buttressing force at their downstream margin. These results suggest that landslide length scales consistently with width, and also provides new expectations for the saturation level required to initiate failures. Ultimately, these experiments provide us with expectations of the flow depth required to transport sediment throughout the entire drainage network, and also allow us to partition the drainage network into river-dominated and landslide-dominated regimes.