Lateral erosion in an experimental channel: the influence of bed roughness on erosion by bedload impact
Lateral bank erosion in bedrock streams is important in setting boundary conditions for landscape evolution, yet little is known about the controls and mechanisms of this process. We conducted a series of flume experiments with erodible ‘bedrock’ walls to investigate the influence of bed roughness on bank erosion in bedrock channels. Bed roughness was varied along the length of the channel by changing the size of particles embedded in a non-erodible bed material. In addition to varying the size of embedded particles, we varied the downstream trend in bed roughness (increasing, decreasing or alternating). Experiments consisted of an initial time period of clear water flow followed by multiple periods in which 4mm gravel was introduced into the channel at a constant rate.
Bank erosion from clear water flow was negligible compared to erosion during periods of gravel supply. Erosion during periods of gravel supply was focused near the base of the channel walls resulting in undercut banks. Reach-average bank erosion rates in ‘rough’ sections were 3 to 5 times greater than those in control sections (no embedded particles). Within a single experiment, erosion rates in ‘rough’ sections of different embedded particle size and concentration were similar. This similarity suggests that reach-average bank erosion rates are insensitive to increases in bed roughness beyond a threshold value. Our results indicate: 1) impacts by transported gravel particles are a viable mechanism of bank erosion; 2) the presence (or absence) of roughness elements on the bed exerts a control on rates of bank erosion.