The Effect of Roughness Spacing and Size on the Lateral Deflection of Bedload Particles

Bedrock river lateral erosion plays a crucial role in landscape evolution, sediment transport and deposition, and the occurrence of geohazards. Fluvial erosion of massive rock is largely driven by impacts of bedload particles, which also drives crack propagation to prepare blocks for plucking. In straight channels, bedload particles generally move parallel to the channel walls and thus need to be deflected laterally to cause wall erosion. Sideward deflection of bedload particles occurs when they interact with roughness elements fixed on the riverbed. Here, we isolated the interaction of moving bedload particles with roughness elements in laboratory experiments. We conducted 21 sets of flume experiments to systematically investigate how spacing (5, 10, 20, 30, 40, 50, and 60 mm) and size (5, 10, and 20 mm) of roughness elements influence sideward deflection of bedload particles in straight channels with relatively low flow depth and a localized roughness zone covering half of the channel width. Velocity changes by the first impact increase with the size of roughness elements. The deflection length and velocity peak at intermediate values of the spacing of roughness elements. The likelihood for a bedload particle to leave the roughness zone decays with the bedload particle's distance to its edge. Our results suggest that lateral erosion rates in bedrock channels are dominantly controlled by the position of the roughness zone within the channel and its relation to the particle path. Plain Language Summary Erosion of the bedrock walls of a river gorge is driven by the impact of sediment particles propelled by the water flow. Both water flow and sediment particles generally move in parallel to the walls, and the particles need to be deflected laterally from their downstream path to impact the walls. Sideward deflection occurs when the particles hit an obstacle on the river bed, such as a boulder. We investigate the sideward deflected length and velocity of moving particles in a laboratory flume experiment for various configurations of obstacles fixed to the bed. The spacing and size of obstacles have a dual control on deflected length and velocity, making deflection possible in the first case, but also limiting the free path until further impacts. Thus, there exists an optimal spacing of obstacles for maximizing sideward deflection length and velocity. Key Points For large roughness elements, the deflection length and velocity show a peak at intermediate values