2015 GSA Annual Meeting in Baltimore, Maryland, USA (1-4 November 2015)

Paper No. 291-29
Presentation Time: 9:00 AM-6:30 PM


WILKINSON, Clare1, HELGANS, Elliott1, HARBOR, David J.1 and KUEHNER, Joel P.2, (1)Department of Geology, Washington and Lee University, Lexington, VA 24450, (2)Department of Physics and Engineering, Washington and Lee University, Lexington, VA 24450, wilkinsonc17@mail.wlu.edu

River channel erosion by plucking or quarrying, although the dominant mechanism in many settings, remains poorly known in laboratory experiments and natural rivers. Experiments in a 250 cm-long, 14 cm-wide flume shed much light on this phenomenon. We modeled bedrock channels in sediment-free flow using immobile < 1 cm-thick ceramic slabs surrounding a pluckable bedrock test zone, where molded and fractured slabs of Plaster of Paris produced irregular blocks that could be eroded. Flow conditions were changed by varying the height of a step on the upstream end of the test zone from 0 to 1.6 cm, varying flume bed slope from 0.5-1.24%, and changing the downstream backwater that controlled the location of a hydraulic jump. Given flows ranging from 50 to 100 cm/s and depths of 3 to 10 cm, the simulated bedrock with a saturated density of 1.86 g/cm3 scales to natural bedrock plucking during floods.

Blocks are plucked in myriad flows and styles of initial motion: blocks plucked in supercritical flow and near hydraulic jumps related to upstream steps and backwater conditions; blocks lift from both the upstream and downstream sides as well as slide both up- and downstream or laterally once neighbors are removed. Initiation of motion from each new bedrock slab began with plucking of blocks exhibiting no protrusion relative to the surrounding bed as well as with and without the presence of an upstream step. The erodibility of blocks increased with step height and channel slope and following the removal of the first block. Before plucking, blocks vibrate in place which abrades the margins of blocks over time. Initiation of plucking occurred with many block shapes and sizes, but was generally located near the toe of hydraulic jumps and particularly along the channel margin. The pressure gradient within the jump likely sets the stage for block uplift through changing pore pressure, but structures in the flow provides the variability that likely initiates lift. Visualizing flow during plucking using Particle Image Velocimetry aids the understanding of the flow conditions required for initiation of motion by lift. After their initial removal, blocks can accumulate downstream and alter flow over the region from which they were eroded. This feedback may be an important control on continued plucking.