EP016-02
Quantifying the geomorphic work of outburst floods

Tuesday, 8 December 2020: 20:34
Virtual
Scott Robert David1, Isaac J Larsen2 and Michael P. Lamb2, (1)University of Massachusetts Amherst, Department of Geosciences, Amherst, MA, United States, (2)California Institute of Technology, Pasadena, CA, United States
Abstract:
Pleistocene outburst floods from Glacial Lake Missoula carved a network of bedrock canyons as they spilled across the Columbia Plateau. Downstream, the flood waters converged and flowed through the Columbia Gorge. Geomorphic evidence, such as scabland topography and dissected lava flows, indicate bedrock erosion enlarged the Columbia Gorge as some of the largest floods in Earth’s history were routed through it. However, the extent which outburst floods reshaped the Columbia Gorge, and the degree to which erosion influences paleo-flood discharge reconstruction, is unknown. To address these unknowns, we used an artificial neural network to estimate the pre-flood topography along a ~170 km long reach of the Columbia Gorge by reconstructing valley walls. Due to lack of constraints, we assumed the bedrock floor of the canyon was unmodified by the floods, making our erosion estimates conservative. In total, we predicted ~7.4 km3 of rock was eroded from the valley walls due to outburst flood erosion. Previous work indicates at least 25 floods were routed through the Columbia Gorge with peak discharges between 1x106 m3s-1 and 1x107 m3s-1. We reevaluated the peak magnitude of these 25 floods using a 2D hydrodynamic model to route floods and match observed high-water marks through the reconstructed, narrower topography. Our results show the peak flood magnitudes ranged from 1x106 m3s-1 and 6x106 m3s-1 ­; the largest floods were ~40% smaller than estimates from previous work that used the modern canyon topography. In addition, the eroded volumes from the neural network are consistent with the total volume of sediment transport predicted by the reconstructed floods, supporting the hypothesis that plucking erosion by megafloods is limited by the sediment-transport capacity of the flows, substantiating a method to estimate canyon formation timescales elsewhere. Results also show that the more frequent intermediate floods with discharges of 4x106 m3s-1 transported the majority of the rock volume eroded from the gorge. Together, our findings indicate that the ‘geomorphic work’ concept used to explain the size and transport dynamics of alluvial rivers also holds in bedrock landscapes subject to repeated outburst floods, in which frequent floods of relatively modest discharge are most responsible in shaping the canyon.