Geomorphic response to large-dam removal: Impacts of a massive sediment release to the Elwha River, Washington

Abstract:

The 2011–2014 staged removals of two nearly century-old dams on the Elwha River in northwest Washington State, the largest dam-removal project in the United States, exposed 21 million m³ of reservoir-trapped sand and gravel to potential fluvial transport. The river downstream from the dams is gravel bedded with a pool-riffle morphology. The river flows 20 km to the marine environment through a riparian corridor lined with large wood and having relatively few anthropogenic alterations. This moderately natural pre-dam-removal condition afforded an unprecedented opportunity to study river response to an anticipated massive sediment release. Four years into the project, 12 million m³ of sediment eroded from the former reservoirs with about 90% of the total load transported to the marine environment. Annualized sediment discharge was as great as 20 times the background natural load. Initial river response to the arrival of the first large sediment pulse was the nearly complete filling of the river’s previously sediment-starved pools, widespread filling of side channels, and increased braiding index. In year 2, during maximum aggradation, the river graded to a plane-bedded system, efficiently conveying sediment to the marine environment. Modest peak flows (<2-yr return period) in year 2 promoted sediment transport but caused little large-scale geomorphic disturbance by channel migration or avulsions. As the river processed the sediment pulse, pools returned and the braiding index decreased in years 3–4. Higher peak flows in year 4 caused localized channel widening and migration but no major avulsions. Gauging indicated sand dominated the first stages of sediment release, but fluvial loads coarsened through time with progressive arrival of larger material. The literature suggests the Elwha River sediment wave should have evolved through dispersion with little translation. However, morphologic measurements and data from a stage-gauge network indicated patterns of deposition, sediment transport, and sediment-wave evolution were heterogeneously complex, challenging our efforts to classify the sediment wave in terms of simple dispersion or translation.