THE FORMATION AND GROWTH OF GRAVEL BARS IN RESPONSE TO INCREASED SEDIMENT SUPPLY FOLLOWING THE MARMOT DAM REMOVAL

What makes some bars grow downstream, others grow upstream, while still others primarily grow in height with a nearly constant planform? The October 2007 removal of the Marmot Dam from the Sandy River, OR, and the subsequent liberation of nearly 750,000 cubic meters of sand and gravel served as an experimental setting to observe these various trajectories of bar growth. In the 2 kilometer reach immediately below the former dam site nearly 350,000 cubic meters were deposited during the first year following the removal. The sediment deposited downstream such that there exists a spatial as well as time variation in the amount of deposition seen by the downstream bars, providing an opportunity to measure different responses to various sediment supplies. This deposit both created bars where there were none the previous year, and increased the size of pre-existing bars. The formation and growth of the bars were analyzed using LIDAR, ground surveys, ground-level photographs, and aerial photography.

Throughout the two year study period, new bars formed in an alternating lateral-bar sequence; a pre-existing point bar migrated downstream from the apex of a bend, and subsequently reset to the apex; and small riffle zones containing large (greater than 1 meter) boulders formed the skeleton for mid-channel bars which filled in and elongated upstream. The observations can be generalized by characterizing the pre-removal topography with a suite of topographic measures and observing relationships between the pre-removal measures, the post-removal measures, and the increased sediment supply. Newly formed bars in the portion of the reach with the highest rate of deposition had the greatest change in relief (measured from the thalweg to the top of the bar) while growth of the pre-existing bars was characterized by greater changes in length than in height or width. This study's observation of a small number of bars over a two-year period provides a base to which further observations from the Sandy and other rivers can be added in order to better understand the controls on various forms of bar growth.