Monitoring and research to describe geomorphic effects of the 2011 controlled flood on the Green River in the Canyon of Lodore, Dinosaur National Monument, Colorado and Utah

Abstract

In 2011, a large magnitude flow release from Flaming Gorge Reservoir, Wyoming and Utah, occurred in response to high snowpack in the middle Rocky Mountains. This was the third highest recorded discharge along the Green River downstream of Flaming Gorge Dam, Utah, since its initial closure in November 1962 and motivated a research effort to document effects of these flows on channel morphology and sedimentology at four long-term monitoring sites within the Canyon of Lodore in Dinosaur National Monument, Colorado and Utah. Data collected in September 2011 included raft-based bathymetric surveys, ground-based surveys of banks, channel cross sections and vegetation-plot locations, sand-bar stratigraphy, and painted rock recovery on gravel bars. As part of this surveying effort, Global Navigation Satellite System (GNSS) data were collected at benchmarks on the canyon rim and along the river corridor to establish a high-resolution survey control network. This survey control network allows for the collection of repeatable spatial and elevation data necessary for high accuracy geomorphic change detection. Nearly 10,000 ground survey points and more than 20,000 bathymetric points (at 1-meter resolution) were collected over a 5-day field campaign, allowing for the construction of reach-scale digital elevation models (DEMs). Additionally, we evaluated long-term geomorphic change at these sites using repeat topographic surveys of eight monumented cross sections at each of the four sites. Analysis of DEMs and channel cross sections show a spatially variable pattern of erosion and deposition, both within and between reaches. As much as 5 meters of scour occurred in pools downstream from flow constrictions, especially in channel segments where gravel bars were absent. By contrast, some channel cross sections were stable during the 2011 floods, and have shown almost no change in over a decade of monitoring. Partial mobility of gravel bars occurred, and although in some locations vegetation such as tamarisk ( Tamarix ramosissima ) was damaged, wholesale bed motion necessary to fully clear these surfaces was not evident. In flow recirculation zones, eddy sandbars aggraded one meter or more, increasing the area of bars exposed during typical dam operations. Yet overall, the 2011 flood resulted in a decrease in reach-scale sand storage because bed degradation exceeded bar deposition. The 2011 response is consistent with that of a similar event in 1999, which was followed by sand-bar erosion and sediment accumulation on the bed during subsequent years of normal dam operational flows. Although the 1999 and 2011 floods were exceptional in the post-dam system, they did not exceed the pre-dam 2-year flood, isolating their effects to the modern active channel with minor erosion or reworking of pre-dam deposits stabilized through vegetation encroachment.