Characteristics of sediment transport at selected sites along the Missouri River, 2011–12

Abstract

Extreme flooding in the Missouri River in 2011, followed by a year of more typical streamflows in 2012, allowed the sediment-transport regime to be compared between the unprecedented conditions of 2011 and the year immediately following the flooding. As part of a cooperative effort between the U.S. Geological Survey and the U.S. Army Corps of Engineers, this report follows up U.S. Geological Survey Scientific Investigations Report 2013–5006 by comparing sediment transport between years and among sampling sites spanning the Garrison Segment in North Dakota, the Gavins Point Segment downstream from Lewis and Clark Lake, and a part of the Channelized Segment along the Nebraska-Iowa border. Suspended sediment, bed material, bedload, and streamflow data from June 2011 through November 2012 were designated as “measured” total loads, wash loads, and bed-material loads; and, alternatively, were applied to the Modified-Einstein Procedure to compute sediment loads that were designated as “estimated” total loads. Beyond the expected result that sediment loads were much lower during typical streamflows than those measured during the flooding, the measured data indicated some localized sediment-transport processes for further examination. Extreme and prolonged flooding can temporarily deplete sediment supplies locally, and evidence indicating such depletion was present at some sites. Unexpectedly high bed-material loads in the Gavins Point Segment may reflect episodic bar erosion just upstream from the sampling site. The relative contribution of bedload was typically 10 percent or less of the total load during the flooding. Following the flooding, this relative amount increased at some sites but not others, the reasons for which are possibly related to differences in stream velocity. Ultimately, the bedload decreased as it entered the Channelized Segment because of increased velocity and the turbulent mixing ability of the river as compared to the Gavins Point Segment. This turbulent mixing may also convert bed-material load into wash load, thereby rendering those sediments unavailable for creating sandbars and other bedforms. Though some of the sampling data support this premise, it was not consistently manifested by differences between the sediment load of the two segments during typical-streamflow conditions. The Modified-Einstein Procedure tended to predict greater total-sediment loads when compared to measured values. These differences may be the result of sediment deficits in the Missouri River that lead to an overprediction by the Modified-Einstein Procedure, the unsampled zone above the streambed that leads to an underprediction by the suspended sampler, or general uncertainty in the sampling approach. The differences between total-sediment load obtained through measurements and that estimated from applied theoretical procedures such as the Modified-Einstein Procedure pose a challenge for reliably characterizing total-sediment transport. Though it is not clear which of the two techniques is more accurate, the general tendency of the two to be within an order of magnitude of one another may be adequate for many sediment studies.