8170 - Laserfiche WebLink

Home Browse Search

1 cross section showed an annual scour and fill cycle of about 3 m durin the assa a of the s rin g P g P g flood (Figure 3). Scour occurred on the ascending limb of the hydrograph, and filling occurred during the descending limb and subsequent low flows. As described below, this pattern of scour ' and fill was similar to the pattern that was observed at some cross sections within the detailed study reach, both in terms of depth of scour and timing of scour and fill. For sand-bedded rivers, bedload is typically 10 to 35 percent of the suspended sediment load [Lane and Borland, 1951 ], so that most sediment is transported by suspension. Bed scour in sand-bedded rivers occurs during floods by the entrainment of bed sediment, and the sediment in ' motion moves at a much slower rate than the water flow. Increased sediment concentrations reduce water velocities, further decreasing the rate of transport. Thus, large volumes of sediment maybe in transport, but the net change in sediment storage within a reach may be small [Leopold ' et al., 1964]. It should be noted that, even for rivers transporting large quantities of sediment, the amount of sediment transported is small compared to the amount of sediment stored on the bed, banks, and in the floodplain. ' Long-Term Channel Response The response of rivers to disturbance is of concern to geomorphologists, ecologists, and , engineers. A "disturbance" to a river maybe either natural, such as the passage of a very large flood, or human induced, such as the closure of a dam. Regime theory considers a river to be an equilibrium expression of the long-term average of the hydrology of a basin [Yu and Wolman, ' 1987], but on a year-to-year scale natural rivers are highly variable. Yu and Wohnan [ 1987] developed a conceptual model to simulate the dynamic adjustment of alluvial river width. They modeled channel width as a function of present ' discharge and past high flow events; the most recent events were given greater weight in the model, and the geomorphic importance of past events decreased with time. Yu and Wolman's ' [1987] model simulated channel widening caused by high flow events and the subsequent recovery, or narrowing, during later lower flows. This model shows that the expected channel form of natural rivers varies over time and is not static; increases in channel width occur when ' certain threshold discharges are exceeded. In addition, channel narrowing continues until the peak discharges are sufficient to maintain or increase channel width. Prior to dam closure, the higher magnitude, but highly variable, flood peaks of the Green ' River formed a channel that was wider than that the current river [Andrews, 1986; Lyons et al., 1992]. Channel widening and narrowing have both occurred since closure of Flaming Gorge Dam [Lyons et al., 1992], but the net trend has been narrowing due to lower mean channel- ' formmg flows, a result consistent with the simulation model results of Yu and Wolman [1987]. Numerical Modeling of Flow and Sediment Transport in Natural Channels ' To design flood flows that will improve and enhance habitats, it is necessary to predict bar and bed response to high flows and bed evolution during passage of a flood. The 3- ' dimensional flow of water in rivers is very complex. Empirical models of river flow reduce this complex system to a simpler 1- or 2-dimensional system with empirically derived coefficients. An often employed example of this type of model is the HEC-2 model developed by the US , Army Corps of Engineers [HEC, 1982]. Cross-sectional data and a channel roughness coefficient are used in the HEC-2 model to calculate water surface profiles for river reaches. This model assumes steady, uniform flow and predicts water surface elevations. However, this , A-8 '