8170 - Laserfiche WebLink

Home Browse Search

1-dimensional model rovides no information on the lateral distribution of h draulic ro erties. P Y P P More recent modeling efforts recognize that 3-dimensional properties such as topographically ' induced convective accelerations, secondary circulation, and the distribution of boundary sheaz stress within the curvilineaz 3-dimensional channel are critical to deternuning patterns of erosion and deposition [e.g., Nelson and Smith, I989a; Smith and McLean, 1984; Engelund, 1974]. An orthogonal streamwise coordinate system appropriate for 3-dimensional stream models was developed by Smith and McLean [ 1984]. The Navier-Stokes equations, the ' equations of motion which consider local and convective accelerations in fluids, were transformed to a streamwise coordinate system and formed the basis of a numerical flow model. Smith and McLean [ 1984] compazed modeled results for bottom sheaz stress and free surface elevation with those measured in Hooke's [1975] flume experiments in a fixed-bed sinusoidal channel. The Smith and McLean [ 1984] model produced the same pattern of boundary shear stress, although the maximum boundary shear stress was underestimated and the area of low boundary sheaz stress extended too faz downstream. Nelson and Smith [ 1989a,b] refined and expanded the model of Smith and McLean [1984]. The revised model allowed variation in channel width, accounted for the presence of bedforms, and predicted sediment transport. Numerical results for boundary shear stress, sediment transport, and vertically averaged velocity compared well with the field measurements ' by Dietrich [ 1982] in Muddy Creek, Wyoming. Nelson and Smith [ 1989b] used sediment transport equations [Malin, 1963] to predict bed evolution . This improved approach allowed the simulation of bed evolution and bar genesis in natural channels. Simulation results of the bed evolution model were compazed with results from Hooke's [1975] flume experiments. The topography predicted by the model was very similar to that produced in the flume, although location of the deepest scour was predicted downstream rather than upstream from the bend apex. ' Overall, the agreement between the sediment fluxes and bottom stresses was good. The expanded model of Nelson and Smith [ 1989b] was applied to a river reach within the Ouray NWR by Andrews and Nelson [1989]. Cross-section topography measured at a discharge ' of 275 m3/s was used as model input, and response of the channel was predicted for steady flow at three discharges (50, 275, and 475 m3/s) for 2-day periods. Model results were not validated, but the calculated distribution of unit dischazge and sediment transport compazed well to field ' measurements [Andrews and Nelson, 1989]. Model results indicated that the river bed adjusted quickly to changes in discharge. Channel form and discharge determine the availability of nursery habitat. Because channel form reflects antecedent flows as well as discharge, nursery habitat availability is, in part, a product of antecedent flows. Although it is not possible to experimentally measure the response of the channel to all flow scenarios, it is desirable to predict channel response and, hence, habitat availability for many scenarios. In the Green River basin, only one "experiment" or flood occurs each year, and the magnitude and timing of releases from Flaming Gorge Dam ' aze regulated by many laws. Therefore, each experiment has different antecedent conditions and it is not possible to conduct an infinite array of scenarios. As an alternative to physical experimentation, aflow and sediment transport model that simulates bed evolution may be used ' to model bed and baz response to flood passage. The model used by Andrews and Nelson [1989] is an example of such a flow and sediment transport model, and may be an appropriate tool to model changes in channel topography in response to varied flow in the Upper Colorado basin. ' A-9