No abstract

A series of laboratory experiments has been carried out in which parallel‐laminated deposits were produced from an upper‐regime plane bed. The laminae had thicknesses of a few mm and could be traced continuously over distances up to the length and width of the depositional area (0–3 m by 1–5m). Fluctuations in bed elevation were measured both during deposition and at equilibrium; much of the bed fluctuation occurs at time scales that are too long to be due directly to turbulence, as most theories for lamina formation would require. We suggest instead that extremely low‐amplitude bed forms are present even on apparently flat beds and that the migration of these bed forms produces laterally continuous lamination. All the lamination produced in the laboratory experiments was normally graded. Using high‐speed photography it was observed that the normal grading results from rapid deposition of a layer of loosely packed coarse sand several grain‐diameters thick followed by the slow sieving‐out of a well packed surface layer of finer sand. The initial deposition is the result of small‐scale turbulent fluctuations in boundary shear stress. The sieving‐out that follows results in a smooth surface whose low friction coefficient temporarily inhibits further deposition; we term this process ‘glazing’. The alignment of small‐scale turbulent scour‐and‐fill structures along the paths traced by migrating bedform troughs produces laterally continuous parallel lamination.

As part of the Glen Canyon Environmental Studies, we have developed a discharge model that routes daily discharge waves released from Glen Canyon Dam to Diamond Creek, 386 km downstream. Owing to the length of the diurnal discharge wave and the sparseness of the available topographic data, the latter were averaged over the entire length of the system. Terms too small to be significant in the momentum equation were identified by scaling arguments based on data from past dam releases and on channel hydraulic geometry. Channel friction results primarily from form drag on large topographic elements and from variations in cross‐sectional area and flow depth, rather than bed roughness, producing a stage‐dependent friction that is not well represented by a constant value of standard channel roughness parameters, such as Manning's n. Channel friction as a function of stage was determined from field data available at high discharge (792 m3/s) and intermediate discharge (425 m3/s) and by using simple kinematic wave theory together with wave speed measurements to determine channel friction at low discharge (about 142 m3/s). Model predictions of wave speed and shape agree well with data from five streamflow gaging stations and 42 stage gaging stations located along this segment of the Colorado River.

A two-dimensional hydrodynamic model was applied to seven study reaches in the Colorado River within Grand Canyon to examine how operation of Glen Canyon Dam has affected availability of suitable shoreline habitat and dispersal of juvenile humpback chub (Gila cypha). Suitable shoreline habitat typically declined with increasing discharges above 226-425 m 3 / s, although the response varied among modelled reaches and was strongly dependent on local morphology. The area of suitable shoreline habitat over cover types that are preferred by juvenile humpback chub, however, stayed constant, and in some reaches, actually increased with discharge.

Methods for computing the volume of sand deposited in the Colorado River in Grand Canyon National Park by floods in major tributaries and for determining redistribution of that sand by main-channel flows are required for successful management of sand-dependent riparian resources. We have derived flow, sediment transport, and bed evolution models based on a gridded topography developed from measured channel topography and used these models to compute deposition in a short reach of the river just downstream from the Little Colorado River, the largest tributary in the park. Model computations of deposition from a Little Colorado River flood in January 1993 were compared to bed changes measured at 15 cross sections. The total difference between changes in cross-sectional area due to deposition computed by the model and the measured changes was 6%. A wide reach with large areas of recirculating flow and large depressions in the main channel accumulated the most sand, whereas a reach with similar planimetric area but a long, narrow shape and relatively small areas of recirculating flow and small depressions in the main channel accumulated only about a seventh as much sand. About 32% of the total deposition was in recirculation zones, 65% was in the main channel, and 3% was deposited along the channel margin away from the recirculation zone. Overall, about 15% of the total input of sand from this Little Colorado River flood was deposited in the first 3 km below the confluence, suggesting that deposition of the flood-derived material extended for only several tens of kilometers downstream from the confluence. concern prompted BOR to initiate an investigation of the sand resources in the river between Lake Powell and Lake Mead, and since 1983, the BOR has coordinated a comprehensive program of investigations, the Glen Canyon Environmental Studies (GCES), to determine the effects of releases from GlenCanyon Dam on the riparian and aquatic resources in the park. As part of GCES, the U.S. Geological Survey (USGS) began in 3579 3580 WIELE ET AL.: SAND DEPOSITION IN THE COLORADO RIVER 1989 to develop a suite of flow and sediment-transport models for predicting the response of sand in the canyon to dam releases. ApproachAccurate evaluation of sand resources by direct measurements of the volumes of sand deposits is not feasible because of the great length of the system (380 km), its inaccessibility, the high cost of making necessary field measurements, and the high degree of variability in channel geometry and sand storage. The frequency and density of field measurements in the Grand Canyon is further limited by the need to limit intrusion into the fragile and valued riparian system and into areas held sacred by Native Americans. As a result, we have made a limited number of field measurements designed to be used in conjunction with a model that provides a physically based method of interpolating changes in bed topography both in time and space. The field measurements consist of initial bed topography from which a gridded model ...