Laserfiche WebLink
<br />'~""'''''', <br /> <br />~ <br /> <br />(77.500 fills), as measured at Phantom Ranch., and the mean daily discharge for the same period <br />was about 478 mJ/s (16,900 fills). In 50 percent of all years between 1922 and 1962, discharges <br />of at least 1410 m3/s (50,000 fills) were sustained for about 30 days. <br />Sediment transport also varied greatly throughout the year. The average annual sediment <br />load of the Colorado River at Lees Ferry was about 6.0 X lO'. kg (Andrews, 1991). At the <br />Phantom Ranch gage, located approximately 170 km downstream from Glen Canyon Darn, the <br />mean annual sediment load for the same period was about 7.8 X 10'0 kg per year, indicating that <br />about 1.8 X 10'0 kg were contributed by tributaries. Andrews (l990) estimated that about 70 <br />perc'ent of this arnount was contributed by the Paria and Little Colorado rivers. The quantity of <br />transported sediment is small for such a highly turbulent river that transports such fine sediment. <br />Sediment coarser than 0.5 mm comprised less than I percent of the pre-dam measured sediment <br />load (Smith et aI., 1960) <br />Availability and equilibrium of sand on the bed of the river have, at times, been <br />estimated Wilson's (1986) side-scan sonar surveys, as summarized by the U. S. Department of <br />the Interior (1988, table A-2), indicate that stream bed composed of bedrock or boulders varied <br />between 30 and 81 percent during three surveys in 1984, possibly indicating bed sand amounts <br />between 70 and 19 percent. Sediment-transport modeling (S. M Wiele and 1. D. Smith, A <br />one-dimensional unsteady model of discharge waves in the Colorado River through the Grand <br />Canyon, unpubl. U. S. Geological Survey manuscript) indicates that less than 50 percent of the <br />bed need be covered with sand to maintain equilibrium transport through Grand Canyon. Video <br />imagery and side-scan surveys since 1990 indicate that large portions of the bed are composed of <br />gravel and cobbles (1. D. Smith, hydrologist, U. S. Geological Survey, Boulder, pers. commun., <br />1994 ). <br />Areas of separated flow (eddies) have low velocity and turbulence. Accordingly, eddies <br />are prominent sites of sediment accumulation. Sand bars deposited in zones of flow separation <br />have distinctive topography and location relative to this flow geometry. Schmidt (1990) <br />classified lateral flow-separation eddy bars as separation bars formed near the flow-separation <br />point and which mantle the downstream parts of debris fans and reattachmem bars that form <br />under the primary recirculating eddy cell. Channe/-/7/Qrgin deposits are those which occur as <br />narrow floodplain-like strips throughout the river corridor. <br />Eddy bars persist in specific zones of recirculation because channel obstructions that give <br />rise to flow separation rarely change. Although bars change shape with discharge, they remain <br />within specific lateral separation eddies and do not migrate from eddy to eddy. Measurements. <br />observations and rematching of historic photos of the Colorado River in Grand Canyon (Turner <br />and Kapisak 198, Howard and Dolan, 1981; Schmidt and Graf,1990) show that the locations of <br />some eddy sand bars have been stationuy during the past century, and observations on the <br />relation between flow geometry and sand-bar location suggest that bars should be persistent over <br />periods consistent with the frequency of events that substantially reshape <br />flow-separation-inducing obstructions (Rubin and others, 1990; Schmidt, 1990). In Grand <br />Canyon, that time scale is on the order of! 0 to 100 yrs (Webbet al., 1989). <br />Large floods may overtop debris fans and cause recirculation zones to dimini~h greatly in <br />size or disappear (e.g. Kieffer and others, 1989, fig. 3.5). Melis (U. S. Geological Survey, <br />hydrologist, Tucson, written commun.. 1993) has shown that most debris fans are overtopped by <br /> <br />6 <br />