|
<br />002348
<br />
<br />There are few large fine-sediment depos-
<br />its in the Colorado River valley, because there
<br />is little space between the channel and the
<br />confining bedrock or colluvium. The amount
<br />of space available for fine-sediment deposition
<br />is reflected in the difference in channel width
<br />between base flow and flood flow. A large
<br />difference in these values indicates large
<br />available areas for alluvium to be deposited.
<br />The base-flow channel width is typically less
<br />than 60% of the flood-flow width near Soap
<br />Creek Rapid (River Miles 11 to 13), near
<br />North Canyon Rapid (River Miles 20 to 22), in
<br />lower Marble Canyon (River Miles 50 to 57),
<br />and in the Big Bend (River Miles 66 to 74).
<br />The base flow channel width is more than 75%
<br />of the flood-flow width in upper Marble
<br />Canyon between River Miles 13 and 16 and
<br />between River Miles 25 and 38.
<br />
<br />3.1 Longitudinal Profile and Bed-Material
<br />Distribution
<br />
<br />The longitudinal profile of the Colorado
<br />River in Marble and Grand Canyon is a series
<br />of long, flat reaches interrupted by short, steep
<br />rapids and riffles that are somewhat less steep
<br />(Fig. 5). Leopold (1969) reported that 50% of
<br />the total elevation decrease of the river, as
<br />surveyed in 1923, took place in only 9% of the
<br />downstream distance, and Magid et al.
<br />("Changes in the water-surface profile of the
<br />Colorado River in Grand Canyon, Arizona,
<br />between 1923 and 2000," unpubl. manuscript)
<br />found that 66% of the total drop, as measured
<br />in 2000, occurred in the same distance. The
<br />spacing between rapids is determined by the
<br />spacing of tributary canyons (Dolan et aI.,
<br />1978), because debris from each tributary
<br />partially blocks the Colorado River. Similar
<br />relationships have been identified on other _
<br />rivers with abundant debris fans (Graf, 1979;
<br />Schmidt and Rubin, 1995).
<br />The bed of the Colorado River includes
<br />shallow areas at rapids and riffles and deep
<br />pools, or scour holes. Deep pools typically
<br />occur downstream from rapids but also occur
<br />
<br />offshore from flow obstructions (Fig. 7, 8).
<br />Measurements of bed topography in the past
<br />decade demonstrate that significant scour and fill
<br />occurs in these pools during post-dam floods.
<br />The size of bed material is wide ranging.
<br />Bedrock occurs as islands in some places and has
<br />been identified on the bed in side-scan sonar
<br />images (Anima, et aI., 1998). Coarse debris
<br />includes boulders that are 10s of meters in
<br />diameter that are delivered to the channel by
<br />rockfall or debris flow. Debris flows deliver a
<br />wide range of sizes; Webb et aI. (2000) estimated
<br />that boulders larger than 256 mm comprise
<br />14% f' 19%, by weight, of each flow, based on
<br />analysis of 41 samples. The percentage of the
<br />bed that is bedrock or boulders varies widely;
<br />Wilson's (1986) side scan sonar data collected in
<br />1984 showed that 30% of the bed of the Big
<br />Bend was boulders and bedrock. The percentage
<br />of the bed that was bedrock or boulders was
<br />36%,62%, and between 42 and 81% in lower
<br />Marble Canyon, Upper Granite Gorge, and upper
<br />Marble Canyon, respectively (D. S. Department
<br />of the Interior, 1988, Table A-2).
<br />Debris flows, as well as stream flow floods,
<br />boulder abrasion, and Pleistocene terraces, are
<br />the sources of cobbles, between 64 and 256 l'Ilri1,
<br />and pebbles, between 2 and 64 mm, to the bed.
<br />Cobbles and pebbles comprise 24% f' 19% and
<br />41 % f' 21 %, respectively, of debris flows.
<br />Cobble bars are common between River Miles 1
<br />and 4, in parts oflower Marble Canyon, and in
<br />the Big Bend. Imaging of the bed using video
<br />cameras has revealed that much of the bed
<br />elsewhere is comprised of cobbles.
<br />Estimates of the proportion of the bed
<br />covered by sand vary widely. Howard and Dolan
<br />(1981) assumed that 75% of the bed of the study
<br />area was sand, but Smith and Wiele (1. D. Smith
<br />and S. M. Wiele, "Flow and sediment transport
<br />in the Colorado River between Lake Powell and
<br />Lake Mead," unpubl. manuscript) predicted that
<br />less than 30% of the bed needed to have been
<br />covered with sand in order to support a uniform
<br />downstream flux of sand during two experimen-
<br />tal flows in 1990 and 1991. Rubin et aI. (1994a)
<br />estimated that the mean thickness of sand in
<br />
<br />10 System-wide Changes in the Distribution of Fine Sediment in the Colorado River Corridor ...
<br />
<br />
|