|
<br />annual snowmeh flood were also associated with coars.
<br />ening or hOlh the suspended sand and the bed sediment.
<br />Coarsening of the su!)pended !)and at the Grand Canyon
<br />gage during the 1996 Rood generally occurred by both a
<br />decrease in the concenLraLion of the finest fractions and an
<br />increase in the concentration of the coarsest sizes. This style
<br />of concentration change associated with coarsening of the
<br />suspended sand occurred al siles located along at least 170
<br />!un of the river, and was observed at the above LCR gage
<br />[42 km upSlfeam from the Grand Canyon gage], the 122-
<br />Mile eddy site [56 km downstream from the Grand Canyon
<br />gage], and the National Canyon gage [127 km downstream
<br />from the Grand Canyon gage] (Figures 3 and 4, Table 2),
<br />
<br />5,2, I 22-Mile Eddy
<br />
<br />During the 1996 controlled Rood, the eddy bar at the
<br />122-Mile eddy site changed dramatically, Initially, a signif-
<br />icant amount of erosion occurred near the upstream end of
<br />the eddy on day I, followed by over 4 m of deposition near
<br />the middle and downstream end of the eddy during the
<br />following 6 days, This panern of erosion and deposition
<br />within the eddy was similar 10 the panern of erosion and
<br />deposition in other eddies during the Rood (AI/drews et aI.,
<br />this volume; Hazel el al., this volume; Schmidt ({ a/., this
<br />volume; Wiele et aI" this volume], and occurred primarily
<br />in response to enlargement of the eddy both in width and
<br />length, At high Row, the reanachment point of the Row
<br />moved downstream such that the length of the eddy almost
<br />doubled relalive to the lower Row preceding the Rood, Data
<br />collected previously at Ihis site [Schmidt al/d Graf, 19901
<br />suggest thai this style of geometric response to the increase
<br />in stage during the 1996 controlled Rood was the Iypical
<br />geometric response of the 122-Mile eddy during previous
<br />Roods. In Figure 5, photographs of Ihe eddy are shown
<br />before, during, and immediately after Ihe 1996 Rood,
<br />Before the Rood, the exisling eddy deposit consisted of a
<br />relatively Rat platfoml about 70 cm above the 238 mJ/s
<br />water surface, Our bathymetric measurements show that
<br />much of this deposil eroded on the first day of the Rood in
<br />response 10 Ihe change in eddy geomelry, When the Rood
<br />receded and the bar was exposed at 238 m)/s on day + I
<br />(Figure 5c), much of the former upstream exLent of the pre-
<br />Rood bar had been eroded, bUllhere had been a greal deal of
<br />deposilion farther downstream.
<br />Comparison of the volume of this har on day .1 and day
<br />+ I shows that sand was redistributed from lower to higher
<br />elevations. hut that the actual increase in sediment stored in
<br />the eddy was modest. The volume of the eddy deposit
<br />ahove the 142 m:\)s water-surface elevalion increased :Oy
<br />only 38% during the 1996 com rolled Rood, while the
<br />
<br />TOPPING ET AL. 77
<br />
<br />volume of the eddy deposit above the 566 mJ/s watcr-
<br />surface <levatlOn increased by 106<;r, [Hazel er aI., 1997),
<br />Net erosion on the channel ward and upstream end of the
<br />pre-Rood bar was nearly balanced hy the very large amount
<br />of deposition at the middle and downstream end of the same
<br />eddy, ThIS observed style of geometric rearrangement of
<br />eddy deposits is in agreement with that observed at 34 other
<br />sites in Marble and Grand Canyons during lhe 1996
<br />controlled Rood [Hazel et ai" 1997J, As shown in Figure
<br />5d, the channel ward margin of the Rood deposit evolved
<br />into a steep cutbank during Ihe first day of the recessional
<br />limb of the flood, and a significant portion of this deposit
<br />was eroded and redeposited in the main channel during the
<br />low Rows following the Rood, However, even after the
<br />erosive effects of the low flows of days + I through +3, thick
<br />Rood deposits remained for a period of many months
<br />[Kaplil/ski et aI" 1998J,
<br />As in the main channel at the Grand Canyon gage, the
<br />suspended-sand concentration in the eddy decreased during
<br />the controlled Rood, and the grain sizes in suspension
<br />coarsened significantly, as expecled from the supply-
<br />limitation effect discussed above. The decrease in concen-
<br />tration was corroborated by the continuous ass measure-
<br />ments. which displayed a near monotonic decay between
<br />the first and last days of the Rood event. In Figures 3 and 4,
<br />the spatially averaged concentrations of the suspended
<br />sediment sampled in the eddy are shown along with the
<br />same quantities from the main channel siles upstream at the
<br />Grand Canyon gage and downstream al the National
<br />Canyon gage, It is surprising that the suspended sedimenl at
<br />the 2 main-channel sites and the 122-Mile eddy sile evolved
<br />so similarly during the 7 days of high discharge, given the
<br />respective 56- and 71-km separations belween these two
<br />main channel sites and the eddy site, and the distinct
<br />difference between main-channel and eddy environments,
<br />Furthennore, coarsening of Ihe suspended sand associaled
<br />with decreasing suspended-sand concentralion was also
<br />observed allhe above LCR gage, 98 km upstream from the
<br />eddy site (Figure 3),
<br />These observations suggest that (he response we
<br />observed in the suspended sediment was representative of
<br />the reach in Marble and Grand Canyons, and that the longi-
<br />tudinal variations in suspended-sedimenl concentration
<br />were relatively small throughout al leasllhe \27-km-long
<br />reach from the Grand Canyon gage to the National Canyon
<br />gage. In cOnlrasl to changes in eddy-bar volume - that
<br />exhibited irregular variahility both from site to site [Hazel
<br />el at., this volumeJ and from day to day at a given site
<br />[AI/drews el aI" this volume] during the 1996 controlled
<br />flood - suspendeo.seo"lment" concentration and grain size
<br />showed little longitudinal variability and showed relatively
<br />
|