WSP05604 - Laserfiche WebLink

Home Browse Search

96 GRAIN-SIZE EVOLUTION DURING fLOODS gage (the only place where bed-sediment grain-size distri- butions were measured during the 1996 controlled flood), coarsening of lhe suspended sediment was observed to be coupled to coarsening of the bed sedinw.n\. Deposils produced during botb the 1996 controlled flood and prc- dam floods in Ihe Colorado River in Marble and Grand Canyons arc inversely graded. 1l1is style of system-wide evolution of sediment grain size coupled to the evolution of suspended.sedimenl conccnlralion during floods is charac- teristic of a supply-limited system like the Colorado River in Marble and Grand Canyons, in which a mismatch in lhe liming of maximum tributary sediment supply and maximum mainstem sedimenllranSport exists. 8, IMPLICATIONS FOR RIVER MANAGEMENT Before the recent conlrolled flood, we thought thai dam operations had only two major impacts on sand in (he canyon: the exchange of sand between bars and the channel bed, and transport down the channel [Rubin el 01., 1994], The approach to building or maintaining bars was to optimize the balance and timing of high flows to transfer sand from the channel bed to the bars and low flows 10 minimize both erosion of bars and sediment transpol1 out of the canyon [US. Departmelll of Ihe Illlerior, 1995], The results of the controlled flood show that the situation is more complicated because dam operations also affecl the grain sizes of sediment retained within the canyon following tributary sediment-input events, The preferred allernative outlined in the Glen Canyon Dam Environ- mental Impact Statement [US, Deportmell/ of Ihe 1/llerior, 1995], based on calculations using stable sediment-rating curves, was designed on the assumption that significant quantities of sediment would be stored in Marble and Grand Canyons during normal powerplant releases from Glen Canyon Dam, Our work, however, suggests that, because sediment rating curves shift over time as a function of the grain sizes of sediment present on the bed and that the grain size of the bed changes significantly over time (Figure 2b), the prediction of net sediment accumulation in the reach between Lees Ferry and the Grand Canyon gage under normal powerplant releases [US, Deparlmenl of Ihe IIl/erior, 1995J may not be valid, Because lhe grain-size distribution of fine sediment evolves as functions of both recent tributary sediment input and dam operation, stable sediment-rating curves art~ precluded in supply-limited rivers such as the Colorado River in Marble and Grand Canyons, For the range of bed grain-size distributions Ihat are possible in the posl-dam Colorado River (Figure 2b), suspended-sand concenlralions at a given discharge of water will vary by over a factor of 10 [Topping, in prep,],Therefore, sediment budgets cannot be conslrucled for reaches of the river using stable sedimenl rating curves. Because sedimenl raling curves shift wilh time in the Colorado River, curves fit to suspended-sand data from time periods when the system is coarse (e.g., 1983-1986) will likely underpredict sediment transpon (export) and overpredicl sediment storage in the mainstem [US. Depor/melll of/he hllenor; 1995], whereas curves fit to suspended-sand data from time periods when the system is fine (e,g" immediately after a significant flood on the Paria or Lillie Colorado Rivers) will likely overpredict sedimenltransport and underpredict sediment storage in the mainslem. The key to managing the fine-sediment resource of the Colorado River below Glen Canyon Dam is to develop a physically based understanding of the processes that control both Ihe short-term fining of the system following large tributary sediment inputs, and the subsequent coars- ening of the system as fines are winnowed from the bed and deposited in the eddies or transpol1ed downstream, This understanding will allow for the design of fulure controlled floods to obtain similar management objectives in the presence of very different supplies of sediment. For ex.ample, because sediment-deposition rates scale approx.i- malely with lhe concentration of sediment in suspension, and Ihe concentration of sediment in suspension will be higher whcn the syslcm is dominaled by finer sedimenl (e.g" immediately after large tributary sediment inputs), sediment-deposition rales in eddies will be much higher when the system is finer. Thus, the response of the system to a shoner-duration controlled flood, in the presence of finer sediment, will be similar to the response of Ihe system to a longer duration controlled flood, in the presence of coarser sediment. The link between floods and grain-size winnowing allows for several additional management capabilities, First, the winnowing process may be used advantageously by managers to manipulate dam releases to keep the main channel relatively coarse, thereby reducing the rate of net downsLream sedimenl transport. If (he channel can be mainlained in a coarsened s(ate, sediment transport down the channel will be reduced, Floods can, therefore, be used to perform a dual funclion by transferring sand from channel 10 bars: (I) cau,ing aggradation on the bars, and (2) coarsening the bed and driving the system toward a "sedimenl-rJting curve" in which sand discharge is reduced (for any given water discharge), Second, dam operations can be used to influencc grain size of surficial sediment on bars, If floods are relatively shol1 and infrequent, the bed will undergo less coarsening during each nood, and the resulting surficial deposits will be finer grained (and contain