WSP05649 - Laserfiche WebLink

Home Browse Search

538 TOPPING ET AL: COLORADO RIVER SEDIMENT TRANSPORT, I "Then, in the weeks or months after the tributary flood the sand on the bed of the Colorado River is winnowed as the sand supply becomes depleled. This causes Ihe sand-Iransport rales in the river to decrease independently of the discharge of water [Topping et al., this issue]. Because the approach used by U.S. Depamnent of Interior (1995J was calibrated to a relatively depleted bed-sediment condition, they probably underpre- dieted sand export and overpredicted sand accumulation in Marble Canyon and upper Grand Canyon [Topping el al., 1999]. Further support for our interpretation of the postdam sed- iment budget comes from both predam and postdam measure- ments of sand-volume change in Marble Canyon and upper Grand Canyon. Geomorphic observations made in the post- dam 'Colorado River do not generally support the system-wide accumulation of fine sediment predicted by Howa.rd and Dolan [1981], Andrews [1990, 1991], and U.S. Depamnenl of the Inte- rior [1995]. Except for during periods of local aggradation (fill) and subsequent degradation (scour) following large tributary sediment inputs, the recent geomorphic studies conducted in Marble Canyon and upper Grand Canyon have documented either no substantiaJ net change in total sand volume over multiyear timescaJes or continued erosion of sand from this reach over multiyear timescales [Webb, 1996; Grof et al., 1995, 1997;Anima elal., 1998; Gromsand Schmidt, 1998; Hazel et al., 1999]. Storage of a fraction of the sand input during large floods on the Uttle Colorado River (in January-February 1993) has been observed to persist for as long as 3 years in the postdam Color8do River [Grof et al., 1995, 1997; Konieczki et al., 1997]. However, the magnitude of this measured longer- tenn storage was less than about 10% of the volume of the sand input during these extreme events [after Grof et al., 1995, 1997; Wudeetal., 1996; KlJnieczkielal., 1997; Roteetat., 1997). Local increases in high-elevation sand volume after high main stem flows have been documented by Beus el al. (1985], Schmidt and Grof [1990], Andrews el al. [1999], Hazel et al. [1999], Schmidl (1999), and Schmidt et at. [1999], but these increases reflect mainly redistribution of sand in a reach and not net gains in sand volume. In the postdam river, only Metis el al. (1995], in their investigation of sediment accumulation in a pool upstream from a recently aggraded debris fan, docu- mented a potentially substantial long-term increase in the vol- ume of local sand storage. The sole example used by Howard and Dolan (1981] to support of their prediction of net system~wide accumulation of fine sediment under nonnal power plant flows was an increase in bed elevation at the Grand Canyon cableway. Uke the example studied by Melis et al. (1995], the bed at the Grand Canyon cableway aggraded by about 2 m in response to the December 1966 flood/debris flow on Bright Angel Creek [Cooley el al., 1977; Burkham, 1986]. This event caused the rapid downstream from the cableway to aggrade and caused a major change in the stage.discharge relationship at the Grand Canyon gage [Cooley et al., 1977; Burkham, 1986). During a high dam release of 2800 m'/s in 1983, the rapid was partially reworked, and the bed at the cableway scoured to its fonner elevation [Burkham. 19861. Data reported by Topping et al. [this issue] show that though the bed at the Grand Canyon cableway temporarily aggraded during the 1996 flood experi- ment, the bed both the day before and 3 weeks after the 1996 flood experiment was at this lower pre-1966 elevation. These observations suggest that the 1967-1983 higher bed elevation a( the Grand Canyon cableway was not indic;\li'.(' of systcm- wide aggradation (as interpreted by Howard and Dolan [1981]) but instead reflected a local change in conditions at the rapid immediately dOWTlstream from the cableway. 7. Conclusions The predam Colorado River in Marble and Grand Canyons was an annually supply-limited system with respect to fine sediment (i.e., sand and finer material). The predam river in Grand Canyon exhibited four effects of sediment supply limi- tation: (I) seasonal hysteresis in sediment concentration, (2) seasonal hysteresis in sediment grain size associated with the hysteresis in sediment concentration, (3) production of in- versely graded flood deposits, and (4) development or modifi- cation of a lag between the time of a flood peak and the time of either maximum or minimum (depending on reach geome- try) bed elevation. Though the predam Colorado River in Glen Canyon also displayed some evidence of supply limitation, none of these effects occurred in Glen Canyon to the degree that they occurred downstream in Marble and Grand Canyons. Thus a predam increase in the degree of sediment supply limitation probably occurred at the change in hydraulic geom- etry near the head of Marble Canyon (where the river steepens and narrows). Sediment budgets provide further evidence that the predam river in Marble Canyon and upper Grand Canyon was annually supply-limited with respect to fine sediment and suggest that the postdam river in this reach. is also annuaUy supply-limited with respect to fine sediment. Given reasonable uncertainties in the annual sediment loads (5% on the main stem Colorado River, 20% on the Paria and Little Colorado Rivers, and a factor of 3 on the ungaged tributaries), the annual supply of fine sediment to Marble Canyon and upper Grand Canyon can be shown to exceed the export in only 1 of the 11 predam years with complete data and in none of the first 7 years of the postdam era. The sole year in which substantial sediment was probably stored for more than I year was sediment year 1963. During this year, Glen Canyon Dam was closed in March, when storage of sand in Marble Canyon and upper Grand Canyon was probably at an annual maximum, and the naturally erosive annual snowmelt flood was replaced by sustained low flows. Though the predam river in Marble Canyon and upper Grand Canyon was annually supply-limited with respect to fine sediment, it was not supply-limited during all seasons. During the average predam year, 0.0625-0.25 mm sand accumulated in this reach from July through March, with the storage of sand being relatively high from September through March. Then, during the higher flows of the annual snowmelt flood (April- June) this stored sand was eroded. During the average post-. dam year with fluctuating flows, however, no season of sand accumulation and storage is evident. Given the uncertainties in lhe sedimenl budgel, slorage of newly input sand can be doc- umented in Marble Canyon and upper Grand Canyon for only 2 months (July through August). Indeed, because the flows in the postdam river are more similar to predam flows during periods of sand erosion than they are to predam flows during periods of sand accumulation and storage, substantial long- term accumulation of sand in the posldam system is unlikely. Thus the sediment-related impacts of the dam have not been only on the volume of the annual tine-sediment supply but, perhaps most significantly, also on the seasonal pauern of sand storage and erosion. Glen ClOyon Dam has converted a system