WSP05649 - Laserfiche WebLink

Home Browse Search

556 TOPPING ET AL: COLORADO RIVER SEDIMENT TRANSPORT. 2 -<>-- MEASURED AT LOWER MARBlE CANYON GAGE _MEASURED UTn.E COLORADO RIVEA (leA) INPUT DURING flOOD ~ - MEASURED AT GRANO CANYON GAGE --0- MEASURED AT NATIONAl CANYON GAGE _ 0- - MEASURED AT ABOVE DIAMOND CREEK GAGE __.RANDLE AND PEMBERTON (1987) PREDICTED -- AT GAAND CANYON GAGE .-...- ---- ~ ./.: ~- .-- 2.0 " o B u '0 1i E 1.5 o ~ I c .. o -' o z .. <II W > >= :5 :J " B 0.5 1.0 0.0 SEPT. OCTOBER a) NOVEMBER DECEMBER Figure 11. (a) Cumulative measured sand loads of the Colorado River at the Lower Marble Canyon gage and at the three gages downstream from the mouth of the Little Colorado River (LCR) and the cumulative measured sand load of the LCR at the highway 89 bridge at Cameron for the period after the beginning of the 1983 LCR flood. The lowest cumulative load was measured at the Lower Marble Canyon gage, because it is located on the Colorado River immediately upstream from the mouth of the LCR. The greatest cumulative sand load during early October was measured in the LCR. Downstream from the mouth of the LCR, progressively smaller loads were measured, since less of the LCR sand input passed each of these more distant sites during the sampling period. Uncertainties (thin dashed lines) of 5% were assigned to the measured sand loads of the Colorado River, and an uncertainty (thin solid lines) of 20% was assigned to the measured sand load of the LCR; see Topping el al. [this issue] for justification of these uncertainties. Also shown is the cumulative sand load predicted by Randle and Pemberton [1987] at the Grand Canyon gage. Because their approach was based on a fixed, coarsened grain-size distribution of bed sediment, Randle and Pemberton (1987) underpredict the sand load at the Grand Canynn gage during this period by about 30%. (b) Sand bndget for the 1983 LCR flood constructed using the data in Figure lOa. Shown are (1) the cumulative measured sand load (with 20% uncertainties) during the LCR flood and (2) plus and minus 5% error envelopes for the adjusted cumulative measured and predicted sand loads at the gages on the Colorado River downstream from the mouth of the LCR. The cross-hatched region indicates the plus and minus 20% error envelope for the sand input during the LCR flood. The loads of the Colorado River downstream from the mouth of the LCR were adjusted by subtracting the measured load (with uncertainties) of the Colorado River at the Lower Marble Canyon gage. See text for further explanation. in the concentration of all sizes of suspended sand is evident 2-3 days after the flood peak passed the LCR near Cameron gage. Following this peak, during the first half of October, the concentration of the finest (0.0625-0.125 mm) sand is higher than that of the coarser (0.25-2.0 mm) sand, whereas by De- cember, the concentration of the coarser sand is much higher than that of the finest sand. This suggests that it takes longer for the bulk of the newly input coarser sand to travel the 222 km from the mouth of the Little Colorado River to the Above Diamond Creek gage than it does for the finest sand to travel this distance. The response of the grain-size distribution of the bed of the Colorado River to the Little Colorado River flood was more complicated than the response of the grain size of the sus- pended sand (Figures lOe and 1Of). The bed at the Grand Canyon gage shows perhaps the clearest behavior. During the Little Colorado River flood the median size of the fine sedi. ment (Le., sand and finer material) on the bed at this site decreased, and the amount of 0.0625- to 0.125-mm sand in- creased (from trace amounts to about 5% of the fine sediment on the bed). Following the Little Colorado River flood, as the suspended sand in the Colorado River coarsened, the finer sand was winnowed from the bed (with the fraction of fine sediment on the bed composed of 0.0625- to O.I25-mm sand decreasing from 5% to about 1 0/0), with no substantial change in the median size of the fine sediment on the bed. This was similar to the observed response of the bed at the Above Diamond Creek gage during the 1997 test flow, where the finer s3nd was winnowed from the bed, but the median grain size of the fine sediment on the bed actually decreased slightly.