WSP05649 - Laserfiche WebLink

Home Browse Search

55S TOPPING ET AL: COLORADO RIVER SEDIMENT TRANSPORT, 2 about late December (i.e., only 3 months after the flood peak passed the LCR near Cameron gage). Because of the short period of record, it is not possible to detennine the latest mass balance at any of the sites with any certainty. However, extrap- olation of the curves in Figure lIb suggests that an amount of sand equivalent to that supplied during the uttle Colorado River flood probably passed the Diamond Creek gage within 6 months to a year after the flood. 5.4. Metbods: Collection and Processing of tbe 1997 Data In anticipation of a large sediment input season from the Paria River and to monitor changes in sediment transport in Marble Canyon during August-September 1997, the USGS Arizona District conducted a daily bed and suspended- sediment measurement program at the Lower Marble Canyon gage from August 25 through September 18. During this sam- pling period, approximaIely 1.2 :!: 0.2 million t of sand [after Topping, 1997] were input from the Paria River to the Colo- rado River, largely during the second and third of the four previously described 1997 Paria River floods. Floods on one or more of the ungaged tnbutaries between the Paria River and the Lower Marble Canyon gage also occurred on or around September 2 and possibly September 11-12. Though the amount of sand supplied during these floods is unknown, it was probably much smaller than that supplied by the Paria River during this period. At Ihe Lower Marble Canyon gage, bed samples were col- lected using a 8M-54 sampler, and cross-sectionally averaged suspended-sediment samples were collected using a D-77 bag sampler. Prior to interpreting these suspended-sand data, the filter described in section 5.1 was applied to these data. Unlike in the 1983 case, however, because of the greater experience of the 1997 sampling personnel, this application resulted in ex- clusion of only one of the 23 suspended-sand samples (i.e., only 4% of the data). The discharge of water in the Paria River and in the Colo- rado River at the Lees Fel1}' and Lower Marble Canyon gages and the bed and filtered suspended-sediment data from the Lower Marble Canyon gage during the August-September 1997 sampling period are shown in Figure 12. Also shown in Figure 12 are the predictions of suspended-sand concentration at the Lower Marble Canyon gage using the method of Randle and Pemberton [1987]. 5.5. Results: 1997 Because of the short duration of the sampling program, the close spacing of multiple floods on the Paria River and un- gaged tributaries, and 3 days of much lower dam releases (Figure 120), the results from the August-September 1997 sampling program are not as clear as those trom September- December 1983. However, coupled changes in grain size and suspended-sand concentration following tributary floods are still evident at the Lower Marble Canyon gage (Figures 12b and 120). For analysis it is useful to divide the 1997 data into two portions, with one period from August 28 through 30 (prior to the decrease in flows) and one period from Septem- ber 3 through 17 (after the return to higher flows). This effec- tively divides the data into a short period with no tributary activity and a period of increased tributary activity (Figure 12a). Because the mean discharge of water is comparable dur- ing these two periods (though the magnitudes of the daily fluctuations were greater after September 3), any substantial difference in suspended-sand concentration and grain size be- tween these two periods is probably due to a difference in the upstream supply of sand resulting from the tnbutary floods. During the period of increased tributary activity (September 3-17), suspended-sand concentration was, on average, higher than that measured prior to August 31 at the Lower Marble Canyon gage (Figure 12b). During September 3-17 the mean suspended-sand concentration was 0.007%, whereas during August 28-30 the mean suspended-sand concentration was 0.002%. Though not as dramatic as that following the 1983 uttle Colorado River flood, slight fining of the suspended sand did occur with this increase in suspended-sand concentration (Figure 120). From August 28 through September 17 the me- dian size of the suspended sand decreased from about 0.13- 0.14 to 0.10-0.11 nun. This factor of 3.5 increase in concen- tration and slight fining of the suspended sand was coupled to a decrease in the median grain size of the fine sediment on the bed. From August 29 through September 14 the median size of the fine sediment on the bed decreased from about 0.50 to 0.33 mm. As during the period following the 1983 utt1e Colorado River flood, the concentrations of suspended sand predicted by Randle and Pemberton [1987J are in good agreement with those measured during the period of no tnbutary activity but are a factor of 2.3 low relative to those measured during the period of increased tributaty activity (Figure 12b). This underpredic- tion occurs because their approach treats bed...,diment grain size as a constant rather than as a variable; such an approach does not predict higher transport rates wben the bed sediment fines as a result of enhancement of the upstream supply of sand. Therefore, as in the 1983 example, the approach of Randle and Pemberton [1987] seems to work reasonably well Figure l2. (opposite) Coupled changes in suspended-sand concentration and grain size in the Colorado River associated with the 1997 Paria River and ungaged tnbutary floods. (a) Computed discharge at the Lower Marble Canyon gage during the August-September 1997 sampling period. Also shown are the computed discharge, at the Lees Ferry gage and the Paria River Lees Fel1}' gage. To allow direct comparison of the records from these upstream gages with those from the Lower Marble Canyon gage, the discharges at these two upstream gages were shifted by the If2-day traveltime of the discbarge waves through Marble Canyon. At the Lower Marble Canyon gage the indicated arrival times of water from the second and third of the four large 1997 Paria River floods (see text), a large ungaged tributary flood and a possible second ungaged tributary flood, were determined by this comparison. (b) Mea- sured suspended-sand and suspended-silt and clay concentra- tions at the Lower Marble Canyon gage during the August- September 1997 sampling period. (c) Spatially averaged median size of the fine sediment on the bed and median size of the suspended sand at the Lower Marble Canyon gage during the August-September 1997 sampling period. The same filters used in Figures 7d and We were applied to these bed-sediment data. Also shown for comparison is the suspended-sand con- centration at the time of each of the suspended-sediment mea- surements predicted by Randle and Pemberton [1987J. The cross-hatched area in Figures 12b and 12c indicates the 3 days of steady low flows in Figure 12a. Measured suspended-sand concentration at the Lower Marble Canyon gage increased 1-2 days after each of the Pari a River and ungaged tributary floods. Because a change in the upstream supply of sand and not a change in the discharge of water caused these increases, these increases were not predicted by Randle and Pembel10n {1987}.