Sediment Characteristics, Supply, and Transport

Development of sediment-transport curves requires an understanding of the sources of sediment and the factors that control transport. Sediment transport at a particular site is a function of the width, depth, velocity, temperature, and energy gradient of the river discharge. Other contributing factors are the size, shape, density, and orientation of the bed materials (Glysson, 1987). Sediment supply in rivers downstream of reservoirs is limited to the storage capacity of sediment in the river reach, and contributions to the river system from tributaries.

Suspended-Sediment Transport

Since the construction of Libby Dam in 1972, the supply of suspended sediment downstream of Libby Dam has been limited largely by the trapping capacity of the Lake Koocanusa. Therefore, sediment discharge from the dam contains an insignificant concentration of suspended sediment. Historical data from 8 sediment samples collected prior to the construction of Libby Dam and 25 samples collected after construction of the dam are shown in figure 5. A power function through each of the two sets of data shows the relation in sediment discharge before and after the construction of Libby Dam. Maximum discharge through Libby Dam without spilling is about 736 m³/s. From historical data, a maximum concentration of 3 mg/L provides a maximum estimated suspended-sediment discharge value of 211 metric tons per day.

Suspended-sediment concentration remains low (less than 211 metric tons per day) within the white sturgeon critical habitat until the spring freshet, when tributary flooding contributes suspended sediment to the system. To account for this effect, the fines part of the suspended-sediment transport curves was developed as a function of tributary discharge. The discharge from Kootenai River below Libby Dam (USGS 12301933) to Kootenai River at Tribal hatchery (USGS 12310100) is shown in figure 3. The tributary discharge is estimated as the difference in discharge between these two stations. To account for the lag time between Libby Dam and the Tribal hatchery, the graphs were adjusted until they aligned in a vertical sequence. A lag time of 12 hours was used to estimate a difference in discharge between the two locations. Estimated tributary discharge and suspended-sediment discharge are shown in figure 6.

Suspended sediment at the Kootenai River below the Moyie River site was split into two phases for the total suspended-sediment transport curve. The change in phase is a result of increased tributary discharge into the system resulting in an increased slope of the transport curve. Figure 7 shows the suspended-sediment concentration as a function of tributary discharge at each sampling station.

Each sampling site follows this pattern of increased suspended-sediment concentration with increasing tributary discharge. Table 3 shows a detailed listing of the river and sediment discharge values for each sampling site. The station farthest downstream, Kootenai River above Shorty’s Island, however, showed some anomalous increases in concentration. These sudden increases may be attributed to the lateral mass wasting of the man-made dikes that occurs when high flows saturate the dikes in the braided and meander reaches, as was observed during the periods of sediment sampling. A complete listing of the particle-size distribution at each sampling site is shown in appendix C (table C1).

Bedload-Sediment Transport

Sieve analysis of the bedload material produced informative trends on the size distribution of the material and the difference in transport characteristics at each sampling location (appendix D, table D1). The results were characterized in terms of total river discharge (independent variable) and sediment discharge (dependent variable) by size class (table 4). Three average particle-size classes were used to describe the bedload-sediment transport loading:

• Class 1 – Diameter less than 0.5 mm (silt to medium sand)
• Class 2 – Diameter greater than or equal to 0.5 mm and less than 2 mm (coarse sand to very coarse sand)
• Class 3 – Diameter greater than or equal to 2 mm (very fine gravel to coarse cobble).

Class 1 bedload sediment represents that amount of collected sediment that typically is considered to be in suspension at velocities greater than 0.61 m/s. This sediment also is smaller than the mesh diameter of the bag used to capture the sediment. The PSD revealed that, in some instances, most bedload sediment was contained within class 1. This discrepancy could be caused by a large amount of suspended sediment and (or) debris plugging the mesh material so that it acts as a trap for fine-grained material that typically would pass through the mesh. This problem was especially pronounced with data from the Helley-Smith sampler because the mesh size was even smaller at 0.25 mm. Emmett (1980) suggested not using a Helley-Smith bedload sediment sampler for estimating transport rates for sediment of particle-sizes that are considered part of the suspended-sediment load. Emmett (1980) also suggested that a particle-size less than 0.25 mm should not be sampled using a Helley-Smith sampler, the mesh size of the bag used in this study. For this study, Class 1 bedload was not considered part of the total bedload, although results are provided in figure 8A.

Class 2 sediment consists of sand in the bedload sediment that is larger than the mesh diameter of the sediment bag. The largest discharge of Class 2 sediment downstream of the Moyie River occurred during the two periods of greatest river discharge, with the greatest single discharge occurring during the spring freshet measured at 71.7 metric tons per day. The peak Class 2 discharge occurred before the peak flow. Another peak in Class 2 sediment was transported during the sturgeon pulse, but the sampled sediment discharge was 34 percent less at only an 8 percent decrease in total river discharge.

Class 2 sediment transport downstream of Fry Creek is dependent on the backwater effects of Kootenay Lake. The greatest quantity of Class 2 sediment was transported when the sampled river discharge was lowest, but during the highest mean water velocity (fig. 8B). This highest sampled sediment discharge was 36.3 metric tons per day on May 7, 2008. The combination of the increased backwater effects and the limited supply of sand to the system probably was the reason for the decrease in Class 2 sediment transport downstream of Fry Creek. The mean velocity was inversely proportionate to the area, depth, and slope indicating an increased influence of backwater on the sampling site (table 4). The total river discharge at this sampling site, therefore, should not be used as the independent variable for estimating sand transport.

The sampling site upstream of Shorty’s Island is a sand-dominated system that is in a backwater-affected area of the reservoir throughout the year. The particle-size distribution of the material was nearly identical for each of the sampling periods. The highest measured sediment discharge was 6.08 metric tons per day. Class 2 sediment discharge increased slightly with increasing water velocity, but due to limited samples, this cannot be confirmed. Samples collected with the Helley-Smith sampler were eliminated from the results because this sampler should not be used in areas dominated by a sandy substrate (Emmett, 1980).

Class 3 sediment consists of bedload material greater than or equal to 2.0 mm in diameter. The greatest bedload sediment transport downstream of the Moyie River occurred during the two highest peaks in total river discharge. A threshold of about 1,130.0 m³/s was determined to be the total river discharge required to begin a significantly larger amount of bedload sediment transport (fig. 8C). Pebble count data (appendix B, fig. B1) indicate that coarse gravels (D₅₀ = 39.9 mm and D₁₆ = 20.8 mm) are the dominant substrate; however, the average particle-size of the sampled material was only 6.5 mm during the greatest total river discharge. This indicates that the local surface material was not mobilized under the sampled flows. The smaller gravels that were collected during sampling indicate a moving bed that is limited in supply. The Helley-Smith sampler was used during two peak flows with the highest sampled Class 3 sediment discharge. The Helley-Smith sampler was considered appropriate for this sampling site because its opening is twice as large as the average particle-size in the channel. On June 1, 2008, the Elwha sampler was used to collect samples at 1,020 m³/s with an estimated Class 3 sediment discharge of 5.0 metric tons per day. During June 2 and 4, 2008, the Helley-Smith sampler was used to collect samples at river discharges of 1,210 and 988 m³/s, respectively, with Class 3 sediment discharge of 107.0 and 1.7 metric tons per day, respectively. This vast difference in Class 3 sediment transport likely was caused by a decrease in total river discharge with respect to water velocity, rather than a difference in the type or size of the bedload sampler.

Class 3 sediment transport at the sediment-sampling site downstream of Fry Creek was dependent on the backwater effect of Kootenay Lake. The greatest Class 3 sediment was transported with the highest mean water velocity measured, which also was the lowest measured discharge. The D₅₀ of the bedload sediment during the highest mean water velocity was about 18.0 mm. Pebble count data (appendix A) support the fact that smaller gravels (D₅₀ = 29.8 mm and D₁₆ = 12.8 mm) are available for transport in this reach, but only while the backwater effect is minimal early in the spring. Although the total discharge increases during spring, backwater decreases the hydraulic gradient to the point at which it can only transport sediment equal to or finer than sand-sized particles. The relation between bedload discharge at the sampling site and the extent of the backwater from Kootenay Lake is shown in figure 9. This relation was developed using three-parameter function (Berenbrock, 2005) to determine the extent in RKM. Nearly all bedload discharge ceases as the extent of the backwater moves past the sampling site. Two ADCP measurements (fig. 10) were made on (1) May 7 when the extent of the backwater was near the sampling site (RKM = 246.7) at a river discharge of 700 m³/s and (2) June 3 when the extent of the backwater was highest (near RKM = 250.7) at a river discharge of 1,100 m³/s. Although the total river discharge increased by 64 percent, the velocity decreased by nearly 1 m/s and the stage increased by 3 m illustrating how the area becomes inundated by backwater and loses the ability to transport bed material greater than coarse sand.

Bedload sediment at the sampling site upstream of Shorty’s Island primarily consists of particles less than 2 mm. Transport of Class 3 sediments in this area was assumed to be zero because 99 percent of the substrate consisted of coarse sand. The results of the class 3 bedload sediment samples are shown in figure 8C.

Total Sediment Transport

During the sediment-sampling period, 52–90 percent of the total transport consisted of sediment less than 0.063 mm. This fine-grained material likely remained in suspension and passed completely through the white sturgeon critical habitat. Total transport for suspended and bedload sediment are presented in table 5. No correlation was determined after estimating total suspended sediment as a function of total river discharge (fig. 11A).

Total suspended-sediment discharge is greatest during peak tributary discharges (fig. 11B). The sands remain in suspension until they settle where water velocities decrease due to backwater. The results of this analysis show no decrease in transport from upstream to downstream, but rather a gain in suspended sand transport. Mass wasting of the banks likely contributed to the increase in suspended sands through the white sturgeon critical habitat. This wasting effect is evident as the riverbed substrate changes from coarse to fine gravel to medium sand from upstream to downstream in the white sturgeon critical habitat.

Total bedload sediment discharge (fig. 12A) consisted of less than 3 percent of the total sediment discharge. Particles less than 0.5 mm in diameter were excluded from the analysis because they were only partially recovered. Sand discharge varied widely at each sampling site, but they seemed to depend on the supply of sediment and the extent of backwater. The relation between total bedload and mean channel velocity is shown in figure 12B. The relation, especially at Kootenai River below Fry Creek, improves because the velocity is a function of the extent of backwater from Kootenay Lake.

Total gravel discharge was nearly equal to the sand discharge at the two upstream sampling sites. In the free-flowing part of the river, the amount of bedload sediment depends on total river discharge. In the backwater-affected areas, bedload sediment discharge primarily is controlled by the extent of backwater into the braided reach.