Restoration Program

Program Administration

Session: Coarse Sediment Measurement and Modeling

Thursday, February 8, 2007 - 9:50 to 11:25 AM

WY 2006 Sediment Transport Monitoring Meeting Flow Objectives Changing Sediment Transport Dynamics

Graham Matthews, Principal, Graham Matthews and Associates, Weaverville , CA 96093 graham@gmahydrology.com
Wesley Smith, Geomorphologist, Graham Matthews and Associates, Arcata , CA 95521 wes@gmahydrology.com

Presentation [PPS - 4.7 mb] (This presentation combines this subject with the one below)

The Trinity River Restoration Program is implementing a long-term fine and coarse sediment management plan below Lewiston Dam. Sediment transport monitoring provides an accounting of the inputs and outputs to the sediment budget cells and therefore plays an integral role in the sediment budgeting process. WY 2006 provided the first opportunity to monitor sediment transport and evaluate flow objectives for an Extremely Wet flow release.

Forty-five bedload and fifty-three suspended sediment samples were collected on Rush and Indian Creeks, and one hundred and forty-three bedload and sixty-five suspended sediment samples were collected at the four mainstem sediment transport monitoring stations. Since data collection began in 1997, Rush and Indian Creek sediment transport rates were the lowest, having decreased by several orders of magnitude. Sediment transport rates for the longer-term mainstem monitoring stations at Lewiston and below Limekiln Gulch were less than predicted. The lower rates were in all probability related to the particle size distribution of the sediment supply. At Lewiston , the decrease in mainstem bedload transport rates was attributable to limited fine sediment transport rates, while the decreases at the Limekiln Gulch station were likely related to the reduction of coarse sediment supply following the elimination of the Grass Valley Creek's coarse bedload.

Daily-averaged peak transport rates occurred during the first day of the 10,000 cfs peak flow at Lewiston , and on the second day above Grass Valley and below Limekiln Gulch. In contrast, the daily-averaged-peak transport rates at the Trinity River near Douglas City peaked six days into the eight day 8,500 cfs bench, after initially falling off after the peak flow bench (open the video file for a short clip of sediment transport on June 1 at Douglas City ). Hales, Pryor, and McBain (2006) identified a drop in the bedload transport rate two to three days after the start of an extended dam release. In general, WY 2006 transport rates follow a similar pattern on the 3,500 cfs rising limb bench and the peak flow bench, with two stations also dropping during the two-day 6,000 cfs bench. However, bedload transport rates increase across the 8,500 cfs bench and at one station during the 5,000 cfs bench. The daily-averaged peak-flow transport rates decreased 6% at Lewiston , 49% above Grass Valley , 25% below Limekiln Gulch, and 16% at Douglas City during the four day peak flow bench.

Total annual bedload estimates for Rush, Grass Valley, and Indian Creeks were 2,160, 7,580, and 1,230 tons, respectively; while bedload estimates for Trinity River at Lewiston, above Grass Valley Creek, below Limekiln Gulch, and at Douglas City were 10,000, 7,570, 11,000, and 27,900 tons, respectively. The fine and coarse sediment transport objectives for the Trinity River were met (and exceeded) during the spring flow release:

• Fine sediment was transported out of the Trinity River reach from Rush Creek to below Limekiln Gulch, at a volume greater than input from Rush Creek and Grass Valley Creek. The cumulative fine bedload total was not reached until June 10, three days into the 5,000 cfs falling limb flow bench; and
• Coarse sediment was transported at a rate equal to tributary inputs as measured at the Limekiln Gulch station on May 4, 19 days before the start of the 10,000 cfs peak flow.

Coarse sediment routing issues created by the tributary deltas require these objectives be evaluated over the four sediment budget cells until full sediment routing is established. Repeat topographic surveys through the Rush Creek pool provide inconclusive information, with a net cut of roughly 4,620 tons of sediment from Nov. 2002, to Dec. 2006. However, 3,200 tons fill were found on the point bar type feature at the upstream end of the pool.

Hales, G.M., C. Pryor, and S. McBain. 2006. Duration of High Flow Releases from Dams: Insights from Recent Bedload Transport Measurements. Poster Presentation, December 14, 2006, American Geophysical Union Fall Meeting, San Francisco, CA.

Presentation notes:

The program goals for sediment dynamics in the upper mainstem of the Trinity River are to achieve a net export of fine sediment out of the system and to maintain steady state transfer of coarse sediment through the system. Transport of sediment seems to have decreased over time. Sediment transport at Lewiston jumps up when flows reach 6,000 cfs. Cobble sizes up to 128 mm can be mobilized at Lewiston . Gravel up to 64 mm can be mobilized at Douglas City . At Grass Valley , transport drops quickly with decreases in flow-does this mean the "bench" is not very useful?

Questions: Could flows over 8,500 cfs be detrimental to the channel if it mines out spawning gravels? No clear answer. Goals for sediment transport were met. We are moving more coarse sediment than expected and more than the inputs. We are moving fine sediment equal to the inputs.

^ TOP

WY 2006 Preliminary Results from the Acoustic Bedload Surrogate Measurements

Wesley Smith and Smokey Pittman, Graham Matthews and Associates, Arcata , CA 95521 wes@gmahydrology.com; smokey@gmahydrology.com

Presentation [PPS - 4.7 mb] (same presentation as above)

Acoustic Doppler Current Profiler

Colin Renee and Rauf Ramooz, from the University of Ottawa, Canada, provided an Acoustic Doppler Current Profiler (ADCP) to conduct field measurements in conjunction with the TR-2 bedload measurements at the Trinity River near Douglas City station. Rauf Ramooz and Dave Gaeuman, from TRRP, assisted the field crews for the initial setup and test runs. The ADCP measures apparent average bed velocities using the instrument's bottom tracking feature. Experiments were performed on May 25 and 26, 2006, and on June 1, 2, 7, and 12, 2006, at flows ranging from approximately 5,000 to 10,000 cfs. Data collected from individual stations with sampler bottom-times of 12 to 60 seconds were processed to obtain transport rates for each station. Initial analysis revealed a poor correlation between bed velocity and bedload discharge measured with the TR2 bedload sampler (r 2 = 0.46). However, when the > and = than 8 mm fractions were compared with bed velocity, the relationship improved for the finer fraction (r 2 = 0.67) and becomes poorer for the coarser fraction (r 2 = 0.14). Other researchers have found bed velocity explained > 70% of observed variability for sand and gravel reaches in the Fraser River, which has smaller maximum particle sizes than the Trinity River (range = 0.25 - 25 mm) (Rennie 2004).

Future analyses will (1) break out the time periods for which the sampler was collecting (rather than a bed velocity averaged over a period of many minutes) which will likely improve relationships; (2) examine stream discharge as an independent variable, as others have hypothesized that the rate of change in bed velocity may not be linearly correlated with increasing streamflow (Gaeuman 2006); (3) explore bed velocity versus < 2 mm bedload discharge, to further isolate sand transport; and (4) isolate sampler down time as an independent variable, as others have founder longer down times necessary when measuring coarser, more episodic transport (Rennie 2004).

Passive Acoustic Hydrophone System

Dr. Rudy L. Slingerland and Dr. Jon Barton from The Pennsylvania State University, University Park , PA. , provided the same passive acoustic sediment sampling equipment used by Barton during the WY 2005Trinity River release. The data collected in WY 2005 provided the field data for Barton's dissertation titled, Passive Acoustic Monitoring of Coarse Bedload in Mountain Streams (Barton, 2006a). Acoustic pressure levels within the river created by interparticle interactions were recorded using a passive acoustic hydrophone, a custom built amplifier, and a laptop running acoustic data acquisition software. One-minute recordings were collected nearly every day from May 18 to June 21, 2006. Future analysis will analyze individual one-minute files for comparison with the TR-2 bedload samples.

Questions: One issue for the poor correlation between the bedload sampler and the Doppler is that the bedload sampler was on the back of the kataraft, while the Doppler was on the front. They may not have been sampling the same bed surface.

^ TOP

HEC-RAS Sediment Model for the Trinity River

Stanford Gibson, US Army Corps of Engineers

Presentation Cancelled

Geomorphic Basis for Gravel Augmentation Design

Dave Gaeuman, Trinity River Restoration Program, dgaeuman@mp.usbr.gov, (530) 623-1813

Presentation [PPS - 2 mb]

Among the foundational hypothesis of the Trinity River Restoration Program is that the quality and availability of physical habitat can be increased by promoting the alluvial processes that create and maintain channel complexity. In a gravel-bed river, the term "alluvial process" is synonymous with gravel transport. Lewiston Dam eliminates the gravel that would otherwise be supplied from the upper basin, making it necessary to artificially replenish the supply of gravel downstream from the dam to support continued downstream transport. In the absence of a gravel augmentation program, ROD releases and other gravel-transporting flows will progressively flush the mobile sizes of material from the bed, resulting in coarsening of the substrate, reduced bed mobility, and reduced channel complexity. To date, most gravel augmentation programs elsewhere have focused on constructing static features designed to provide spawning or rearing habitat rather than on restoring alluvial dynamics. These efforts provide little guidance exists regarding the quantities and size gradation needed to restore alluvial dynamics. Three basic approaches that consider different components of the sediment-transport system have been suggested developing augmentation specification for the Trinity River . These involve defining targets for and managing 1) the volume of gravel stored in the system, 2) gravel transport rates, and 3) bed surface texture. Each approach has different implications for the development of specifications for an augmentation program and the assessment of its success. An analysis based on consideration of bed surface texture suggests that annual additions of approximately 10,000 tons of gravel less than 4" in sieve diameter are needed to support fluvial dynamics in downstream from Lewiston Dam.

Presentation notes:

Finer bed material in stream may be better than coarse as fine moves more readily and allows for fluvial restoration to occur. The 4-inch or smaller gravel he recommends to be added at Lewiston comes from the observation that the D50 below Deadwood Creek is 90 mm, and that graphs showing relationships of gravel size to width suggest the Trinity should have 65 mm gravel.

^ TOP

Discussion Session: Coarse Sediment Measurement and Modeling

Participants: Graham Matthews, Wesley Smith, and Dave Gaeuman.

Decreased bedloads across the bench suggests the bed is becoming armored. Perhaps we could vary the flow to break up the armor? This has been contemplated but has not been done.

Change in bedload transport over time could be an artifact of different measurements and attendant errors? Does more sand move more bedload? Not necessarily. Lewiston had more bedload but less sand. Each sediment cell (river reach) acts differently and probably has to be managed differently. There is less sand than in 1996 due to lots of big landslides then. Indian Creek has a big deposit that is being metered out; it could let loose though.

How to implement gravel augmentations? Riffles are designed at the Lewiston site and 9,000 tons are planned; 10,000 tons are planned for Bucktail (Dark Gulch). It will take several gravel injection sites; over the short term most gravel injections will occur at rehabilitation sites.

Drop in discharge doesn't necessarily cause a drop in velocity if the channel is narrower at the deeper sections. Could this be the cause for the spike in transport at 8500 cfs? Most channels are uniform and 10,000 cfs flows did not generally reach the floodplain, so probably not. If we released high flows without gravel augmentation, there is a fear of mining out the channel. We are transporting high rates relative to inputs.

Bottom stream velocities? Not measured.

^ TOP