CALFED has identified gravel augmentation as a restoration action to be applied on several Bay-Delta tributaries regulated by dams, including Clear and Putah Creeks, and Feather, Yuba, Bear, American, Mokelumne, Calaveras, Stanislaus, Tuolumne, and Merced Rivers (CALFED 2000). The general goal of gravel augmentation projects is to improve spawning and rearing habitat by modifying grain-size distributions within the channel bed and increasing the frequency of bed movement (CALFED 2000). Channels requiring gravel augmentation are typically armored by coarse, relatively immobile sediments, usually as a result of upstream dams that block sediment transport to downstream reaches and reduce the frequency of high flow events that mobilize the channel bed.

When designing a gravel augmentation project, an implementer must consider several factors, including: the grain size distribution of the sediment to be added; the volume of gravel to inject; the frequency of gravel addition that will be required in light of sediment transport; and how the added gravel will interact with potential modifications to the flow regime and/or channel geometry.

Questions

The gravel augmentation experiments will be designed to address the implementation issues discussed above and answer the following questions:

• How much gravel needs to be added to provide the desired results?
• How often should gravel be added to the channel?
• How much benefit (change in bed texture) will the augmentation result in?
• How long will the benefits last?
• How far downstream will the effect of the augmentation occur?

Hypotheses

The following hypotheses have been developed to guide the experimental design:

1. The beneficial effects of episodic gravel augmentation, expressed as the increase in fraction of the bed area composed of a desired grain size distribution, will scale with the mass of gravel added relative to the mass of bedload material that could be transported by the suite of discharges occurring during the interval between gravel additions.
2. To stimulate bed mobility, the most effective grain size distribution for augmentation should be finer than the desired bed distribution because of the potential for introduced finer sediments to stimulate mobilization of the armor layer, and should scale with the difference between the mean grain sizes of the existing armor and the target distribution.
3. The time scale of bed adjustment and subsequent relaxation to the pre-augmentation state should depend on the residence time of the added gravel, which should also be a function of the potential transport rate and the extent of armor mobilization.
• Small gravel additions should create a small improvement in bed texture for a relatively short time.
• Moderate gravel additions should produce a more significant change in bed texture, which should also persist for a considerably longer time.
• The relationship between mass of gravel added and the extent of bed texture improvement will not be linear, because topographically favorable deposition sites will eventually become saturated. Rather, large gravel additions beyond some threshold amount should primarily extend the duration of bed fining.
4. For a constant rate of gravel augmentation, more frequent additions should maintain a more constant bed texture.

Description of the Experiments

We will test these hypotheses with a series of flume experiments, which are detailed below. In a long flume with a pre-existing armored bed, we will systematically vary the size distribution and volume of added gravel, as well as the interval between gravel additions, and we will measure the spatial and temporal response of the bed. We will compare our experimental results with predictions from a numerical sediment transport model. From the physical and numerical results we will generalize a set of practical guidelines for determining the optimal size distribution, gravel mass and augmentation frequency for a range of field conditions.

The gravel augmentation experiments will focus on changes in the simulated channel bed that result from the manipulation of discharge and sediment delivery. Prior to each individual experiment, the channel bed will be mapped and characterized using a variety of instruments and techniques. At the conclusion of each trial, the simulated channel bed will be re-mapped and re-characterized using the same instruments and techniques, thereby providing the basis for a comparison with the reference channel. The gravel augmentation experiments will focus on three key aspects of sediment delivery: the grain size distribution of injected gravel; the volume of injected gravel; and the frequency of gravel augmentation.

Initial bed condition experiments: The first set of experiments will involve developing the initial channel bed within the flume. Sediment will be delivered to the flume and water will be circulated through the flume to allow the flow to form a channel bed composed of the added sediment. This subtask will require trial-and-error to refine the operation of the sediment delivery and flow control systems. Once the initial bed is established, it will be mapped and characterized.

Single pulse experiments: The second set of gravel augmentation trials will focus on the manipulation of two variables: the volume and grain size distribution of gravel added to the channel. Variable volumes of gravel, with variable grain size distributions, will be added to the simulated channel in single pulses. For this set of trials, steady-state flows of uniform discharge will be circulated through the flume, so that the resultant changes in bed texture will reflect only the changes in sediment supply. Prior to each individual experiment, the channel bed will be mapped and characterized using a variety of instruments and techniques. At the conclusion of each trial, the simulated channel bed will be re-mapped and re-characterized using the same instruments and techniques, thereby providing the basis for a comparison with the reference channel.

Multiple pulse experiments: The third set of gravel augmentation trials will focus on the manipulation of two variables: the volume and frequency of gravel additions. The results of the single pulse experiments will be used to help define the volume and frequency of gravel additions for the multiple pulse experiments. Prior to each individual experiment, the channel bed will be mapped and characterized using a variety of instruments and techniques. At the conclusion of each trial, the simulated channel bed will be re-mapped and re-characterized using the same instruments and techniques, thereby providing the basis for a comparison with the reference channel.

Variable hydrograph experiments: The final set of gravel augmentation experiments will examine the effects of a variable hydrograph on channel bed mobilization by varying the magnitude and duration of peak discharge. The results of the single-pulse experiments and multiple-pulse experiments will be used to define the parameters of the gravel additions to be simulated. Prior to each individual experiment, the channel bed will be mapped and characterized using a variety of instruments and techniques. At the conclusion of each trial, the simulated channel bed will be re-mapped and re-characterized using the same instruments and techniques, thereby providing the basis for a comparison with the reference channel.