Project Summary

Successful application of the restoration strategies of gravel augmentation, dam removal, and floodplain and channel reconstruction, to rivers of the Bay-Delta watershed is limited by large gaps in the scientific understanding of the underlying fluvial geomorphic processes. In particular, little is currently known about how river bed texture and mobility are influenced by episodic sediment delivery, or how floodplain channel geometry and stability are influenced by changes in the discharge and sediment supply regimes. With funding from CALFED, University of California, Berkeley, San Francisco State University, and Stillwater Sciences are implementing an experimental program to investigate these fundamental questions, with the goal of developing practical science-based methodologies for designing and implementing these restoration strategies. We will use the hydraulic laboratories at the University of California’s Richmond Field Station to test mechanistic hypotheses regarding the processes underlying each restoration strategy. For gravel augmentation, we will vary the timing and amount of model gravel supplied to an armored alternating bar channel, and will measure the temporal and spatial variation in the fraction of the bed composed of a target spawnable gravel size. For dam removal, we will allow channel incision into model reservoir deposits of varying width and grain size and document the rate of gravel evacuation and downstream channel evolution for various discharges. For floodplain and channel redesign, we will conduct the first systematic experiments on the influence of variable discharge and sediment supply on channel width and meander migration rate, using an innovative experimental model substrate recently shown to be capable of supporting self-formed, freely migrating meandering channels. Experimental results will be used to calibrate and extend existing numerical models, in order to provide improved tools to restoration practitioners. We will build strong linkages to existing and future CALFED supported research and restoration projects through project reviews, annual workshops, a scientific advisory panel, and published restoration guidelines.

Project Goals

The primary goal of this project is to assist the design of future gravel augmentation, dam removal, and channel-floodplain reconstruction projects by addressing two areas of inquiry in the field of fluvial geomorphology:

1. The project will seek to develop a mechanistic understanding of river channel response to episodic delivery of bedload size sediments, as occurs in both gravel augmentation and dam removal projects. We will test hypotheses for the spatial and temporal evolution of channel bed texture and geometry following pulses of sediment delivery, and for the potential mobilization of coarse bed armor due to the injection of finer gravel.
   
   
   Physical modeling experiments for this project will be conducted at the hydraulic modeling facilities at the University of California's Richmond Field Station.
2. The project will seek to establish quantitative relationships between equilibrium channel geometry and a range of discharges and sediment supply events, whether natural or regulated. Such quantitative relationships would better guide the process of designing and reconstructing “scaled-down” channels in restoration projects that aim to restore active fluvial geomorphic processes (e.g., bedload sediment transport, bar formation and migration, bank erosion and lateral channel migration) on regulated streams. Progress on both sets of questions would represent significant advances in geomorphic science as well as applied river restoration practice. The results of the experiments will also be used to calibrate and refine numerical models to serve as tools for designing and assessing river restoration actions.

Project Results and Outreach

Since the physical modeling experiments will address many of the key scientific questions underlying gravel augmentation, dam removal, and channel-floodplain reconstruction projects, it will be important to disseminate the results of this research to the many restoration practitioners that will be involved in implementing these projects on behalf of CALFED. The experimental results from the physical modeling will be used to calibrate, refine and extend existing numerical models for predicting sediment transport and bed texture response to changes in sediment supply, sediment mobilization and transport following dam removal, and meander planform evolution. These refined numerical models represent an important bridge between the laboratory and the field, because the numerical models will serve as tools that project implementers can use to design and implement restoration projects in the field.

Workshops will be organized for restoration practitioners and managers and other scientists working on fluvial geomorphic research, combining technical presentations covering experimental design, methods and results to-date with facilitated discussion by participants of their field experiences and implementation concerns. Through these workshops, we intend for the experimental project to provide a forum and catalyst for restoration practitioners, managers and researchers to debate and build consensus on restoration methods.

The implications of the experimental results for restoration design and river management will be synthesized in a set of practical guidelines, one for each of the three restoration strategies (gravel augmentation, dam removal, and channel and floodplain redesign). The guidelines will be written for restoration practitioners and managers, and they will describe methods for evaluating the restoration potential of a site, data collection needs and priorities, appropriate use of numerical models, optimization (in terms of both implementation cost and restoration benefit) of restoration opportunities regarding modifying channel morphology, sediment supply and flow regime, and adaptive management techniques for monitoring and refinement of project implementation.

Project Conception

The Physical Modeling Experiments project was developed in response to CALFED’s identification and funding of a number of strategies for restoring fluvial geomorphic processes in Bay-Delta tributaries. Three such strategies include:

1. injecting gravel to compensate for the loss of coarse sediment trapped behind dams and mined from the channel;
2. removing diversion dams to open access to upstream habitats and restore fluvial geomorphic continuity; and,
3. reconstructing river channels and floodplains to be more in balance with a regulated flow regime as a means of restoring fluvial geomorphic processes.

CALFED has provided funding for several projects that employ these restoration strategies on the Tuolumne, Merced, and Stanislaus Rivers, as well as Clear Creek. Experience with these projects highlights several significant gaps in the scientific understanding of fluvial geomorphic processes, particularly concerning how river bed texture and mobility are influenced by episodic sediment delivery, and how floodplain and channel geometry and stability are influenced by changes in the discharge and sediment supply regimes. The lack of a strong scientific basis for design decisions has often forced project implementers to rely on their professional judgment, which is typically based on qualitative conceptual models and site-specific past experience.

The quantitative methods that are available to guide implementation of these restoration strategies were generally developed to solve more narrowly focused engineering problems. For example, in floodplain restoration projects on the Merced and Tuolumne Rivers (e.g. DWR 2001), one-dimensional water surface profile modeling was used to select the channel width that will produce bed mobility and overbank flooding at specific discharges. This method cannot predict, however, if this is the channel width that would naturally occur under the regulated flow regime, whether the channel can be expected to widen or narrow, or whether the channel will migrate laterally and at what rate. These are cutting-edge questions in fluvial geomorphology, and they illustrate how river restoration practice is currently ahead of the science.

CALFED’s adaptive management approach will help promote learning from river restoration projects. However, an adaptive management approach alone will be insufficient to address many of the key scientific questions underlying river restoration strategies. For example, if a reconstructed channel and floodplain do not perform as expected under high flow conditions, it may be difficult to explain why, because such large-scale restoration projects usually involve changing many variables simultaneously. As a result, it will be more difficult to identify causal relationships, and the ambiguous results will limit the scientific insight that can be gained. Incorporating adaptive management into river restoration projects can yield valuable information, but these lessons can be comparatively expensive. For example, channel-floodplain reconstruction projects in the Central Valley typically costs tens of millions of dollars to implement, owing to the large scale of earth-moving required. If such a project does not perform as predicted, then the lessons derived from the project will come at a hefty price, and the cost of re-designing and re-implementing the project to provide the desired benefits may be prohibitive. The adaptive management approach to river restoration also produces a very slow learning process, relative to CALFED’s 30-year planning horizon. Many of CALFED’s proposed restoration strategies for tributaries require high flows to provide the energy necessary to test the design of a project. In California’s semi-arid climate, several years may elapse before a significant high flow occurs. This delay in the learning feedback loop could prevent projects that are implemented early in the restoration program from informing the design of similar Bay-Delta projects.

To accelerate the pace of learning, we will conduct a series of physical modeling experiments to address some of the fundamental scientific questions underlying the river restoration strategies of gravel augmentation, dam removal, and channel-floodplain reconstruction. Physical models have been at the heart of the basic scientific research that has provided the analytical tools for numerical models of fluvial processes. For example, insight from flume studies underlies our understanding of bedload sediment transport (e.g. Gomez and Church 1989; Buffington and Montgomery 1997), controls on channel width (e.g. Ikeda et al. 1988), downstream fining (e.g. Seal et al. 1997), and the mechanics of flow through meander bends (e.g. Hooke 1975). Physical modeling has also been used to guide large-scale ecosystem restoration projects. For example, a scale model of the Kissimmee River system was developed at the University of California’s Richmond Field Station, and the modeling was instrumental in defining a restoration program for the river (Shen et al. 1994). Similarly, Parker et al. (2002) conducted laboratory studies to examine the effects of modifying flow diversions on mountain channels, with the results informing restoration efforts in Idaho. Also, Wilcock (1998) conducted physical experiments that helped to define channel maintenance flow releases for the Trinity River.