What do we know?

1. A pulse of sediment will move downstream after a dam is removed.
2. Numerous case studies provide examples from which we can draw qualitative inferences.
3. Existing 1D models provide approximate estimates for low gradient, simple (?) channels.
4. Organisms have differing tolerances to the strength and duration of sediment disturbances based on their life histories.
5. Contaminants travel with sediments (fine sediments carry more), and their biological availability depends on redox potential.
6. Existing models describe the routing of flow through static channels very well.
7. Fine sediments will generally move more downstream more quickly.

What do we need to know?

1. How does the downstream system respond to increasing the load of sediments of differing size classes
   • coarse (bedload)
   • fines? (suspended load – sand, silt, clay?)
2. How does variation in the sediment input from the reservoir affect downstream response?
3. Routing: how do we predict sediment routing for laterally and longitudinally variable channels? In boulder-bedrock channels?
4. Local geomorphic response issues
   • What affects pools and bars?
   • What controls storage residence time?
5. Biological response issues:
   • What are the boundaries and characteristics of biological oases?
   • What is the dispersal range?
   • At what scale do we need to answer biological response questions?
   • Is a statistical description adequate?
   • What is the size, spacing and duration of “oases” (where fundamental niche conditions are met)?
   • watershed scale sources of nutrients, toxicity
6. How generalizable are predictive models
   • Conceptual models
     • Regional
     • Geometrical
     • Sediment characteristics
   • Process models
7. Under what circumstances can potential sediment effects be neglected?

A proposed research agenda: 

The downstream group came up with the following research agenda, organized into nine issues, each with a sets of Hypotheses regarding the response of the downstream system to dam construction and removal, and suggested methods of testing them:

Issue 1: Scaling

Hypotheses

• There are quantitative geomorphic scaling relationships that define the sensitivity of the downstream channel to dam operations (sediment release from dams; hydrologic changes)
• There are scaling relationships between downstream channel and reservoir characteristics that would give an estimate of the magnitude and duration of the response.
• Magnitude of response to dam removal will scale with initial response of channel to dam construction
• In cases where sediment release is small (little sediment or excavated), downstream channel response will scale with the change in magnitude of flood peaks

Tests:

• Empirical characterization of responses to dam removal in a variety of settings;
  • Dimensionless parameterization to predict categories of response
  • Exploit dam removal and other sediment releases (mining, landsliding, dam releases)

Issue 2: Channel response

Hypotheses:

• Channel response to dam construction is reversible
  • depends on magnitude of disturbance
  • depends on vegetation/biology…
  • depends on cumulative impacts of multiple disturbances
• Channel response could lead to a state change (vegetation)

Tests:

• Field investigation of past and future dam removals, dam failures
• Physical, numerical and conceptual modeling to gain insight to ecological and physical thresholds (planform, community ecology, etc.)
• desktop watershed model to include re-colonization dynamics

Issue 3: 1D models

Hypotheses:

• Planform change and lateral textural, topographic adjustments and storage lead to significantly different responses to dam released sediment than predicted by one-D models

Tests:

• Flume studies to identify lateral effects on mean (1D) response
• Flume studies to build understanding of lateral variability in response
• Flume studies of model floodplain systems hit by fine sediment pulse
• Numerical approaches in incorporating lateral effects into 1D models
• Numerical approaches to partitioning mean response across channel
• Field studies to conduct intensive study of downstream response to dam removal (pick places carefully – prioritize…)
  • Sand bed stream (big meandering)
  • high sediment supply stream (steep bedrock/boulder dominated)
• Mine data from past studies of sediment pulses (Navarro, Ok Tedi, Fall, North Fork Poudre, East Fork River)
• Field studies of storage transport exchanges

Issue 4:   Near-dam vs. far-dam response

Hypotheses

• Downstream response will be different in the ‘near-dam’ and ‘far-dam’ regions.
  • aggradation near dam, textural lateral response far from dam
  • little response in steep reach (near), large response in low gradient reach (far)
  • sand rich release will cause biggest response in sand reach downstream (really far)

Tests:

• Flume studies with really looooong flume
• Lab studies on abrasion
• Field studies of downstream fining by abrasion and selective transport

Issue 5: Sediment storage

Hypotheses

• Different sediment storage units in channel-floodplain system will respond differently to grain size classes in sediment release
• Aggradation and degradation due to changes in coarse sediment storage can influence storage of fine grain sediments on the floodplain
• Fine grain storage reservoirs in channel will respond more quickly than coarse grain storage sites

Tests:

• Field and lab studies of develop models of production, transport and storage of washload (silt and clay)
• Studies of high suspended concentration effect on coarse sediment transport
• Monitor downstream of dam removal where tracer is available to map deposition/infiltration/storage

Issue 6: Biological response

Hypotheses

• Biological responses will scale to the magnitude and duration of the disturbance
• Biological responses will be damped if refugia and oases are maintained for key and sensitive species
• State changes in channel geomorphology will lead to state changes in community structure and ecosystem processes

Tests:

• Conceptual and numerical models to predict responses of organisms with various life history characteristics
• Exploiting field studies to document how organisms respond
• lab studies to measure short term parameters (inundate with sediments, see how organisms respond)

Issue 7: Fines and hydraulic conductivity

Hypothesis:

• High fine sediment concentrations can result in reduced hydraulic conductivity of the river bed

Tests:

• Flume study to test how high concentrations of suspended sediment, including washload and suspendable bed material infiltrate a coarse bed with variable pressure gradient in bed
• Study
• Meta data analysis

Issue 8: Hydrograph effects

Hypotheses

• Low flow adjustment influences response to (first) flood flow
• Character of first flush matters

Tests:

• Flume experiments with a range of discharges and a range sediment supplies to create a range of agradational conditions in downstream channel

Issue 9: The lab-field connection

Hypothesis

• Scaling from lab to field can be improved by new analytical, experimental and field calibration efforts