Stream Channel Design I-III-2011

Stream Restoration
Design I and II
• Stream Channel Design USDA-NRCS Stream
Restoration Design, National Engineering
Handbook, 2007
Five different approaches
Rosgen’s Natural Channel Design
Criticism of Rosgen’s approaches
ASCE Stream Restoration Task Committee
Indices for Meandering Planforms
Approaches to Stream Channel Design (1)
• Analogy Method: Channel dimensions from a reference reach can
be transferred to another location
• Regime Method: Dependent channel dimensions can be
determined from regression relationships with independent variables
• Hydraulic Geometry Method: Dependent channel dimensions can
be determined from regression relationships with independent
• Extremal Hypotheses: Alluvial channels will adjust channel
dimensions so that some parameter is optimized
• Analytical Method: Depth and sediment transport can be calculated
from physically-based equations
Analogy Method (1)
• Estimates for stable channel design width, depth, and slope in an
alluvial channel can be made using channel dimensions from a
similar stable channel (reference reach)
• The concept is that alluvial streams will evolve to the same stable
channel dimensions, given the same independent driving hydraulic
variables; bed and bank materials, sediment inflow, slope, valley
type, and annual discharge hydrograph should be close to the same
in both the design and reference reaches
• A reference reach is a site that is able to transport sediments and
detritus from its contributing watershed drainage area, while
maintaining a consistent profile, dimension, and plan view, over time
• Several reference reaches with relatively similar channel-forming
discharges may be used to develop a range of solutions for a single
dependent design variable, typically, width
Analogy Method (2)
1. Very difficult to find a stable alluvial reference reach with
characteristics physiographically similar to the reach to be restored
2. The independent driving variables of sediment inflow, bed and bank
material, and channel-forming discharge must be similar
3. The dependent design variables of slope, depth, and width must be
taken together as a set
Regime Method (1)
• Introduced by British engineers in the late 19th C. to design and
operate extensive irrigation systems in India, excavated into fine
sand-bed material and carried their design discharge within the
• The objective of channel design was to set the channel dimensions
so that the inflowing sediment load would be passed without
significant scour or deposition
• Regime formulations include relationships to calculate channel
width, depth, and slope as functions of channel-forming (or
dominant) discharge and bed-material
Regime Method (2)
1. Stable channel dimensions may be calculated using the Blench (1957) regime equations;
data came from Indian canals with sand beds and slightly cohesive to cohesive banks
2. Three channel dimensions—width, depth, and slope—are calculated as a function of bedmaterial grain size, channel-forming discharge, bed-material sediment concentration, and
bank composition.
Regime Method (3)
1. Equations often contain empirical coefficients that must be
estimated primarily using judgment and experience
2. Regime equations are typically regression equations; should not be
used in cases where the discharge, sediment transport, bed
gradations, and channel characteristics of the project channel are
significantly different from those used in the development of the
regime relationships
Hydraulic Geometry Method (1)
• Based on the concept that a river system tends to develop in a
predictable way, producing an approximate equilibrium between the
channel and the inflowing water and sediment
• Typically relates a dependent variable, such as width, to an
independent or driving variable, such as discharge or drainage area
• May be useful for preliminary or trial selection of the stable channel
• Hydraulic geometry depends on: watershed characteristics, geology,
vegetation, land use, sediment load and gradation, runoff
characteristics, woody debris, and composition of the bank
Hydraulic Geometry Method (2)
1. From gaging stations with long-term records, calculate an annual
peak frequency curve and a flow-duration curve
2. Survey stable alluvial channel reaches in association with gaging
stations; determine average channel top width and depth at bankfull
flow, estimate channel hydraulic roughness; determine average
channel bed slope and bankfull discharge
3. Note channel characteristics such as bank material composition,
bed-material gradation, and bank vegetation
4. Determine the channel-forming discharge (2-year peak discharge
from the annual peak frequency curve; flow-duration curve and
sediment transport curve; bankfull discharge from field
5. Plot the measured channel top width versus the channel-forming
discharge, and develop a power regression curve through data
(implicit: determine similar relations for depth, velocity, and slope)
wBF = aQBFb
a=5.19 (T1)
a=3.31 (T2)
Hydraulic geometry relation for width
based on 32 sand-bed rivers with less
or more than 50% tree cover on banks
Hydraulic Geometry Method (3)
• Relations are for the channel-forming discharge only
• Difficult to determine water surface elevation at channel-forming
• Equations must be developed from physiographically similar
• Assumption that channel dimensions are related only to one or two
independent variables is simplistic
• Relationships are assumed to be power functions
• Relationships should not be extended beyond the range of the data
used to develop them
Extremal Hypotheses (1)
• River channel design can be optimized to maximize or minimize a
critical parameter
• Such hypotheses include:
Minimum stream power (Chang, 1979)
Minimum unit stream power (Yang, 1976)
Maximum sediment transport efficiency (Kirkby, 1977) or capacity (White et al.,
Maximum friction factor (Davies and Sutherland, 1983)
• N.B. Under appropriately defined circumstances, some of these
conditions turn out to be equivalent
• Solve the resistance, sediment transport, and extremal equations
simultaneously and obtained a unique solution for the dependent
variables of width, depth, and slope
Extremal Hypotheses (2)
Given equations for flow continuity, flow resistance, sediment transport
capacity and bank strength, and independent variables Q, Sv, Qb, D
and φ′, one can predict S, w, d, and t (Eaton et al., 2004)
Solution curves subject to the constraints Q = 100 m3/s, S = 0.003. For constant φ′ (50°),
bedload concentration predicted using the Meyer-Peter and Muller (1948) equation for constant
variable D50 (22 mm, 32 mm, 45 mm) plotted against bed width and aspect ratio
Extremal Hypotheses (3)
• Channels can be stable with widths, depths, and slopes different
from those found at the extremal conditions
• Sensitivity of energy minima or sediment transport maxima to
changes in driving variables may be low, so that the channel
dimensions corresponding to the extremal value are poorly defined
• Project constraints (anthropogenic or geologic) may limit the
theoretical variability in channel geometry
Analytical Method (1)
• Calculate stable channel dimensions that will pass a prescribed
sediment load without deposition or erosion
• Stable channel analytical method in the COE SAM program
(Copeland, 1994; Thomas et al., 2003) that simultaneously solves
resistance and sediment transport equations (also available in
• Determines dependent design variables of width, slope, and depth
from the independent variables of discharge, sediment inflow, and
bed-material composition; it solves flow resistance and sediment
transport equations simultaneously, leaving one dependent variable
optional (e.g., extremal hypothesis)
• Family of solutions from which the unique solution for depth and
slope can be determined using width from geomorphic principles or
project constraints
Analytical Approach (2)
For subcritical, sand-bedded channel (abridged SAM)
Analytical Approach (3)
For subcritical, sandbedded channel (abridged
Analytical Method (4)
1. Determine the channel-forming discharge—bankfull discharge,
effective discharge, or a specific peak frequency
2. Determine sediment inflow for reach—calculate a sediment transport
rating curve for the upstream supply reach (typical upstream cross
section using a normal depth equation and a sediment transport
3. Develop a stability curve—Calculate a family of slope-width-depth
solutions that satisfy resistance and sediment transport equations
for the channel-forming discharge (stable design channel)
4. Determine channel width—A channel top width for the channelforming discharge is selected from the stability curve using
geomorphic principles or project constraints; depth and slope for
selected width are determined from the stability curve
Analytical Approach (5)
5. Conduct an analytical sediment budget analysis—Using the design
channel dimensions, calculate a sediment-transport rating curve in
the project reach
6. Determine channel planform—Sinuosity determined; also need
meander wavelength, channel length, and trace of the channel
7. Natural variability in cross-sectional shape—Variability in channel
width and depth can either be allowed to develop naturally or can
be part of the project design (e.g., riffles and pools)
8. Instream structures—Successful stream restoration often includes
the use of bank protection, grade control, and habitat features
Analytical Approach (6)
Implications of USDA-NRCS
Stream Restoration Design
• Stream Restoration Handbook: First time such
methods were presented as a technical handbook
• There are advantages and disadvantages with each
• Primary limitations
No assessment of success
Ambiguity of design dimensions
Choice of equations
Rosgen’s Natural Channel Design (1)
• Restoring the dimension, pattern, and profile of a
disturbed river system by emulating the natural, stable
• A channel design technique based on the morphological
and morphometric qualities of the Rosgen classification
• Based on measured morphological relations associated
with bankfull flow, geomorphic valley type, and
geomorphic stream type
• Est. 14,000 “students” spent $28 to $40M
USDA-NRCS, Part 654 Stream Restoration Design National Engineering
Handbook, Chapter 11, Rosgen Geomorphic Channel Design, 2007.
Rosgen’s Natural Channel Design (2)
The methodology is divided into eight phases:
1. Define specific restoration objectives associated with physical,
biological, and/or chemical process
2. Develop regional and localized information on geomorphic
characterization, hydrology, and hydraulics
3. Conduct a watershed/river assessment to determine river
potential, current state, and the nature, magnitude, direction,
duration, and consequences of change
4. Initially consider passive restoration recommendations based
on land use change in lieu of mechanical restoration
Rosgen’s Natural Channel Design (3)
5. Initiate natural channel design with subsequent analytical
testing of hydraulic and sediment transport relations
(competence and capacity).
6. Select and design stabilization/enhancement/vegetative
establishment measures and materials to maintain
dimension, pattern, and profile to meet stated objectives.
7. Implement the proposed design.
8. Design a plan for effectiveness, validation, monitoring, and
maintenance to ensure stated objectives are met, prediction
methods are appropriate, and the construction is
implemented as designed.
River restoration using Rosgen geomorphic channel design approach (Rosgen, 2007)
1. Restoration Objectives
• Clear and concise statements of restoration objectives to
appropriately design the solution(s)
• Common objectives are: flood level reduction, streambank
stability, reduce sediment supply, land loss, and attached
nutrients, improve visual values, improve fish habitat and
biological diversity, create a natural stable river, withstand
floods, be self-maintaining, be cost-effective, improve water
quality, improve wetlands
(Rosgen, 2007)
2a. Local and Regional Relations
• Developing local and regional relations in geomorphic
characterization, hydrology, and hydraulics
• Geomorphic characterization: valley types and stream types
(see table and figures)
• Often advantageous to have an undisturbed and/or stable
river reach immediately upstream of the restoration reach
• Specific design variables use reference reach data for
extrapolation purposes, assuming the same valley and stream
type as represented
(Rosgen, 2007)
Valley types used in geomorphic characterization
(Rosgen, 2007)
Broad-level stream classification delineation showing longitudinal,
cross-sectional, and plan views of major stream types
(Rosgen, 2007)
Classification key for natural rivers
(Rosgen, 2007)
2b. Local and Regional Relations
• Hydrology often determined from regional curves constructed
from long-term stream gage records
• Bankfull discharge and dimensions are plotted as a function of
drainage area for extrapolation to ungaged sites in similar
hydro-physiographic provinces
• Regional curves of bankfull discharge versus drainage area
and at-a-station hydraulic geometry relations are developed
• Hydraulic relations are validated using resistance equations
for velocity prediction at ungaged sites (such as Manning’s n)
(Rosgen, 2007)
Regional curves from stream gaging stations showing bankfull
discharge (ft3/s) vs. drainage area (mi2)
(Rosgen, 2007)
Regional curves from stream
gaging stations showing
bankfull discharge (ft3/s) vs.
drainage area (mi2)
(Rosgen, 2007)
Prediction of Manning’s n roughness coefficient
(Rosgen, 2007)
3a. Watershed and River Assessment
• A stream channel stability analysis is conducted along with
riparian vegetation inventory, flow and sediment regime
changes, limiting factor analysis compared to biological
potential, sources/causes of instability, and adverse
consequences to physical and biological function
– Streamflow alteration (magnitude, duration, and timing) must be
– Sediment competence and capacity—Potential aggradation,
degradation, and channel enlargement are predicted for the disturbed
reach (based on shear stress, maximum grain size, and regional
sediment transport relations)
– Streambank erosion—Streambank erosion rate is predicted as part of
the river stability assessment
(Rosgen, 2007)
3b. Watershed and River Assessment
• A stream channel stability analysis (continued)
– Successional stages of channel evolution—identify the present stage of
the stream and predict the direction and consequence of change
– Base-level change—degree of channel incision determined by the
lowest bank height divided by maximum bankfull depth
– Direct disturbance and riparian vegetation—disturbance of stream
channels must be offset by correcting dimension, pattern, profile, and
often channel materials. Riparian vegetation reestablishment should
contain the correct species
– Biological assessments—Fish species, food chains, diversity with broad
categories of ecoregions, and stream types (habitat units) are collected
for identifying biological potential
(Rosgen, 2007)
Various stream type
succession scenarios
(Rosgen, 2007)
4. Passive recommendations
for restoration
• A change in management strategies can be very effective in
securing stability and function
• The alternative of self-stabilization is always a key
consideration in any stability assessment
• If natural recovery potential is poor and/or does not meet
specific objectives, phase 5 would be appropriate
(Rosgen, 2007)
5. Natural Channel Design
• Mixture of analogue, empirical, and analytical methods in the
design procedure
• Determine existing valley type and potential stream type of
the stable form
• The proposed channel type must be converted to a
dimension, pattern, and profile to initially test whether the
hydraulic and sediment relations associated with the
watershed are compatible prior to advancing through all of the
procedural steps
• Watershed and river assessment given in the following steps
(Rosgen, 2007)
Generalized flowchart representing Rosgen geomorphic
channel design utilizing analogue, analytical, and empirical
(Rosgen, 2007)
(Rosgen, 2007)
(Rosgen, 2007)
6. Stabilization and Enhancement
• River stabilization and enhancement structures used in the
Rosgen geomorphic channel design methodology
• Utilize native materials such as natural boulders, logs,
rootwads, and vegetative transplants
• Grade control—cross vane (+/- logs and rootwads) and Wweir
• Streambank stabilization—J–hook vane, and others
(Rosgen, 2007)
Cross section, profile, and
plan view of a cross vane
(Rosgen, 2007)
(Rosgen, 2007)
Cross section, profile, and plan view of a W-weir
(Rosgen, 2007)
(Rosgen, 2007)
Log vane/J-hook combo with rootwad structure
(Rosgen, 2007)
(Rosgen, 2007)
Longitudinal profile of
proposed C4 stream
type showing bed
features in relation to
structure location
(Rosgen, 2007)
7. Design Implementation
• Layout—making necessary onsite adjustments to the design
based on constraints that may have been previously
• Construction supervision (oversight)—critical
• Water quality controls—sediment detention basins,
diversions, silt fences, and pump sites must be located to
prevent onsite and downstream sediment problems as
8. Monitoring and Maintenance
(Rosgen, 2007)
Criticisms of Rosgen’s NCD (1)
(Simon et al., 2008, JAWRA)
• Complete lack of technical peer review for
the ‘‘Natural Channel Design’’
• Observations of channel form (stream
types) placed in a series of temporal
sequences is not the same as being able
to predict these sequences and, therefore,
stable morphologies
Criticisms of Rosgen’s NCD (2)
(Simon et al., 2008, JAWRA)
Fundamental flaws include:
Multiple stream-type succession end points for a given initial type
Flaws in the sediment-transport equations and their application
Inability to predict changes in channel width
Inability to predict stable morphologies within a currently unstable
5. Inconsistent and erroneous determination of stream types
6. Mixing of bed and bank-material properties
7. Sediment-supply calculations based on a single snapshot in time
and space
Rosgen’s Natural Channel Design
• A channel design technique based on the morphological
and morphometric qualities of the Rosgen classification
• Based on measured morphological relations associated
with bankfull flow, geomorphic valley type, and
geomorphic stream type
• Not without criticism (primarily from form indices rather
than process indices)
Implications of Rosgen’s NCD
to Stream Restoration
• Probably the most widely-used approach
in SR today
• Method now part of EPA’s website of tools
and technology
• Short-courses and software
• Not without serious criticism
ASCE Channel Design (1)
(Shields et al., JHE, 2003)
1. Determine “channel-forming” or “dominant”
(a) Effective discharge (Qeff) determined as the crossproduct of flow frequency and sediment load
(b) Bankfull discharge (Qbf) may or may not be the
same as Qeff
(c) Discharge for a given return period (Qri), 1 to 2.5
years, if data are available
2. Bed material size—based on sampling
ASCE Channel Design (2a)
For threshold channels: channel boundary is immobile at
3. Compute a preliminary channel width
(a) Hydraulic geometry: w = aQb
(b) Regime relation: w = kQlD50m
4. Using D50, estimate maximum shear stress tmax
5. Using D50 and estimated channel sinuosity, bank
vegetation, and flow depth, determine flow resistance
coefficient (may be based on grain size, i.e., n  D50)
6. Using flow continuity and uniform flow equation (i.e.,
Manning’s equation), compute the average depth and
bed slope
7. Stability checks (typically models such as SAM, HEC-6)
ASCE Channel Design (2b)
For active-bed channels: channel boundary is mobile at Qeff
3. Compute a sediment inflow, based on supply reach and
sediment transport equation
4. Develop a family of slope-width curves that satisfy
resistance and sediment transport relations (regional
curves, hydraulic geometry, SAM)
5. Reduce range of solutions to meet site constraints such
as S, w, or d
6. Compute sediment transport out of section and ensure
no spatial gradient
7. Stability checks (typically models such as SAM, HEC-6)
Characterization of Stream
Meanders for Restoration (1)
Approaches to stream meander development:
Construction of asymmetric cross-sections
Techniques to induce point-bar development in
discrete locations
Two-stage channel design
Pool-and-riffle re-creation
Sinuosity and meander restoration
Rinaldi and Johnson (1997) examined the utility of common
indices for meander characterization
Characterization of Stream
Meanders for Restoration (2)
l = 10.95W1.01
A = 4.48W1.02
Rc = 2.59W1.01
(in SI units)
Characterization of Stream
Meanders for Restoration (3)
Compared to 18 streams in MD
Problems due to:
1. MD streams are smaller
2. MD streams have lower Sn
3. Vegetation differences
4. Effect of urbanization
Stream Restoration Approaches II
• Various approaches to stream channel design
– Regime, Analogy, Hydraulic Geometry, Extremal Hypothesis,
– Limitations with each method
• Rosgen’s Natural Channel Design combines hydraulic
geometry with a stream channel classification and
evolution scheme; not without criticism
• ASCE Channel Design supports simplified analytic
• Meander indices need to be carefully applied in river
restoration design

similar documents