Stream Restoration

Prep8AICI by the North Carolina Stream Restonltlon Institute

and North Carolina Sea Grant

*554360*
SDMS Doc ID 554360
INC STATE UNIVERSITY I
North Carolina State University and North Carolina A&T State University commit themselves to positive
action to secure equal opportunity regardless of race, color, creed, national origin, religion, sex, age or
disability. In addition, the two Universities welcome all persons without regard to sexual orientation.
 Contents

Introduction to Fluvial Processes 1

Stream Assessment and Survey Procedures 2

Rosgen Stream-Classification Systems/

Channel Assessment and Validation Procedures 3

Bankfull Verification and Gage Station Analyses 4

Priority Options for Restoring Incised Streams 5

Reference Reach Survey 6

Design Procedures 7

Structures 8

Vegetation Stabilization and

Riparian-Buffer Re-establishment 9

Erosion and Sediment-Control Plan 10

Flood Studies 11

Restoration Evaluation and Monitoring 12

References and Resources 13

Appendices

 Preface

Streams and rivers serve many purposes, including water supply, The authors would like to thank the following people for reviewing
wildlife habitat, energy generation, transportation and recreation. the document:
A stream is a dynamic, complex system that includes not only Micky Clemmons

the active channel but also the floodplain and the vegetation Rockie English, Ph.D.

along its edges. A natural stream system remains stable while Chris Estes

transporting a wide range of flows and sediment produced in its Angela Jessup, P.E.

watershed, maintaining a state of "dynamic equilibrium." When Joseph Mickey

changes to the channel, floodplain, vegetation, flow or sediment David Penrose

supply significantly affect this equilibrium, the stream may Todd St. John

become unstable and start adjusting toward a new equilibrium

state. This transition may take a long time and cause big Funding for this guidebook was provided by:
changes to water quality, habitat and adjacent property. N.C. Department of Environment and Natural Resources,

Stream restoration is the re-establishment of the general structure, EPA 319 Grant Program

function and self-sustaining behavior of the stream system that N.C Department of Transportation
existed prior to disturbance. It is a holistic process that requires
an understanding of all physical and biological components of
the stream system and its watershed. Restoration includes a
broad range of measures, including the removal of the watershed
disturbances that are causing stream instability; installation of
structures and planting of vegetation to protect streambanks and
provide habitat; and the reshaping or replacement of unstable North Carolina
stream reaches into appropriately designed functional streams
and associated floodplains.
This document promotes a natural channel design approach
to stream restoration. It is intended primarily as a reference for
natural resource professionals who plan, design, review and
implement stream-restoration projects. This document is not a
substitute for training and experience. Users should take advantage
of training opportunities and work closely with experienced
stream-restoration professionals to learn more about natural
channel-design principles. Users must recognize that all stream-
restoration projects are different and require applications of specif­
ic techniques to meet project objectives. This document provides
a general framework and some design aids to help planners and
designers address complex stream-restoration projects.
The techniques and methodologies described in this document
are evolving rapidly. New design aids are being developed that
will improve design efficiency and confidence. We encourage
stream-restoration professionals to carefully document their
experiences—including project successes and failures—so that
the restoration community can better understand the appropriate
techniques for various conditions.

From the Authors (listed in alphabethical order):

Barbara A. Doll
Garry L. Grabow
Karen R. Hall
James Halley
William A. Harman
Gregory D. Jennings
Dani E. Wise

2 Preface Stream Restoration Stream Restoration Preface 3

 Introduction to Fluvial Processes Chapter 1
Bankfull Discharge and Stage 1.1
Natural Channel Stability 1.2
Channel Dimension 1.3
Channel Pattern 1.4
Channel Profile 1.5
Channel Features 1.6
Biological Considerations of Stream Restoration 1.7
Conclusions 1.8

4 Preface Stream Restoration

 Chapter 1: Introduction to Fluvial Processes Figure 1.2
Stream order classification
Streams and rivers are integral parts of the landscape that The Federal Interagency Stream
Restoration Working Group,
carry water and sediment from high elevations to downstream 1998, 1-26
lakes, estuaries and oceans. The land area draining to a stream
or river is called its watershed. When rain falls in a watershed, it
runs off the land surface, infiltrates the soil or evaporates (Figure
1.1). As surface runoff moves downslope, it concentrates in low
areas and forms small stream channels. These are referred to
as ephemeral channels, which carry water only when it rains.
Downstream from ephemeral channels are intermittent streams,
which carry water during wet times of the year. These streams
are partially supplied by groundwater that rises to the surface
Figure 1.1 as stream base flow. They dry up
The Hydrologic Cycle when groundwater levels drop.
The Federal Interagency Stream Farther downstream, where
Restoration Working Group,
1998, 2-3. base flow is large enough to

its longitudinal zone and the relative size and depth of its channel.
sustain stream flow through­
The uppermost channels in a drainage network (i.e., headwater
out the year, perennial

streams are formed.

channels with no upstream tributaries) are designated as first-order
The size and flow of
streams down to their first confluence (Strahler, 1957). A second-order
a stream are directly
stream is formed below the confluence of two first-order channels.
related to its watershed
Third-order streams are created when two second-order channels
area. Other factors that
join, and so on (Figure 1.2).
affect channel size and
In addition to transporting water and sediment, natural streams
stream flow are land use,
provide habitat for many aquatic organisms, including fish,
soil types, topog­ amphibians, aquatic insects, mollusks and plants. Trees and
raphy and climate.
shrubs along the banks provide a food source and regulate water
The morphology
temperatures. Channel features such as pools, riffles, steps and
—or size and
undercut banks provide diversity of habitat, oxygenation and Figure 1.3
shape—of the
cover. For these reasons natural resource managers increasingly Bankfull bench below top of
channel reflects
use natural channel design to restore impaired streams. bank in an incised channel
all of these factors.

Though streams
1.1. Bankfull Discharge and Stage
and rivers vary in size,
The most important stream process in defining
shape, slope and bed
channel form is the bankfull discharge, which is essentially
composition, all streams
the same as the effective—or dominant—discharge.
share common characteris­ Bankfull discharge is the flow that transports the
tics. Streams have left and
majority of a stream's sediment load over time and
right banks and beds consisting
thereby forms and maintains the channel. Any flow
of mixtures of bedrock, boulders,
that exceeds the stage of the bankfull flow will move
cobble, gravel, sand or silt/clay.
onto the floodplain; therefore bankfull stage is consid­
Other physical characteristics shared
ered the incipient point of flooding. This may or may
by some stream types include pools,
not be the top of the streambank. If the stream has
riffles, steps, point bars, meanders,
become incised due to changes in the watershed or
floodplains and terraces. All of these streamside vegetation, the bankfull stage may be a
characteristics are related to the interactions among climate, small bench or scour line on the streambank. In this
geology, topography, vegetation and land use in the watershed. case the top of the bank, which was formerly the
The study of these interactions and the resulting streams and floodplain, is called a terrace. A stream that has
rivers is called fluvial geomorphology. terraces close to the top of the banks is considered
Streams are classified—or ordered—according to the hierarchy an incised—or entrenched—stream (Figure 1.3). If the
of natural channels within a watershed. The order of a stream stream is not entrenched, then bankfull is near the
can provide clues about other stream characteristics, including top of the bank (Figure 1.4). For examples of bankfull indicators,
refer to River Course Fact Sheet Number 3 (Appendix A). On aver-
6 Chapter 1 Stream Restoration Stream Restoration Chapter 1 7
 age, bankfull discharge occurs every 1.5 years. In
other words, each year there is about a 67 percent
chance of a bankfull discharge event. The Rosgen
stream-classification system (Rosgen, 1996) uses
bankfull stage as the basis for measuring the width-
to-depth and entrenchment ratios. Therefore, it is
critical to correctly identify bankfull stage when
classifying streams and designing stream-restoration
measures. The Rosgen stream classification is
discussed in detail in Chapter 3 and in River Course
Fact Sheet Number 2 (Appendix A).

Figure 1.4 1.2. Natural Channel Stability

A naturally stable stream channel maintains its dimension,
Bankfull is at the top of the
streambank on this reference pattern and profile such that the stream does not degrade or
reach stream aggrade. Stable streams migrate across the landscape slowly
over geologic time while maintaining their form and function.
Naturally stable streams must be able to transport the sediment
load supplied by the watershed. Instability occurs when scouring
causes the channel bed to erode (degrade) or excessive deposi­
tion causes the channel bed to rise (aggrade). A generalized rela­ 1.3. Channel Dimension
tionship of stream stability is shown as a schematic drawing in The dimension of a stream is its cross-sectional view or perspective. Figure 1.6
Figure 1.5. The drawing shows that the product of sediment load Specifically, it is the bankfull cross-sectional area (bankfull width Pattern measurements of a
and sediment size is proportional to the product of stream slope multiplied by bankfull mean depth) measured at a stable riffle in meander bend
and discharge—or stream power. A change in any one of these the stream. The width of a stream generally increases in the Rosgen, 1996, 2-6
variables causes a rapid physical adjustment in the stream channel. downstream direction in proportion to the square root of discharge.
Stream width is a function of discharge (occurrence and magni­
tude), sediment transport (size and type) and the streambed and
bank materials. North Carolina has a humid subtropical climate
with abundant rainfall and vegetation throughout the year.
Because vegetation along streambanks provides resistance to
erosion, our streams are often narrower than those in more arid
regions. The mean depth of a stream varies greatly from reach
to reach depending on channel slope and riffle/pool or step/pool
spacing.

1.4. Channel Pattern

Stream pattern refers to the "plan view" of a channel as seen
from above. Natural streams are rarely straight. They tend to
follow a sinuous path across a floodplain. The sinuosity of a
stream is defined as the channel length following the deepest
point in the channel (the thalweg) divided by the valley length,
which is measured along the direction of fall of the valley. In
general, channel sinuosity increases as valley gradient decreases.
A meander bend increases resistance and reduces channel gradient
relative to a straight reach. The geometry of the meander and
spacing of riffles and pools adjust so that the stream performs
minimal work. Stream pattern is qualitatively described as
Figure 1.5
straight, meandering or braided. Braided channels are less sinuous
Factors affecting channel degradation and aggradation than meandering streams and possess three or more channels
Reproduced with permission from the American Society of Civil Engineers from
Lane, E.W. 1955. The importance of fluvial morphology in hydraulic engineering.
on a given reach. Quantitatively, stream pattern can be defined
Proceedings from the American Society of Civil Engineers. 81(745): 1-17. by measuring meander wavelength, radius of curvature, amplitude
and belt width (Figure 1.6).
8 Chapter 1 Stream Restoration Stream Restoration Chapter 1 9
 1.5. Channel start of the glide, but the glide usually begins where coarser
Profile materials have been deposited.) The inside of the meander bend
The profile of a is a depositional feature called a point bar, which also helps
stream refers to its maintain channel form. Step/pool sequences are found in high-
longitudinal slope. At gradient streams. Steps are vertical drops often composed of
the watershed scale, large boulders, bedrock knick points, downed trees, etc. Deep
channel slope gener­ pools are found at the bottom of each step. The step serves as a
ally decreases down­ grade control, and the pool dissipates energy. The spacing of
stream. The size of step pools shortens as the channel slope increases.
the bed material also
typically decreases in 1.7. Biological Considerations of Stream Restoration
the downstream Stream restoration may be undertaken for a number of reasons,
direction. Channel including to repair erosion problems or to improve fish and
slope is inversely wildlife habitat. When the project is done correctly, using natural
related to sinuosity. channel design, biological enhancements will always be a side
This means that steep benefit. This is because a natural channel design utilizes a
streams have low reference reach, which provides a template for restoring a stable
sinuosity and flat and biologically diverse stream channel (see Chapter 6). Biologically,
streams have high stream channels include the area below bankfull as well as the
sinuosity. The profile floodplain. A restored stream reach should provide enhancements
Figure 1.8
of the streambed can that are demonstrated at the reference reach. For example,
be irregular because establishing and protecting a vegetated buffer that includes all or Location of features in a
part of the floodplain will provide a number of benefits. Trees and step-pool system
of variations in bed Rosgen, 1996, 5-10
material size and shrubs growing within the buffer will produce a root mass that
shape, riffle/pool
spacing and other
variables. The water-
surface profile mimics
the bed profile at low
Figure 1.7 flows. As water rises in a channel during storms, the water-sur­
Features of natural streams face profile becomes more uniform (Figure 1.7).
From Hey, R.D. and Heritage, G.L.
(1993). Draft guidelines for the 1.6. Channel Features
design and restoration of flood
alleviation schemes. National
Natural streams have sequences of riffles and pools or steps
Rivers Authority, Bristol, UK, R&D and pools that maintain channel slope and stability. These fea­
Note 154 tures are shown in figures 1.7 and 1.8. The riffle is a bed feature
that may have gravel or larger rock particles. The water depth
is relatively shallow, and the slope is steeper than the average
slope of the channel. At low flows, water moves faster over riffles,
which removes fine sediments and provides oxygen to the
stream. Riffles enter and exit meanders and control the streambed
elevation. Pools are located on the outside bends of meanders
between riffles. The pool has a flat surface (with little or no slope)
and is much deeper than the stream’s average depth. At low
flows, pools are depositional features and riffles are scour features.
At high flows, however, the pool scours and the bed material
deposits on the riffle. This occurs because a force applied to the
streambed, called shear stress, increases with depth and slope.
Depth and slope increase rapidly over the pools during large
storms, increasing shear stress and causing scour. Runs and
glides are transitional features between riffles and pools. A run is
the transitional feature between a riffle and a pool. A glide is the
upward sloping area of the bed from the pool to the head of the
riffle. (A flattening of the negative slope sometimes marks the
10 Chapter 1 Stream Restoration Stream Restoration Chapter 1 11
 will greatly increase bank stability. Leaves from these trees will
shade the stream through the hottest part of the year, and when
they drop in the fall, provide organic detritus that fuels food
chains in lower-order streams. Riparian vegetation also provides
food and hiding places for many wildlife species. Since stream
corridors may be the only undeveloped areas within a watershed
or the only linkage between woodlands, they are important travel
routes for animals. The stems and root mass of the riparian vege­
tation benefit water quality by filtering sediment and other pollu­
tants from surface and subsurface flow so these substances
won’t enter the stream and harm aquatic organisms. Restoration
projects should provide these benefits by replacing or enhancing
riparian vegetation. Use of native plants is encouraged because
they are less invasive and better for wildlife (see Section 2.10).
Restoration of proper dimension, pattern and profile will create
a channel that moves water and sediment through the reach
without causing aggradation or degradation. Restored streams
enable the sorting of bed material, which results in habitat diver­
sity. This is particularly important to such fish species as trout,
which require clean gravel for reproduction. Sorting benefits
aquatic organisms by providing stable habitats. In high-gradient
Stream Assessment and Survey Procedures Chapter 2
streams, fish and other aquatic organisms use the space Office Procedures
between gravel, cobble and boulders for resting and feeding. Watershed Drainage Area Measurement 2.1
These sites provide an escape from swift currents higher in the Land-Use Survey 2.2
water column. In many degraded streams the absence of pool
habitat may limit gamefish populations. Structures used in natu­ Field Procedures
ral channel design, such as vanes, cross-vanes, weirs and root- Bankfull Identification 2.3
wads, create and maintain pool habitat, thereby improving the Dimension 2.4
quality of the fishery (see Chapter 8). Restoration of the proper Pattern 2.5
dimension will ensure that the stream is connected to the flood­ Profile 2.6
plain. As a result, riparian vegetation and other components Substrate Analysis 2.7
that roughen the channel will mitigate damage from floodwaters. Bar, Pavement and Subpavement Sampling
This guidebook provides examples of how to enhance the Methods and Scour Chains 2.8
biological benefits of a restoration project (see Chapter 8). Estimating Bankfull Discharge and Velocity 2.9
Assessing Riparian Condition 2.10

1.8 Conclusions
A stream and its floodplain comprise a dynamic environment
where the floodplain, channel and bedform evolve through natu­
ral processes that erode, transport, sort and deposit alluvial
materials. The result is a dynamic equilibrium in which the stream
maintains its dimension, pattern and profile over time, neither
degrading nor aggrading. Land-use changes in the watershed,
channelization, culverts, removal of streambank vegetation,
impoundments and other activities can upset this balance. As a
result, large adjustments in channel form, such as extreme bank
erosion and/or incision, will happen. A new equilibrium may
eventually result, but not before the associated aquatic and ter­
restrial environment are severely damaged. Understanding natu­
ral stream processes and applying this knowledge to stream-
restoration projects will help create a self-sustaining stream with
maximum physical and biological potential.

12 Chapter 1 Stream Restoration

 Chapter 2: Stream Assessment and Survey Procedures
An existing-condition survey is an important first step in the
stream assessment and restoration process. Data and information
collected from the existing-condition survey are used to determine
the stability of the project stream reach and the need for restoration.
They also are used to determine the potential for restoration,
and later they are essential to developing a restoration plan. The
existing-condition survey is accomplished through a quantitative
and qualitative investigation of the stream corridor and its watershed.
A morphological investigation of the stream is a key component
of the survey and includes assessment of channel dimension,
pattern, profile and substrate materials. Data collected during
the survey are used to determine if the stream is evolving toward
stability or instability and if the cause of any instability is localized
or system-wide (see Chapter 3). Examples of localized instability
include removal of riparian vegetation and trampling of the
streambanks by livestock or people. System-wide instability is
often caused by channel incision, which causes headward ero­
sion that continues upstream until it is stopped by a knick point.
At a minimum, the following steps should be completed for the
existing-condition survey.

Office Procedures

2.1. Watershed Drainage Area Measurement

Delineate the project watershed boundary and calculate the 2.4. Dimension Figure 2.1
drainage area. Most people use a geographic information system The permanent cross section is the location for measuring Watershed delineation for the
(GIS) with a topographic map layer such as the U.S. Geological channel dimensions (width, depth and cross-sectional areas), East Prong of the Roaring
Survey (USGS) digital line graphs (Figure 2.1). A topographic map stream discharge, particle size distributions and other long-term River restoration project in
and planimeter work fine as well. Depending on the length of the work. Establish at least one permanent cross section over a riffle Stone Mountain State Park,
project reach, it may be necessary to calculate the drainage area and another over a pool. Ideally, it is best to measure the dimen­ Wilkes and Alleghany coun­
at both upstream and downstream ends of the project. ties, North Carolina
sion of several riffles and pools.
Step 1: Establish permanent markers for cross-section endpoints
2.2. Land-Use Survey by driving a 4-foot-by-1/2-inch-diameter piece of rebar vertically
Complete a survey of land use in the watershed. This should into the ground, leaving one-half inch above the ground if it is
include both historical (when available) and present land uses. acceptable to the landowner. Attach colored plastic caps to
Resources may include aerial photographs, topographic maps the top of the rebar for identification. Drive a wooden stake
or zoning maps. As part of the survey, calculate the Natural beside the rebar and mark the cross-section identification
Resource Conservation Service (NRCS) curve number and the on the stake (usually the location of the cross section on the
percentage of impervious surface (roads, parking lots, etc.) in the longitudinal survey, such as XSEC 4+05).
watershed. The method for calculation is included in Appendix B. Step 2: Measure and note the cross-section endpoint locations
This information will help determine which curve—rural or urban — with a tape. Triangulate between a benchmark, the nearest
to use for bankfull verification (see Appendix D for North Carolina regional curves). cross-section endpoint and another permanent feature, such
as a large tree. Record the measurements in the field book so
Field Procedures the cross section can be relocated for future surveys.
(Adapted from Harrelson, 1994) Step 3: Attach the zero end of the tape to the stake that is on the
2.3. Bankfull Identification left when looking downstream (use a second piece of rebar or
River Course Fact Sheet Number 3 (Appendix A) offers a stepwise another stake to hold the tape). Stretch the tape so it is tight
procedure for verifying bankfull through development and use of and level above the water from the left endpoint to the right
regional curves. Complete and verify bankfull identification before endpoint.
proceeding with the existing-condition survey. Maintain the bankfull Step 4: Set up the surveyor’s level. Start with the surveyor’s rod
and inner-berm flags used in the bankfull identification procedure on the benchmark to establish the height of instrument (HI).
because they will be needed for the longitudinal-profile survey (see Starting with the left endpoint stake as zero, begin the channel
Section 2.6).
14 Chapter 2 Stream Restoration Stream Restoration Chapter 2 15
Figure 2.2 cross-section survey (left and right are always determined Figure 2.4
Example of surveyed
looking downstream). Record the rod reading for the top of the
left pin and the ground at the left pin if they are different. Along Pattern measurements
cross section of a meander bend
Rosgen, 1996, 5-24 the tape, shoot the elevation of each important feature and Rosgen, 1996, 2-6
break in slope, such as top of bank, bankfull, inner berm, edge
of water (water surface and ground) and thalweg. Note these
features in the field book. For cross-section survey methods
and examples of how to
set up field-survey notes,
see Harrelson, 1994
(available for download at
www.stream.fs.fed.us/PDFs/
Cross-section
RM245.PDF).
survey is shown in figures
2.2 and 2.3. Sample field
data sheets are in
Appendix B.
Step 5: Close the survey
by setting the survey rod Step 3: Measure channel belt
back on the benchmark and width, Wblt (figures 2.4 and 2.5).
verifying that the foresight for Belt width for a particular
this shot subtracted from the meander bend is the straight-line
instrument height is equal to the distance from the crest of the bend
known elevation of the benchmark. being evaluated to the crest of the
Step 6: Calculate bankfull cross- next downstream bend. Overall belt
sectional area for all riffles (Abkf) and width for a stream is the straight-
pools (Apool) using the procedures line distance from the two outer­
outlined in River Course Fact Sheet most bends of the channel.
Number 3 (Appendix A). Step 4: Measure meander wave­
Step 7: Calculate bankfull width, length, Lm (figures 2.4 and 2.5).
Wbkf, as the horizontal distance Meander wavelenth for a particu­
between the left and right bankfull lar meander bend is the straight-
stations. line distance from the crest of the
Step 8: Calculate mean depth, upstream meander to the next
Dbkf=Abkf/Wbkf. downstream meander.
Figure 2.3 Step 9: Calculate max depth, Dmax, as the vertical distance Figure 2.5
Cross-section survey between bankfull elevation and the thalweg elevation. 2.6. Profile Plan-form measurements
The longitudinal-profile survey establishes the elevation of the and dimensionless ratios
2.5 Pattern existing streambed, water surface, inner berm, bankfull, and top
Complete plan-form measurements—including sinuosity (K), of bank or terrace features. It helps the designer determine and
meander wavelength (Lm), radius of curvature (Rc) and belt width monitor the lengths, depths and slopes of all the stream features
(Wblt))—using aerial photos if available (figures 2.4, 2.5 and 2.6). (or facets), including riffles, runs, pools and glides.
Step 1: Measure sinuosity. Step 1: Establish a benchmark for the project site. If possible use
Sinuosity is a measure of how crooked a stream is. an existing USGS or Federal Emergency Management Agency
Specifically, it is the channel length divided by a straight-line (FEMA) benchmark. Permanent structures such as a concrete
valley length (Figure 2.6). The greater the number, the higher the headwall or manhole cover also can be used. If permissible,
sinuosity. Sinuosity is related to slope. Natural streams with install a permanent benchmark. Methods for establishing a
steep slopes have low sinuosity, and streams with low slopes benchmark are discussed in Harrelson, 1994 (available for download
typically have high sinuosity. Sinuosity should be measured from at www.stream.fs.fed.us/PDFs/RM245.PDF).
large-scale aerial photographs; do not use topographic maps Step 2: Start the survey at a stable, upstream riffle and continue
with scales of 1:24,000 or less. through the reach to a stable downstream riffle. The first station
Step 2: Measure radius of curvature at several meander bends. should be at the upstream edge (head) of the riffle. This point
Radius of curvature, Rc=C2/8M +M/2, is the degree of curvature is the highest elevation of the riffle.
for an individual meander bend (figures 2.7 and 2.8).
16 Chapter 2 Stream Restoration Stream Restoration Chapter 2 17
 Step 3: Set up the level so that 2.7. Substrate Analysis

the benchmark and as much of The composition of the streambed and banks is an important
the site as possible are visible. facet of stream character. It influences channel form and
This is a difficult task in North hydraulics, erosion rates, sediment supply and other parameters.
Carolina because of the dense Each permanent reference site should include a basic characteri­
vegetation. A lot of turning zation of bed and bank material. For more information on
points may be necessary. The substrate sampling, see Bunte and Abt, 2001 (Section 13.3).
best survey locations often are You may download this report, RMRS-GTR-74, from the U.S. Forest Service’s Rocky
in the channel, which usually Mountain Research Station Web site, http://www.fs.fed.us/rm/main/pubs/
means the survey instrument electronic/rmrs_gtr.html. Studies of fish habitat, riparian ecosystems or
may be below the top of the stream hydraulics may require more detailed characterization of
streambank. In this case, use a substrates and bank materials than is provided in this manual.
hand level and survey rod to See papers by Dorava (2001), Gore (1988), Merritt (1984), Frothingham (2001),
measure the distance between Montgomery (2001) and Statzner (1988) referenced in Section 13.3.
the survey instrument and the The composition of the streambed (substrate) influences how
top of the bank. This distance streams behave. Steep mountain streams with beds of boulders
is referred to as a negative and cobbles act differently than low-gradient streams with beds
foresight. When possible, set of sand or silt. This difference may be documented by a quantitative
the instrument atop the bank description of the bed material called a pebble count. Figure 2.8
or terrace and clear any limbs There are three methods of pebble counts, each with different Field measurement of
and leaves from the line of purposes. The first and most efficient method, a reachwide radius of curvature
sight. Stretch a tape along the pebble count (developed by Wolman, 1954, and modified by
thalweg, starting at the Rosgen, 1996), samples a
upstream riffle. Make sure that total of 100 pebbles from
the thalweg distance surveyed cross sections throughout
is at least 20 times the bankfull the longitudinal reach of
Fig. 2.6 width or encompasses a minimum of two full meander wave­ the stream. This count is
lengths. If flags are not still in place from the bankfull identifi­ used for stream classifica­
Sinuosity measurement
Rosgen, 1998b, 211 cation, replace them for verification during the survey. tion. The second method
Step 4: Record the longitudinal station from the tape, then use samples 100 pebbles at a
the level to read a foresight for the thalweg, water surface, single cross section. This
inner berm (if present), bankfull and top of the low bank. is for cross-section analysis.
Collect this data at the head of every feature (riffle, run, pool The third method also
and glide) and the maximum pool-depth location. Note the samples 100 pebbles at a
channel feature associated with the longitudinal station, i.e., riffle, but includes only the
riffle, run, pool or glide. See Harrelson, 1994 (available for download at pebbles from the wetted
www.stream.fs.fed.us/PDFs/RM245.PFD) for longitudinal-survey methods perimeter (anywhere the
and examples of how to set up field-survey notes. Sample field water is in contact with
data sheets are in Appendix B. Be sure to collect information the channel bed) at normal
for each bed feature. For a long feature such as a riffle or run, flow. This count is used
take a measurement at least every bankfull width. to calculate entrainment
Step 5: Once the longitudinal survey is finished, close the and velocity.
survey back to the benchmark. The longitudinal profile data
will allow calculation of the length and slopes for all the stream ● Reachwide characterization of the substrate (Wolman Pebble
features. Count)
Step 1. This technique requires two people—an observer with a
metric ruler to wade the stream and a note-taker to wade or
remain on the bank with a notebook. For stream characteriza­
tion, sample pools and riffles in the same proportion as they
Figure 2.7 occur in the study reach. Once the longitudinal profile is complete,
Radius of curvature compute the percentage of the total length of the profile that is
Rosgen, 1998b, 186 riffle/run and the percentage that is pool/glide. For example,
the reach may be 60 percent riffle/run and 40 percent pool/
glide. Use these percentages to determine the number of
18 Chapter 2 Stream Restoration Stream Restoration Chapter 2 19
 samples to take from these features. If six riffles exist in the measurements is taken. The note-taker should keep count.

longitudinal profile, sample 10 pebbles (left bankfull to right Continue traversing the stream until all areas between the

bankfull) at each riffle. This will give a total of 60 pebbles in left and right edges of water are representatively sampled.

riffles, or 60 percent of the 100 pebbles sampled from the

entire reach. Similarly, collect 40 pebbles from the pools. At 2.8 Bar, Pavement and Subpavement Sampling Methods
each riffle and pool, sample the particles in a transect perpen­ and Scour Chains
dicular to the flow of water, working from left bankfull to right
bankfull. Averting the eyes, pick up the first particle touched ● Bar Sample
by the tip of an index finger at the toe of a wader. Step 1. Collect a bar sample from the lower (downstream) third
Step 2. Measure the intermediate axis of each particle collected of a well-developed point bar in the stream. If significant bank
(Figure 2.9). Measure embedded particles or those too large to be erosion or watershed disturbance has caused sedimentation
moved in place by using the smaller of the two exposed axes. of the lower third of the bar, sample the middle of the bar.
Call out measurements for the note-taker to tally by size class. Step 2. Place a 5-gallon bottomless bucket on the lower third of
Sample pebble count data sheets are in Appendix B. the bar, halfway between the thalweg and the bankfull elevation.
Step 3. Take one step across the channel in the direction of the Place the bucket in an area that contains a representative
opposite (right) bank and repeat the process, continuing to grouping of the maximum particle sizes found on the lower
pick up particles until the requisite number of measurements is third of the bar. Remove the two largest particles from the sur­
taken. The note-taker should keep count. Traverse the stream face covered by the bottomless bucket. Measure and record
perpendicular to the flow. Continue to an indicator of bankfull the intermediate axis (median diameter) and weigh the parti­
stage on the opposite bank so that all areas between the cles individually. The largest particle obtained from the bar is
bankfull elevations are representatively sam­ the di.
pled. If necessary, duck under vegetation or Step 3. Push the bottomless bucket into the bar material.
reach through brush to get an accurate count. Excavate the material within the sample area to a depth equal
Move upstream or downstream to appropriate to twice the length of the intermediate axis of the di. Place
features (riffles or pools) and make additional these materials in a bucket or bag for sieving and weighing.
transects to sample at least 100 particles. For fine bar material: Push the bottomless bucket into the bar
After counts and tallies are complete, plot material. Excavate the material within the sample area to a
the data by size-class and frequency. Figure depth of 4 to 6 inches. Place these materials in a separate
2.10 is an example of a pebble-count form. bucket or bag for sieving and weighing.
A sample pebble count plot is shown in Step 4. Wet-sieve the collected bar materials, using a standard
Figure 2.11. sieve set with a 2-millimeter screen size for the bottom sieve.
(The standard sieve set should include the following sizes in
● Cross-section analysis of the substrate millimeters: 2, 4, 8, 16, 32, 64, 128 and 256.) Place a bucket
Step 1. For cross-section characterization, below the 2-millimeter sieve to catch the smaller material.
sample pools and riffles separately with 100 (Materials in the 256-512 millimeter range should be measured
counts per feature. Sample the pebbles at and weighed individually rather than sieved.) Weigh the sieved
the cross section, moving from left bankfull materials and record weights (less tare weight) by size-class.
elevation to right bankfull elevation, sampling Weigh the bucket with fine materials after draining off as much
at intervals that will equal 100 counts across water as possible. Subtract the tare weight of the bucket to
the bankfull width of the stream. For example, obtain the net weight of the sand and fine material. Include the
if the bankfull width of the stream is 50 feet, individual intermediate axis widths and weights of the two
Figure 2.9 sample a pebble every six inches to equal 100 samples. largest particles that were collected.
Different axes of a pebble
Step 2. Follow Step 2, Reachwide Characterization. Step 5. Determine a material size-class distribution for all of the
Harrelson, et al., 1994, 50 Step 3. Follow Step 3, Reachwide Characterization (disregard collected materials. The data represent the range of channel
last sentence in first paragraph). materials subject to movement or transport as bed-load sediment
materials at bankfull discharge.
● Wetted Perimeter Cross-Section Substrate Analysis Step 6. Plot the cumulative frequency of each sediment-size­
Step 1. Collect 100 pebbles from a riffle cross section, zigzaging range fraction. From the cumulative frequency plot, determine
from the left water’s edge to the right water’s edge at normal flow. size-class indices, i.e., d16, d35, d50, d84 and d95. The d100 should
Step 2. Follow Step 2, Reachwide Characterization. represent the actual intermediate axis width of the largest
Step 3. Take a step forward and collect a pebble, then take a particle when plotted. The intermediate axis measurement of
step backward to collect a pebble, moving across the channel the largest particle will be the top end of the catch range for
in a direction perpendicular to the flow. Repeat the process, the last sieve that retains material. Note: d100=di.
continuing to pick up particles until the requisite number of
20 Chapter 2 Stream Restoration Stream Restoration Chapter 2 21
Figure 2.10 Pebble count field-data form (Rosgen, 1996)

Pebble Count Pebble Count Pebble Count

Site: Date: Date: Date:
Party: Reach: Reach: Reach:
INCHES PARTICLE MILLIMETERS PARTICLE COUNT TOTAL# ITEM % %CUM TOTAL# ITEM % %CUM TOTAL# ITEM % %CUM
SILT/CLAY < .062 S/C 1 2 3

Very Fine .062 – .125 S

Fine .125 – .25 A

Medium .25 – .50 N

Coarse .50 – 1.0 D

.04 – .08 Very Coarse 1.0 – 2 S

.08 – .16 Very Fine 2–4

.16 – .24 Fine 4–6 G

.24 – .31 Fine 6–8 R

.31 – .47 Medium 8 – 12 A

.47 – .63 Medium 12 – 16 V

.63 – .94 Coarse 16 – 24 E

.94 – 1.26 Coarse 24 – 32 L

1.26 – 1.9 Very Coarse 32 – 48 S

1.9 – 2.5 Very Coarse 48 – 64

2.5 – 3.8 Small 64 – 96 C

3.8 – 5.0 Small 96 – 128 O

5.0 – 7.6 Large 128 – 192 B

7.6 – 10 Large 192 – 256 L

10 – 15 Small 256 – 384 B

15 – 20 Small 384 – 512 L

20 – 40 Medium 512 – 1024 D

40 – 160 Lrg-Very Lrg 1024 – 4096 R

BEDROCK BDRK
TOTALS
 ● Pavement Sample

Figure 2.11 Sample pebble count cumulative frequency plot

Step 1. To sample the pavement of the stream, select a repre­
sentative riffle that has been surveyed in cross section. To
define the sampling area, use a bottomless 5-gallon bucket
to isolate a section of the riffle that is the most depositional.
Locate the sample in the coarsest part of the riffle but not in
the thalweg.
Step 2. Push the bucket into the riffle, being sure to eliminate
flow of water through the sample area.
Step 3. Carefully remove the top veneer (surface layer only) of
the particles within the sample area by picking the particles
off the top, smaller particles first. Continue removing the small
and then the large particles, working from one side of the
sample area to the other. Weigh the largest and second largest
particles and measure the length of their intermediate axes.
Place the pavement material in a bag or bucket for sieving
and weighing.

● Subpavement Sample
Step 1. Collect the subpavement sample beneath the pavement
sample. The bucket should continue to define the boundaries
of the sampling area. Excavate and remove the material below
the pavement sample to a depth equal to twice the intermediate
axis of the largest particle that was collected from the pavement
sample. If an armored layer is reached, do not continue to
excavate below this layer, even if a depth equal to twice the
median diameter of the largest particle in the pavement layer
has not been reached. Place all subpavement sample material
in a separate bucket or bag for weighing and sieving. The sub-
pavement sample is the equivalent of the bar sample; therefore,
the largest particle from the subpavement sample is used in
lieu of the largest particle from the bar sample for entrainment
calculations (see Section 7.2). Note: If larger particles are collected
from the subpavement than from the pavement layer, discard
the sample and select a new sampling location.
Step 2. Wet-seive both the pavement and subpavement samples
separately using a standard sieve set with a 2-millimeter
screen size for the bottom sieve. (The standard sieve set
should include the following sizes in millimeters: 2, 4, 8, 16, 32,
64, 128 and 256.) Place a bucket below the 2-millimeter sieve
to catch the smaller material. (Materials in the 256-512 millime­
ter range should be measured and weighed individually rather
than sieved.) Weigh and record each sieve fraction (less the
tare weight). Weigh the bucket with fine materials after draining
off as much water as possible. Subtract the tare weight of the
bucket to obtain the net weight of the sand and fine material.
Include the intermediate axis widths and weights of the two
largest particles collected from the pavement and the largest
collected from the subpavement.
Step 3. Determine size-class distribution for the materials by
plotting the cumulative frequency of each fraction. From the
cumulative frequency plot, determine size-class indices, i.e.,
d16, d35, d50, d84, d95, for both the pavement and subpavement
samples. The d100 should represent the actual intermediate axis
24 Chapter 2 Stream Restoration Stream Restoration Chapter 2 25
 width of the largest particle when plotted. The intermediate Hydraulic Radius is determined using equation 2:
axis measurement of the largest particle will be the top end of A
the catch range for the last sieve that retains material. Note: d100=di. R= WP Equation (2)

● Scour Chains Where:

Scour chains installed in the bed of the stream can measure WP = Wetted Perimeter of the channel bottom at bankfull

the depth of scour or bed deposition. Bulk substrate samples stage (ft)

collected in conjunction with scour-chain monitoring can help A = Cross-Sectional Area of the riffle at bankfull stage (sq. ft.)

characterize bed material. Bury scour chains in the stream at a

permanent cross-section riffle, with the chain extending vertically Cross-Sectional Area and Wetted Perimeter can be calculated

through the stream substrate. Include the location of the chain in using the cross-section survey data. Wetted Perimeter, WP, can

the cross-section survey data so it can be found later. Place the also be approximated using equation 3. Equation 3 assumes a

top of the chain level with the existing streambed surface. When rectangular channel shape.

a bankfull or other sediment-transporting flow passes through

the stream, material will be lifted off the bed, transported and WP = 2D + W Equation (3)
deposited farther downstream.
The scour chain will be folded over in the downstream direc­ Where:

tion at the depth to which the bed material was scoured and D = Average Bankfull Depth of the riffle cross-section (ft)

newly deposited during the flow. To measure the depth of scour W = Bankfull Width at the riffle (ft)

(length of chain from the top to the point where the chain was
folded over), dig through the newly deposited material to the Manning’s Roughness Coefficient can be estimated by using

chain. If deposition (or aggradation), rather than scour, has Chow’s coefficients for various channel substrate and vegeta­
occurred, the chain will remain in a vertical position, concealed tion characteristics (1959). Velocity, v, can then be determined

by the deposited sediment load. To obtain a scour-chain sub­ using the Continuity Equation (Equation 4):

strate sample: Q
Step 1: Place a 5-gallon bottomless bucket over the location of V= A Equation (4)
the chain prior to excavating. The bucket will define the sub­
strate sampling area. Where:

Step 2: Collect a pavement and subpavement sample over the V = Bankfull Velocity (fps)

chain. For the subpavement sample, continue to excavate Q = Bankfull Discharge (cfs)

down to the chain (rather than to the depth equal to twice the A = Bankfull Cross-Sectional Area at the riffle cross-section

intermediate axis of the largest particle found on the pave­ (sq. ft.)

ment).
2.10 Assessing Riparian Condition
2.9. Estimating Bankfull Discharge and Velocity Compose a general description of the topography or promi­
Discharge is the volume of water flowing through a stream nent topographic features in the floodplain, as well as soil texture
channel cross section per unit time. If the stream has a USGS and type. Important features may include ditches, old crop rows,
gage, use the stage-discharge rating table to determine the sloughs and pools, wetlands, knolls or steep banks. Note the
discharge for the specific elevation of the field-determined bank- length and width of the valley. If the project is in an urban setting,
full stage (see Chapter 4 for more information on gage-station note obvious constraints, such as location of utilities, structures
analyses). However, most stream reaches are not gaged, so it and roads.
probably will be necessary to estimate the bankfull discharge Examine and describe soils throughout the floodplain. County
and velocity using other methods. Bankfull discharge, Qbkf, can soil-survey classifications are useful in preparing descriptions.
be estimated using Equation 1, which is Manning’s equation (Chow, During this initial assessment, appropriate labs, including the
1959). N.C. Department of Agriculture (NCDA) Agronomic Division’s
2/3 1/2
soil-testing lab, can perform soil-fertility tests. This information
Q = ( 1.49AR S ) Equation (1) will help determine the nutrient needs of vegetation planted at
n
the project site. An example of a soil-sample form is found in
Where:
Appendix H.
Q = Discharge (cfs)
Next, take a plant inventory. Note the type, size and relative
R = Hydraulic Radius of the riffle cross-section at bankfull
abundance of each species in the project area. Also note and
stage (ft)
flag potential vegetation for transplanting. Utilizing on-site vege­
s = Average Channel Slope (ft/ft)
tation that might otherwise be destroyed by construction is an
n = Manning’s Roughness Coefficient
excellent way to save money and to maintain locally adapted
26 Chapter 2 Stream Restoration Stream Restoration Chapter 2 27
 plant ecotypes. Note invasive and exotic plants that occur within
the project area. Throughout much of North Carolina, stream-
banks and floodplains are infested with invasive and exotic
plants that include kudzu (Pueraria lobata), English ivy (Hedera
helix), Chinese privet (Ligustrum sinense) and multiflora rose
(Rosa multiflora). This vegetation can outcompete native riparian
plants, leading to a decrease in wildlife habitat and food diversity
along the streambanks. Also, non-native vegetation often is less
nutritious for native fauna. If invasive exotic plants inhabit the
project area, take measures to control them before restoring
native vegetation.

Rosgen Stream-Classification System/

Channel Assessment and Validation Procedures Chapter 3
Level I 3.1
Level II 3.2
Level III 3.3
Level IV 3.4

28 Chapter 2 Stream Restoration

 Chapter 3: Rosgen Stream-Classification Figure 3.1
Classification inventory and
System/Channel Assessment and Validation assessment
Procedures Rosgen, 1996, 3-5

The Rosgen stream-classification system categorizes streams

based on channel morphology so that consistent, reproducible
and quantitative descriptions can be made. Through field meas­
urements, variations in stream processes are grouped into distinct
stream types. Rosgen (1996) lists four specific objectives of
stream classification:
1. To predict a stream’s behavior from its appearance.
2. To develop specific hydraulic and sediment relationships for a
given stream type.
3. To provide a mechanism to extrapolate site-specific data to
stream reaches having similar characteristics.
4. To provide a consistent frame of reference for communicating
stream morphology and condition among a variety of disciplines
and interested parties.

The Rosgen stream classification consists of four levels of

detail ranging from broad qualitative descriptions to detailed
quantitative assessments. Figure 3.1 shows the hierarchy (Levels
I through IV) of the Rosgen classification inventory and assess­
ment. Level I is a geomorphic characterization that categorizes
streams as "A," "B," "C," "D," "DA," "E," "F" or "G." Level II is
called the morphological description and requires field measure­
ments. Level II assigns a number (1 through 6) to each stream
type that describes the dominant bed material based on the d50
of the reachwide pebble count. Level III is an evaluation of the
stream condition and its stability; it requires an assessment and
prediction of channel erosion, riparian condition, channel modifi­
cation and other characteristics. Level IV is the verification of
predictions made in Level III and consists of sediment transport,
stream flow and stability measurements.
A hierarchical key to the Rosgen stream-classification system
is shown in Figure 3.3. Use the steps outlined in Level II (Section
3.2) to determine the Rosgen classification for the project stream.

3.1 LEVEL I
Level I is a broad-level description of Rosgen’s major stream
types (Figure 3.2). This description is based on general map and
visual assessment of valley types; landforms; and the stream’s
shape, slope and channel patterns. Valley morphology has a
profound influence on stream type (See Rosgen 1996, Chapter 4).

3.2 LEVEL II
Step 1. Determine single or braided channel. A braided channel
consists of three or more distinct channels. Anything less is
considered a single channel. The only stream types for braided
channels are "D" and "DA." Single or braided channel determi­
nation can be made from aerial photograph or field observation.

30 Chapter 3 Stream Restoration Stream Restoration Chapter 3 31

Figure 3.2 Step 2. Calculate entrenchment ratio. The entrenchment ratio is
Broad-level stream classifi­ a field measurement of channel incision. Specifically, it is the
cation delineation showing flood-prone width divided by the bankfull width. The flood-prone
longitudinal, cross-sectional width is measured at the elevation of twice the maximum depth
and plan-views of Rosgen’s of the channel at bankfull (Figure 3.4). Lower entrenchment ratios
major stream types
Rosgen, 1996, 4-4
indicate channel incision; large entrenchment ratios indicate a
well-developed floodplain. An example of this measurement is
shown in Figure 3.4. The following stream types are
entrenched (low entrenchment ratio): "A," "F" and "G."
2a: Obtain a rod reading for an elevation at the max (bankfull)
depth location at a riffle.
2b: Obtain a rod reading for an elevation at the bankfull stage
location.
2c: Subtract the Step 2 reading from the Step 1 reading to
obtain a max (bankfull) depth value; then multiply the max depth
value times 2 for the 2x max depth value.
2d: Subtract the 2x max depth value from the Step 1 reading for
the Flood-prone Area (FPA) location rod reading. Move the rod
upslope, online with the cross section, until a rod reading for the
FPA location is obtained.
2e: Mark the FPA locations on each bank. Measure the distance
between the two FPA locations.
2f: Determine the distance between the two bankfull stage
locations.
2g: Divide the FPA width by the bankfull width to calculate the
entrenchment ratio.

Step 3. Calculate width-to-depth ratio. The width-to-depth ratio

is a field measurement of the bankfull width divided by the Figure 3.3
mean bankfull depth. To calculate width-to-depth ratio, first
determine the bankfull cross-sectional area and average bank- Rosgen classification of natural rivers
Rosgen, 1996
full depth (see River Course Fact Sheet Number 2, Appendix A). The bankfull
32 Chapter 3 Stream Restoration Stream Restoration Chapter 3 33
 average depth is the 3.3 LEVEL III
cross-sectional area Restoration projects often fail because the designers did not
(Abkf) divided by the incorporate the existing and future channel morphologies into the
bankfull width (Wbkf). design. As mentioned in Chapter 2, data and information collect­
The primary break ed from the existing-condition survey are used to determine the
between various stability of the project stream reach, the need for restoration and
stream types in the the potential for restoration (if needed). Therefore, it is imperative
Rosgen classification that the designer complete morphological analyses upstream and
system is 12, meaning downstream of the project reach. Data collected during the exist-
that the bankfull width ing-condition survey are used to determine if the stream is mov­
is 12 times greater ing toward stability or instability and if the cause of instability is
than the mean bankfull localized or streamwide.
depth. Stream types Watershed-Scale Instability
with width-to-depth Various factors can disrupt the equilibrium of a watershed. In
ratios greater than 12 are "B," "C" and "F." Stream types less North Carolina, modification of the channel (channelization) and
Figure 3.4
than 12 are "A," "E" and "G." The "D" stream types have a development of the watershed are the most common causes of
Measurement technique for width-to-depth ratio greater than 40, and the "DA" stream watershed-scale instability. The designer must address these factors
flood-prone area and types have less than 40. before installing bank-stabilization or habitat-improvement struc­
entrenchment ratio
Rosgen, 1996, 5-20
Step 4. Determine sinuosity (see section 2.5). tures. During watershed-scale adjustments, channel evolution
Step 5. Measure water-surface slope. Measure the water usually progresses from downstream to upstream. For example,
surface from the top of one riffle to the top of another at least an incised stream might have a downstream reach that is develop­
20 bankfull widths downstream. This can be done using the ing a new floodplain at a lower elevation. The rate of bank erosion
data collected from the longitudinal-profile survey (Section 2.6). decreases as the channel dimension, pattern and profile become
The channel slope is calculated by dividing the difference in stable for the given slope and drainage area. However, the distur­
elevation between the water surface at the most upstream bance can have effects that move upstream (in the form of a
head-of-riffle and the most downstream head-of-riffle by the head-cut), causing degradation, widening and deposition.
length of the channel between the two riffles, as measured Local (Reach)Instability
along the thalweg. This is considered the average slope. "A" Local, or reach, instability refers to erosion and deposition
and "B" stream types have the steepest slopes, and "E" and processes not caused by instability in the watershed. Perhaps
"DA" stream types have the lowest. However, slope varies the most common form of local instability is erosion along the
greatly among stream types. outside bank in a meander bend. Local instability also can occur
Step 6. Determine median size of the bed material. A pebble in isolated locations as the result of channel constriction, flow
count is used to determine the median particle size, or d50, of obstructions (ice, debris, structures, etc.), trampling by livestock
the bed material. The d50 means that 50 percent of the material or geotechnical instability (high banks, loss of riparian vegetation,
is smaller and 50 percent is larger. First, conduct a reachwide soil structure, etc.). Local instability problems usually respond to
pebble count by collecting 100 pebbles from a stream reach local bank-protection measures, but stabilization treatment
with a minimum of 20 bankfull widths (see Section 2.7). A cumulative should begin and end at stable riffles.
frequency plot of the particle-size distribution will provide the Channel Stability Assessment
d50. The d50 will provide the Level II classification as shown in Rosgen’s stream-channel assessment methodology includes a
Table 3.1. field assessment of the following variables:
● Stream-channel condition or "state" categories
● Vertical stability—degradation/aggradation
Material Classification Size Range (mm) ● Lateral stability
● Channel pattern
Bedrock 1 >2,048 ● River profile and bed features
Boulder 2 256-2,048 ● Channel dimension relations
Cobble 3 64-256
● Stream channel scour/deposition potential (sediment
Gravel 4 2-64
competence)
Sand 5 0.062-2
● Dimensionless ratio sediment-rating curves
Silt/Clay 6 <0.062
● Channel evolution

Table 3.1 Substrate Material Classification For more information, see Rosgen, 2001b (Section 13.1), available for download at
www.wildlandhydrology.com.

34 Chapter 3 Stream Restoration Stream Restoration Chapter 3 35

 Channel Evolution Figure 3.5
A common sequence of physical adjustments happens in Channel-evolution model
many streams following a disturbance. This adjustment process FISRWG, 1998, 7-35
is often referred to as channel evolution. Disturbance can result (based on Simon, 1989)
from channelization, urbanization, removal
of streamside vegetation or other changes that negatively affect
stream stability. Several models have been used to describe this
process of physical adjustment.
Two models (Schumm et al., 1984 and Simon, 1989, 1995) are most widely
accepted (Figure 3.5 and Table 3.2).
According to Simon’s channel-evolution model, the channel
evolution process is initiated once
a stable, well-vegetated stream that frequently interacts with its
floodplain is disturbed. Disturbance commonly results in an
increase in stream power that causes degradation, often referred
to as channel incision. Incision eventually leads to oversteepen­
ing of banks; when critical bank heights are exceeded, the banks
begin to fail and mass wasting of soil and rock leads to channel
widening. Incision and widening continue upstream. Eventually
the mass wasting slows and the stream begins to aggrade. A
new low-flow channel begins to form in the sediment deposits.
By the end of the evolutionary process, a stable stream with a
dimension, pattern and profile similar to those of undisturbed
channels forms in the deposited alluvium. The new channel is
at a lower elevation than its original form with a new floodplain
constructed of alluvial material. The old floodplain has now
become a dry terrace (FISRWG, 1998).
Channel-evolution models can illustrate the current trends in
a disturbed or constructed channel and show the direction in
which they are moving (Figure 3.6). Evaluate the current stage of
evolution for the project stream before selecting the appropriate
restoration actions.
Streambank Erosion
Streambanks can be eroded by collapse or by moving water.
Collapse or mass failure occurs when the bank is too weak to
resist gravitational forces. Banks that are collapsing or about to
collapse are referred to as being geotechnically unstable. The
physical properties of the streambank should be evaluated to
determine potential stability problems and to identify the dominant
sources of bank instability. Factors to consider include bank and profile and installing root wads and rock vanes can stabilize the Figure 3.6
height, bank angle, surface protection, soil material and soil streambanks. The role of in-stream structures is discussed further in Channel-evolution model
stratigraphy. Whenever possible, the streambank-stabilization Chapter 8. FISRWG, 1998, 7-35
measure should reconstruct the bank so that bankfull is the top Estimates of streambank erosion rates are valuable for evaluating (based on Simon, 1989)
of the bank. This often means building a bankfull bench (Figure 3.8). stream impairment and the need for restoration (FISRWG, 1998; Rosgen, 1996).
Shear stress is a measure of the force of water against the Techniques for estimating streambank erosion rates include cross-
channel boundary (i.e., bed and banks) of the stream. section surveys, bank-erosion pins, photography and photoelectronic
Determining mean shear stress and critical dimensionless shear systems. Recent studies in Wyoming (Troendle et al., 2001) and Oklahoma
stress provides a means for evaluating the stress required to (Harmel et al., 1999) showed correlation between bank erosion rates and
entrain and move sediment in a stream. Changes to the stream various field-measured erodibility factors. By taking relatively simple field
that increase slope or water depth can increase shear stress, measurements, one can use these relationships to predict annual erosion
thus increasing erosion of the banks and bed. Evaluation of shear rates for stream reaches. By conducting these measurements at many
stress and sediment transport are discussed in Section 7.2. locations, one can estimate the expected annual sediment load due
Whether streambank erosion is a localized problem or part of a to streambank erosion for a watershed. This information is valuable in
larger restoration project, restoring the proper dimension, pattern prioritizing restoration projects and targeting resources.
36 Chapter 3 Stream Restoration Stream Restoration Chapter 3 37
 Fig. 3.7
Six evolutionary stages
Class Dominant Process Characteristic Forms Geobotanical
Evidence I of channel evolution

No. Name Fluvial Hillslope

Stable, alternate chan­ Vegetated banks to
nel bars; convex top- flow line
I Premodified Sediment transport—mild
aggradation; basal erosion on
bank shape; flow line II
high relative to top
outside bends; deposition on
bank; channel straight
inside bends
or meandering

Removal of vegetation
Constructed Trapezoidal cross sec­
II (Channelized) tion; linear bank sur­
faces; flow line lower
relative to top bank

Degradation Degradation; basal erosion on Pop-out Heightening and steep­ Riparian vegetation
III banks failures ening of banks; alter­ high relative to flow
nate bars eroded; flow line and may lean
line lower relative to toward channel
top bank

Threshold Degradation; basal erosion on Slab, Large scallops and Riparian vegetation
IV (Degradation banks rotational bank retreat; vertical high relative to flow IV
and Widening) and pop-out face and upper-bank line and may lean
failures surfaces; failure blocks toward channel
on upper bank; some
reduction in bank
angles; flow line very
low relative to top bank

V
Aggradation
and Widening
Aggradation; development of
meandering thalweg; initial
Slab, rotational
and pop-out
Large scallops and
bank retreat; vertical
Tilted and fallen ripari­
an vegetation; re­
III
deposition of alternate bars; failures; low- face, upper bank and establishing vegetation
reworking of failed material angle slides of slough line; flattening on slough line; deposi­
on lower banks previously of bank angles; flow tion of material above
failed material line low relative to top root collars of slough-
bank; development of
new floodplain
line vegetation V
Restabilization Aggradation; further develop­ Low-angle Stable, alternate chan­ Re-establishing vege­
VI (Quasi-equilibrium) ment of meandering thalweg; slides; some nel bars; convex-short tation extends up
further deposition of alternate pop-out vertical face on top slough line and upper
bars; reworking of failed failures near bank; flattening of bank bank; deposition of
material; some basal erosion flow line angles; development of material above root
on outside bends; deposition new floodplain; flow collars of slough-line
on floodplain and bank sur­ line high relative to top and upper-bank vege­
faces bank tation; some vegeta­
tion establishing on
bars

Table 3.2. Channel-evolution model description VI

Simon, 1989, 24, and FISRWG, 1998, 7-36
(photographic examples of each of the six evolutionary stages are provided in Figure 3.7)

38 Chapter 3 Stream Restoration Stream Restoration Chapter 3 39

 Stability Assessment
Procedures
Adjective Hazard or Risk Bank Height/ Root Depth/ Root Bank Angle Surface Totals
Bank Erodibility Hazard Index Rating Categories Bankfull Height Bank Height Density % (Degrees) Protection %
Rosgen (1996) developed the
Bank Erodibility Hazard Index VERY LOW Value 1.0-1.1 1.0-0.9 100-80 0-20 100-80
(BEHI) as a quick way to estimate Index 1.0-1.9 1.0-1.9 1.0-1.9 1.0-1.9 1.0-1.9 5-9.5
the potential for bank erosion along
a stream reach. The BEHI assess­ LOW Value 1.11-1.19 0.89-0.5 79-55 21-60 79-55
ment requires field-determination of Index 2.0-3.9 2.0-3.9 2.0-3.9 2.0-3.9 2.0-3.9 10-19.5
five factors: (1) the ratio of bank MODERATE Value 1.2-1.5 0.49-0.3 54-30 61-80 54-30
height to bankfull height, (2) the Index 4.0-5.9 4.0-5.9 4.0-5.9 4.0-5.9 4.0-5.9 20-29.5
ratio of vegetative-rooting depth to
bank height, (3) the density of HIGH Value 1.6-2.0 0.29-0.15 29-15 81-90 29-15
roots, (4) the streambank angle and Index 6.0-7.9 6.0-7.9 6.0-7.9 6.0-7.9 6.0-7.9 30-39.5
(5) the vegetative bank protection.
VERY HIGH Value 2.1-2.8 0.14-0.05 14-5.0 91-119 14-10
Convert the data to a BEHI index Index 8.0-9.0 8.0-9.0 8.0-9.0 8.0-9.0 8.0-9.0 40-45
and adjust depending on the bank
Figure 3.8 materials and the stratification of the bank (see Table 3.2). EXTREME Value >2.8 <0.05 <5 >119 <10
Bankfull bench on For a general indication of BEHI ratings for various streambank Index 10 10 10 10 10 46-50
restoration site conditions, see Figure 3.9. BEHI data sheets are in Appendix B.
If the banks are made of bedrock or boulders, the BEHI rating in Adjust points with respect to the specific nature of bank materials and stratification, as follows:
most cases should be "very low" and "low," respectively, in spite Bank Materials: bedrock (very low rating), boulders (low rating), cobble (subtract 10 points unless gravel/sand >50
of the lack of vegetative surface protection and root mass. percent, then no adjustment), gravel (add 5-10 points depending on percentage sand), sand (add 10 points), silt/clay
Therefore, the numerical index may need to be reduced substan­ (no adjustment).
tially to reflect this situation. Similarly, cobble banks with less Stratification: Add 5-10 points depending on the number and position of layers.
than 50 percent sand would also be resistant to erosion; subtract
10 points from the total numerical index. In contrast, gravel, sand Table 3.2 Bank Erodibility Hazard Index (BEHI) rating guide
Rosgen, 2001a
and gravel-and-sand mixed banks would be much more likely to
erode and require a higher numerical index. The presence and
position of stratified layers can affect bank erodibility also. Many
layers focused near the bankfull elevation, where the highest
shear stress occurs, would create the most erodible bank, requir­
ing an increase in the numerical index. In contrast, fewer layers
located at the bottom or top of the bank away from bankfull
would necessitate a smaller increase to the index. The BEHI rat­
ing requires visual evaluation of the streambanks, and so it is
subjective. Use consistent rating procedures from site to site.
Two different assessors likely would report a different numerical
index, but probably would report the same overall rating.

Permanent Cross Sections

Establish three to six permanent cross sections at each reach
perpendicular to the direction of flow at points that represent
varying degrees of erosion in straight reaches and bends. Install
left and right survey pins well beyond the top of the bank to
ensure that the pins will not erode with the streambank. Survey
each cross section to identify the channel thalweg, edge of
water, bankfull stage, top of bank and permanent survey pins.
Collect data from enough stations to accurately characterize the
shape of the channel. Repeat cross-section surveys after major
storm events and at least once per year. Plot and overlay the sur­
Fig. 3.9
vey data to determine the amount of erosion over time. Calculate
streambank erosion rates using changes in cross-section area Streambank erodibility factors
Rosgen 1996, 6-40
over time. See Figure 3.10 for an example of how to monitor
bank erosion using a permanent cross section.

40 Chapter 3 Stream Restoration Stream Restoration Chapter 3 41

 Bank Pins and Bank Profile and is thus more important in channel formation and natural Figure 3.11
Where possible, install toe pins vertically into the bed of the channel design.
Students monitor bank
channel at the location of each permanent cross-section. Locate The ability of the stream to transport its total sediment load is profile on Rocky Branch,
the pins close to the bank to enable multiple bank-profile surveys quantified through two measures: sediment-transport competen­ Raleigh, N.C.
in the same location. Depending on the bank height, install two cy and sediment-transport capacity.
to five bank pins with a vertical spacing of 1 to 2 feet into the Competency is a stream's ability to move
outside bank most prone to erosion. If it is not possible to install a particles of a given size; it is a measurement
2
toe pin in the channel bed, install a pin at the top of the bank. of force, often expressed as units of lbs/ft .
Place the pin far enough from the top edge of the bank that the Sediment-transport capacity is a stream's
pin won’t be lost if the bank erodes. Drive half-inch steel rods (3.5 ability to move a given quantity of sediment; it
to 7 feet long) horizontally into the bank, leaving 1 to 2 inches is a measurement of stream power, often
exposed. To calculate the bank profile, place the survey rod on expressed as units of lbs/ ft•sec. Sediment-
Figure 3.10 top of the toe pin to take vertical measurements. Place a second transport capacity is also calculated as a
Permanent cross- surveying rod perpendicular to that rod beginning at the toe pin sediment-transport rating curve, which pro­
section monitoring (Figure 3.11). Record both vertical and horizontal measurements each vides an estimate of the quantity of total
results time the horizontal rod is moved to another feature on the bank. Be sediment load transported through a cross
sure to take measurements often section per unit time. The curve is provided
to obtain a detailed characteriza­ as a sediment-transport rate in lbs/sec ver­
tion of the streambank. sus discharge or stream power.
Calculate streambank erosion Sediment-transport studies are needed in
rates using changes in cross- North Carolina and the rest of the Southeast.
section area over time and Many designers now rely on equations and
changes in bank-pin extensions. data produced by Andrews (1983) and Shield (Leopold, 1994) to vali­
An example of bank-erosion date sediment-transport competency of streams (see Chapter 7).
monitoring using bank profile is However, this data has never been validated for the Southeast.
shown in Figure 3.12. Validation would require both the establishment of gage stations
to monitor discharge and long-term sampling of bedload at sev­
eral reference-reach streams with similar bed material. This moni­
3.4 LEVEL IV toring could help generate sediment-capacity curves, which
Sediment Transport could be used to develop dimensionless sediment-transport
Natural-channel designs are curves (Troendle et al., 2001).
based on the premise that a sta­ If long-term study is possible, a
ble stream maintains its dimen­ designer can develop a sediment-
sion, pattern and profile and rating curve for a stream and
does not degrade or aggrade over a long period. All natural validate models used to predict
channel designs are based on the bankfull discharge and corre­ sediment transport or to predict
sponding floodplain elevation. Bankfull discharge is assumed to changes in sediment load due
be the effective discharge, which is the flow that transports the to changes in watershed land-
bulk of the sediment over a long period. Effective discharge is use. For instance, clear-cutting
calculated as the product of the flow-duration curve and the sed­ or urbanization in the watershed
iment-transport rating curve. Because sediment-rating curves are may increase the proportion of
lacking in the southeastern United States, designers rely suspended sediment to total sedi­
on the bankfull stage and corresponding discharge; sediment- ment load. To develop a sedi­
transport competency; and capacity calculations to ensure that ment-rating curve for either bed-
channels do not aggrade or degrade. These calculations are load or sediment load, obtain
used to predict the size-class and quantity of bedload transport. field measurements of material
Bedload, combined with suspended load, makes up the transported in the stream during
stream’s total sediment load. Bedload is defined as those parti­ different flow events. Several devices and methods are available Figure 3.12
cles that slide, roll and saltate (hop) along the streambed during for collecting samples of suspended sediment and bedload. For Results of bank-profile
storm flows; such material does not start moving until the dis­ more information, see Edwards and Glysson (1999) in Section 13.3. This publication can monitoring on the South
charge amount is at least 40 percent of bankfull discharge. be viewed online at http://water.usgs.gov/pubs/twri/twri3-c2/. Fork of the Mitchell River
Suspended sediment includes the sediment particles that are Stream Stability Validation
transported in the water column; it is an important water-quality In 1998, NC State University initiated a study to develop relation­
parameter. Bedload forms the bed features within the channel ships between bank erosion rates and field-measured erodibility
42 Chapter 3 Stream Restoration Stream Restoration Chapter 3 43
Figure 3.13
Measured streambank erosion factors in the Piedmont and mountains of North Carolina (Patterson
rates in relation to near-bank et al., 1999). Bank-erosion pins were established at 27 cross sections
stress and BEHI score along seven stream reaches representing various land uses.
Bankfull Verification and Gage Station Analyses Chapter 4
Source: Natural Resource
Conservation Service, NC State The NRCS and the N.C. Division of Soil and Water Conservation
Stream Gage Survey Procedure 1

University and NC Division of Soil monitored and validated the bank erosion at three of the university’s
and Water Conservation. study reaches (14 cross sections) and established six cross sections
at three new study sites (Jessup, 2002). Figure 3.13 shows the
results of the follow-up study. Sites with moderate BEHI ratings
exhibited bank erosion rates ranging from 0.04 to 0.74 ft/yr; sites
with high BEHI ratings exhibited 0.11 to 0.45 ft/yr of erosion;
sites with very high BEHI ratings exhibited 0.48 to 1.7 ft/yr; and
sites with extremely high BEHI ratings exhibited 2.19 to 11.15
ft/yr. Additional bank erosion monitoring is needed to expand the
data set and increase the length of the sampling period to
accommodate potential climatic influences.

44 Chapter 3 Stream Restoration

 Chapter 4: Bankfull Verification and Gage
When identifying bankfull or developing regional curves in
Station Analyses urban areas, quantify the level of development in each watershed
using land-use maps or data. Use impervious-cover percentage
Whether assessing the existing condition of a stream or devel­ or NRCS runoff-curve numbers. See NRCS, 1986 (available for
oping a restoration design, it is important to validate the bankfull download at http://www.wcc.nrcs.usda.gov/hydro/hydro-tools-models-tr55.html) for
stage for the stream channel. River Course Fact Sheet Number 3 the method to calculate the curve number. Streams with similar
(Appendix A) explains how to find the bankfull stage in North levels of development within the same hydrophysiographic region
Carolina. The easiest way to validate bankfull is through a gage can be grouped together for a single regional curve.
station reading. However, stations rarely exist along project
reaches. Therefore, it is important to develop a relationship of Stream Gage Survey Procedure: (From Leopold, 1994)
bankfull area and discharge (i.e., hydraulic geometry) to water­ Note: Sample USGS station data is provided in Appendix C.
shed area in the region. These hydraulic geometry relationships
are often referred to as regional curves. Developing a regional Step 1. Obtain the following information from the stream gage:
curve requires a survey of streams and analysis of gage data a. Location (including location of current meter measurement
for several gage stations within the same hydrophysiographic sites)
region. It is recommended that the gage station have a minimum b. Drainage area (in square miles)
of 10 years of record. Use the gage station survey procedure c. Stage/discharge curve (gage height/discharge rating table)
described here to develop regional curves and establish the Call USGS at (919) 571-4000 or visit the Web site
return period of the flows that shape and maintain the channel. www.usgs.gov.
This information is critical when designing a stream where stream d. Stream discharge notes (9-207 forms) for the previous 10
flow records are not available. Because hydraulic geometry rela­ years or widest range of measured discharge (data for
tionships for streams vary with hydrology, soils and extent of depth, width, velocity and cross-sectional area/discharge)
development within a watershed, it is necessary to develop e. Flood-frequency data (Log-Pearson III) if previously published
curves for various levels of development in each hydrophysio­ (If not, obtain the listing of highest momentary maximum
graphic region. Regional curves for various hydrophysiographic flows for period of record and ranking of flood peaks, high­
regions of North Carolina are provided in Appendix D. est to lowest. Then calculate (m/N+1) x 100, where m =
On gaged streams, determine the bankfull discharge and rank, N = total number of years of record. This calculation
return period by matching the field-determined bankfull indicators gives exceedence probability for a respective flood peak,
to the corresponding USGS stage elevation at the gage (see which allows a determination of return period of the various
Appendix C). Then determine the bankfull discharge that corresponds peak flows.)
to the bankfull stage by using the USGS stage-discharge rating Step 2. Travel to gage site and observe bankfull indicators along
table. Determine the return period by applying a Log-Pearson the stream reach. Measure a longitudinal profile upstream of
Type III distribution to the annual peak discharges recorded for the gage, locating elevations of thalweg, water surface and
the period that the gage has been in operation (USGS, 1982). bankfull stage. Mark bankfull stages along profile with tempo­
Calculate the annual exceedence probability as the inverse of rary flags, then measure this stage at the gage-height staff
the recurrence interval. On log-probability paper, plot the excee­ reference at the stream-gage cross section. Record the gage
dence probabilities as functions of corresponding calculated height (staff plate) reading that corresponds with the bankfull
discharge measurements. Fit a regression line to the data. Then elevation.
determine the bankfull discharge recurrence interval from the Step 3. Read discharge from the stage-discharge rating table for
graph, using the steps in this section. the stream gage corresponding to the gage height of the field-
It is often necessary to supplement data from gaged streams estimated bankfull stage.
with data from non-gaged, stable streams. Stable streams have Step 4. Determine exceedence probability associated with field-
little or no bank erosion, and bankfull stage is located at the top determined bankfull discharge (from Step 1e). To convert excee­
of the streambank. For non-gaged streams, calculate bankfull dence probability (P) to return period in years, inverse P and
discharge using Manning’s equation (Chow, 1959). Determine cross- multiply by 100 (1/P x 100).
sectional area from cross section survey data using the average- Step 5. If the return period of the field-determined bankfull dis­
end area method (see River Course Fact Sheet Number 3, Appendix A.) charge is between one and two years, the bankfull indicators
Estimate a roughness coefficient using Manning’s equation or by are within the range of acceptability for use.
using the d84 particle size of the bankfull channel-bed material Step 6. Plot bankfull discharge versus drainage area for the
with the method described by Rosgen, 1998b. The d84 is appropriate hydrophysiographic province associated with the
defined as the particle size in which 84 percent of the material stream gage. All North Carolina regional curve information is
from the pebble count is finer than this particle. A reachwide located in Appendix D.
pebble count should be used to determine the d84 particle size Step 7. Plot bankfull values of depth, width and cross-sectional
(Section 2.7).
46 Chapter 4 Stream Restoration Stream Restoration Chapter 4 47
 Step 8. Calculate Manning’s roughness coefficient n or other
resistance equations from actual velocity for bankfull stage
and/or other flows.
Step 9. Obtain the following information to classify the stream at
the gage site:
● Description
1. Describe valley type, landform/land type.
2. Photograph upstream/downstream.
3. Delineate watershed using topographic map.
4. Determine drainage area in square miles (usually provided
by USGS).
5. Evaluate watershed land use/land cover and compute per­
centage of watershed that is impervious.
6. Calculate bankfull-discharge return period in years.

● Riffle Cross-Section Dimension

1. Bankfull width (Wbkf)
2. Bankfull mean depth (Dbkf)
3. Bankfull maximum depth (Dmax)
4. Width-to-depth ratio (Wbkf/Dbkf)
5. Bankfull cross-sectional area (Abkf)
Priority Options for Restoring Incised Streams Chapter 5
6. Width of flood-prone area (Wfpa) Priority 1: Establish Bankfull Stage at
7. Entrenchment ratio (Wfpa/Wbkf) the Historical Floodplain Elevation 5.1
8. Bank height (DTOB) Priority 2: Create a New Floodplain and
9. Bank height ratio (DTOB/Dmax) Stream Pattern with the Stream
10. Bankfull velocity (Vbkf) Bed Remaining at the Present Elevation 5.2
11. Bankfull discharge (Qbkf) Priority 3: Widen the Floodplain at the
Existing Bankfull Elevation 5.3
● Plan View (Pattern) Priority 4: Stabilize Existing Streambanks in Place 5.4
Measure sinuosity. (K=stream length/valley length) Priority 1 Case Study:
Yates Mill Pond Tributary 5.5
● Longitudinal-Profile Survey Priority 2 Case Study:
1. Measure average water-surface slope from the head of one Pine Valley Golf Course Tributary 5.6
riffle to the head of a downstream riffle (or from max pool to Priority 3 Case Study: Cove Creek 5.7
max pool) at a distance of at least 20 times the bankfull Priority 4 Examples 5.8
width.
2. Locate bankfull stage along the longitudinal profile.
Note: The elevation difference between bankfull and the water surface at various
locations in the reach should not vary more than 6 inches.

● Materials
1. Particle size of channel material (riffles and pools)
(Reachwide pebble-count frequency distribution): d15, d35,
d50, d84, d95.

48 Chapter 4 Stream Restoration

 Chapter 5: Priority Options for Restoring typically an E or C
stream with bankfull
Incised Streams stage located at the
ground surface of the
Incision of stream channels is caused by straightening of chan­ original floodplain.
nels, loss of riparian buffers, changes in watershed land-use or The increase in
changes in sediment supply. Because incised streams typically streambed elevation
are unstable and function poorly, they are good candidates for also will raise the
stream-restoration projects. Rosgen (1997) presents four priority water table, in many
options for restoring incised channels. This chapter describes cases restoring or
those four options—with the first priority being the most preferred enhancing wetland
and the last being the least optimal. conditions in the
An incised stream has a bank height ratio greater than 1.0 ft/ft, floodplain.
meaning that the bankfull stage is at a lower elevation than the If designed and
Figure 5.1
top of either streambank (Figure 5.1). Severely incised streams with constructed properly,
Cross section of an incised bank height ratios greater than 1.8 ft/ft are usually classified as a Priority 1 project
channel Rosgen stream produces the most
types G or F. Shear long-term stable
stress at high flows stream system. It
in these streams may also be the
may become very least expensive and
high, increasing the simplest to construct
potential for stream- depending on surrounding land-use constraints. Priority 1 proj­ Figure 5.2
bank erosion and/or ects usually can be constructed in dry conditions while stream Cross section of a Priority 1
streambed down- flow continues in its original incised channel. The new channel restoration project
cutting. Moderately can be stabilized with structures and bank vegetation before water
incised streams with is directed into the new stream. A special consideration with Priority
bank height ratios 1 projects is the unbalanced cut/fill requirements. Typically, the
between 1.4 and 1.8 amount of soil excavated in constructing the new channel will be
ft/ft may be classified much less than that required to completely fill the existing incised
as Rosgen stream channel. The designer has the option of bringing additional fill to
types E, C or B, but the site or creating floodplain ponds and/or wetlands to support
they are at increased habitat and recreation.
risk of instability. Surrounding land uses can limit the use of a Priority 1
Slightly incised approach if there are concerns about increased flooding or
streams with bank widening of the stream corridor. Most Priority 1 projects will
height ratios result in higher flood stages above bankfull discharge in the
between 1.1 and 1.3 immediate vicinity of the project and possibly downstream. The
ft/ft are often stable; Priority 1 approach also requires sufficient land area on one or
however, they may become unstable if land use in the watershed both sides of the existing incised stream to construct the new
changes or riparian buffers disappear. meandering channel on the floodplain. It also may be necessary
Designers should consider each restoration option in priority to raise the existing channel at the beginning of the project reach
order before settling on a final design. The options are described and/or lower the new channel at the end of the project reach to
in the following sections and compared in Table 5.1. This chapter connect with the existing channel.
also discusses several recent North Carolina case studies that
illustrate the application of these restoration approaches. 5.2 Priority 2:

Create a New Floodplain and Stream Pattern with the

5.1 Priority 1:
Stream Bed Remaining at the Present Elevation.

Establish Bankfull Stage at the Historical Floodplain

The objective of a Priority 2 project is to create a new, stable
Elevation.
stream and floodplain at the existing channel-bed elevation.
The objective of a Priority 1 project is to replace the incised This is accomplished by excavating a new floodplain and stream
channel with a new, stable stream at a higher elevation. This is channel at the elevation of the existing incised stream (Figure 5.3).
accomplished by excavating a new channel with the appropriate The new channel is designed with the appropriate dimension,
dimension, pattern and profile (based on reference-reach data) to pattern and profile (based on reference-reach data) to fit the
fit the watershed and valley type (Figure 5.2). The new channel is
50 Chapter 5 Stream Restoration Stream Restoration Chapter 5 51
 watershed. The new floodplain. The designer may elect to raise the bed of the stream
channel is typically slightly in an attempt to balance cut and fill. Further, surrounding
an E or C stream land uses can limit the use of a Priority 2 approach if there are con­
with bankfull stage cerns about widening of the stream corridor. This approach requires
located at the eleva­ sufficient land area on one or both sides of the existing incised stream
tion of the newly to construct the new floodplain and meandering channel.
excavated floodplain.
A Priority 2 proj­ 5.3 Priority 3:

ect can produce a Widen the Floodplain at the Existing Bankfull Elevation.

stream system with Priority 3 is similar to Priority 2 in its objective to widen the
long-term stability if floodplain at the existing channel elevation to reduce shear
designed and con­ stress. This is accomplished by excavating a floodplain bench on
structed properly. It one or both sides of the existing stream channel at the elevation
may be more expen­ of the existing bankfull stage (Figure 5.4). The existing channel
sive and complex to may be modified to enhance its dimension and profile based on
construct than a reference-reach data. The resulting channel is typically a B or Bc
Priority 1 project, (low slope) stream with bankfull stage located at the elevation of
depending on valley the newly widened floodplain. Priority 3 projects typically do not
conditions. Priority 2 increase sinuosity to a large extent because of land constraints.
projects usually can A Priority 3 project can produce a stream system with long­
be constructed in dry term stability if it is designed and constructed properly. But it
Figure 5.3 conditions while may require more structural measures and maintenance than
stream flow continues in its original channel or is diverted around Priority 1 or 2 projects. It may be more expensive and complex
Cross section of a Priority 2 the construction site. Typically, water is diverted into the new to construct, depending on valley conditions and structure
restoration project
channel as soon as all or part of it is constructed and stabilized requirements. Priority 3 projects are constructed in wet conditions
with structures and temporary bank-protection measures. unless stream flow is diverted around the construction site.
Because the new floodplain is excavated at a lower elevation, These projects typically have little impact on flooding potential
Priority 2 projects do not increase—and may decrease—the unless there are large changes in channel dimension. Priority 3
potential for flooding. Also, the stream corridor created by the projects typically do not produce large quantities of extra cut
excavated floodplain may enhance riparian wetlands. material or require extensive changes to surrounding land uses.
Unlike Priority 1 projects, which are normally short on material They also do not typically affect riparian wetlands or elevation of
to fill the old channel, Priority 2 projects typically produce a sur­ the water table.
plus of cut material. Designers must consider the expense and In-stream structures are important to the success of Priority 3
logistics of managing extra soil material excavated from the projects. In many projects, a channelized stream must remain in
Figure 5.4 its current location because of surrounding land uses or utilities.
The resulting stream may be classified as a B or Bc channel even
Cross section of a Priority 3
restoration project though the valley conditions support a more meandering E or C
channel. In this case, boulder cross-vane structures should be
used to protect streambanks, provide grade control and support
scour pools for habitat (see Chapter 8).

Section 5.4 Priority 4:

Stabilize Existing Streambanks in Place.

Priority 4 projects use various stabilization techniques to armor

the bank in place. These projects do not attempt to correct problems
with dimension, pattern or profile. Priority 4 projects often use
typical engineering practices to harden (armor) one or more
streambanks. Projects may use riprap, concrete, gabions, bio­
engineering or combinations of structures to protect streambanks.
Both the upstream and downstream impacts of the project
should be carefully evaluated. Because these projects do not
correct dimension, pattern and profile, they are likely to continue
being susceptible to extreme shear stress, which can erode
streambanks in spite of armoring.
52 Chapter 5 Stream Restoration Stream Restoration Chapter 5 53
 A Priority 4 project can stabilize streambanks if designed and
constructed properly, but inspection and maintenance may be
necessary to ensure long-term success. For these reasons, the
long-term cost may be more.

Priority 4 projects are con­

Option Advantages Disadvantages
structed in wet conditions
1 Results in long-term stable stream Increases flooding potential unless stream flow is diverted
Restores optimal habitat values Requires wide stream corridor around the construction site.
Enhances wetlands by raising Unbalanced cut/fill
These projects typically
water table May disturb existing vegetation
Minimal excavation required have no impact on flooding
potential and do not require
changes to surrounding land
2 Results in long-term stable stream Requires wide stream corridor
uses. They also do not typi­
Improves habitat values Requires extensive excavation
Enhances wetlands in stream corridor. May disturb existing vegetation cally affect riparian wetlands
May decrease flooding potential Possible imbalance in cut/fill or elevation of the water table.

3 Results in moderately stable stream May disturb existing vegetation 5.5 Priority 1 Case Study:
Improves habitat values Does not enhance riparian wetlands Yates Mill Pond Tributary
May decrease flooding potential Requires structural stabilization measures
The Yates Mill Pond
Maintains narrow stream corridor May require maintenance
Tributary project is located
Figure 5.5
in a rural watershed in Wake
4 May stabilize streambanks Does not reduce shear stress County just south of Raleigh. Cross-section survey of
Maintains narrow stream corridor May not improve habitat values Yates Mill Pond tributary-
May not disturb existing vegetation May require costly structural measures The existing intermittent
restoration project
May require maintenance stream was incised due to
historic straightening and
Table 5.1 Advantages and disadvantages of restoration options for incised streams removal of riparian vegetation.
The upstream end of the
project reach was not incised, meaning that the new channel
could be connected with the existing channel at its current eleva-
tion. At the downstream end of the first phase of construction in
2000, the existing channel was six feet below the new streambed Figure 5.6
elevation. A temporary boulder-drop-structure connected the Yates Mill Pond tributary-
new and old channels until the second phase of construction restoration project before
was completed in 2002. construction
Table 5.2 lists physical param­
Parameter Existing Design
eters for the existing and new Figure 5.7
Watershed Area (sq mi) 0.12 0.12 stream channels. A cross-sec-
Yates Mill Pond tributary-
Stream Classification E6-G5 C5 tion survey depicting the exist- restoration project after
ing and as-built stream channels construction
Bankfull Cross-Sec Area (sq ft) 8 8
is shown in Figure 5.5. Before
Width/Depth Ratio (ft/ft) 5-12 14 and after photos of the project
Entrenchment Ratio (ft/ft) 0.6-4.0 15 are shown in Figures 5.6 and
Bank Height Ratio (ft/ft) 1.0-3.7 1.0 5.7. The project design called
Length (ft) 750 800 for constructing a new, stable
Sinuosity (ft/ft) 1.1 1.2 C5 stream on the floodplain west
of the existing channel. All of the
Riparian Buffer Width (ft) 5-10 50-80
construction was completed in
Table 5.2. Parameters of Yates Mill Pond tributary-restoration project dry conditions before water was
turned into the new channel.
Because the excavated soil didn’t completely fill the existing
incised channel, several small ponds were created to provide
habitat. To help stabilize the new channel, several log vanes and
log weirs were installed along the streambank in addition to root
wads, transplants and erosion matting.
54 Chapter 5 Stream Restoration Stream Restoration Chapter 5 55
 5.6 Priority 2 Case Study:
Pine Valley Golf Course Tributary
The Pine Valley Golf Course tributary project is located in an
urban watershed in New Hanover County in Wilmington. The
existing perennial stream was incised due to historic ditching and
draining for construction of the golf course and surrounding resi­
dential community. The upstream end of the project reach was a
drainage culvert that prevented
Parameter Existing Design a Priority 1 approach. Project
constraints included a sewer
Watershed Area (sq mi) 0.5 0.5 line along the left streambank,
Figure 5.9
Stream Classification F E two permanent golf-cart bridges,
two irrigation-line crossings and Pine Valley Golf Course
Bankfull Cross-Sec Area (sq ft) 10 10 restoration project before
vegetation concerns at three
Width/Depth Ratio (ft/ft) 15 10 construction
golf holes crossing the stream
Entrenchment Ratio (ft/ft) 1.5 5 reach.
Bank Height Ratio (ft/ft) 2 1 Table 5.3 lists physical
Length (ft) 789 906 parameters for the existing and
Sinuosity (ft/ft) 1.04 1.2 design stream channels. A cross- Figure 5.10
Riparian Buffer Width (ft) 10 50 section survey depicting the
Pine Valley Golf Course
existing and as-built stream restoration project after
Table 5.3 Parameters of Pine Valley Golf Course restoration project channels is shown in Figure construction
5.8. Before and after photos of
the project are shown in figures 5.9 and 5.10. The project design
called for constructing a new, stable E5 stream and floodplain at
the elevation of the existing channel. Stream flow was diverted
through a pump during construction, after which water was
turned into the new channel. Because the excavated soil exceed­
Figure 5.8
ed the amount needed to fill the existing channel, excess soil
was hauled to a stockpile area on the golf course property. To
Cross-section survey of Pine help stabilize the new channel, several log cross-vanes and log
Valley Golf Course restoration
project
weirs were installed along the streambank in addition to root
wads, transplants and erosion mats.

56 Chapter 5 Stream Restoration Stream Restoration Chapter 5 57

 5.7 Priority 3 Case Study:
Cove Creek
The Cove Creek project is located in a rural watershed in
Watauga County, west of Boone. The existing perennial stream
was incised due to a head-cut advancing from a downstream mill
dam that was removed in 1989. The upstream end of the project
reach was a bridge that pre­
vented a Priority 1 approach.
Parameter Existing Design Adjacent landowners were not
able to provide sufficient prop­
Watershed Area (sq mi) 15 15
erty to construct a new mean­
Stream Classification F4 B4c dering stream, which ruled out Figure 5.12
Bankfull Cross-Sec Area (sq ft) 175 164 a Priority 2 approach. The
Width/Depth Ratio (ft/ft) 16 15 resulting project goals were to Cove Creek restoration
project before construction
Entrenchment Ratio (ft/ft) 1.1 1.7 change stream types from F4 to
Bank Height Ratio (ft/ft) 2.0-2.2 1.0 B4c by excavating floodplain
benches and to enhance habi­
Length (ft) 1200 1200
tat using in-stream structures.
Sinuosity (ft/ft) 1.1 1.1 Table 5.4 lists physical
Riparian Buffer Width (ft) 5-10 25-40 parameters for the existing
and design stream channels.
Table 5.4. Parameters of Cove Creek restoration project
A cross-section survey depict­
ing the existing and
as-built stream channels is
shown in Figure 5.11. Before and after photos of the project are
shown in figures 5.12 and 5.13. The project design called for
constructing floodplain benches at the bankfull elevation of the
existing channel and installing boulder cross-vanes. Construction
was completed during low flow. Cross vanes, root wads, trans­
plants and erosion mats were used along the streambank to help
stabilize the channel and floodplain.

Figure 5.11 5.8 Priority 4 Examples

Cross-section survey of Cove Examples of Priority 4 stabilization and armoring projects are
Creek restoration project shown in figures 5.14-5.17.

Figure 5.13
Cove Creek restoration
project at bankfull flow
after construction

58 Chapter 5 Stream Restoration Stream Restoration Chapter 5 59

Figure 5.14
Streambank stabilization
using riprap at the toe of
the bank and bioengi­
neering on the slopes

Figure 5.16
Streambank armoring
using gabion baskets

Figure 5.15
Channel armoring using
riprap at the toe of the
streambank

Figure 5.17
Armoring of streambank
using log-crib wall

60 Chapter 5 Stream Restoration Stream Restoration Chapter 5 61

 Notes:

Reference Reach Survey Chapter 6

Field Procedures
Bankfull Identification 6.1
Longitudinal Profile 6.2
Pool and Riffle Cross-Section Survey 6.3
Pebble Count 6.4
Rosgen Stream Classification 6.5
Plan-Form Measurements 6.6

Office Procedures
Profile Data Summary 6.7
Dimension Data Summary 6.8
Pattern Data Summary 6.9
Reference Reach Summary Table 6.10
Dimensionless Ratio Calculations 6.11
Vegetation Reference Reach 6.12
North Carolina Reference Reach Data 6.13

62 Chapter 5 Stream Restoration

 Chapter 6: Reference Reach Survey 6.5 Rosgen Stream Classification
Classify the stream using the procedures outlined in Section 3.2.
Successful stream restoration requires an understanding of the Figure 6.1
causes of degradation; specific knowledge of the stream’s current 6.6 Plan-Form Measurements Plan-form measurements
state; and an understanding of the stream's most stable dimen­ Measure radius of curvature, Rc, meander wavelength, Lm, and and dimensionless ratios
sion, pattern and profile based on its present valley type and belt width, Wblt, at several meander bends in the reference channel
flow regime. In addition, quantitative knowledge of stable (see Section 2.5 and Figure 6.1). Measure sinuosity, K, for the reference
meander length
streams is necessary to determine the stable dimension, pattern reach (see Section 2.5). Draw a schematic map of the reference
ratio=Lm/Wbkf;
and profile that can be applied in a restoration design. A refer­ reach. Show plan view of stream, bed forms, large woody debris,
cross sections, valley width, plan-form measurement locations, meander width
ence reach is a stable river segment that represents a stable
landmarks, benchmark, etc. ratio=Wblt/Wbkf;
channel within a particular valley morphology (Rosgen, 1998a).
Reference reaches provide the numerical template that can be radius of curvature
ratio=Rc/Wbkf
applied to unstable reaches. Morphology relationships for refer­
ence stream channels are valuable tools for stream-restoration
professionals. Designers and reviewers should use reference
reaches to determine appropriate stream-channel dimension,
pattern and profile for various stream types and watershed
conditions.
The reference stream is not necessarily pristine (completely
unimpaired). It instead is a reach that characterizes a stable
morphology within its setting. Factors that affect reference reach
selection include watershed land-use, valley and stream mor­
phology, and flow regime. Reference reach streams should have
stable watersheds without significant land-use changes within
the past five years; a channel with bankfull stage at the top of
bank and without apparent signs of incision or head-cutting;
stable, well-vegetated, gently sloping streambanks; and well-
defined and properly located bed features. Channels that should
not be used as reference reaches include streams with changing
or recently modified watershed land-use; active streambank
erosion and undercutting; leaning trees with undermined root Office Procedures
systems; channel incision; and poorly functioning or improperly
located channel features (i.e., no pools or riffles located in the 6.7 Profile Data Summary
meander bends). For each restoration design, survey at least one Step 1: Plot the longitudinal profile with the longitudinal station
stream of the appropriate type; same hydrophysiographic region; on the horizontal axis and thalweg, water surface, inner berm,
and similar valley type, watershed type and size, and bed-materi­ bankfull and top of bank on the vertical axis.
al distribution. This will supply morphologic relationships that can Step 2: Calculate the length and slopes for the following bed-
be applied in the design. A body of data from several reference form features: riffles, runs, pools and glides. Length is calculat­
reach channels is preferable. Once a reference reach has been ed using the longitudinal thalweg station from the head of the
identified, follow the field and office procedures described here. feature (i.e., riffle, run, pool or glide) to the head of the next
downstream feature. Slope is then calculated as the length of
Field Procedures the feature divided by the water-surface elevation change over
6.1. Bankfull Identification the thalweg distance for that feature. Pool-to-pool spacing (p­
Follow the procedures outlined in Section 2.3. p) should also be calculated as the distance from max pool to
max pool thalweg stations (see Figure 6.2).
6.2. Longitudinal Profile
Follow the procedures outlined in Section 2.6. 6.8 Dimension Data Summary
Step 1: Plot riffle-and-pool cross sections with the cross-section
6.3. Pool and Riffle Cross-Section Survey stations on the horizontal axis and the elevation on the vertical
Follow the procedures outlined in Section 2.4. axis.
Step 2: Calculate bankfull cross-sectional area for all riffles (Abkf)
6.4 Pebble Count and pools (Apool) using the procedures outlined in River Course
Conduct a reachwide pebble count. Follow the procedures Fact Sheet Number 3 (see Appendix A).
outlined in Section 2.7.
64 Chapter 6 Stream Restoration Stream Restoration Chapter 6 65
 Step 3: Calculate bankfull width, Wbkf, for all riffles (Wbkf) and pools 6.9 Pattern Data Summary
(Wpool) as the horizontal distance between the left and right Consolidate the pattern data—including sinuosity, radius of
bankfull stations (Figure 6.3). curvature, meander wavelength and belt width. Record the minimum,
Step 4: Calculate mean depth, Dbkf=Abkf/Wbkf, for all pools and maximum and mean values for radius of curvature, meander
riffles (Figure 6.3). wavelength and belt width.
Step 5: Calculate max depth, Dmax, for all riffle cross-sections and
Dpool for all pool cross-sections as the vertical distance between 6.10 Reference Reach Summary Table
bankfull elevation and thalweg elevation (Figure 6.3). Summarize the reference reach data and record in the summa­
Step 6: Calculate the bank height ratio, BHR, at all riffle cross ry table (Table 7.1, Chapter 7). Report the maximum, minimum and
sections. Divide the difference in elevation between the top of mean values for each parameter.
Figure 6.2
the low bank and the thalweg by the difference in elevation
Stream profile features and between the bankfull elevation and the thalweg (BHR=DTOB/Dmax). 6.11 Dimensionless Ratio Calculation
dimensionless ratios If bankfull is the top of the bank, then the BHR is 1 (Figure 6.3). Dimensionless ratios are design parameters that are tools for
scaling the data from the reference stream to the design stream,
which may have a different bankfull dimension and discharge.
The measured reference reach data are divided by a bankfull
dimension, Wbkf, Dbkf or Abkf, to create the dimensionless ratios.
Ratios should be calculated for the maximum, minimum and
riffle slope ratio=Srif/Savg;

mean values for each morphologic parameter. After ratios are

run slope ratio=Srun/Savg;
calculated, record them in the summary table (Table 7.1).
pool spacing ratio=pool

spacing/Wbkf;
6.12 Vegetation Reference Reach
pool slope ratio=Spool/Savg;
Riparian and floodplain restoration should be based on a ref­
glide slope ratio=Sglide/Savg
erence area found within close proximity of the project site. This
should be chosen based on the initial riparian assessment of the
project site, if possible. Choose a site that has topographic and
vegetative characteristics similar to the project site. Reference
sites should be as pristine as possible. Ideal areas will not have
been disturbed recently and will be free of exotic vegetation
(see Figure 6.4). If the project site has no native riparian characteristics
(i.e., it is urbanized or farmed), look upstream or downstream of
the project site to determine the stream’s riparian characteristics.
Once the riparian reference site has been chosen, follow the
Figure 6.3 riparian assessment process for describing topography, soil and Figure 6.4
vegetation as discussed in Section 2.10.
Riffle and pool cross- Example of a reference
sections and dimensionless reach for vegetation
ratios 6.13 North Carolina Reference Reach Data
NC State University conducted a
study of reference reach streams (Clinton et
bankfull mean depth
al., 1999) that included detailed morpholog­
(Dbkf)=Abkf/Wbkf;
ic surveys of 14 streams from the Blue­
entrenchment ratio
Ridge/Piedmont physiographic regions
(ER)=Wfpa/Wbkf;
of North Carolina (Table 6.1). The reference
width to depth ratio=Wbkf/Dbkf;
reaches included in the study were
max depth ratio=Dmax/Dbkf;
stable streams with: consistent land use
bank height ratio
over the past 60 years, no channelization,
(BHR)=DTOB/Dmax;
and no severe bank erosion. The bankfull
pool max depth ratio=Dpool/Dbkf;
width of the channels ranged from 8.7 to
pool area ratio=Apool/Abkf;
69 feet. The data from the stream reaches
pool width ratio=Wpool/Wbkf
were analyzed to develop channel pattern
and profile relationships. These relation­
ships are described in Table 6.2 and figures
6.5-6.10. Williams (1986), Leopold and
Wolman (1960) and Rinaldi and Johnson
66 Chapter 6 Stream Restoration Stream Restoration Chapter 6 67
 ty
t /ft) osi
(1997) also developed interrelationships for river meander and (f in u

1.02
1.1

1.1

1.2

1.1

1.1

1.2

1.5

1.2

1.2

1.1

1.1

1.7

1.1
S
channel size. Their data are presented in Table 6.2 for comparison. eac
h

Channel-pattern relationships from stable reference reaches R

are important in designing naturally stable, meandering streams )
/ft t
(ft en
that will replace previously straightened streams. The relationships

18.7

26.3

30.3
tio hm

2.1

1.9

1.4

1.3

1.3

3.1

3.4

3.0

2.8

3.2

6.0
for belt width, radius of curvature and meander wavelength as Ra enc
tr
functions of bankfull channel width are shown in Table 6.2 and En
figures 6.5-6.7. All three data sets indicate high variability with )
the best regression fit occurring for belt width and worst for /ft h
(ft ept

21.8

22.3

30.7

25.1

17.8

19.3

16.3

18.0

16.4

23.8

10.8
7.3

7.1

6.6
io
radius of curvature. The relationship for pool-to-pool spacing as t
Ra h to
D
a function of bankfull channel width is shown in Figure 6.8. id
t
Channel-profile relationships are described in figures 6.9 and
W

6.10. These relationships are important in designing stable

100/380

100/370

135/512

144/512
streams that dissipate energy through changing bed features and

33/370

40/210

75/362

45/145

38/130

45/400

40/110

45/125
9.5/30

8.6/77
)
provide stable aquatic habitat. They also can be used to estimate (m
m
maximum depth of riffles and maximum depth of pools for a 84
/d
given stream-type and watershed condition. Regression relation­
0
d5
ships (figures 6.9 and 6.10) provide a good fit to the measured data for )
both of these parameters. /ft

0.055

0.019

0.016

0.033

0.010

0.009

0.005

0.013

0.014

0.004

0.011

0.013

0.008

0.009
(ft ace
Channel-morphology relationships on reference streams are e f
op ur
Sl er S
valuable tools for engineers, hydrologists and biologists involved at
in stream restoration and protection. They also can help evaluate W
.)
the relative stability of a stream channel. This study created a (ft n
h Mea
good fit for most regression equations, indicating strong correla­ t

1.1

1.8

1.8

1.1

2.0

3.4

3.2

0.9

1.9

2.9

1.2

1.2

2.1

2.0
p l
De kful
tion between morphology relationships in reference stream n
channels in the rural Piedmont of North Carolina. However, Ba
users must consider the natural variability represented by these
h
relationships. The data and relationships from the NC State .) dt

24.6

39.1

54.0

28.8

35.0

64.9

52.3

16.7

30.7

69.0

13.7

21.5
(ft Wi

8.7

8.7
University study can be useful for comparing additional reference kf
ul
l
reach data collected in North Carolina’s Piedmont region. Ba
n
t.) al
However, the availability of this data does not replace the need q. f tion
for a reference reach survey that is specific to each individual (s ec

218.0

168.8

199.0
ea X-S

27.8

68.5

95.1

33.0

68.8

15.5

57.4

10.4

10.6

28.3

42.8
r
restoration project. A ll
fu
a nk
B

i) ea
m r
q eA

24.8

29.0
s

1.1

7.9

9.2

1.8

6.5

6.8

0.2

0.5

4.7

1.9
(

23
g

1
na
rai
D

pe
Ty

B3a

B3c

B3c

B4c
am

C3

C3

C4

C4

C4
B4

E4

E4

E4

E4
re
St

Table 6.1 Reference Reach Survey Data

North Fork of the New River

Mitchell River Headwaters

Lost Cove (at Edgemont)

Lost Cove (Upstream)
Stackrock Creek

Richland Creek

Spencer Creek
Barnes Creek
Steels Creek
Clinton et al., 1999

Sal's Branch
Basin Creek
Craig Creek
STREAM NAME

Big Branch
Mill Creek
68 Chapter 6 Stream Restoration
 Table 6.2 Reference-reach relationships
Clinton et al., 1999

Rinaldi and Johnson (1997)

Leopold and Wolman (1960)

Williams (1986)

North Carolina
Figure 6.5 Figure 6.6

as a function of Channel Width

Belt width as a function of bankfull width Radius of curvature as a function of bankfull width

Meander Wavelength
4.9 ≤ W ≤ 13,000 ft
Applicable Range:

Applicable Range:

Applicable Range:

Applicable Range:
Clinton et al., 1999 Clinton et al., 1999
9.8 ≤ W ≤ 14.8 ft

8.7 ≤ W ≤ 69 ft
Lm=2.86W

Lm=6.5W

Lm=1.8W
Lm=9.7W
n/a

1.12

1.54
1.1
1.13

as a function of Channel Width

Meander Beltwidth
Applicable Range:

Applicable Range:
4.9 ≤ W ≤ 7,000 ft

8.7 ≤ W ≤ 69 ft

Wblt =0.57W
Wblt =4.4W
1.12

1.49

Figure 6.7 Figure 6.8

Meander wavelength as a function of bankfull width Pool-to-pool spacing as a function of bankfull width
Clinton et al., 1999 Clinton et al., 1999
as a function of Channel Width
Radius of Curvature
Applicable Range:

Applicable Range:

Applicable Range:
4.9 ≤ W ≤ 7,000 ft

8.7 ≤ W ≤ 69 ft
Rc=2.5W

Rc=1.3W

Rc=2.4W
n/a

1.02

1.12

0.94

Figure 6.9 Figure 6.10

Max pool depth as a function of riffle mean bankfull depth Max riffle depth as a function of mean bankfull depth
Clinton et al., 1999 Clinton et al., 1999

70 Chapter 6 Stream Restoration Stream Restoration Chapter 6 71

 Notes:

Design Procedures Chapter 7

Design Steps 7.1
Sediment Transport 7.2

72 Chapter 6 Stream Restoration

 Chapter 7: Design Procedures Step 6. Calculate average slope. (Save=Sval/K) based on the sinuosity
hoped to be achieved with the design.
Step 7. Validate whether the design will sufficiently transport its
7.1 Design Steps
sediment load (see Section 7.2). Validation may require adjustment
The design process may begin after completion of the exist-
of the width-to-depth ratio and/or the sinuosity. If width-to-depth
ing-condition survey (Chapter 2); validation of bankfull (Chapter 4);
ratio must be adjusted, return to Step 2.
determination of restoration goals and selection of stream type
Step 8. Calculate mean bankfull velocity, Vbkf=Qbkf/Abkf.
to be built (Chapter 5); and identification and survey of a reference
Step 9. Calculate bankfull max depth at the riffle (Dmax= [Dmax/Dbkf]
reach stream (Chapter 6). The natural channel design process is an
x Dbkf). Obtain the max depth ratio, [Dmax/Dbkf], from the refer­
iterative approach to fitting proper dimension, pattern and profile
ence reach information (Table 7.1).
to the stream based on reference reach data, restoration goals
Step 10. Calculate flood-prone area width (from cross section of
and the existing site constraints. The reference reach data have
stream and valley), Wfpa=width of the valley at an elevation of 2
been converted into dimensionless ratios so that they can be
x Dmax (see Figure 3.4).
applied to the stream even if the watershed area and associated
Step 11. When flooding is a concern or the project is subject to
channel size are different. Reference reach data should be collected
FEMA requirements, compute the flood-stage levels with HEC
from a stream that is stable, in the same hydrophysiographic
2 or HEC-RAS procedures (see Chapter 11). These procedures pro­
region and similar in watershed size (see Chapter 6).
vide only an approximate flood-stage level; they are not intended
Three key steps in the natural channel design process include
as substitutes for the FEMA procedures. At gage stations, it is
determining the new dimension, repatterning the stream and
necessary to plot various return-period floods and their corre­
developing the longitudinal profile. Complete these steps in
sponding depths on the flood-prone area.
order. Afterward, evaluate shear stress and flood studies as a
Step 12. Calculate meander wavelength (Lm=[Lm /Wbkf] x Wbkf).
design check. These checks will ensure that the design causes
Obtain the meander length ratio, [Lm /Wbkf], from the reference
neither erosion, excessive deposition of sediment nor flooding of
reach data (Table 7.1).
nearby homes, businesses or roads.
Step 13. Calculate radius of curvature (Rc=[Rc /Wbkf] x Wbkf).
Obtain the radius-of-curvature ratio, [Rc /Wbkf], from the refer­
Design Steps (Adapted from Rosgen, 1998c)
ence reach information (Table 7.1).
Note: Values in bold typeface within brackets represent the dimensionless ratios
Step 14. Calculate belt width, (Wblt=[Wblt /Wbkf] x Wbkf). [Wblt /Wbkf]
obtained from the reference reach.
is the meander width ratio (MWR) from the reference reach. If
the river is confined, use available belt width for the design
Step 1. Select Abkf and Qbkf based on existing-condition survey,
stream and back-calculate meander width ratio
regional curve, build-out scenarios and reference reach infor­
(MWR=Wblt/Wbkf). Make sure MWR is within the acceptable
mation. If the river is regulated by a storage reservoir or diver­
range for the design stream type.
sion, obtain the operational hydrology of the installation.
Step 15. Sketch or draw the proposed stream alignment (plan
Compare the hydrograph with the field evidence of bankfull
view) over the existing aerial photo or channel map with the
discharge. Using morphological evidence, back-calculate the
appropriate bankfull width; pool width; and appropriate range
stream flow from the cross-sectional area of the bankfull chan­
of values for meander wavelength, radius of curvature and belt
nel. Verify that the estimated bankfull discharge is appropriate
width. Adjust pattern to account for existing vegetation and
for the watershed size. The reservoir or diversion may cause a
landform changes and to avoid high banks such as terraces or
reduction in bankfull discharge.
alluvial fans. Vary the stream alignment to simulate natural vari­
Step 2. Select a width-to-depth ratio for the design, considering
ability, avoiding a symmetrical layout. Measure stream length
the width-to-depth ratio [W/D] of the reference reach. Do not
by delineating a thalweg in the new channel; measure valley
select an extremely low ratio that will result in very steep
length along the fall line of the valley. Calculate sinuosity.
streambanks; these are difficult to build and may erode during
Sinuosity (K)=stream length/valley length.
the early stages of the project when vegetation is immature
Step 16. Calculate average slope (Save= Sval/K).
and rootmass is insubstantial. As a rule of thumb, don’t design
Step 17. If the actual sinuosity and associated average slope are
a width-to-depth ratio of less than 9 unless bank soils are
not equal to the targeted values determined in Steps 5 and 6,
cohesive and consolidated.
validate that the design stream is competent to transport its
Step 3. Calculate proposed bankfull width, Wbkf= √(Abkf) x [W/D].
sediment load (see Section 7.2). This validation may require adjust­
Step 4. Calculate proposed bankfull mean depth, Dbkf=Wbkf/[W/D],
ment of the width-to-depth ratio and/or the sinuosity. If width-
or Abkf/Wbkf.
to-depth ratio must be adjusted, return to Step 2. If sinuosity
Step 5. Select the design stream’s target sinuosity, K, based on
must be adjusted, return to Step 15.
the sinuosity of the project’s reference reach and valley.
Step 18. Calculate riffle slope, (Srif=[Srif / Save] x Save), where [Srif /
Consider such constraints as large trees, utilities, buildings and
Save] is the riffle-slope ratio from the reference reach (Table 7.1).
other infrastructure.

74 Chapter 7 Stream Restoration Stream Restoration Chapter 7 75

 Step 19. Calculate pool slope, (Spool = [Spool / Save] x Save), where Figure 7.2
[Spool / Save] is the pool-slope ratio from the reference reach (Table
Pool cross-section
7.1).
dimension detail
Step 20. Calculate pool area, (Apool = [Apool / Abkf] x Abkf), where [Apool
/ Abkf] is the pool-area ratio from the reference reach (Table 7.1).
Step 21. Calculate max pool depth, (Dpool = [Dpool / Dbkf] x Dbkf),
where [Dpool / Dbkf] is the pool-depth ratio from the reference
reach (Table 7.1).
Step 22. Calculate pool width, (Wpool = [Wpool / Wbkf] x Wbkf), where
[Wpool / Wbkf] is the pool-width ratio from the reference reach (Table
7.1).
Step 23. Calculate pool length, (Lpool = [Lpool / Wbkf] x Wbkf), where
[Lpool / Wbkf] is the pool-length ratio from the reference reach (Table
7.1).
Step 24. Calculate sequence of pool-to-pool spacing, (p-p = [p­
Figure 7.1 p/Wbkf] x Wbkf), where [p-p/Wbkf] is the pool spacing ratio from
Riffle cross-section the reference reach (Table 7.1)) for riffle-pool or step-pool stream
dimension detail types.
Step 25. Plot typical cross sections for riffles, pools, steps, glides cross sections overtop the existing channel cross sections.
or other features. Scale the Step 28. Select specific stabilization devices such as grade-con­
dimensions properly and trol structures, streambank revetment and riparian vegetation.
show point-bar slopes (C Locate these features on the plan, profile and section views.
channels only), entrench­ Step 29. Develop detailed design drawings for such specific
ment ratio and side-slope stabilizing features as cross-vanes for grade control and bank
gradients (figures 7.1 and 7.2). stabilization (see Chapter 8). Develop a plan, profile and section
Step 26. Establish stations view for each stabilization feature. In the design details and speci­
along the thalweg of the fications, show all dimensions and describe the materials and
new stream channel installation procedures (see River Course Fact Sheet Number 4, Appendix A).
alignment. Locate the Step 30. Develop a planting plan for the project reach (see Chapter 9).
position for each riffle, Step 31. Develop a construction sequence and erosion-control
run, pool and glide along plan (see Chapter 10).
the new thalweg, remem­ Figure 7.3
Step 32. If flooding is a concern or the project is in a FEMA-
bering that pools are mapped area, produce hydraulic models to determine changes Existing versus proposed
located on the outside of longitudinal profile
in flooding (see Chapter 11).
the meander bends.
Determine the station for
the head of each
riffle, run, pool and glide
and the max pool. When
establishing the stations for these features, refer to the appro­
priate pool, riffle, run and glide lengths and pool-to-pool spac­
ing established from the reference reach data (Table 7.1). Then
plot the new longitudinal profile for the proposed stream align­
ment, including the thalweg and bankfull elevation using the
feature stations. When constructing the new profile, first set
the bankfull elevation using pool riffle, run and glide slopes
from the reference reach data (Table 7.1). Then set the thalweg
elevation using the maximum depths for riffles, runs, pools and
glides determined from the reference reach data (Table 7.1).
Overlay the new longitudinal profile with the existing profile for
comparison (Figure 7.3).
Step 27. Calculate earthwork (cut-and-fill) volumes from the
cross sections, and use stream length that is appropriate for
the persistence of a particular cross section. Plot proposed
76 Chapter 7 Stream Restoration Stream Restoration Chapter 7 77
 Existing Stream Reference Reach Design Stream Table 7.1 Design, reference and existing-condition information
Parameter MIN MAX MEDIAN MIN MAX MEDIAN MIN MAX MEDIAN Modified from Rosgen, 1998c

Drainage area (sq mi)

Stream type (Rosgen)

Bankfull XSEC area, Abkf (sq ft)

Bankfull width, Wbkf (ft)

Bankfull mean depth, Dbkf (ft)

Width-to-depth ratio, Wbkf/Dbkf

Width flood-prone area, Wfpa (ft)

Entrenchment ratio, Wfpa/Wbkf

Bankfull velocity (vbkf)

Bankfull discharge (Qbkf)

Max depth @ bkf, Dmax (ft)

Max depth ratio, Dmax/Dbkf

Bank height, DTOB (ft)

Bank height ratio (BHR), DTOB/Dmax

Meander length, Lm (ft)

Meander length ratio (BHR) , Lm/Wbkf

Radius of curvature, Rc (ft)

Radius-of-curvature ratio, Rc/Wbkf

Belt width, Wblt (ft)

Meander width ratio, Wblt/Wbkf (ft)

Sinuosity, K

Valley slope, Sval (ft/ft)

Channel slope, Save (ft/ft)

Riffle length, Lrif (ft)

Riffle length ratio, Lrif/Wbkf

Riffle slope, Srif (ft/ft)

Riffle slope ratio, Srif/Save

Pool slope, Spool (ft/ft)

Pool slope ratio, Spool/Save

Run slope, Srun (ft/ft)

Run slope ratio, Srun/Save

Glide slope, Sglide (ft/ft)

Glide slope ratio, Sglide/Save

Pool max depth, Dpool (ft)

Pool max depth ratio, Dpool/Dbkf

Pool area, Apool (sq ft)

Pool area ratio, Apool/Abkf

Pool width, Wpool (ft)

Pool width ratio, Wpool/Wbkf

Pool length, Lpool (ft)

Pool length ratio, Lpool/Wbkf

Pool-pool spacing, p-p (ft)

Pool-pool spacing ratio, p-p/Wbkf

Reach-wide pebble count:

d16 (mm)

d35 (mm)

d50 (mm)

d84 (mm)

d95 (mm)
Stream Restoration Chapter 7 79
 7.2 Sediment Transport Step 4. Once τ∗ci is determined, calculate the minimum bankfull­
A stable stream has the capacity to move its sediment load mean-depth required for entrainment of the largest particle in
without aggrading or degrading. The total load of sediment can the bar sample (or subpavement sample) and the bankfull
be divided into bedload and suspended load. Suspended load is water-surface-slope required for entrainment of the largest
normally composed of fine sands, silts and clay and transported particle using equations 3 and 4, respectively.
in suspension. Bedload moves by rolling, sliding or hopping
(saltating) along the bed. At higher discharges, some portion of Dr=( 1.65 τ∗ci di ) (Equation 3)
se
the bedload can be suspended, especially if it contains sand.
Sr=( 1.65 τ∗ci di ) (Equation 4)
The movement of particles depends on their physical properties— De
notably size, shape and density. Grain size directly influences the
mobility of a given particle. Where: Dr=bankfull mean depth required (ft)
1.65=sediment density (submerged specific weight)=
Gravel Bed Streams (d50 > 2 millimeters):
density of sediment (2.65g/c3)–density of water (1.0g/c3)
Sediment transport in streams with gravel and/or cobble beds τ∗ =critical dimensionless shear stress
ci

is usually analyzed by estimating the shear stress or the compe­ di=largest particle from bar sample (or subpavement
tency of the stream to move a particular-size particle. Critical sample) (ft)
dimensionless shear stress τ∗ci is a measure of the force required Se=existing bankfull water surface slope (ft/ft)
to mobilize and transport a given-size particle resting on the Sr=bankfull water surface slope required (ft/ft)
channel bed. It can be calculated using a bar sample and a wet­ De=existing or design mean bankfull depth (ft)
ted-perimeter cross-section pebble count or the pavement and
subpavement particle sample from a representative riffle in the If the design mean-riffle-depth is significantly larger or smaller
reach (see Section 2.7-2.8 for pebble count, pavement, subpavement and bar sampling than the depth needed to move the largest particle, the width-to­
methods).
depth ratio may need to be adjusted up or down, respectively, to
Step 1: Collect bar samples from several key points along the correct the depth.
stream reach that is being restored and the reference reaches. Step 5. Check the bankfull shear stress at the riffle using Shield’s
Key points include anywhere there are changes in stream type, curve (Figure 7.4) to ensure sediment-transport competence using
bed-material composition or stability. For example, two or more Equation 5 (for wetted perimeter equations and information on
samples may be needed to represent a 1,000-foot reach of calculating hydraulic radius, see Section 2.9). The shear stress
stream. Collect pavement and subpavement samples from any placed on the sediment particles is the force that entrains and
areas of the design channel and reference reaches at which a moves the particles, given by:
bar sample is not possible. Also collect a wetted-perimeter
cross-section substrate analysis (pebble count) for both the τ=γ RS (Equation 5)

design channel and the reference reaches. See the substrate

sampling procedures in sections 2.7 and 2.8 for methods of Where: τ=shear stress (lb/ft2)
collecting bar, pavement, subpavement and wetted-perimeter γ =density of water (62.4 lb/ft3)
pebble counts. R=hydraulic radius of the riffle cross-section at bankfull
Step 2: Calculate the existing and proposed average bankfull stage (ft)

slopes for the design reach from the longitudinal profile. s=average stream slope (ft/ft)

Step 3: Calculate critical dimensionless-shear-stress, τ∗ci

a. Calculate the ratio d50/d5̂0, where d50=median diameter of the If Shield’s curve reveals that the shear stress can move a parti­
riffle bed (from 100 count in the riffle or the pavement sample) cle size that is significantly larger or smaller than the di of the bar
and d5̂0=median diameter of the bar sample (or subpavement or subpavement sample, the sinuosity may need to be increased
sample). If the ratio d50/d5̂0 is between the values of 3.0 and 7.0, or decreased, respectively. Decreasing the sinuosity would
calculate τ∗ci using Equation 1 (Andrews, 1983). increase the average channel slope, thus increasing the shear
stress. Increasing the sinuosity would decrease the average
d50

τ∗ =0.0834 ( dˆ
ci ) – 0.872
(Equation 1)

channel slope, thus decreasing the shear stress. It is important to

50
note that in field studies of rivers in Colorado, Rosgen reported
b. If the ratio d50/d5̂0 is not between the values of 3.0 and 7.0, transport of larger particles than Shield’s tested at the upper
then calculate the ratio of di/ d50, where: di =largest particle from range of shear stress (Rosgen, 2002).
the bar sample (or from the subpavement sample) and d50=medi­
Sand and Silt/Clay Bed Streams
an diameter of the riffle bed (from 100 count in the riffle or the
pavement sample). If the ratio di/ d50 is between the values of 1.3 In the case of sand-bed streams, evaluate sediment-transport
and 3.0, then calculate τ∗ci, using Equation 2 (Andrews, 1983). capacity, including stream power and sediment discharge. This
type of analysis ensures that the stream has the ability to move
di
τ∗ =0.0384 (
ci
d50 )
– 0.887
(Equation 2)
the total sediment load through a cross section. Unit stream
80 Chapter 7 Stream Restoration Stream Restoration Chapter 7 81
 Structures Chapter 8
Root Wads 8.1
Vanes 8.2
Stream Crossings 8.3
Structures and Design Features
Figure 7.4 power and/or a sediment-transport model, such as HEC 6 or for Habitat Enhancements 8.4

Shield’s Diagram
SAM, can be used to model the design channel and compare the
Leopold, 1994, 194 sediment-discharge rates to a section of reference stream,
preferably upstream and downstream of the restoration reach. In
this way, a sediment budget can be created in which the inflow
of sediment is equal to the outflow. In addition, individual stream
sections can be modeled to show localized competency and
capacity. The same procedure can be applied to streams whose
beds are sand/silt. In a stream with a cohesive-clay bed, little
bed load transport would be expected. Clay-bed streams are
typically stable or erode at very low rates; however, bed load
could move through a stream reach. For example, sand and silt
may pass through the stream reach as a result of low cohesion
between sand and clay.

82 Chapter 7 Stream Restoration

 Chapter 8: Structures 8.1. Root Wads
A root wad is the root mass or root ball of a tree, including a
Selecting the methods for stabilizing a streambank is one of portion of the trunk. Root wads armor a streambank by deflect­
the last steps in designing a natural channel. This chapter pro­ ing stream flows away from the bank. They also support the
vides natural channel designers with specifications and sugges­ streambank structurally, provide habitat for fish and other aquatic
tions for installing rock and log structures (some structures and animals and supply food for aquatic insects. A few examples of
methods for stabilizing streambanks are not presented here). root wads are shown in figures 8.1 and 8.2.
River Course Fact Sheet Number 4 (Appendix A) and Appendix E
provide additional information and some design diagrams for Figure 8.1
structures. The designer must complete a thorough morphological Root wad placed on
assessment of the stream reach and watershed before using these outside of meander
techniques. Designers are encouraged to use a variety of tech­ bend
niques, depending on site conditions and the supply of native
materials. Materials native to the region vary, so the materials
chosen also will vary. Boulders are appropriate for streams with
substrate of gravel and larger rocks; log structures are more
appropriate for the low-sloping Coastal Plain sand-bed streams,
where woody debris plays a significant role.
In-stream structures in restoration projects control the grade
and protect the bank. Rock and log structures force the flow of
water away from vulnerable streambanks that lack vegetation or
have high bank-height ratios. Log vanes, root wads and similar
structures add woody debris to the stream, enhancing habitat.
Rosgen (2001c) has published helpful information on placement
considerations, rock sizing, specifications and applications.
Rosgen also has noted that in-stream structures should achieve
the following goals:
ALL SITES:
● Maintain stable width-to-depth ratio
● Maintain enough shear stress to move the large particles
(competence)
● Decrease near-bank velocity, shear stress or stream power
● Maintain channel capacity
● Maintain fish passage at all flows
SITE-SPECIFIC:
● Provide safe passage for or enhance recreational boating
● Improve fish habitat
● Be visibly compatible with natural channels
● Cost less than traditional structures
● Create maintenance-free diversion
● Reduce bridge pier/footer scour
● Reduce road-fill erosion and prevent sediment deposition.

Rosgen also notes that when sizing and choosing placement

for in-stream structures, the project designer must:
● Base the rock size on bankfull shear-stress
● Use footers, in the absence of bedrock, to the depth of scour
● Consider using and locating these structures after completing Figure 8.2
the proper design of the dimension, pattern and profile for the Track hoe with hydraulic
restored channel thumb inserting root
● Ensure stability of structure during high flows (floods) wad into streambank
See Rosgen, 2001c (available for download at http://www.wildlandhydrology.com).

84 Chapter 8 Stream Restoration Stream Restoration Chapter 8 85

 Design Criteria 8.2. Vanes

Ideally, the trunk of the tree above the root wad should have a Vanes come in four types: single vane, J-hook vane, cross
10- to 24-inch basal diameter. Root wads with larger diameters vane and W-weir. Vanes can be constructed from large tree
are more expensive to install and disturb more soil and vegeta­ trunks or boulders, but most are built using boulders. Single and
tion. Regardless of diameter, the trunk length should be 10 to 15 J-hook vanes protect the streambank by redirecting the thalweg
feet. Install root wads where the primary flow vectors intercept away from the streambank and toward the center of the channel.
the bank at acute angles. It generally is not necessary to place They also improve in-stream habitat by creating scour pools and
root wads against each other for the entire length of a meander providing oxygen and cover. Cross vanes serve a similar purpose
bend. Install root wads at the toe of the bank, as low as possible. and also may control the grade in both meandering and step-
Generally, one-third to one-half of the root wad is placed below pool streams.
the base-flow elevation. Where scour depths are high, install
footer logs below the root wads. Where bank heights are low—1 Design Criteria
to 1 1/2 times bankfull height—place boulders at least 1 ton or All four vanes are oriented upstream at 20- to 30-degree
heavier behind the root wad. If banks are high and have plenty of angles off the bank. Single and J-hook vanes are located just
vegetation and root mass, footer logs and boulders may not be downstream of where the stream flow encounters the stream-
needed (Figure 8.1). Boulders and transplants prevent back-eddy bank at acute angles. Vanes should be highest next to the bank,
scour that may be caused by the root wad during high flow. In generally starting at or slightly below bankfull. Rock vanes along
North Carolina, root wads are most successful on the outside of the outside of a meander bend are shown in Figure 8.3. If the
gentle meanders (high ratio of radius-of-curvature to bankfull potential for bank erosion is not too high, start the structures
width) and upstream of streambank vegetation, where they will between bankfull and the inner berm. In either case, slope the
help prevent erosion from any back eddies that occur during structures downward, pointing them upstream. The size of rock
high flow. will depend on the size of the stream, the dominant bed material
and the depth of scour in the channel at high flow. In streams
Installation with substrate of gravel or larger rock, the boulders should be
Root wads are installed by either the drive-point method or generally 1 to 2 tons. Flat rocks are preferable. In a newly creat­
trenching methods. The drive-point method is preferred because ed channel (i.e., Priority 1 restoration), consider using sills on the
it disturbs the least amount of soil and adjacent vegetation and is vane structures. Sills extend into the bank where the highest rock
more cost-effective. The drive-point method uses a track hoe meets the streambank. The purpose of the sill is to prevent water Figure 8.3
with a hydraulic thumb to insert the root wad directly into the from cutting around the boulders next to the bank during high
bank (Figure 8.2). Sharpen the end of the log with a chainsaw before Rock vanes on outside
flow. This is especially important on newly excavated channels
of meander bend imme­
driving it into the bank. A loader or second track hoe may be that may have unconsolidated materials on the banks and little or diately after installation
used to hold the root wad in place while the track hoe with the no vegetation for a while. All structures (diagrams) shown in this (looking upstream)
hydraulic thumb grasps the root fan and drives the trunk into the section include sills.
bank. To prevent destruction of the root fan, don’t ram the track- The length of a
hoe bucket into the root wad excessively (if the streambank is single-vane struc­
resistant to the root wad and trunk, consider the trenching ture may span up to
method or substitute another structure). If vegetation exists on one-half of the
the streambanks, avoid destroying these plants during installa­ base-flow channel
tion. Orient root wads upstream so that the stream flow meets width. The slope of
the root wad at a 90-degree angle, deflecting water away from the structures may
the bank (Figure 8.1). If a back eddy is formed by the structure, range from 2 to 20
place a transplant or boulder on the downstream side of the root wad. percent; the longer
If the root wad cannot be driven into the bank or the bank and flatter the struc­
needs reconstruction, use the trenching method. For this tures, the more
method, excavate a trench for the log portion and install a footer streambank protect­
log underneath the root wad. Place the footer log in a trench ed and habitat
excavated parallel to the bank and well below the streambed. enhanced. The
Place the root wad on top of the footer. Keep at least one-third of rocks in all three
the root wad below normal base-flow conditions. Once the root structures (except
wad is installed, backfill the trench and rebuild the bank with the last two rocks of
transplants or sod mats. Grade the upper bank or terrace scarp a J-hook) must
to a maximum slope of 1.5:1, seed it with a rye grain or other touch each other,
native seed material, and cover it with an erosion-control fabric. and footer rocks
must be placed at
86 Chapter 8 Stream Restoration Stream Restoration Chapter 8 87
 the depth of scour. One to two rocks underneath and down­ This placement sets the elevation of the upstream pool and holds Figure 8.6
stream of the top rock usually will suffice. To prevent the struc­ the elevation of the downstream riffle. A cross vane at the head
J-hook placement in
ture from toppling into a scour hole, place the footer rock down­ of a riffle is typical in small streams that have a short distance meander bend
stream of the top rock. between features. In larger streams, the cross vane is placed in
J-hook vanes are built like single vanes except for the last the glide (Figure 8.9).
two to three rocks. Space these rocks apart about one-half the With cross vanes
Figure 8.4 diameter of the rock to create flow-convergence (figures 8.4 and 8.6). and log structures,
Looking downstream at This flow-convergence creates a large scour hole to dissipate geotextile material is
a J-hook vane energy and provide aquatic habitat. used on the upstream
Cross vanes side of the boulders
provide grade or logs. Footers help
control, keep the prevent movement of
thalweg in the the structure during
center of the channel high flow, but spaces
and protect between the boulders
streambanks from can allow material to
erosion. A cross move through, creat­
vane has three ing a "hole" in the
components: two cross vane. Even if
rock vanes and the rocks are touch­
one center structure ing, these holes still
placed perpendicu­ may appear (Figure 8.10).
lar to the flow. The If the hole is large
center structure enough, the majority,
sets the grade of if not all, of the flow
the streambed. at base-flow level
Installed cross may move through it. If the chan­
vanes are shown nel has a variety of substrate
sizes (small gravel to cobble),
back-filling on the upstream side
of the structure may close these
gaps. But if the material is too
Figure 8.5 uniform or the gaps too large,
Rock vane and J-hook the structure may eventually
vane (looking upstream) be compromised. To prevent this,
place geotextile fabric on the
upstream side of the structure
during construction and bury it
to the depth of the footers (this
is strongly recommended for
structures that provide critical
grade-control on a project). The
Figure 8.7
fabric will help to prevent water from piping between or under­
neath the rocks or logs. Once the backfill material is placed Cross-vane structure with
upstream of this, no material should move through at all. Figure woody debris for habitat
enhancement
8.11 shows fabric being used on a log vane. To ensure stability
of important grade-control structures, such as in a step-pool
system, minimize the drop in elevation for each structure. The
larger the difference in elevation from immediately upstream to
downstream of the structure, the more stress is placed on the
structure itself. In this case, grade control may fail and jeopardize
in figures 8.7 and 8.8. Since a cross vane raises or holds the bed the entire project.
elevation, it is often placed within the glide or at the head of The W-weir structure is very similar to the cross vane, in that
the riffle. it maintains the grade of the streambed and provides excellent
aquatic habitat. W-weirs can be used only on large rivers
88 Chapter 8 Stream Restoration Stream Restoration Chapter 8 89
 because they span a signifi­
cant distance across the
channel. Their design is
described as a W formation
in the downstream direction.
From the plan-view perspec­
tive, the weir is similar to
two cross-vanes joined in
the center of the channel.
Figure 8.12 shows a
schematic of the W-weir.
Due to the double cross-
vane effect of the W-weir,
two thalwegs are created.
This design helps to
enhance fish habitat.
The W-weir also can
be designed to maintain
recreational boating, stabilize streambanks, facilitate irrigation
diversions, reduce scour of a bridge’s center pier and foundation, Figure 8.11
Figure 8.8
and increase sediment transport at bridge crossings. Two W-weirs Use of geotextile fabric on the
Cross vane showing place­ may be constructed together on very wide rivers and/or where upstream side of log structure
ment and measurements
* The lowest slope is most
two bridge center piers (three cells) require protection (Rosgen, 2001c).
desirable, but in small streams
a narrow channel may neces­ 8.3 Stream Crossings
sitate higher slopes (10 to 20
percent). Design road crossings to minimize negative impacts on stream Figure 8.12
stability, sediment transport, aquatic habitat and fish passage
Plan, cross section and profile
while meeting prescribed hydraulic and structural criteria. The views of the W-weir
ultimate goal is to construct a stable stream system that neither Rosgen 2001c

Figure 8.9
Placement of cross-vane
structure in a meandering
stream

Figure 8.10
Hole formed on
cross vane due to
gap in structure

90 Chapter 8 Stream Restoration Stream Restoration Chapter 8 91

 scours nor aggrades. This approach includes maintaining the stream reaches. Designs often are patterned after reference-
consistency of dimension, pattern and profile of the stream with reach streams and focus on reducing bank erosion and providing
particular attention to maintaining bankfull width and width-to­ effective sediment transport. Restoration and enhancement
depth ratio. Where feasible, use bridges or arch culverts to mini­ projects generally also address the restoration of the riparian
mize floodplain restrictions. For culvert systems, use floodplain buffer. However, the restoration of in-stream habitat has not been
culverts, where appropriate, to relieve the hydraulic load on the addressed as thoroughly as channel stability and riparian vegetation.
main-channel culvert; this will limit downstream scour and ero­ Many benthic organisms prefer one type of microhabitat,
sion (see Figure 8.13). depending on season. For example, certain species of caddisflies
are typically found only in riffles, which are the most productive
Figure 8.13 habitat for many benthic (bottom-dwelling) organisms. Successful
Floodplain culverts on restoration projects should therefore provide proper riffle/pool
Rocky Branch, NC sequences to ensure recolonization. Different fish species require
State University different habitat types. Good in-stream habitat is structurally
complex and is composed of both inorganic (i.e., boulder,
cobble, fine sediment particles) and organic components. Pools
and riffles of varying sizes and placements are important too.
Additional important habitat features that can be included in a
stream restoration design are listed below.
Overhanging woody vegetation provides food, shade and
cover for aquatic organisms. Installing transplants and live stakes
of alder, silky dogwoods and willows around rootwads will help
to establish overhanging vegetation quickly.
Erratic rocks with ledges and shelves provide cover and habi­
tat. Use that "odd" rock that won’t fit into a structure as part of a
boulder cluster. Stacking rocks can be used to create a "cubbyhole"
feature.
Boulder clusters create multiple points of flow-convergence
and eddies. Upwelling from subsurface flows around boulders
pulls material into the water column. Fish can hold behind the
clusters in eddies and feed in the upwelling. Currents also
cleanse the substrate and provide better spawning habitat.
Specific design recommendations are: Boulder clusters and other structures (such as large woody
1. Maintain the natural stream-gradient and meander-pattern. debris) can catch and hold limbs and debris that will snag leaf-
Avoid overly steep or perched culverts that will block fish packs. Leaf-packs accumulate in streams and provide habitat
passage. and food for a number of benthic insects. Therefore, adding large
2. Cross the stream at a perpendicular angle. woody debris can enhance the habitat of boulder clusters.
3. Size the main culvert to match the natural channel bankfull Large woody debris placed in pools or lodged under boulders,
width. Provide for the unobstructed flow of the bankfull combined with other structures, can provide "snag" habitat for
storm-event in the main culvert without changing velocity. fish and will help trap leaf-packs, which are important to produc­
4. Design the culvert openings to maintain base flow at its nor­ tivity. Logjams between vane structures can be incorporated to
mal width, depth and velocity. This may require low weirs or improve pool habitat. Logs should not be incorporated into a
multiple openings to carry base flow and avoid sediment vane structure because it may create a gap in the structure that
buildup in the system. could cause a failure. Large woody debris can be placed in the
5. Use bankfull culvert openings on the floodplain to carry flows floodplain and will later be available to the stream during high
exceeding bankfull discharge. flows. However, too much large woody debris in the floodplain
6. Where appropriate, use boulder cross-vanes upstream and could cause a downstream debris jam and extensive bank erosion.
downstream of the culvert to maintain desired flow direction Deep pools provide great cover and holding areas (places with
and grade, improve sediment transport through culverts, and little or no current) for fish. Large woody debris anchored in the
improve habitat. pool also will provide snag habitat. Designers are often reluctant
to dig the thalweg at the outermost edge of the meander bend
8.4 Structures and Design Features for Habitat for fear that the bank might collapse. However, installation of root
Enhancement wads, live stakes or fabric anchors in the meander bend should
Stream restoration work historically has concentrated on prevent instability.
redesigning the dimension, pattern and profile of impacted
92 Chapter 8 Stream Restoration Stream Restoration Chapter 8 93
 Floodplain pools provide excellent habitat not only for
amphibians but also for certain species of insects such as drag­
onflies and damselflies. To ensure that these pools continue to
provide good amphibian habitat, it is important to design and
build them so that they dry out every two to three years. This
prevents a large population of predators (i.e., fish) from becoming
established in the pool. Amphibian organisms are adapted to
periods of drought—adults can burrow in the substrate for pro­
tection, and many eggs and small larvae also can survive.
Coarse substrate harvested from the existing stream channel
can be reintroduced into newly constructed riffles to speed habitat
development. Substrate harvesting can be particularly beneficial in
Priority 1 stream-restoration projects that involve constructing a
new channel and abandoning the existing stream channel.
Other microhabitats, which presumably will develop over time,
often are not specifically considered as part of a restoration project.
Although these habitat components are hard to construct, project
monitors should note their development or lack thereof. Some
examples of these complementary habitats are described as follows:
Fine particulate organic material: Over time, fine particulate
organic matter collects in the interstitial spaces between the
dominant substrate material. This material is food for many
benthic organisms. All collector-gatherer organisms will feed on
this type of organic material at some point in their life cycle. The
increase in habitat heterogeneity should also improve the
Vegetation Stabilization and Riparian-Buffer Re-establishment Chapter 9
streambed/hyporheic zone connection and movement of animals Salvaging On-Site Vegetation 9.1
between zones under different flow conditions. Live Staking 9.2
Aquatic plants: Very little consideration has been given to how Bare-Root Plantings 9.3
important aquatic plants (including macrophytes and attached Container Plant Material 9.4
algae) are to the benthic fauna of restoration reaches. Many ben­ Permanent Seeding 9.5
thic insects are collected only in this type of habitat. Caddisflies
(Micrasema, Brachycentrus), mayflies (Ephemerellidae) and
chironomids are commonly collected in aquatic macrophytes
(specifically Podostemum, commonly known as river weed, in
North Carolina). Living plants provide structural habitat. When
they die, they are colonized by bacteria and fungi, becoming
food for aquatic macroinvertebrates.
Fine streambank root material: Rootwads provide habitat for
fish and stabilize eroding streambanks. However, most of them
do not mimic streambank plants, which usually extend fine roots
into the current along the outside bends in a stable stream. Many
leptocerid caddisflies (Triaenodes, Oecetis) and odonata (dragon­
flies and damselflies) are found primarily in this habitat.

94 Chapter 8 Stream Restoration

 Chapter 9: Vegetation Stabilization and
Riparian-Buffer Re-establishment
A combination of planting methods improves the chances for
successfully fulfilling the restoration objectives of bank stabilization,
flood attenuation and habitat enhancement. Appendix F lists
appropriate species from the three physiographic regions of North
Carolina to incorporate into restoration plans.

9.1. Salvaging On-Site Vegetation

Potential transplants may include small trees up to 3 inches in
diameter. Sycamores are an easily salvaged species. Prune these
trees to about 6 feet and scoop the entire root mass with the
bucket of a track hoe. Keep the root balls and surrounding soil
intact. Don’t rip limbs or bark from the transplants. Such native
shrubs as alder (Alnus spp.), elderberry (Sambucus canadensis)
and spicebush (Lindera benzoin) also are good transplants. Prune
shrubs to 3 or 4 feet and harvest like the trees. Herbaceous
plants can be salvaged as well. Rushes (Juncus spp.), sedges
(Carex spp.) and other tender plants can be harvested and placed 9.2. Live Staking Figure 9.2
at the toe slope along the water’s edge, where woody vegetation As with transplants, it may be possible to harvest stake material Successfully planted
is not appropriate. from the site. Stakes are branches or small limbs cut from a larger black-willow stake
If salvaged vegetation cannot be installed immediately, stockpile tree or shrub. If material is not available on-site, check with sur­
it in a relatively moist area or keep it continually moist. This is rounding landowners or nurseries. Silky dogwood (Cornus
especially important during summer. amomum) and willow (Salix spp.) are good candidates for staking.
Place woody transplants at bankfull elevation or above. If soil is Some species of shrubs and trees can be propagated from cuttings
compacted in the planting area, loosen it to a depth of at least 1 and root stems, although this technique is labor-intensive. Stakes
foot. Plant transplants the same depth at which they were origi­ should range from one-half inch to 2 inches in diameter with an
Figure 9.1 nally growing. Replace soil around the transplants and tamp it average length of 3 feet. Cut stakes with an angle on the bottom
down to eliminate air pockets. Spacing will depend on availability and flush on tops, with buds oriented upward. Trim all side
Salvaging vegetation at
construction site of material. If transplants are limited, start in critical areas, such branches cleanly so the cutting is one stem. Keep stakes cool
as along meander bends or near in-stream structures. and moist to keep them alive and dormant. Plant stakes in late
fall to early spring while they are dormant. Install stakes in areas
where erosive forces are greatest, such as along meander bends
and behind in-stream structures. Stakes usually are installed 2
to 4 feet apart using triangular spacing along the streambanks.
Different sites may require slightly different spacing. Drive stakes
into the ground with a rubber hammer, or make a hole using a
metal bar and slip the stake into it. Tamp each stake in at a right
angle to the slope, keeping one-half to four-fifths of the stake
below the ground surface. At least two buds (lateral and/or terminal)
should remain above the ground surface. Pack the soil firmly
around the hole afterward. Do not use split stakes.

9.3. Bare-Root Plantings

Bare-root material is recommended on large restoration sites
requiring many trees. Bare-root plantings are more economical
than container plants, although survival rates may be lower.
Choose plants from local nurseries or growers that offer plants
suited to the site. Refer to Appendix F for a list of appropriate
species to plant in North Carolina.
Late fall to early spring is the best time for planting. Early fall
planting allows more time for root establishment. If bare-root

96 Chapter 9 Stream Restoration Stream Restoration Chapter 9 97

Figure 9.3 plants can’t be installed right away, heel them into moist soil or
Demonstration of bare-
sawdust, according to general horticultural practice. Use wet
root seedling installation canvas, burlap, straw or other suitable material at all times to
with dibble bar prevent drying. The method selected should be appropriate to
the weather conditions and the length of time the roots will
remain out of the ground.
Loosen soil in the planting area to a depth
of at least 5 inches. Make planting holes with
a mattock, dibble, planting bar, shovel or
other appropriate tool. Plant rootstock in a
vertical position with the root collar about
one-half inch below the soil surface. Make
sure the planting trench or hole is deep and
wide enough to permit the roots to spread
out and down. Keep the plant stem upright.
Replace soil and tamp firmly around each
transplant to eliminate air pockets. See
Appendix F for an installation diagram.
Spacing guidelines for rooted shrubs and
trees are provided in Table 9.1.

Figure 9.4
9.5. Permanent Seeding
Installing potted plants
For maximum habitat diversity and ground cover, include
Type Spacing # Per 1,000 sq ft seeds among the planted material. Permanent (perennial) seeding
Shrubs mixtures are available from nurseries and can vary widely. A site-
(<10 ft) 3-6 ft 25-110 specific combination of herbaceous species and grasses based
on surrounding native flora is recommended. Site conditions and
Shrubs and trees project requirements will determine the vegetation needs and
(10-25 ft) 6-8 ft 15-25 installation methods. Appendix F lists appropriate herbaceous
Trees species for North Carolina. Follow nursery recommendations
(>25 ft) 8-15 ft 4-15 for appropriate planting times and methods. Before planting the
permanent seed mix, see the site-preparation and soil-amend­
ment procedures in Section 10.2.
Table 9.1. Spacing guidelines for shrubs and trees

Figure 9.5
9.4. Container Plant Material Example of permanent
Some projects may require container, or potted, plants. These seed mix
come in many different sizes and shapes. Check with local nurseries
and growers for availability. When installing potted plants, dig a
hole that is twice the diameter of the pot. Remove the plant from
the container and tease roots apart if the plant is root-bound.
Place plant in hole, making sure the root collar is even with the
ground surface and the stem is upright. Back-fill with potting soil
or fill from the hole. Make sure the fill is free of clods and stones,
loose and evenly distributed around the plant. Tamp firmly
around the plant to eliminate air pockets. Add mulch to retain
moisture. Refer to Section 9.3 for installation techniques and
spacing requirements. Appendix F lists appropriate species for
North Carolina.

98 Chapter 9 Stream Restoration Stream Restoration Chapter 9 99

 Notes:

Erosion and Sediment Control Plan Chapter 10

Pollution Control:
Construction Sequence and Structures 10.1
Pollution Control:
Seeding 10.2

100 Chapter 9 Stream Restoration

 Chapter 10: Erosion and Sediment Control erosion. All pollution-control measures and works must be kept
functional as long as needed during the construction operation.
Plan Remove all temporary measures and restore the site as closely
as possible to original conditions (see the N.C. Sedimentation and Erosion-
10.1. Pollution Control: Construction Sequence and Control Manual).
Structures Diversions – Diversion structures divert water away and collect
All restoration work should comply with the requirements of runoff from work areas for treatment by sediment traps, such as
the North Carolina Sedimentation Pollution Control Act and the check dams. If possible, diversions should be constructed along
federal Clean Water Act. During construction, measures must be a contour so that they have a near-flat slope. Diversions should
taken to control erosion and minimize the production of sediment be seeded and lined with erosion-control fabric, if necessary, or
and other pollutants of water and air. otherwise stabilized so they do not erode.
Stream Crossings – Equipment should cross streams at fords
Construction Sequence or temporary culverts. To construct a ford, grade a ramp into the
The construction sequence is a critical component of the stream channel on both banks. These ramps should be 5:1 or
erosion and sediment control plan for a stream-restoration project. flatter and lined with stone. Install filter fabric combined with
First, it is important to divide the stream into segments or reaches stone in the bed of the stream. Any temporary culvert should be
for construction. Each segment can be completed and stabilized sized to carry at least the bankfull discharge. Place stone on the
before moving on to the next. This will minimize the exposed soil upstream and downstream sides of the culvert to prevent erosion
that is vulnerable to erosion at any given time during the project. of the streambanks, and fill soil around the culverts. Also, place
Schedule the excavation and moving of soil materials so areas stone on top of the fill on which heavy equipment will be driven.
will be unprotected from erosion for the shortest time feasible. Sediment Filters – Geotextile sediment fences will trap sedi­
Stockpile any soil excavated from the new channel in locations ment from areas with limited runoff (never use them in areas of
shown on construction plans/drawings. Install silt fences around concentrated flow). Install these fences on the contour along the
all stockpiles. entire downstream perimeter of the area being disturbed. To
Three basic approaches can be used to address potential effectively trap sediment, these filters should be trenched into the
sediment and erosion associated with stream restoration: 1) ground and properly anchored. Make sure support stakes are
construct the new channel in the dry (absent of stream flow), 2) properly spaced; if heavy-duty filter fabric is not being used,
pump or divert the water around each project stream reach, or 3) install wire support behind the filter.
work in the active channel. Constructing a new channel in the dry Waterways – Waterways can be used for the safe disposal of Figure 10.1
is preferred, and it is often possible in many Priority 1 and some runoff from fields, diversions and other structures. Stabilize
Priority 2 (see Chapter 5) stream-restoration projects. Because water waterways with grass, erosion-control fabric or stone, depending Properly installed
erosion-control blanket
continues to flow in the old channel, this approach allows the on the slope of the waterway. Make sure the outlet for the water-
new channel to be built and stabilized on dry ground before it is
exposed to stream flow. Pumping or diverting the water around
the active construction project is feasible in small watersheds
with low to moderate base flow; it generally is not feasible in
streams with large base flow. Even in smaller streams, pumping
usually cannot be maintained during storm flows, so precautions
must be taken to minimize exposed soil and associated erosion.
The least-preferred option is working in the active stream channel,
though it is necessary in many cases. When working in the active
channel, it is important to start and finish each element of the
project in a single day. For example, if construction of a boulder
cross-vane begins in the morning, the vane should be completed
and erosion-control matting installed on disturbed streambanks
the same day. Sediment-control measures should be taken
below the construction project to prevent sediment from traveling
downstream. Such measures might include check dams and
various sediment-trapping fabrics as described in the North
Carolina Erosion and Sediment Control Planning and Design
Manual (available from the North Carolina Division of Land Resources,
http://www.dlr.enr.state.nc.us/eropubs.html).
Following are examples of temporary measures commonly
used in stream-restoration projects to reduce sedimentation and
102 Chapter 10 Stream Restoration Stream Restoration Chapter 10 103
 way is stable and equipped with stone or other material that will Choose an annual seed that will not outcompete native vegeta­
dissipate the energy of water being discharged. tion. Apply the following vegetation at the listed rates.
Coconut/Straw-Fiber Blanket – Coconut/straw-fiber blankets
should be used only on streambanks with little or no established Fall, Winter, Spring Seeding:
vegetation at the time the stream flow is directed into the newly Rye grain/winter wheat mix, winter wheat or barley
constructed channel. Project specifications will determine the 3 lbs per 1,000 sq ft or 130 lbs per acre.
type of erosion-control blanket to use. For wildlife and habitat
purposes, it is best to use completely biodegradable blankets. Summer Seeding:
Blankets with plastic components often trap animals. Lay the Browntop millet, Sudan grass
coconut/straw-fiber blanket when grading is complete. Provide a 1 lb per 1,000 sq ft or 45 lbs per acre
smooth soil surface free of stones, clods or other debris that will
prevent the contact of the blanket with the soil. Apply fertilizer, 4. Mulching – Mulch temporarily protects soil from erosion.
seed and lime prior to installing blankets. Follow manufacturer’s Apply mulch within 48 hours of seeding. Apply straw mulch on
guidelines for installation. The engineer/ project manager may seeded areas at a rate of 3 bales per 1,000 sq ft (130 bales per
need to adjust the trenching or stapling requirements to fit indi­ acre). Apply mulch uniformly. Anchor with biodegradable netting.
vidual site conditions.
Other – Additional erosion-control measures may be required
by the federal, state or local government agency that is responsible
for reviewing and inspecting the site’s erosion-control plan.

10.2. Pollution Control: Seeding

Seed any disturbed areas, including streambanks, access
areas and stockpile locations. Immediately after construction
activities are completed, plant seeds of both permanent and
temporary vegetation. This work includes preparing the area;
furnishing and placing the seed, mulch, fertilizer and soil amend­
ments; and anchoring mulch.
1. Seedbed Preparation – On sites where equipment can be
operated safely, loosen the seed bed mechanically. Compacted
soil may require disking. Steep banks may require roughening,
either by hand-scarifying or equipment. The engineer/ project
manager should determine the condition needs on-site. If seeding
is done immediately after construction, seedbed preparation may
not be necessary. Exceptions would be in compacted, polished or
freshly cut areas.
2. Fertilizing/Liming – In disturbed areas, fertilizer and lime
will help seeds establish more quickly. If possible, test the soil’s
fertility. The N.C. Department of Agriculture tests soil samples at
no charge. These tests help determine proper distribution rates
for fertilizer and lime in the sampled area. See Appendix F for the
department’s Soil Sample Information Sheet and contact information.
Distribute fertilizer and lime evenly over the area to be seeded.
Mix the fertilizer and lime uniformly into the top 3 inches of soil;
if the bed is gravelly or cobbled, incorporation is not necessary.
Fertilizer and lime should be applied at the following rates:

10-10-10 Fertilizer: 10 lbs per 1,000 sq ft or 435 lbs per acre

Lime: 50 lbs per 1,000 sq ft or 2,200 lbs per acre

3. Temporary Seeding – Temporary seeding is useful for erosion-

control when permanent vegetation cannot be established due to
planting season and where temporary ground cover is needed to
allow time for native or woody vegetation to become established.

104 Chapter 10 Stream Restoration Stream Restoration Chapter 10 105

 Notes:

Flood Studies Chapter 11

Introduction 11.1
FEMA Maps and Nomenclature 11.2
FEMA Requirements and Flood Modeling 11.3
Case Study 11.4

106 Chapter 10 Stream Restoration

 Chapter 11: Flood Studies When a new project is proposed, the designer must consider
two potential impacts. The first potential impact is the change in
flood levels. During flood flows, flooding may occur in areas that
11.1. Introduction
have never flooded. The second impact is any change in the
The regulations administered by the Federal Emergency Figure 11.2
floodway. A floodway is the area around a stream in which devel­
Management Agency (FEMA) are applicable to projects in areas
opment is prohibited (Figure 11.2). Increasing the floodway could Illustration of floodway
that have been mapped by FEMA. FEMA’s National Flood
have significant impacts on the insurance rates for any affected and floodway fringe
Insurance Program (NFIP) promotes sound land-use practices U.S. Army Corps of
property owner(s) and decrease the amount of developable property.
within the floodplain. The NFIP limits the impact of flooding by Engineers, HEC-RAS User
Contact the local FEMA administrator early in the process to Manual documentation
restricting development, buying property and using flood-control
ensure that the project is acceptable to the community.
structures in the floodplain. One of the most important functions
of the NFIP is establishing flood insurance rates through the use
11.2. FEMA Maps
of risk data.
and Nomenclature
Section 60.3(d)(3) of the National Flood Insurance Program
When a project will
(NFIP) regulations states that a community shall "prohibit
change the existing
encroachments, including fill, new construction, substantial
floodway and 100-year
improvements, and other development within the adopted regulatory
flood elevations, an
floodway unless it has been demonstrated through hydrologic and
application must be
hydraulic analyses performed in accordance with standard engineering
Figure 11.1 submitted to FEMA
practice that the proposed encroachment would not result in any
containing the model­
Flooding from Hurricane increase in flood levels within the community during the occurrence
Floyd
ing results from the
of the base (100-year) flood discharge."
Photograph courtesy of J. proposed project and
Jordan, U.S. Army Corps of the proposed map revi­
Engineers
sions. If approved,
FEMA will issue a "con­
ditional letter of map
revision" (also known
as a CLOMR). Once the
project is completed,
new cross sections
must be generated from the "as-built" survey
information. The new cross sections are then Figure 11.3
used to develop a new hydraulic model. The
Sample FEMA map
new maps and modeling results generated from
the as-built information are then submitted to
FEMA. Once these are approved by FEMA, a
"letter of map revision" (LOMR) is issued. One of
the most important aspects of a map revision is
the change in the floodway (Figure 11.2).
FEMA offers NFIP flood maps that show the
extent of flooding for a 100-year flood (Figure 11.3).
Reports called "Flood Insurance Studies," which
contain data and results, accompany the flood maps. Flood
maps provide the community name, community number (six-digit
number) and effective date in the lower right corner of the map.
This information is required for ordering a hydraulic model and
completing the MT-2 form for a CLOMR.
A FEMA map (or plate as it is sometimes called) depicts sever­
al zones that indicate flood boundaries and shows whether they
were derived from modeling or from approximate methods.
Zones A and AE indicate the 100-year floodplain using approximate
methods and modeling, respectively. Zones V and VE indicate
the 100-year floodplain plus hazards from storm waves (for
coastal projects) using approximate methods and modeling,
108 Chapter 11 Stream Restoration Stream Restoration Chapter 11 109
 if there are structures in the floodplain. In a Priority 1 project,
the channel is raised and reconnected to the floodplain, which
results in increased water-surface elevations and potential dam­
age to existing structures. A Priority 2 or Priority 3 restoration
may be more appropriate if structures are located in the flood­
plain. A project is not allowed to cause an increase in predicted
flood elevations for existing structures.
To determine if the project site is in a mapped area, contact
the local city/county planning office. The local planning office
may have the Flood Insurance Rate Map (FIRM) needed for the
project site; otherwise it can be ordered from FEMA. FEMA does
not map drainage areas less than 1 square mile. Internet
resources for FEMA maps and information are:
http://www.fema.gov/maps/ (maps)
http://www.fema.gov/about/regoff.htm (regional and state offices)
http://msc.fema.gov/MSC/toc.htm (index for map data and user guides)

Many stream-restoration projects may increase water-surface

Figure 11.4
respectively. FEMA Zone X has multiple meanings, including but elevations at low to moderate discharges but have little or no
Sample excerpt from
not limited to: outside the 500-year floodplain (not regulated); impact on flood flows. This is because at 100-year flows, there
FEMA map usually isn’t much difference between pre- and post-restoration
within the 500-year floodplain (not regulated); within the 100-year
floodplain with flood depth less than 1 foot; and areas protected water levels. However, water-surface elevation (and therefore
from the 100-year flood by dikes. Zone D indicates areas where flood extent) is difficult to predict because it varies depending on
flood studies haven’t been conducted but are possible. The the geometry, roughness and vegetation in the channel or flood­
floodway is indicated by the crosshatched areas over the stream plain and conditions and structures downstream. This is why
channel (Figure 11.4). flood studies incorporate a hydraulic model.
If a no-impact certification is possible, it will almost always
11.3. FEMA Requirements and Flood Modeling require hydraulic modeling. Hydraulic modeling must be under­
If the project is not in a FEMA-mapped area, no federal taken if a map revision is needed. A hydraulic model will gener­
requirements apply. The only requirements may be those of local ate data to show the impact the project will have on the new
authorities (city, town or county). It is the designer’s responsibility floodway and floodplain. The hydraulic model currently used to
to comply with any local regulations. If the area floods after a calculate flood elevations is the U.S. Army Corps of Engineers
project is built, the designer may be required to prove the flooding Hydrologic Engineering Center’s HEC-2 hydraulic model. A
was not a result of the new project. Windows-interfaced version of this model called HEC-RAS (River
If the project is in a FEMA-mapped area, there are two options: Analysis System) also is available.
(1) submit a no-impact certification or (2) submit the necessary
application for a map revision. A no-impact certification is commonly The basic modeling steps include:
granted for sewer-line installation. Sewer-line installation takes Step 1. Obtain the FEMA map for the project area
(http://www.fema.gov/maps/). From the FEMA map, obtain the community
place inside the floodway, although the impact of the new sewer
line is usually negligible on flood elevations. A no-impact certifi­ name, community number (six-digit number) and effective date
cation for a proposed stream-restoration project is unlikely to be (all in the lower right corner of the front cover of the map).
approved and would have to be handled on a case-by-case Step 2. Obtain the HEC-2 model. Call 1-(877) 336-2627 or
basis by the local FEMA administrator. FEMA Region IV’s procedures for 1-(877) FEMA MAP.
no-rise certification for proposed developments in regulatory floodways are available for
Step 3. Obtain the forms needed to document modeling.
download at http://www.msema.org/forms/nfip/No-Rise_certification.pdf.
For the CLOMR, visit the Web site
A map revision is needed if a no-impact certification is not http://www.fema.gov/mit/tsd/ dl_mt-2.htm. For a no-impact
applicable and the project is in a FEMA-mapped area. There are certification form, contact the local FEMA administrator directly.
two types of FEMA-mapped areas: detailed study areas and Step 4. Develop a series of models in the following order:
those mapped using approximate methods. Detailed study areas a. Duplicate Effective Model – Take the model that was provid­
are those that have a mapped floodway; approximate areas do ed, get it running, and make sure the output matches the origi­
not have a mapped floodway. Hydraulic modeling is required nal output used to generate the FEMA map.
when a proposed project exists in either area. b. Corrected Effective Model – Add any new topographic data
If the proposed project is in a FEMA-mapped area, give serious to the model. The model must not reflect any man-made
thought before proposing a Priority 1 stream restoration, especially changes that have occurred since the date of the original
110 Chapter 11 Stream Restoration Stream Restoration Chapter 11 111
 model (check results with original and look for any errors asso­
ciated with the model).
c. Existing or Pre-Project Model—Add any new changes (man­
made) within the floodplain made after the date of the original
model (not including the proposed project). Insert new cross
sections at this stage.
d. Revised or Post-Project Conditions Model—Change the
pre-project model to reflect the proposed project. These
changes will include modification of cross sections and possi­
bly channel roughness.

Figure 11.6
Cove Creek after
restoration

Figure 11.5 11.4. Case Study

Cove Creek prior to The Cove Creek restoration project illustrates several important
restoration aspects of FEMA requirements for stream-restoration projects.
Cove Creek is located in Watauga County within a FEMA-
mapped area. A Priority 3 restoration project was completed at
this site; therefore, pattern was not changed and structures were
installed to create a step-pool system. The cross-sectional area
of the channel below bankfull stage remained unchanged,
although the cross-sectional area above bankfull stage was
increased. The project was modeled using HEC-RAS, and the
results along with a project description were submitted to the
local FEMA administrator. Based on hydraulic modeling results
and description of the project, the local FEMA administrator
determined that a map revision was not necessary. Figures 11.5
and 11.6 show before and after photographs of the channel
cross section at the project location.

112 Chapter 11 Stream Restoration Stream Restoration Chapter 11 113

 Notes:

Restoration Evaluation and Monitoring Chapter 12

Methodology 12.1
Morphology 12.2
Success Criteria 12.2.1
Photo Documentation 12.3
Vegetation 12.4
Plot Location 12.4.1
Plot Size 12.4.2
Timing 12.4.3
Additional Monitoring Opportunities 12.5
Bank Stability Monitoring 12.5.1
Shading and Temperature 12.5.2
Fish and Invertebrate Data 12.5.3

114 Chapter 11 Stream Restoration

 Chapter 12: Restoration Evaluation and bank-pins to mark the location of each cross section surveyed.
Drive each pin (left and right) vertically all the way into the ground
Monitoring on each side of bank to establish the outer limits of each surveyed
cross section. Place wooden stakes (wrapped with surveyor’s
Monitoring and evaluation help determine whether the design tape) adjacent to the rebar marker to aid in locating the cross
objectives have been met. They also reveal the need for adjustments sections in the field. Show the locations of all cross sections
to design parameters, installation procedures and/or stabilization on the plan-view drawings. Use as much detail as possible, as it
methods. Information collected should be made available to is very difficult to find the markers once vegetation becomes
other restoration professionals to ensure continued improvement established.
in the field of stream restoration, design and construction. Complete the following steps to ensure successful replication
Each stream-restoration design should have a monitoring plan to: of cross section location and survey parameters. Also, see
a. Determine if stabilization and grade-control structures are Section 2.4 for cross section survey instructions.
functioning properly.
b. Check channel stability by measuring dimension, pattern General procedure for permanent cross-section survey:
and profile; particle-size distribution of channel materials; ● Locate cross section on plan-view drawing and in field.
sediment transport; and streambank erosion rates. ● Locate end points on banks and mark them with rebar.
c. Determine biological response (i.e., vegetation, macroinver­ ● Pull a survey tape from left bank to right bank looking down­
tebrates and fish). stream at the cross-section location between the two rebar
d. Determine if the specific objectives of the restoration have endpins. The zero end of the tape should be directly over the
been met. left rebar stake.
● Set up level/surveying equipment in location with the fewest
12.1 Methodology visual obstacles.
A monitoring plan should include items presented in levels III ● Survey any permanent/temporary benchmarks (refer to plan-
(Section 3.3) and IV (Section 3.4) of Rosgen’s stream hierarchy that pre­ view drawings).
dicts and validates natural channel stability (Figure 3.1). Classify the ● Survey from left to right bank.
geomorphology of the stream using the Rosgen (1996) system; ● Survey distinctive points (i.e., top of bank, edge of water,
assess it using the results of the survey data. Current agency bankfull features, thalweg) and any other breaks in slope.
stream-mitigation monitoring requirements include morphology,
photo-documentation and vegetation. Monitor these parameters Survey elevations in the area can be based on any of the rebar
at least once a year for five years after construction. In addition, pins (benchmark) set in the field. The relative elevation at each
it may be useful to monitor shading and temperature; fish and pin is located on the cross section survey data.
invertebrates; and/or stream stability. Prepare a monitoring report Measure all significant breaks in slope that occur across the
that organizes data in a format that is easy to replicate annually. channel. Outside the channel, measure important features,
including the active floodplain and terraces.
12.2 Morphology
Complete a geomorphic survey. Include in the monitoring plan Longitudinal Profile
an assessment of streambank stability as well as stream mor­ The longitudinal profile measures points along the thalweg of
phology. During field reconnaissance, establish permanent cross- the stream channel. The profile indicates the elevations of water
sections at riffles and pools, survey the longitudinal profile and surface, channel bed, floodplain (bankfull) and terraces. The
conduct pebble counts. Select distinctive areas (upstream to elevations and positions of channel-defining indicators and in-
downstream) along the stream corridor as individual sections or stream structures also can be monitored with this profile.
reaches for reference; survey and monitor them. Denote these Take longitudinal profiles for each reach of the project along
areas on the plan-view drawings. the corridor of the restored stream. Survey the longitudinal profile
For all of the following, collect data once a year for at least five and cross sections at the same time. Place the beginning of the
years after construction. Plot cross sections over those of previous longitudinal-profile tape at the established station-zero point
years for comparison and evaluation. (STA 0) and continue downstream to the end of the restored
stream reach. At each station along the profile, survey the thalweg,
Cross-Section Geometry water surface, bankfull and, if appropriate, top of low bank. The
For monitoring purposes, establish permanent cross sections in start and end points of each longitudinal profile should be locat­
each of the reaches along the restored stream corridor. Survey at ed on the plan-view drawings. Extend each profile from upstream
least one riffle and one pool cross section for each reach. Note the to downstream along the entire length of the restored channel.
location of each cross section to establish the exact transect Also, see Section 2.6 for longitudinal-profile survey instructions.
location along the longitudinal profile.
Use rebar to mark each cross section. Install left and right

116 Chapter 12 Stream Restoration Stream Restoration Chapter 12 117

 Pebble-Count Data Take photographs standing at the approximate location of the
The composition of the stream bed and banks is a good indi­ established photo-point, cross-section location, and/or refer­
cator of changes in stream character, channel form, hydraulics, enced stream feature/structure. Take photographs throughout the
erosion rates and sediment supply. A pebble count gives a quan­ monitoring period at the same locations. Compare to photos
titative description of the bed material. Pebble counts should be from previous years to evaluate vegetative growth and channel
performed at permanent cross sections within each reach of the stability.
project. Each count should include 100 pebbles collected from Use photographs to subjectively evaluate channel aggradation
left bankfull to right bankfull. Follow the procedures for cross- or degradation, bank erosion, success of riparian vegetation and
section analysis of the substrate outlined in Section 2.7. Perform effectiveness of in-stream structures and erosion-control meas­
a pebble count at each of the reaches along the stream channel. ures. Photos will indicate the presence or absence of developing
Record the count on a tally sheet and plot the data by size-class bars within the channel or an excessive alteration in channel
and frequency (see Figure 2.10). depth or width. Photos also will indicate the presence of any
excessive bank erosion or continuing degradation of the bank.
12.2.1 Success Criteria The series of photos over time should indicate successional mat­
Using this data to judge success or failure of restoration activi­ uration of riparian vegetation.
ties is somewhat subjective. There likely will be minimal changes
in the cross sections, profile and/or substrate composition. 12.4 Vegetation
Evaluate changes that occur during the monitoring period to Survival of vegetation should be evaluated using survival plots
determine if they represent a movement toward a more unstable and/or direct counts along the entire corridor of the restored stream.
condition. When analyzing monitoring results, physical parame­ Survival of vegetation inside the riparian buffer may be docu­
ters of particular concern include: width-to-depth ratio, entrench­ mented for the monitoring period through stem-counts and pho­
ment ratio, bank height ratio, radius-of-curvature ratio, feature tographic documentation of the entire length of the buffered cor­
slopes and substrate composition. Deviations from the design ridor. Document the data from stem-counts and photographs at
values on these parameters may lead to significant channel pre-established stations/plot areas. If the initial (year-one) survey
instability. For example, analysis of changes in the width-to­ doesn’t show 80 percent survival, plant supplemental vegetation
depth ratio and/or channel slope may determine if any changes the next winter.
will lead to problems with sediment transport. In a stable condi­
tion, the monitoring results should show only a slight adjustment 12.4.1 Plot Locations
in width-to-depth ratio, which is expected as vegetation and the Locate plots adjacent to the stream and survey them for future
associated root mass create a narrowing of the channel. With replication. Plots should be located in areas large enough to
regard to the substrate material and expected adjustments during obtain a representative sample of the planted population. Ideally,
the monitoring period, a coarsening of the bed is normal because a sample size of 10 percent of the planted area should be sur­
fine material moves downstream and is not replaced. The stabiliza­ veyed. In some cases, plots will be located in areas such as out­
tion of eroding banks, for example, decreases the amount of fine side meander bends or atop bankfull benches and extended into
material in the stream. Profile measurements consist of the facet the riparian buffer.
slopes for each of the features in the channel (riffle, run, pool and
glide). Stability of the channel depends on maintaining these 12.4.2 Plot Size
slopes, especially the riffle slopes. Significant adjustments to the Two different types of plots need to be established to determine
facet slopes may indicate such processes as channel down-cut­ survivorship of stakes and bare-root seedlings. Sizes and numbers
ting and increased channel slope. Because each restoration proj­ of plots will depend on site conditions, particularly buffer width
ect will have its own critical values, the values that determine the and project size. Ideally, rectangular plots as large as 100 square
geomorphic threshold for a particular stream must be determined meters will be used in determining survivability for bare-root
on a case-by-case basis. Adjustments that do not exceed the trees. These should be linear and parallel to the stream channel.
critical values may be attributed to changes within or along the Count stakes from beginning to end of outside meander bends
channel that signal increased stability, such as added vegetation if this is the sole location of stakes. If stakes are planted along
on the banks. runs, riffles or glides, use rectangular plots as with the bare-root
trees. Plot size will depend on site conditions and project size.
12.3 Photo Documentation Herbaceous plants are neither stakes nor bare-root trees. If
Establish photographic points at distinguishing locations along development of herbaceous cover is desired, include counts of
the stream, including in-stream structures. this material (establish subplots) in either the stake or bare-root
Take photos at points along the stream corridor (i.e., standing tree plot counts. The plot size for herbaceous cover should be no
upstream, looking downstream). Mark each photo point in the more than 1 square meter.
field with a wooden stake, or reference it by cross section or
stream feature/structure (i.e., rock vane). Place all photo-point
locations on the plan-view drawings for future reference.
118 Chapter 12 Stream Restoration Stream Restoration Chapter 12 119
 12.4.3 Timing tions can provide insights on the overall health of the stream and
Sample vegetation during the growing season. Ideally, this the need for habitat improvement. When restoration work can be
would be mid-summer in June or July. The growing season ends done throughout the watershed, these populations are a valuable
between Aug. 1 and Oct. 31 depending on location. tool for assessing the success of the work. When populations
can be evaluated on a watershed basis and at the restoration
12.5 Additional Monitoring Opportunities site, a marked difference at the site might indicate that local con­
ditions are limiting populations. In this case, on-site work may
12.5.1 Bank Stability Monitoring improve the populations, and monitoring of important popula­
The newly constructed or repaired streambanks can be moni­ tions may be warranted.
tored and assessed for their stability. This monitoring can be When sampling fish and invertebrate populations, use standard
accomplished through BEHI rating, bank pins, bank profile and procedures so that results can be compared with other studies.
permanent cross section. See Section 3.3 for instructions. Post- Quantitative fish-population samples can be evaluated using the
restoration stability assessment and bank-erosion monitoring 3-pass depletion method that the N.C. Wildlife Resources
results can be compared to preconstruction data to determine if Commission uses to evaluate trout populations (Armour et al., 1983).
the restoration work has improved the stability and thereby less­ Population estimates can be computed using Microfish 3.0
ened streambank erosion. (Deventer and Platts, 1989). Population estimates and biomass estimates
can then be easily converted to densities and standing crops.
12.5.2 Shading and Temperature The Index of Biotic Integrity used by the N.C. Division of Water
Monitoring of water and air temperatures will show how well Quality (Department of Environment and Natural Resources)
the planted vegetation is providing thermal stability in the riparian is a good method for qualitative fish-population sampling.
zones. Water temperature may be sampled using recording ther­ Invertebrate sampling should follow the methods
mometers such as the StowAway‚ XTI made by Onset Computer prescribed by the Division of Water Quality (available for download at
Corporation or a similar device. These thermometers may be http://h2o.enr.state.nc.us/ncwetlands/dave.pdf). Monitoring reports should
placed in the stream at the beginning and end of each site and explain the need for the fish and invertebrate data and how they
set to record the water temperature every hour. Water temperature will be used to evaluate any restoration work.
recording can continue each year until the desired stream-shading
is accomplished. Evaluate shading effects on air temperature by
recording air temperature along each reference transect established
for lateral photo reference (upstream and downstream of the
photo points to the extent of the photographs). Record air tem­
perature at each location in which the shading effect is measured;
measure 1 meter above the ground or water surface. Determine
temperature stability by measuring air temperature in the shade
for seven consecutive days. This temperature stability measure­
ment may be done within the easement or buffer area at the top
of the streambank as well as outside of the easement, both along
one of the established photo-point transect lines.
Comparisons of air temperature and shading along each transect
(from edge of buffer to midstream) should indicate a lower tempera­
ture and increased shading. Water temperature should decrease or
at least be constant as it moves through the restoration site.
Decreased temperature might not be observed until riparian vegeta­
tion grows enough to shade the stream and riparian zone.
Temperature stability data should indicate that the riparian zone has
a more stable (less varied) temperature regime than a site outside of
the vegetated buffer. Reference data from existing riparian zones in
excellent condition need to be developed to provide targets for
shading and thermal buffering.

12.5.3 Fish and Invertebrate Data

Information on fish and aquatic macroinvertebrate populations
(density and diversity) may be used to guide decision-making in
the restoration planning and monitoring process. These popula­

120 Chapter 12 Stream Restoration Stream Restoration Chapter 12 121

 Notes:

References and Resources Chapter 13

122 Chapter 12 Stream Restoration

 Leopold, L.B. and M.G. Wolman. Rosgen, D.L. 1998a. The reference
Chapter 13: References and Resources
1960. River Meanders. Geological reach—a blueprint for natural chan-
Society of America Bulletin 71, pp. nel design. Draft Presented at ASCE
769-794. Conference on River Restoration,
13.1 References Restoration: Principles, Processes Denver, Colorado, March 1998.
Andrews, E.D. 1983. Entrainment Of and Practices. Springfield, Va: Leopold, L. B., G. M.. Wolman and J. Reston, Va.: American Society of
Gravel From Naturally Sorted National Technical Information P. Miller. 1992. Fluvial Processes in Civil Engineers.
Riverbed Material. Geological Service. Geomorphology. Mineola, New York:
Society of America Bulletin 94, pp. Dover Publications Inc. Rosgen, D.L. 1998b. The Reference
1225-1231. Harman, W. A., D. E. Wise, M. A. Reach Field Book. Pagosa Springs,
Walker, R. Morris, M. A. Cantrell, M. Leopold, Luna. A View of the River. Colo.: Wildland Hydrology Inc.
Armour, C. L., K. P. Burnham, and W. Clemmons, G. D. Jennings, D. 1994. Cambridge, Mass.: Harvard
S. Platts. 1984. Field Methods and Clinton, and J. Patterson. 2000. University Press. Rosgen, D.L. 1998c. River
Statistical Analyses for Monitoring Bankfull regional curves for North Restoration and Natural Channel
Small Salmonid Streams, Publication Carolina mountain streams. In Proc. Patterson, J. M., D. R. Clinton, W. A. Design. Course Handbook. Pagosa
FWS/OBS-83/33, Western Energy AWRA Conf. Water Resources in Harman, G. D. Jennings, and L. O. Springs, Colo.: Wildland Hydrology.
and Land Use Team, Division of Extreme Environments, Anchorage, Slate. 1999. Development of stream-
Biological Services, Research and Alaska, ed. D. L. Kane, pp. 185-190. bank erodibility relationships for Rosgen, D.L. 2001a. A practical
Development. Washington, D.C.: Middleburg, Va.: American Water North Carolina streams. In Wildland method of computing streambank
U.S. Fish and Wildlife Service. Resources Association Hydrology, Proc. AWRA Specialty erosion rate. In Proceedings of the
Conf., Bozeman, Montana, ed. D. S. Seventh Federal Interagency
Chow, V.T. 1959. Open Channel Harmel, R. D., C. T. Haan, and R. C. Olson and J. P. Potyondy, 117-123. Sedimentation Conference, vol. 2,
Hydraulics. New York: McGraw-Hill Dutnell. 1999. Evaluation of Middleburg, Va.: American Water Reno, Nevada, March 25-29.
Inc. Rosgen’s streambank erosion poten­ Resources Association.
tial assessment in Northeastern Rosgen, D.L. 2001b. A stream chan-
Clinton, D. R., G. D. Jennings, W. A. Oklahoma. Journal AWRA 35(1):113­ Radford, A.E., H.E. Ahles, and C.R. nel stability assessment methodolo-
Harman, J. M. Patterson, L. O. Slate, 121. Bell. 1968. Manual of the Vascular gy. In Proceedings of the Seventh
and J. Williams. 1999. North Carolina Flora of the Carolinas. Chapel Hill, International Sedimentation
reference stream channel morpholo­ Harrelson, C. C., C. L. Rawlins, and N.C.: The University of North Conference, Reno, Nevada, March
gy relationships. In Wildland J. P. Potyondy. 1994. Stream
Carolina Press. 25-29.
Hydrology, Proc. AWRA Specialty Channel Reference Sites: An

Conf. Bozeman, Montana, ed. D.S. Illustrated Guide to Field Technique.

Rinaldi, M. and P.A. Johnson. 1997. Rosgen, D.L. 2001c. The cross-vane,
Olson and J. P. Potyondy, 393-400. General Technical Report RM-245.
Stream meander restoration. Journal w-weir, and j-hook vane structures
Middleburg, Va.: American Water Fort Collins, CO:
of the American Water Resources … their description, design and
Resources Association. USDA Forest Service, Rocky
Association 33:855-866. application for stream stabilization
Mountain Forest and Range
and river restoration. In Proceedings
van Deventer, J. S. and W. S. Platts. Experiment Station.
Rosgen, D. L. 1994. A classification of the Wetlands Engineering and
1989. Microcomputer Software of natural rivers. Catena 22:169-199. River Restoration Conference, Reno,
System For Generating Population Jennings, G. D., and W. A. Harman. Nevada, Aug. 27-31. Reston, Va:
Statistics From Electrofishing Data: 2000. Stream Corridor Restoration Rosgen, D.L. Applied River American Society of Civil Engineers.
User’s Guide For MicroFish 3.0. Experiences In North Carolina. ASAE Morphology. 1996. Pagosa Springs,
General technical report INT-254. Paper 002012, ASAE Annual Colo.: Wildland Hydrology Books. Rosgen, D.L. 2002. River
Ogden, Utah: Intermountain International Meeting, Milwaukee, Assessment and Monitoring Field
Research Station, U.S. Department Wis. St. Joseph, Mich.: American Rosgen, D.L., 1997. A geomorpho- Guide. Pagosa Springs, Colo.:
of Agriculture, U.S. Forest Service. Society of Agricultural Engineering. logical approach to restoration of Wildland Hydrology Inc.
incised rivers. In Proceedings of the
Dunne, T. and L. B. Leopold. 1978. Jessup, Angela (Natural Resource Conference On Management of Schumm, S. A., M. D. Harvey, and C.
Water in Environmental Planning. Conservation Service). 2002. E-mail Landscapes Disturbed by Channel C. Watson. 1984. Incised Channels:
New York: W. H. Freeman and to Barbara Doll, 10 July. Incision, ed. S.S.Y. Wang, E.J. Morphology, Dynamics and Control.
Company. Langendoen and F.B. Shields, Jr. Littleton, Colo.: Water Resources
Knighton, David. Fluvial Forms and Oxford, Miss.: University of Publications.
Federal Interagency Stream Processes. 1984. New York: Mississippi.
Restoration Working Group (FISR­ Rutledge, Chapman and Hall Inc.
WG). 1998. Stream Corridor

124 Chapter 13 Stream Restoration Stream Restoration Chapter 13 125

 Simon, A. 1989. A model of channel N.C. Division of Water Quality, Index Dorava, J. M., D. R. Montgomery, B. Leopold, Luna. 1994. A View of the
response in disturbed alluvial chan­ of Biotic Integrity B. Palcsak and F. A. Fitzpatrick. 2001. River. Cambridge, Mass.: Harvard
nels. Earth Surface Processes and http://h2o.enr.state.nc.us/ncwet- Understanding Geomorphic Processes University Press.
Landforms 14(1):11-26. lands/dave.pdf and Riverine Habitat. Water and
N.C. Division of Water Quality, Science Application Volume 4: 3-4. Merritt, R. W. and K. W. Cummins,
Simon, A. and P. W. Downs. 1995. Wetlands/401 Certification Unit Washington, D.C.: American eds. 1984. An Introduction to the
An interdisciplinary approach to eval­ http://h2o.enr.state.nc.us/ncwet- Geophysical Union. Aquatic Insects of North America. 2nd
uation of potential instability in allu­ lands/ edition. Dubuque, Iowa: Kendall/Hunt
vial channels. Geomorphology Edwards, T. K. and G. D. Glysson. Publishers.
12:215-232. North Carolina Stream Restoration 1999. Field methods for measurement
Institute of fluvial sediment In Techniques of Montgomery, D. R. 2001.
Strahler, A.N. 1957. Quantitative http://www.ncsu.edu/sri/ Water Resources Investigations, Book Geomorphology, river ecology, and
analysis of watershed geomorpholo­ 3, Chapter C2. U.S. Geological ecosystem management In
gy. American Geophysical Union Stream Systems Technology Center Survey. Understanding Geomorphic Processes
Transactions 38:913-920. http://www.stream.fs.fed.us/ and Riverine Habitat. Water and
Federal Interagency Stream Science Application Volume 4: 247­
Thorne, Colin R., Hey, Richard D., U.S. Army Corps of Engineers, Restoration Working Group (FISRWG). 253. American Geophysical Union.
and Malcolm D. Newson, eds. 1997. Wilmington District Regulatory 1998. Stream Corridor Restoration:
Applied Fluvial Geomorphology for Division Principles, Processes and Practices. Rosgen, D. L. Applied River
River Engineering and Management. http://www.saw.usace.army.mil/wet- Springfield, Va.: National Technical Morphology. 1996. Pagosa Springs,
New York: John Wiley and Sons. lands/regtour.htm Information Service. Colo.: Wildland Hydrology Books.

Troendle, C. A., D. L. Rosgen, S. E. U.S. Environmental Protection Frothingham, K. M., B. L. Rhoads and Schiechtl, H. M. June 1986. FAO
Ryan, L. S. Porth and J. M. Agency, Office of Water, River E. E. Herricks. 2001. Stream geomor- Watershed Management Field Manual:
Nankervis. 2001. Developing a refer- Corridor and Wetland Restoration phology and fish community structure Vegetative and Soil Treatment
ence sediment transport relationship. http://www.epa.gov/owow/wetlands/ in channelized and meandering reach- Measures. Rome, Italy: Food and
In Proceedings of the 7th restore/ es of an agricultural stream. In Agriculture Organization of the United
International Sedimentation Understanding Geomorphic Processes Nations.
Conference, Reno, NV, March 25-29 U.S. Geological Services, Water and Riverine Habitat. Water and
Resources of North Carolina Science Application Volume 4: 105­ Schiechtl, H. M. and R. Stern.
U.S. Department of Agriculture, Natural http://nc.water.usgs.gov/ 117. Washington, D.C.: American December 1996. Water
Resources Conservation Service. 1986. Geophysical Union. Bioengineering Techniques for
Urban Hydrology for Small Watersheds: Wildland Hydrology Consultants Watercourse Bank and Shoreline
Technical Release 55. 210-VI-TR-55, http://www.wildlandhydrology.com/ Gore, J. A. 2001. Models of habitat Protection. Cambridge, Mass:
Second Ed. use and availability to evaluate anthro- Blackwell Science Inc.
pogenic changes in channel geometry.
13.3 Recommended Reading
U.S. Geological Survey. 1982. In Understanding Geomorphic Schiechtl, H. M., R. Stern and L.
Guidelines for Determining Flood Bunte, K. and S. R. Abt. 2001. Processes and Riverine Habitat. Laklitsch (translator). July 1996.
Flow Frequency. Bulletin 17B of the Sampling Surface and Subsurface Water and Science Application Ground Bioengineering Techniques for
Hydrology Subcommittee, Water Particle-Size Distributions in Volume 4: 27-36. American Slope Protection and Erosion Control.
Resources Council, pp. 14-20. Wadable Gravel- and Cobble-Bed Geophysical Union. Cambridge, Mass.: Blackwell Science
Streams for Analyses in Sediment Inc.
Williams, G.P. 1986. River meander Transport, Hydraulics, and Harrelson, C. C., C. L. Rawlins, and J.
and channel size. Journal of Streambed Monitoring. General P. Potyondy. 1994. Stream Channel Statzner, B., J. A. Gore and V. H.
Hydrology 88:147-164. Technical Report RMRS-GTR-74. Reference Sites: An Illustrated Guide Resh. 1988. Hydraulic stream ecolo-
Fort Collins, Colo.: U.S. Department to Field Technique. General Technical gy: observed patterns and potential
Wolman, M.G., 1954. A method of sam­ of Agriculture Forest Service, Rocky Report RM-245. Fort Collins, Colo.: applications. J. N. Am. Benthol. Soc.
pling coarse river-bed material. Mountain Research Station. USDA Forest Service, Rocky 7(4)307-360.
Transactions of American Geophysical Mountain Forest and Range
Union 35: 951-956. Calow, P. and G. E. Petts, eds. Vol.1, Experiment Station. Thorne, C. R., R. D. Hey, and M. D.
1992; Vol. 2, 1994. The Rivers Newson, eds. 1997. Applied Fluvial
13.2 Web Information Handbook: Hydrological and Leopold, L. B., G. M.. Wolman and J. Geomorphology for River Engineering
Federal Interagency Stream Ecological Principles. Cambridge, P. Miller. 1992. Fluvial Processes in and Management. New York: John
Restoration Working Group: Mass.: Blackwell Science, Inc. Geomorphology. Mineola, N.Y.: Dover Wiley and Sons.
http://www.usda.gov/stream_restoration/ Publications Inc.
126 Chapter 13 Stream Restoration Stream Restoration Chapter 13 127
 Notes:

Appendices

River Course Fact Sheets A

Field Data Sheets B
Gage Station Data C
Regional Hydraulic Geometry Relationships D
Stream Structure Details E
Vegetation Specifications and Plant Lists F

128 Chapter 13 Stream Restoration

 1 Introduction
Introduction to Fluvial Processes Chapter 1

Bankfull Discharge and Stage 1.1

Natural Channel Stability 1.2
Channel Dimension 1.3
Channel Pattern 1.4
Channel Profile 1.5
Channel Features 1.6
Biological Considerations of Stream Restoration 1.7
Conclusions 1.8
 2 Survey Procedures
Stream Assessment and Survey Procedures Chapter 2

Office Procedures
Watershed Drainage Area Measurement 2.1
Land-Use Survey 2.2

Field Procedures
Bankfull Identification 2.3
Dimension 2.4
Pattern 2.5
Profile 2.6
Substrate Analysis 2.7
Bar, Pavement and Subpavement Sampling
Methods and Scour Chains 2.8
Estimating Bankfull Discharge and Velocity 2.9
Assessing Riparian Condition 2.10
 3 Rosgen Classification
Rosgen Stream-Classification System/
Channel Assessment and Validation Procedures Chapter 3

Level I 3.1
Level II 3.2
Level III 3.3
Level IV 3.4
Bankfull Verification and Gage Station Analyses Chapter 4

Stream Gage Survey Procedure 1

4 Gage Analyses
 Priority Options for Restoring Incised Streams Chapter 5

Priority 1: Establish Bankfull Stage at

the Historical Floodplain Elevation 5.1
Priority 2: Create a New Floodplain and
Stream Pattern with the Stream
Bed Remaining at the Present Elevation 5.2
Priority 3: Widen the Floodplain at the
Existing Bankfull Elevation 5.3
Priority 4: Stabilize Existing Streambanks in Place 5.4
Priority 1 Case Study:
Yates Mill Pond Tributary 5.5
Priority 2 Case Study:
Pine Valley Golf Course Tributary 5.6
Priority 3 Case Study: Cove Creek 5.7
Priority 4 Examples 5.8

5 Restoration Options
 6 Reference Reach
Reference Reach Survey Chapter 6

Field Procedures
Bankfull Identification 6.1
Longitudinal Profile 6.2
Pool and Riffle Cross-Section Survey 6.3
Pebble Count 6.4
Rosgen Stream Classification 6.5
Plan-Form Measurements 6.6

Office Procedures
Profile Data Summary 6.7
Dimension Data Summary 6.8
Pattern Data Summary 6.9
Reference Reach Summary Table 6.10
Dimensionless Ratio Calculations 6.11
Vegetation Reference Reach 6.12
North Carolina Reference Reach Data 6.13
 7 Design Procedures
Design Procedures Chapter 7

Design Steps 7.1

Sediment Transport 7.2
 Structures Chapter 8

8 Structures
Root Wads 8.1
Vanes 8.2
Stream Crossings 8.3
Structures and Design Features
for Habitat Enhancements 8.4
Vegetation Stabilization and Riparian-Buffer Re-establishment Chapter 9

Salvaging On-Site Vegetation 9.1

Live Staking 9.2
Bare-Root Plantings 9.3
Container Plant Material 9.4
Permanent Seeding 9.5

9 Vegetation
 Erosion and Sediment Control Plan Chapter 10

Pollution Control:
Construction Sequence and Structures 10.1
Pollution Control:
Seeding 10.2

10 Erosion Control
 11 Flood Studies
Flood Studies Chapter 11

Introduction 11.1
FEMA Maps and Nomenclature 11.2
FEMA Requirements and Flood Modeling 11.3
Case Study 11.4
 12 Monitoring
Restoration Evaluation and Monitoring Chapter 12

Methodology 12.1
Morphology 12.2
Success Criteria 12.2.1
Photo Documentation 12.3
Vegetation 12.4
Plot Location 12.4.1
Plot Size 12.4.2
Timing 12.4.3
Additional Monitoring Opportunities 12.5
Bank Stability Monitoring 12.5.1
Shading and Temperature 12.5.2
Fish and Invertebrate Data 12.5.3
 13 Reference/Resources
References and Resources Chapter 13

 Appendices

River Course Fact Sheets A

Field Data Sheets B
Gage Station Data C
Regional Hydraulic Geometry Relationships D
Stream Structure Details E
Vegetation Specifications and Plant Lists F

Appendices