Tardir Tiffs A383261.tiff

US Army Corps
of Engineers®
Engineer Research and
Development Center
Effective Discharge Calculation:

A Practical Guide
David S. Biedenharn, Ronald R. Copeland, August 2000
Colin R. Thome, Philip J. Soar, Richard D.Hey,
and Chester C. Watson
Approved for public release; distribution is unlimited.

20001023 028
The contents of this report are not to be used for advertising,
publication, or promotional purposes. Citation of trade names
does not constitute an official endorsement or approval of the use
of such commercial products.
The findings of this report are not to be construed as an

official Department of the Army position, unless so desig-
nated by other authorized documents.
® PRINTED ON RECYCLED PAPER

ERDC/CHLTR-00-15
August 2000
Effective Discharge Calculation:

A Practical Guide
by David S. Biedenharn, Ronald R. Copeland
Coastal and Hydraulics Laboratory
U.S. Army Engineer Research and Development Center
3909 Halls Ferry Road
Vicksburg, MS 39180-6199
Colin R. Thome, Philip J. Soar

University of Nottingham
Nottingham, NG7 2RD, UK
Richard D. Hey
University of East Anglia
Norwich NR4 7TJ, UK
Chester C. Watson
Colorado State University
Department of Civil Engineering
Fort Collins, CO 80523
Final report
Approved for public release; distribution is unlimited
Prepared for U.S. Army Corps of Engineers

Washington, DC 20314-1000
Engineer Research and Development Center Cataloglng-in-Publlcatlon Data
Effective discharge calculation : a practical guide / by David S. Biedenham ... [et al.];
prepared for U.S. Army Corps of Engineers.
63 p.: ill.; 28 cm. - (ERDC/CHL ; TR-00-15)
Includes bibliographic references.
1. Stream measurements - Guidebooks. 2. Hydraulic measurements - Guidebooks.
3. Alluvial streams. I. Biedenham, David S. II. United States. Army. Corps of Engineers.
III. U.S. Army Engineer Research and Development Center. IV. Coastal and Hydraulics
Laboratory (U.S.) V. Title. VI. Series: ERDC/CHL TR ; 00-15.
TA7 E8 no.ERDC/CHL TR-00-15
Contents
Preface vi
Conversion Factors, SI to Non-SI Units of Measurement vii
Summary viii
1—Channel Forming Discharge Concept 1
2—Effective Discharge Calculation 4
Hydrological Data 4
Gauged sites 4
Ungauged sites 7
Sediment Transport Data 10
Nature of the sediment load 10
Sediment transport data: gauged sites 10
Sediment transport data: ungauged sites 11
Computational Procedure 11
Flow frequency distribution 12
Bed-material load rating curve 13
Bed-material load histogram 15
3—Evaluation and Problem Solving 19
Problems with Lowest Discharge Class 19

Problem with Outliers 19
Checking the Effective Discharge 20
Guidance on return periods for the effective discharge 20
Drainage area- flow duration curve 20
Check effective discharge against bankfull discharge 20
4—Applications 22
Effective Discharge Calculation for the Mississippi River at Vicksburg 22

Flow frequency distribution 22
Bed-material load rating curve 23
Bed-material load histogram 24
in
Effective discharge determination 24
Check if effective discharge is reasonable 24
Channel Management: Case Study of the River Blackwater, UK 26
Channel Restoration Hydraulic Design 27
Introduction 27
Hydraulic design methodology 27
Effective Discharge Calculation: Case Study 31
Introduction 31
Restoration of Willful Creek 32
Post-project channel change 33
Channel restoration design procedure 34
Supply reach magnitude-frequency analysis 35
Simulated channel restoration design 35
Sediment budget assessment: Capacity Supply Ratio (CSR) 36
References 39
Glossary 45
List of Figures
Figure 1. Derivation of total sediment load-discharge histogram (iii) from

flow frequency (i) and sediment load rating curves (ii) 5
Figure 2. Daily mean flow-duration curve: Sevier River, Hatch UK

(Hey 1997) 5
Figure 3. Downstream daily flow-duration curves: River Wye, UK

1937-1962 (from Hey 1975) 9
Figure 4. Comparison of sediment relationships for Abiaca Creek,

Site no. 6 11
Figure 5. Computational procedure for generating a flow-frequency

distribution 14
Figure 6. Computational procedure for generating a bed-material load

rating curve 16
Figure 7. Computational procedure for generating a bed-material load

histogram 18
Figure 8. Effective discharge duration versus drainage area 21
IV
Figure 9. Flow-duration curve for mean daily discharge: Lower
Mississippi River at Vicksburg, 1950-1982 (Biedenharn and
Thorne 1994) 22
Figure 10. Sand load rating curve: Lower Mississippi River at Vicksburg
1969-1979 (Biedenharn and Thorne 1994) 23
Figure 11. Bed-material load histogram: Lower Mississippi River at

Vicksburg (Biedenharn and Thorne 1994) 24
Figure 12. Long-channel variation in bank top elevations: Lower Mississippi

in study reach (Biedenharn and Thorne 1994) 25
Figure 13. Design cross sections and post-project channel change: i) above:
bend apex; ii) below: meander crossing (200 percent vertical
exaggeration) 34
Figure 14. Supply Reach Magnitude-Frequency Analysis derived from

stable reaches in the tributary and main channel upstream of
the tributary confluence , 36
Figure 15. Stable width-depth-slope diagram. The range of stable slope

(primary axis) and depth (secondary axis) are shown between
95 percent confidence intervals of the mean response 37
Figure 16. Comparison of sediment supply and capacity for the enhanced
design based on 30 arithmetic discharge classes and increased
slope of 0.00324 38
Preface
The work described in this report was authorized by Headquarters, U.S. Army
Corps of Engineers (HQUS ACE) as part of the Flood Damage Reduction
Research Program. The guidelines developed herein were products of Work
Units 32776 "Channel Response and Channel-Forming Discharge" and 32878
"Channel Restoration Design." The Program Monitor was Mr. Richard J.
DiBuono, HQUSACE. The Program Manager was Ms. Carolyn Holmes, Coastal
and Hydraulics Laboratory (CHL), U.S. Army Engineer Research and
Development Center (ERDC). Principal investigators for the work units were
Drs. David S. Biedenharn and Ronald R. Copeland, CHL.
The report was prepared by Dr. David S. Biedenharn, and Dr. Ronald R.
Copeland, CHL, Dr. Colin R. Thorne, and Mr. Philip J. Soar, University of
Nottingham, Dr. Richard D. Hey, University of East Anglica, and Dr. Chester C.
Watson, Colorado State University. Data review and analysis was provided by
Ms. Dinah N. McComas, CHL.
The study was performed under the supervision of Mr. Michael Trawle,
former Chief of the River Sedimentation Branch, Dr. Phil G. Combs, former
Chief of the Rivers and Structures Division, and Dr. James R. Houston, former
Director, CHL.
At the time of publication of this report, Dr. James R. Houston was Director
of ERDC, and COL James S. Weiler, EN, was Commander.
The contents of this report are not to be used for advertising, publication,
or promotional purposes. Citation of trade names does not constitute an
official endorsement or approval of the use of such commercial products.
VI
Conversion Factors,
SI to Non-SI Units of
Measurement
SI units of measurement used in this report can be converted to Non-SI units

as follows:
Multiply By To obtain
meters 3.281 feet
cubic meters 35.32 cubic feet
square kilometers 0.39 square miles
tonne 1.1 US ton
VII
Summary
An alluvial river adjusts the dimensions of its channel to the wide range of
flows that mobilize its boundary sediments. However, in many rivers it has been
demonstrated that a single representative discharge can be used to determine a
stable channel geometry. The use of a single representative discharge is the
foundation of "regime" and "hydraulic geometry" theories for determining
morphological characteristics of alluvial channels and rivers. This representative
discharge has been given several names by different researchers including
dominant discharge, channel-forming discharge, effective discharge, and bankfull
discharge. This has led to some confusion. In this report the channel forming
discharge and the dominant discharge are equivalent and are defined as a
theoretical discharge that if maintained indefinitely would result in the same
channel geometry as the existing channel subject to the natural range of flow
events. Although conceptually attractive, this definition is not necessarily |
physically feasible, because bank line vegetation, bank stability and even the bed \
configuration would be different in a natural stream than in a constant discharge
stream. Channel-forming discharge concepts are applicable to stable stream
systems, i.e., streams that are neither aggrading or degrading.
Channel-forming discharge can be estimated using one of three prescribed

methodologies. One such deterministic discharge is the bankfull discharge,
which is the discharge that fills the channel to the top of its banks. Another
deterministic discharge used to represent the channel-forming discharge is a
specified recurrence interval discharge, typically between the mean annual and
five-year peak. This report focuses on a third approach to determine the channel-
forming discharge, known as the effective discharge. The effective discharge
transports the largest fraction of the bed material load. Because of this, the
effective discharge can be a good estimator for channel-forming discharge.
All three methodologies for estimating channel-forming discharge provide

challenges. In practice, identification of bankfull elevation along a reach of
channel in the field is often problematic. If nearby gauge data are not available, .
bankfull discharge must be calculated using an assumed hydraulic roughness.
Likewise, correlating recurrence interval and bankfull discharge has been shown
to vary widely for different types of streams. The calculation of effective
discharge requires historical hydrologic and sediment data. Without nearby
gauge data, effective discharge calculations require using an assumed hydraulic
roughness and determining a reliable sediment transport equation.
VIM
Equivalence of bankfull and effective discharges has been demonstrated for
naturally stable alluvial channels in a wide variety of river types (sand-bed,
gravel-bed, cobble-bed and boulder-bed) and in different hydrological
environments (perennial, humid and slightly ephemeral, and semiarid). However,
this equivalence may not hold for truly ephemeral streams in semi-arid or arid
areas.
The procedure for effective discharge calculations presented herein has been
developed for a range of river types. It is a systematic method designed to have
general applicability, and to integrate the effects of the physical processes
responsible for determining stable channel dimensions. The effective-discharge
should not be assumed to be the channel-forming discharge a priori without
confirmation using field indicators of geomorphic significance. It is good
practice to also determine the bankfull discharge where possible, and to develop a
frequency curve, and to cross-check the estimates for channel-forming discharge
to reduce the uncertainty in the final estimate.
Applications of the effective discharge concept outlined in the report include

a fully worked example of the calculations, a channel stability assessment, a
channel management application, and an application to restoration design.
IX
1 Channel-Forming
Discharge Concept
Alluvial rivers have the potential to adjust their shape and dimensions to all
flows that transport sediment, but Inglis (1941) suggested that for rivers that are
in regime, a single steady flow could be identified which would produce the same
bankfull dimensions as the natural sequence of events. He referred to this flow as
the dominant discharge, which is equivalent to the term channel-forming
discharge used in this report.
Based on field observations, Inglis (1947) determined that the channel-

forming discharge was approximated by flows at or about bankfull stage. This
finding has been supported by subsequent research (Nixon 1959, Simons and
Albertson 1960, Kellerhals 1967, Hey and Thome 1986). These studies have not,
however, explained why bankfull flow controls channel form.
To explain this phenomenon it is necessary to recognize that any local

imbalance in the sediment budget must generate change in the morphology of an
alluvial river through either erosion or deposition. Hence, to remain dynamically
stable, the regime dimensions of the channel must be adjusted so that, over a
period of years, sediment input and output are balanced. During this time,
Wolman and Miller (1960) showed that rivers adjust their bankfull capacity to
the flow transporting the most sediment. That flow was named the effective
discharge (Andrews 1980).
Wolman and Miller (1960) found that the effective discharge corresponds to
an intermediate flood flow since frequent minor floods with both shorter
durations and smaller peaks transport too small a sediment load to have a marked
impact on the gross features of the channel, while catastrophic events, which
individually transport large sediment loads, occur too infrequently to be effective
in forming the channel. The potential for large floods to disrupt the regime
condition and cause major channel changes is recognized by this concept, but
large floods are not the channel-forming events, provided that the return period
of these extreme events is longer than the period required for subsequent, lesser
events to restore the long-term, average condition (Wolman and Gerson 1978).
In humid environments, perennial rivers usually recover their long-term,

average morphology within 10 to 20 years following a major event, principally
because riparian and floodplain vegetation limits the impacts of major floods and
Chapter 1 Channel-Forming Discharge Concept

because vegetation regrowth encourages the processes of siltation involved in
morphological recovery (Gupta and Fox 1974; Hack and Goodlett 1960). In
semiarid regions, the recovery period tends to be longer, reflecting both the
reduced effectiveness of sparse vegetation in increasing the channel's resilience
to change and the sensitivity of the channel to the occurrence of wet and dry
periods (Schumm and Lichty 1963, Burkham 1972). In arid areas, the largest
floods leave long lasting imprints on the channel because of the lack of
vegetation and because lesser events capable of restoring a regime condition
rarely occur (Schick 1974).
Equivalence between bankfull and effective discharges for natural alluvial

channels that are in regime has been demonstrated for a range of river types
(sand, gravel, cobble and boulder-bed rivers) in different hydrological
environments (perennial, humid and slightly ephemeral, and semiarid) provided
that the flow regime is adequately defined and the appropriate component of the
sediment load is correctly identified (Andrews 1980; Carling 1988; Hey 1997).
The equivalence of bankfull and effective discharges for stable channels suggests
that either one could be used to define the channel-forming discharge. Also, in
theory, channel-forming discharge could be determined indirectly from an
estimate of the return period for either bankfull flow or the effective discharge.
In practice, major problems are likely to arise when attempting to use

bankfull discharge to determine the channel-forming discharge. Problems center
on the wide range of definitions of "bankfull stage" that exist (Williams 1978).
Although several criteria have been identified to assist in field identification of
bankfull stage, ranging from vegetation boundaries to morphological breaks in
bank profiles, considerable experience is required to apply these in practice,
especially on rivers which have undergone aggradation or degradation.
In many studies channel-forming discharge is estimated from the recurrence

interval for bankfull discharge. Leopold and Wolman (1957) suggested that the
bankfull flow has a recurrence interval of between one and two years. Dury
(1973) concluded that the bankfull discharge is approximately 97 percent of the
1.58-year discharge, or the most probable annual flood. Hey (1975) showed that
for three British gravel-bed rivers, the 1.5-year flow in an annual maximum
series passed through the scatter of bankfull discharges measured along the
course of the rivers. Richards (1982) suggested that, in a partial duration series,
bankfull discharge equals the most probable annual flood, which has a one-year
recurrence interval.
In practice, bankfull discharge is often assumed to have a recurrence interval

of about 1.5 years and recently Leopold (1994) stated that most investigations
have concluded that the bankfull discharge recurrence intervals range from 1.0 to
2.5 years. However, there are many instances where the bankfull discharge does
not fall within this range. For example, Pickup and Warner (1976) determined
that bankfull recurrence intervals may range from 4 to 10 years in the annual
maximum series. Therefore, extreme caution must be used when estimating the
channel-forming discharge using a flow of specific recurrence interval.

Problems associated with field identification of the bankfull elevation and the
choice of an appropriate recurrence interval are eliminated when the effective
discharge approach is employed. To determine the effective discharge, both an
annual flow-duration curve and a sediment-discharge rating curve are required.
The basic approach is to divide the range of discharges into a number of equal
arithmetic classes and then calculate the total sediment load for each flow class.
This is achieved by multiplying the frequency of occurrence of each flow class
by the median sediment load for that flow class. The mean of the flow class
with the greatest sediment load is the effective discharge.
Although the channel-forming discharge concept is not universally accepted,

most river engineers and scientists agree that the concept has merit, at least for
perennial and slightly ephemeral streams in humid and semiarid environments.
On the basis of this review, it is concluded that there are three approaches to
determining the channel-forming discharge: a) bankfull discharge, b) flow of a
given recurrence interval, and c) effective discharge.
Ideally, the method used to determine the channel-forming discharge should

have general applicability, should have the capability to be applied consistently,
and should integrate the physical processes responsible for determining the
channel dimensions. There are limitations for each of these methods that the user
must recognize. The selection of the appropriate method will be based on data
availability, physical characteristics of the site, level of study and time and
funding constraints. If possible, it is recommended that all three methods be used
and cross-checked against each other to reduce the uncertainty in the final
estimate.
This report outlines the best practical procedure for performing the necessary
calculations to determine the effective discharge.

2 Effective Discharge
Calculation
Hydrological Data
A standardized procedure is required to ensure that effective discharge
calculations are accurate and that results from different sites can be compared. To
be practical, the procedure must use only data which are readily available from
gauging stations, or which can be synthesized using limited additional
computations.
The basic approach is to divide the range of riverflows during the period of
record into a number of arithmetic classes and then calculate the total sediment
quantity transported by each class. This is achieved by multiplying the frequency
of occurrence of each flow class by the median sediment load for that flow class
(Figure 1). The initial data required are flow-duration data and a sediment
transport rating curve.
The calculated value of the effective discharge depends to some extent on the
steps used to manipulate the input data to define the flow regime and sediment
transport function. The procedure described represents "best practice" in this
regard, based on extensive firsthand experience in using flow and sediment
transport data to determine the effective discharge.
Gauged sites
The first step in an effective discharge calculation is to group the discharge

data into equal arithmetic flow classes and determine the number of events
occurring in each class during the period of record. Logarithmic or non-equal
width arithmetic classes introduce systematic bias into the calculation of effective
discharge and should not be used. Grouping the discharge data is usually
accomplished using a flow-duration curve, which is a cumulative distribution
function of observed discharges at the gauging station. For example, Figure 2 is a
flow-duration curve for the Sevier River, UK. The flow duration curve defines
the percentage of time a particular discharge is equaled or exceeded. The
frequency of occurrence of each discharge class is calculated from this curve.
Three critical components must be considered when developing a flow-duration
Chapter 2 Effective Discharge Calculation

X
<t>
CD
Ü 1M.
TO
c tl
m M
•o
<r> .«_,
O) C
&_ <n
OS
E
o ■n
m m
V)
Tfc T3
O ■*— ©
c 1_ >
sV FT3
0)
"5
LL W Ü
Figure 1. Derivation of total sediment load-discharge histogram (iii) from flow

frequency (i) and sediment load rating curves (ii)
m2
_
!A -
101-
0)
wi
-
o
.52
T5
10u^
m-1 !IIIIIII ! 11mil! mi mi;:iirn'iiiiiijiiiüinijmiimMhiiüiiiiiiiiiuihiii linn I i IIIIIII i i um in

0.01 1 10 30 50 70 90 99 99.99
0.1 5 20 40 60 80 95 99.9
Percentage of time discharge equalled or exceeded
Figure 2. Daily mean flow duration curve: Sevier River, Hatch, UK (Hey 1997)

curve: the number of discharge classes; the time base for discharge averaging;
and the length of the period of record. It is important that the historical record is
homogeneous, i.e. watershed conditions are unchanged.
Class interval and number of classes. The selection of class interval can
influence the effective discharge calculation. Intuitively, it may be expected that
the smaller the class interval and, therefore, the greater the number of classes, the
more accurate would be the outcome. However, when too small an interval is
used, discontinuities appear in the discharge frequency distribution. These in
turn produce a rather irregular sediment-load histogram, which has multiple
peaks. Therefore, the selected class interval should be small enough to
accurately represent the frequency distribution of flows, but large enough to
produce a continuous distribution.
There are no definite rules for selecting the most appropriate interval and
number of classes, but Yevjevich (1972) stated that the class interval should not
be larger than s/4, where .v is an estimate of the standard deviation of the sample.
For hydrological applications he suggested that the number of classes should be
between 10 and 25, depending on the sample size.
Hey (1997) found that 25 classes with equal arithmetic intervals produced a
relatively continuous flow frequency distribution and a smooth sediment-load
histogram with a well defined peak, indicating an effective discharge which
corresponded exactly with bankfull flow. A smaller interval, and correspondingly
larger number of classes, produced anomalous results. Experience has shown
that in some cases, 25 classes produce unsatisfactory results and that up to
250 classes may be required. Particular care has to be exercised on rivers where
there is a high incidence of very low flows as, for example, on the Little Missouri
River at Marmarth and Medora (Hey 1997). Under these circumstances, the
effective discharge may be biased towards the lowest discharge class.
Time base. Mean daily discharges are conventionally used to construct the
flow-duration curve. Although this is convenient, given the ready availability of
mean daily discharge data from the United States Geological Survey (USGS), it
can, in some cases, introduce error into the calculations. This arises because
mean daily values can underrepresent the occurrence of short-duration, high
magnitude flow events that occur within the averaging period, while
overrepresenting effects of low flows.
On large rivers, such as the Mississippi, the use of the mean daily values is
acceptable because the difference between the mean and peak daily discharges is
negligible. However, on smaller streams, flood events may last only a few hours,
so that the peak daily discharge is much greater than the corresponding mean
daily discharge. Excluding the flood peaks and the associated high sediment
loads can result in underestimation of the effective discharge. Rivers with a high
flashiness index, defined as the ratio of the instantaneous peak flow to the
associated daily mean flow, are more likely to be affected. To avoid this problem
it may be necessary to reduce the time base for discharge averaging from
24 hours (mean daily) to 1 hour, or even 15 minutes on flashy streams. For
example, an investigation of discharge data for 11 USGS gauging stations in the

Yazoo River Basin, Mississippi revealed that the annual yields of bed-material
calculated using mean daily discharge data were approximately 50 percent less
than the yields calculated using 15 minute data (Watson, Dubler, and Abt 1997).
These are relatively small basins (drainage areas less than 1,000 km2) with high
rainfall intensities and runoff characteristics that have been severely affected by
land-use change and channel incision. Consequently, hydrographs are
characterized by steep rising and falling limbs, with events peaking and returning
to base flow in much less than 24 hours.
In practice, mean daily discharge data may be all that are available for the
majority of gauging stations and these data may be perfectly adequate.
However, caution must be exercised when using mean daily data for watersheds
with flashy runoff regimes and short-duration hydrographs. The use of 15-
minute data to improve the temporal resolution of the calculations should be
seriously considered whenever the available flow records allow it.
In the absence of 1 hour or 15-minute data, recorded hydrographs from

USGS gauging records can be used to refine the high discharge portion of the
flow-duration curve. Actual instantaneously recorded hydrographs can be used
to determine durations of the highest discharges in the historical record.
Period of record. The period of record must be sufficiently long to include

a wide range of morphologically-significant flows, but not so long that changes
in the climate, land-use or runoff characteristics of the watershed produce
significant changes with time in the data. If the period of record is too short,
there is a significant risk that the effective discharge will be inaccurate due to the
occurrence of unrepresentative flow events. Conversely, if the period is too long,
there is a risk that the flow and sediment regimes of the stream at the beginning
of the record may be significantly different to current conditions.
A reasonable minimum period of record for an effective discharge calcula-

tion is about 10 years, with 20 years of record providing more certainty that the
range of morphologically significant flows is fully represented in the data.
Records longer than 30 years should be examined carefully for evidence of
temporal changes in flow and/or sediment regimes. If the period of record at a
gauging station is inadequate, consideration should be given to developing an
effective discharge based on regional estimates of the flow duration.
Ungauged sites
At locations where gauging records are either unavailable or are found to be

unrepresentative of the flow regime, it will be necessary to synthesize a flow-
duration curve. There are two possible methods of doing this. The first method is
by using records from nearby gauging stations within the same drainage basin.
The second is developing a regionalized flow-duration curve.

It must be recognized that these methods simply provide an approximation of
the flow-duration characteristics, and that there can be considerable uncertainty
in the results. The reliability of these methods is a function of the quality of the
existing gauge data and the morphologic similarity between the gauged and
ungauged locations. Caution is advised whenever the existing gauge data are
limited, or the site in question has a significantly different morphologic character
than the gauged site.
Drainage area, flow-duration curve method. This method relies on the

availability of gauging station data at a number of sites on the same river as the
ungauged location. First, flow-duration curves for each gauging station are
derived for the longest possible common period of record. This guarantees
comparability between the data, as all the gauging stations have experienced the
same flow conditions, and ensures that the curves represent the longer period.
Provided there is a regular downstream decrease in the discharge per unit
watershed area, then a graph of discharge for a given exceedance duration against
upstream drainage area will produce a power function with virtually no scatter
about the best-fit regression line. For example, Figure 3 shows this relationship
for the River Wye, UK (Hey 1975). The equations generated by this method
enable the flow-duration curve at an ungauged site on that river to be determined
as a function of its upstream watershed area.
For sites on streams where there is only one gauging station, flow duration
curves can be estimated at ungauged locations provided the streams are
tributaries to rivers where the relation between discharge and drainage area
conforms to a known power function. Estimates of the contributing flow to the
mainstem can be obtained from the difference between discharges on the
mainstem above and below the tributary junction. Discharge - drainage area
relations can then be derived for the tributary given the flow-duration curve at the
gauging station and the predicted curve at its confluence with the mainstem.
However, this technique should not be used if there are distinct and abrupt
downstream changes in the discharge per unit area for the watershed. This could
occur if portions of the drainage area consisted of different hydrological regions.
In this case it would be preferable to use the regionalized duration curve method
described in the next paragraph.
Regionalized duration curve method. An alternative to the use of

watershed area to generate a flow duration curve for an ungauged site is to use a
regional-scaling method based on data from watersheds with similar
characteristics. For example, Emmett (1975) and Leopold (1994) suggest using
the ratio of discharge to bankfull discharge (QIQh) as a non-dimensional index to
transfer flow-duration relationships between basins with similar characteristics.
However, bankfull discharge does not necessarily have either a consistent
duration or return period (Williams 1978).
To avoid this problem, a non-dimensional discharge index was proposed by

Watson, Dubler, and Abt (1997) using the regionalized 2-year discharge to
normalize discharges (QIQ2). For ungauged sites the 2-year discharge may be
estimated from regionalized discharge frequency relationships developed by the

Figure 3. Downstream daily flow-duration curves: RiverWye, UK 1937-1962
(Hey 1975)
USGS (1993) on the basis of regression relationships between the drainage area,
channel slope, and slope length. These relationships are available for most states.
The dimensionless discharge index {QIQi) can be used to transfer a flow-

duration relationship to an ungauged site from a nearby gauged site. The gauged
site may be within the same basin, or an adjacent watershed.
To transfer a flow-duration relationship within a watershed use the following

steps:
a. Develop the regionalized flow-duration curve. Using a flow-duration

curve from a gauged site in a physiographically similar watershed, divide
the discharges in the flow-duration relationship by the Q2 for the gauged
site. This creates a dimensionless flow-duration curve. If more than one

gauge site is available an average dimensionless flow-duration curve for
all the sites can be developed.
b. Compute the Q2 for the ungauged site.
c. Calculate the flow-duration curve for the ungauged site. Multiply the
dimensionless ratios from the regionalized flow-duration curve by the
ungauged Q2.
Sediment Transport Data
Nature of the sediment load
The total sediment load of a stream can be broken down on the basis of
measurement method, transport mechanism or source as shown in the following
diagram. When discussing the sediment load of a stream, it is important to keep
track of the terminology adopted and the nature of the load being discussed.
Sediment Source
Measurement Transport Mechanism
Method
Unmeasured Load Bed Load

Bed Material Load
Measured Load
Suspended Load Wash Load
Sediment transport data: gauged sites
In most alluvial streams the major features of channel morphology are princi-
pally formed in sediments derived from the bed-material load. It is, therefore, the
bed-material load which should be used in an effective discharge calculation.
At gauged sites the measured suspended load usually contains most of the
suspended load, but excludes the bed load. Under these circumstances, the
coarser fraction of the measured suspended load (generally the sand load - i.e.
particles larger than 0.062 mm) should be used to derive a bed-material load
rating curve. If available, bed load data should be combined with the coarse frac-
tion of the measured suspended load to derive a bed-material load rating curve.
Where a significant proportion of the bed-material load moves as bed load

(such as in gravel-bed rivers) but no measurements of bed load are available, it
may be necessary to estimate the bed load. This may be achieved using a suit-
able bed-load transport equation. The U.S. Army Corps of Engineer hydraulic
design package, SAM, (Copeland, et al. 1997) can be used to make these
calculations.
10 Chapter 2 Effective Discharge Calculation

Similarly, at gauging stations with no measured sediment load data at all, a
bed-material sediment rating curve may be generated using appropriate sediment
transport equations.
Sediment transport data: ungauged sites
At ungauged sites it will be necessary to generate a bed-material sediment

rating curve. The application of a suitable sediment transport equation is vital
and the SAM package is helpful because it includes guidance on the selection of
equations best suited to the type of river and bed-material in question (Raphelt
1990). An example of this approach is given in Watson, Dubler, and Abt (1997).
Figure 4 shows the resulting calculated sand discharges for several water
discharges together with the measured sand fraction load, and a rating curve
based on regression of the measured data. Close agreement is apparent between
the Brownlie computation of the bed-material load and the regression line based
on the observed USGS sand fraction data.
° 1.00E-03
'S 1.00E-04
Figure 4. Comparison of sediment relationships for Abiaca Creek, Site no. 6

(Watson, Dubler, and Abt 1997)
Computational Procedure
The recommended procedure to determine the effective discharge is executed
in three steps. In Step 1, the flow-frequency distribution is derived from available
flow-duration data. In Step 2, sediment data are used to construct a bed-material
load rating curve. In Step 3, the flow-frequency distribution and bed-material
load rating curve are combined to produce a sediment-load histogram, which
Chapter 2 Effective Discharge Calculation 11

displays sediment load as a function of discharge for the period of record. The
histogram peak indicates the effective discharge.
Flow frequency distribution
a. Evaluate flow record. The flow record is a historic record of discharges

at a gauging station. The record from a single gauging station can be
used to develop the flow-frequency distribution if the gauge is in close
proximity to the project reach and the discharge record at the gauge is
representative of the flow regime in the project reach. It is also
important that watershed conditions have remained unchanged during the
selected flow period. If a gauging record is either unavailable or
unrepresentative, then the flow-frequency distribution can be derived
using either the basin-area method or the regionalized duration curve
method.
b. Check the period of record. It is recommended that the length of the

period of record be at least 10 years and that measurements be
continuous to the present day. Discharge data can still be used if there
are short gaps in the record, but caution must be exercised when
collecting data from a discontinuous record. The flow-frequency curve
will not be representative of the natural sequence of tlows over the
medium term if the length of record is less than 10 years or if the record
has been influenced by changes in the watershed runoff regime. If this is
the case, a flow-duration curve should be developed using the methods
described in "ungauged sites."
c. Determine the discharge-averaging time base. To construct the flow-

frequency distribution, the time base should be sufficiently short to
ensure that short-duration, high magnitude events are properly
represented. If 15-minute data are unavailable, then either 1 hour or
mean daily data can be used, but caution must be exercised when using
mean daily data to develop a flow frequency distribution for a stream
exhibiting a flashy regime.
d. Calculate discharge range. The discharge range is calculated by

subtracting the minimum discharge in the flow record from the
maximum discharge.
e. Calculate discharge class size. It is recommended that the initial attempt

to construct the flow-frequency distribution should use 25 classes and an
arithmetic scale class interval. Therefore, the class interval is the
discharge range, calculated in Step (d), divided by 25. The class interval
should not be approximated by rounding. The relative proportions of the
bed-material load moving in suspension and as bed load should be
estimated during site reconnaissance. For rivers in which the bed-
material load moves predominantly as suspended load, the first discharge
class goes from zero to the class interval; the second class is determined
by adding the class interval to the upper value of the previous class and

so on until the upper limit of the discharge range is reached. For gravel-
bed rivers, where bed-material load moves predominantly as bed load,
the minimum discharge used in generating the flow-frequency
distribution should be set equal to the critical discharge for the initiation
of bed-load transport.
/ Calculate flow frequency distribution. The frequency of occurrence for

each discharge class is determined from the record of observed flows. It
is more convenient to convert the actual frequency to percentage
frequency. To do this multiply the actual value for each discharge class
by 100 and divide by the total number of recorded flows in the discharge
record. The percentage frequency represents the proportion of the flow
record falling within the upper and lower limits of each discharge class.
If a regional flow-duration curve has been developed for an ungauged
site, the frequency for each discharge class must be calculated using the
equation for the curve. This can be achieved by calculating the
geometric mean discharge of each discharge class and deriving the
percentage frequency from the equation. On the flow-duration curve, this
is the probability of each discharge being equaled or exceeded.
g. Check for extreme flow events. It is recommended that all discharge

classes display flow frequencies greater than zero and that there are no
isolated peaks in individual classes at the high end of the range of
observed discharges. If this is not the case, it is likely that either the
class interval is too small for the discharge range, or the period of record
is too short. Both zero frequencies and extreme flow events (outliers)
can be eradicated by reducing the number of classes. Steps e and f are
repeated to generate the flow-frequency distribution for the new class
intervals. The computational procedure for generating a flow frequency
distribution is outlined in Figure 5.
Bed-material load rating curve
a. Determine sediment data availability. Sediment data are required to

generate the bed-material load rating curve. These data may be obtained
from measurements at a gauging station if the gauge is in close proximity
to the project reach and if size-class fractions are provided so that the
bed-material portion of the measured load can be determined. A bed
gradation from the project reach is required to determine the division
between wash load and bed-material load, and to calculate sediment
transport if necessary.
b. Define composition of bed-material load. The wash load should be

excluded from the data set used to develop the rating curve. If the bed-
material load moves both as bed load and suspended load, then both bed
load and suspended-load measurements are required to determine the
bed-material load. If measured load data are insufficient, appropriate

START
' ■
(i) Basin-Area Method
(ii) Regionalized Duration

Curve Method
c Flow record
B
Calculate range of
discharge
Divide discharge range

into 25 arithmetic class
intervals
Calculate frequency of
each discharge class
Reduce number
of discharge classes
>
Figure 5. Computational procedure for generating a flow-frequency distribution

equations in the hydraulic design package SAM (Copeland et al. 1997),
can be used to generate bed-material loads for selected discharges.
c. Determine bed-material load. In streams dominated by suspended load, a

best-fit regression curve, fitted to the data, may be adequate to produce a
bed-material load function. Frequently this takes the form of a power
function:
Q=aOh
where Qs is the bed-material load discharge, Q is the water discharge, a is a

regression coefficient, and b is a regression exponent. However, a straight line
power function may not be appropriate in all cases. Sometimes, at high
discharges the rate of increase in sediment concentration with discharge begins to
decrease, especially for the finer sand sizes. In this case it may be necessary to
use a different curve fitting function. In coarse bed streams it is likely that a
coarse surface layer will develop at lower discharges, significantly reducing
sediment transport potential. This process involves both hydraulic sorting of the
stream bed and hiding of small particles behind bigger particles. Typically,
calculated sediment transport rating curves developed from a single-bed
gradation will overestimate sediment transport at low discharges. This is
probably the most important reason for too much sediment being calculated in
the lower discharge class intervals. The computational procedure for generating
a bed-material load rating curve, is diagrammed in Figure 6.
Bed-material load histogram
a. Calculate representative discharges. The discharges used to generate the

bed-material load histogram are the mean discharges in each class in the
flow frequency distribution.
b. Construct the bed-material load histogram. The histogram is generated

by using the representative discharges and the bed-material load rating
curve to find the bed-material load for each discharge class and
multiplying this load by the frequency of occurrence ofthat discharge
class. The results are plotted as a histogram representing the total
amount of bed-material load transported by each discharge class during
the period of record. This calculation can be completed using the
sediment yield routine in the SAM hydraulic design package.
c. Calculate effective discharge. The bed-material load histogram should

display a continuous distribution with a single mode (peak). If this is the
case, the effective discharge corresponds to the mean discharge for the
modal class (the peak of the histogram). If the modal class cannot be
readily identified, the effective discharge can be estimated by drawing a
smooth curve through the tops of the histogram bars and interpolating the
effective discharge from the peak of the curve.

z START
Define composition
of bed material load
c Measured load
~E
Remove wash load
component
Plot sediment load, Qs, Use appropriate

(ordinate) as a function of sediment transport
discharge, Q, (abscissa) equation(s) to calculate
using logarithmic axes bed material load
Use multiple sedime Use single sediment

rating curves rating curves
> >
Calculate best-fit regression

Calculate best-fit regression
power functions through
power function through all
plotted data for different
plotted data
ranges of discharge
Multiple sediment
rating curves
Single sediment
rating curve
7
c END
J
Figure 6. Computational procedure for generating a bed-material load rating curve

d. Evaluate the bed-material load histogram. If the modal class of the bed-
material load histogram is the lowest discharge class, it is likely that the
indicated effective discharge is erroneous. In this case, it may be
necessary to modify the procedure by increasing the number of discharge
classes or modifying the bed-material rating curve, noting the cautions to
be exercised in each case.
e. Check if the calculated effective discharge is reasonable. At the end of

the procedure, it is important to check that the effective discharge is a
reasonable value for the project reach. Suitable ranges of the channel
forming flow return period have been reported in the literature and
guidance is given in Chapter 3. A further check is to compare the
duration of the effective discharge with basin area-flow duration curves.
Finally, a morphological check should be undertaken to compare the
effective discharge to the bankfull discharge. This is best performed by
identifying the bankfull stage at a stable cross-section and calculating the
corresponding discharge either from the stage-discharge relationship at a
nearby gauging station, or using the slope-area method. The computa-
tional procedure for generating a bed-material load histogram, is
diagrammed in Figure 7.

z START
J
GH Calculate arithmetic mean
discharge within each class
interval
Sediment rating
curves
Calculate bed-material load #

Flow frequency
transported at each mean
distribution
discharge
For each mean discharge

multiply bed-material load
by discharge frequency
Bed-material load
function(s)
7
Bed-material load
histogram
3
If necessary, adjust the
number of classes to yield
a smooth sediment
frequency histogram with
a single peak frequency
that if possible is not in
the first class
Adjust number of classes
L
Recalculate flow
frequency distribution *0
Bankfull discharge —
Identify the mean discharge
corresponding to the peak
frequency
Effective
discharge, Qe
7
Basin area-flow duration Check effective discharge
curves is reasonable
Suitable return period —'

c END
J
Figure 7. Computational procedure for generating a bed-material load histogram

3 Evaluation and Problem
Solving
Problems with Lowest Discharge Class

When a significant proportion of the recorded discharges fall within the first
arithmetic class interval of the flow-duration curve, the range of discharges is
inadequately represented and it is likely that the computed effective discharge
will be significantly underestimated. This is likely the case for streams
displaying a highly skewed distribution of flow events, for example, rivers in
semiarid environments, channelized streams, or incised channels. Under these
circumstances, it is advisable to increase the number of class intervals to better
define the flow-duration curves.
Problem with Outliers

Discharge records, especially those based on mean daily values, can contain
distinct gaps in the higher discharge categories due to the averaging process.
This is likely on rivers with flashy hydrographs or if there has been an extremely
large flood event during a short period of record. The use of arithmetic discharge
class intervals can produce a discontinuous flow-frequency distribution which in
turn generates an irregular bed-material load histogram with outliers that reflect
the transport associated with individual flood events. The danger is that the
wrong peak may be selected to represent the modal class, leading to serious
overestimation of the effective discharge.
When there are only a few high discharges in the record it is possible to end
up with too many high discharge events in one class and no events in adjacent
classes. This results in unrealistic gaps in the bed-material load histogram. The
problem can usually be solved by increasing the class interval and reducing the
number of classes. This reduces the number of discharge classes with zero bed-
material load and removes the outliers, smoothing the histogram. For example,
on the White River upstream of Mud Mountain Dam, when 40 class intervals
were used, the effective discharge was indicated to be unrealistically high
(387 mV1). This problem was solved by reducing the number of classes to 25,
resulting in an effective discharge of 81 mV1, which corresponded to bankfull
flow (Hey 1997).
Chapter 3 Evaluation and Problem Solving 19

Checking Effective Discharge
Guidance on return periods for the effective discharge
The return period for the effective discharge is expected to vary between
sites depending on the flow and sediment-transport regime of the individual river
or reach. For sites where annual maximum series flood-flow data are available,
the return period of the calculated effective discharge may be checked to ensure
that it lies within acceptable bounds.
Unfortunately, there is very limited information available regarding the

return period of the effective discharge for stable rivers. Experience indicates that
it lies within the range 1.01 and 3 years with a preponderance between 1.01 and
1.2 years, irrespective of the type of river (Hey 1997). There is more information
on the return period of bankfull flow and, as this is equivalent to the effective
discharge for rivers that are in regime, it may be taken to define the general range
of likely return periods for the effective discharge. Despite the lack of consensus
regarding the definition of bankfull stage (Williams 1978), which can
significantly affect calculated return periods, available data indicate that
78 percent of bankfull flow return periods are in the range 1.01 to 3.0 years, with
a modal value in the 1.01- to 1.20-year category (Hey 1994). Predicted effective
discharge return periods outside the range of approximately 1- to 3-years should
be queried.
Drainage area - flow-duration curve
The percentage of time the effective discharge is equaled or exceeded should

be compared to the expected range of values reported in the literature.
For example, Figure 8 presents a log-log plot of the flow duration of

effective discharge as a function of drainage area for several U.S. rivers (Watson,
Dubler, and Abt 1997). The graph can be used to assess whether the duration of
the effective discharge computed using the method described in this manual is
comparable to the results of other studies. It is not intended that this graph be
used to predict effective discharge as a function of drainage area, as large errors
are likely to result from this application.
Check effective discharge against bankfull discharge
The effective discharge should be compared to the bankfull discharge. This

can be accomplished by identifying the bankfull stage during stream
reconnaissance of the project reach and calculating the corresponding discharge,
as outlined in Step e of "Bed-material load histogram."
20 Chapter 3 Evaluation and Problem Solving

100 ;,,„,. . „ , , ,, ,,„ . ., . ,.,,.,,,,, , ,,„«. . T ., -, , rt „
$5
-,
-•- ..
•o
:!:'"::"":" i"
I10 . . „ ..■.•....!.<■- -
•; ; -' - '-_'-' ';";"; - * •_' ;

. .' -
;"■*'::'::: v.'" :A
■
•v..-.-
"*'*''*"" * '" *' *

'Jk-:■•':-::: :•:• ......
. V .' . <* >- **
4 *.' *i
CB -. n ■•••'H|- •■ ■•■■ jg " - '^^JSk •! ','.'-.'.- ■ r -*. - x'Z'r

3
0*
IX!
e : ::»::::
... .,.,.,.,,.,.,,. . , -, . l .-. ,1, r - ,- -» t .-. f. .«• * " - ■ -. r. f.-- - r -, - * ."> <*
o
.-.-,,-.,*■ - .,
*s -m::'- '■> •-•'•" •: - ' •-•-..-• .-. ., ,, „,. - , -. -

CO * - " -
::::
3
D
:v:::!S:!::: Y:'.'.:'.' '':'.'.':'.:',':, ~ ~ - ;;-;;-;;;- - •; -;•; "' ':;■"" "';*!"
0.1 1 i (.„BWl 1-4-t-H+H) 1 1 1 1 HH|—t— i i mill—(-4-444+4+j t-H-H-H+t-
1E1 1E2 1E3 1E4 1E5 1E6 1E7

Drainage Area (sq. kilometers)
Gravei Sand
Figure 8. Effective discharge duration versus drainage area
Chapter 3 Evaluation and Problem Solving 21

4 Applications
Effective Discharge Calculation for the

Mississippi River at Vicksburg
Flow frequency distribution
Discharge data were obtained from the Vicksburg gauge for the period 1950
to 1982. This period of record was selected as it encompasses the period when
sediment loads were routinely measured at the gauging station. The record
contains a wide distribution of flows including both low and high runoff years
and with discharges ranging from about 4,200 to just over 56,600 mV. On this
large river, mean daily discharges do not differ significantly from instantaneous
discharges so the use of mean daily values was acceptable in the production of
the flow-frequency distribution (Figure 9).
SC-
54-
- *8-
I 42-
36-
30-
« 24-
18-
12-
10 20 30 40 50 60 70 80 90 100
Percent of time equalled or exceeded
Figure 9. Flow-duration curve for mean daily discharge: Lower Mississippi River
at Vicksburg, 1950-1982 (Biedenharn and Thorne 1994)
22 Chapter 4 Applications
Bed-material load rating curve
Sediment-transport data were also obtained from the Vicksburg gauging

station. The period of record was 1969 to 1979, as this was the only period for
which measured sediment-transport records were available. Typically, sediment
load was measured weekly. Robbins (1977) provides a detailed description of
the sediment measurement program on the Lower Mississippi. The period of
record includes both low runoff years and several events of high magnitude and
long duration, so that the full range of sediment transporting flows is represented
in the measured data.
The measured sediment loads were divided into two components: silt load
consisting of particles less than 0.062 mm, and sand load consisting of particles
coarser than 0.062 mm. The bed of the Lower Mississippi River is sand, so the
sand fraction of the measured load was taken to represent the bed-material load.
The silt load was taken to represent "wash load" and was excluded from the
analysis.
There are no measurements of bed load, but according to the calculations of

Toffaleti (1968) the bed load comprises less than 5 percent of the total sand load.
Hence, it was deemed to be acceptable to ignore the bed load and to take the
measured sand load as indicative of the bed-material load. The measured sand
load data were used to construct a sand load rating curve for the study site
(Figure 10).
Regression analysis of sand load as a function of discharge produced a

coefficient of determination (r2) of 0.82 and defined the bed-material load rating
curve as:
Qs = 0.00000513 QAAZ tonnes/day (1)
10000000•
>>
to
T5
fe 1000000'
CL
M
©
C 100000-s
C
o
10000
CO
o
T3
C 1000-
CO
100- i—i—r—i—IT] 1 1—i—r-n—r

1000 10000 100000
Discharge (cumecs)
Figure 10. Sand load rating curve. Lower Mississippi River at Vicksburg 1969-
1979 (Biedenharn and Thome 1994)
Chapter 4 Applications 23
Bed-material load histogram
The data in the flow-duration curve was divided into 50 equal classes ranging
from 5 to 55,000 m3/s and with a class width of 1,000 mVs. The bed-material
transport rate for each discharge class (Qs) was found from equation (1), with
Q equal to the arithmetic mean discharge for that class. The quantity of bed-
material load (in tonnes) transported by each discharge class was calculated by
multiplying the frequency of each class (in days) by the bed-material transport
rate for the average discharge (in tonnes per day). The resultant histogram is
plotted in Figure. 11.
SO
45
40
or
o
£ 35
o
o 30
o
£25
■o
§ 20
_i
■g 15
ID
ir
<" 10
nil IIIIIII IIIIIIIII mnvn
5
0
iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiH..
10 15 20 25 30 35 40 45 50 55
Discharge (1,000 cumecs)
Figure 11. Bed-material load histogram. Lower Mississippi River at Vicksburg (Biedenham and
Thorne 1994)
Effective discharge determination
The peak of the histogram in Figure 11 is defined by the mean discharge of

the modal class, which is 30,000 nrVs. This defines the effective discharge.
Check if effective discharge is reasonable
In the Biedenharn and Thorne study, the effective discharge calculation was
also performed for gauging stations at Arkansas City, AR (upstream of
Vicksburg) and Natchez, MS (downstream of Vicksburg). No major tributaries
enter the Mississippi River between these stations. Hence, it would be expected
that the effective discharge should be the same at all three sites. This was in fact
the case, illustrating consistency in the effective discharge analysis using three
separate flow-duration and sediment-transport records.
Comparison of the water surface profile at the effective discharge
(30,000 mV) to the long-channel distribution of bank top elevations is
illustrated in Figure 12. The graph shows that bank top elevations are highly
variable and can differ by 3 m or more between adjacent cross sections. This
makes it difficult to assign a value to bankfull discharge for the reach. However,
comparison of the water surface profile for the effective discharge to the bank top
data indicates that the effective discharge forms a good lower bound to the
scatter, indicating that the capacity of the channel is adjusted just to contain flows
up to and including the effective flow. As discharge increases beyond the
effective flow, water begins to spill over the bank tops at more and more
locations.
The return period for the effective discharge (equal to or just less than one
year) is consistent with the ranges given in "Guidance on return periods for the
effective discharge," and its flow duration (equalled or exceeded on 13 percent of
the time) is as expected for a river with a drainage area of approximately
3 million km2 (Figure 8).
40-
ed
OQ.'
o
g
30-
a ..•**
a cP .-•■"'
erb .UP
a
B jaw
&
so-
tfa^-d■<i
m Top bank
30,000 cumecs
P-
°jjP
lo- —I , 1—I1 ) i • r" I1 i |
soo 600 700 800 300 1000
Distance above me head of the passes (kilometres)
Figure 12. Long-channel variation in bank top elevations: Lower Mississippi in

study reach (Biedenharn and Thome 1994)
These checks indicate the calculated effective discharge is reasonable and

support the accuracy of the analysis including the necessary assumptions
concerning the use of measured sand load to represent the bed-material load.
Channel Management: Case Study of the River
Blackwater, UK
The need for geomorphological studies to support sustainable engineering
and management of rivers is now generally accepted (Gardiner 1991; Downs and
Thorne 1996). Calculation of the channel- forming discharge and application of
the principles of hydraulic geometry analysis can be useful in developing a sound
understanding of the stability of an alluvial stream. The utility of this approach
can be illustrated by brief reference to a case study of the River Blackwater,
England. The study is reported in detail in a report by Hydraulic Research,
Wallingford (1992) and in summary in a paper by Thorne, Simon, and Allen
(1996).
The River Blackwater is a lowland stream in southeast England. In the

1960s and 1970s the channel was modified by engineering works installed as part
of a series of flood defense and land drainage schemes. Subsequently, the
channel required heavy maintenance to maintain a sufficient capacity for flood
flows. More recently, steps have been undertaken to improve stream habitats by
softening the impacts of engineering works, modifying the maintenance regime,
and enhancing the channel environment.
As part of plans to restore the environmental function of the river, a

geomorphological study was performed to establish how the existing, engineered
channel differed morphologically from a natural, regime channel. An effective
discharge calculation was performed using data from a nearby gauging station
and the "expected" morphology for a natural, regime channel was established
by applying the hydraulic geometry equations of Hey and Thorne (1986). Stream
reconnaissance was performed to establish the morphology and bankfull
dimensions of the existing, engineered channel using the method reported by
Thorne (1998).
The effective discharge calculation showed that the channel-forming

discharge was 3.65 mV1, compared to a bankfull capacity observed in the
engineered channel of 16.4 mV'. The main morphological parameters calculated
and observed were:
Regime Channel Engineered Channel

width 6.3 m 12 m
depth 0.59 m 1 m
2
x-section area 3.7 m 12 m2
mean velocity 1.0 m/sec 0.3 m/sec
These contrasts between the regime and engineered channels were used to
support the conclusion that the engineered channel was over-large in width, depth
and area and that in-channel velocities were insufficient to transport the sediment
load supplied from upstream. This explained its tendency for siltation and need
for frequent maintenance.
On the basis of the geomorphological assessment, initial recommendations
were proposed for morphological restoration of the channel to support the
enhanced river environment. It was further proposed that the viability of these
initial recommendations should be examined further to determine their feasibility
for a restoration project.
The effective discharge was found in this application to be the key to

deriving values for the dimensions of a natural regime channel appropriate to the
current flow regime. This information was useful in highlighting the problems
with the engineered channel and indicated the starting point for detailed design of
a restored channel.
Channel Restoration Hydraulic Design
Introduction
The U.S. Army Engineer Research and Development Center is currently

developing a systematic methodology for design of channel restoration projects.
This methodology will be applicable to (but not limited to) channels that have a
flood control component.
When hydrologic and sediment conditions are steady and the existing
channel is stable, the existing channel wavelength and sinuousity should be
maintained in any channel restoration scheme. The proposed methodology is
intended for cases where a historically stable channel has been realigned creating
instability, or where hydrologic and/or sediment inflow conditions have changed
so much that the channel is currently unstable. Stability is defined as the ability
to pass the incoming sediment load without significant degradation or
aggradation. Bank erosion and bankline migration are natural processes and may
continue in a stable channel. When bank line migration is deemed unacceptable,
engineering solutions must be employed to prevent bank erosion. Bank
protection technology is addressed by Biedenham et al. (1997).
Hydraulic design methodology
Identify objectives. The first step in the methodology, as with any

engineering project, is to clearly define project objectives. Is the objective to
create an aesthetic setting, to create a natural setting, to enhance fish or wildlife
habitat, to prevent bank erosion or channel degradation, or is it something else?
Although these objectives could all conceivably be considered channel
restoration, they are not necessarily compatible.
Geomorphic analysis. The hydraulic design of a channel restoration project

should be preceded by a study of the watershed in order to determine the
relationship of the stream to the rest of the hydrologic and geomorphologic
system. If the stream is unstable, trends and stages of geomorphic evolution
should be identified. A determination of the existing and desired ecosystem is
also a necessary component of a channel restoration project.
Determine hydrologic regime. To provide long-term stability, a historical

hydrograph or a flow-duration curve is required. These may be obtained from
stream gauges or from regional analysis. A channel-forming discharge can be
developed from this data. Current research from the Flood Damage Reduction
Research Program at ERDC suggests that the effective discharge can be used to
estimate the channel-forming discharge. The effective discharge is the increment
of discharge that transports the most sediment on an annual basis and can be
determined by integrating a sediment-transport rating curve with the annual flow-
duration curve. This calculation requires a knowledge of the flow-duration
characteristics, bed-material size distribution, and a sediment rating curve (either
measured, calculated, or a combination thereof)- It is important to attempt to
verify this channel-forming discharge with field indicators of bankfull discharge.
This may be difficult or impossible in an incised stream. This verification step
may be complicated because it requires an estimate of channel roughness unless
measured stage-discharge data are available. It is also prudent to check the
recurrence interval of the effective discharge. In most cases, it should be
between the one- and three-year peak frequency events.
Determine the design width of the channel. When channel width is not
constrained by right-of-way limitations, the preferred method for determining the
width is to use geomorphic principles. Several techniques are available for
determining the width of a stable alluvial stream. In order of preference they are:
a. Analogy method A width can be determined for the project stream by

assigning a measured average width from a reference reach. The refer-
ence reach must be stable and have the same channel-forming discharge
as the project reach. The reference reach may be in the project reach
itself, upstream and/or downstream from the project reach, or in a
physiographically similar watershed. Streambanks in the project and
reference reaches must be composed of similar material, and there should
be no significant hydrologic, hydraulic, or sediment differences in the
reaches. This technique is inappropriate for streams where the reference
reaches are in disequilibrium.
b. Hydraulic geometry method. Hydraulic geometry theory is based on the

concept that a river system tends to develop in a predictable way, pro-
ducing an approximate equilibrium between the channel and the inflow-
ing water and sediment (Leopold and Maddock 1953). The theory
typically relates the dependant variable, width, to an independent or
driving variable, such as drainage area or discharge. Hydraulic geometry
relationships have also been applied to other dependent variables such as
depth, slope, and velocity. Hydraulic geometry relations are sometimes
stratified according to bed-material size or other factors.
Hydraulic geometry relations can be developed for a specific river,

watershed, or for streams with similar physiographic characteristics.
Data scatter is expected about the developed curves even in the same
river reach. The more dissimilar the stream and watershed characteris-
tics are, the greater the expected data scatter is. So-called "regional
curves" would be expected to have a wide band of scatter. It is important
to recognize that this scatter represents a valid range of stable channel
configurations due to variables such as geology, vegetation, land use,
sediment load and gradation, and runoff characteristics. The composi-
tion of the bank is important in the determination of a stable channel
width. It has been shown that the percentage of cohesives in the bank
and the amount of vegetation on the bank significantly affect the stable
channel width.
c. Analytical methods. If a reliable width vs. effective discharge relationship

cannot be determined from field data or in cases where there is signifi-
cant sediment transport, analytical methods may be employed to obtain a
range of feasible solutions. These are discussed in the next section.
General guidance is available in U.S. Army Engineer Manual EM-1110-2-

1418 (1994), and Hey and Thorne (1986). Currently under development at
ERDC are hydraulic geometry predictors for various stream types with different
bank characteristics. These predictors include confidence limits and may be used
for general guidance when site specific data cannot be obtained.
Calculate a stable channel slope and depth. Depth and slope should be
calculated using analytical techniques. Analytical techniques are more reliable
than hydraulic geometry relationships for establishing the relationships between
channel-forming discharge and depth, slope or velocity. Analytical approaches
calculate the design variables of width, slope, and depth from the independent
variables of discharge, sediment inflow, and bed-material composition. Three
equations are required for a unique solution of the three dependent variables.
Flow resistance and sediment transport equations are readily available. Several
investigators propose using the extremal hypothesis as the third equation (Chang
1980; White, Beltess, and Paris 1982; Millar and Quick 1993). However,
extensive field experience demonstrates that channels can be stable with widths,
depths, and slopes different from extremal conditions. Therefore, others propose
using a hydraulic geometry width predictor as the third equation. The stable-
channel analytical method (Copeland 1994) in the U.S. Army Engineer hydraulic
design package SAM may be used to determine a depth and slope for the width
selected in the previous section. This method is based on a typical trapezoidal
cross section and assumes steady uniform flow. The method is especially
applicable to small streams because it accounts for sediment transport, bed form
and grain roughness, and bank roughness. This method uses the Brownlie (1981)
sediment transport and roughness equations for sand bed streams and the Meyer-
Peter and Müller (1948) sediment transport equation with the Limerinos (1970)
bed resistance equation for gravel bed streams. This procedure assumes a fully
mobile sand bed.
In coarse bed streams, where sediment transport is small, or in streams with

bedrock outcrops or with cohesive beds, threshold design methods (e.g. Lane
1955; USDA 1977) may be used to calculate depth and slope. However, in sand
bed streams, sediment transport is typically significant and an analytical
procedure that considers both sediment transport and bed form roughness is
required.
Determine a stable channel planform. This step involves determining a

meander wavelength, an appropriate channel length for one meander wavelength,
and then laying out a planform. Meander wavelength, can be determined using
hydraulic geometry techniques. The most reliable hydraulic geometry
relationship is wave length vs. width. As with the determination of channel
width, preference is given to wavelength predictors from stable reaches of the
existing stream either in the project reach or in reference reaches. Lacking data
from the existing stream, general guidance is available from several literature
sources (e.g., Leopold, Wolman and Miller 1964). The channel meander length is
simply the meander wavelength times the valley slope divided by the channel
slope.
One way to lay out the planform is to cut a string to the appropriate length
and lay it out on a map. Another, more analytical approach, is to assume a sine-
generated curve for the planform shape as suggested by Langbein and Leopold
(1966) and calculate x-y coordinates for the planform. This rather tedious
numeric integration can be accomplished using a computer program such as the
one in the SAM hydraulic design package. The sine-generated curve produces a
uniform meander pattern. A combination of the string layout method and the
analytical approach would produce a more natural looking planform.
In streams that are essentially straight (sinuosity less than 1.2) riffle and pool
spacing may be set as a function of channel width. The empirical guide of 6-10
channel widths applies here, with the lower end for steeper channels and the
higher end for flatter channels. Two times this riffle spacing gives the total
channel length through one meander pattern.
Conduct a sediment impact assessment. The purpose of the sediment

impact assessment is to assess the long-term stability of the restored reach in
terms of aggradation and/or degradation and determine the magnitude of future
maintenance problems. This can be accomplished using a sediment budget
approach for relatively simple projects or by using a numerical model which
incorporates solution of the sediment continuity equation for more complex
projects. With a sediment budget analysis, average annual sediment yield with
the design channel is compared to the average annual sediment yield of the
existing channel, if the existing channel is stable, or of the upstream supply
reach, if the existing channel is unstable. Large differences in calculated
sediment yield indicate channel instability. This step is especially important if
the restored reach is part of a flood damage reduction project. In such cases it
may be necessary to design a channel that is less than ideal in terms of channel
stability in order to achieve flood control benefits. Typically, a compound
channel design provides the best combination of benefits.
The most reliable way to determine the long-term effects of changes in a

complex mobile bed channel system is to use a numerical model such as HEC-6.
River systems are governed by complicated dependency relationships, where
changing one significant geometric feature or boundary condition affects other
geometric features and flow characteristics both temporally and spatially.
Changes at any given location in a stream system are directly related to the
inflow of sediment from upstream. This makes the application of the sediment
continuity equation essential to any detailed analysis. The most significant of
these relationships and the continuity of sediment mass are accounted for in the
numerical model approach. The fact that application of a numerical sediment
model requires knowledge of sediment transport and river mechanics should not
be a deterrent to its use, and that knowledge is required for any responsible
design work in a river system. It should be expected that an analysis of system
response in a complicated system, such as a mobile bed river system, will require
some engineering effort. That effort should be based on analysis of the physical
laws that govern the system. The system cannot be expected to adhere to
constraints placed on it in violation of natural physical laws, no matter how well
intentioned or frugally those constraints were developed. The critical decision
with respect to using a numerical model should be based on whether or not
"significant" changes are expected to occur in the system as a result of the
proposed design work. In the U.S. Army Corps of Engineers, this decision
typically is reached in the reconnaissance level planning study using the sediment
impact assessment approach.
Effective Discharge Calculation in Support of a

Channel Restoration Evaluation: Case Study
Introduction
Modifications to a river system that are inconsistent with natural processes

can lead to a series of complex responses throughout the system. If the imposed
channel dimensions or environmental features are not commensurate with the
position of the restored reach in the fluvial system, the channel may adjust to a
more stable form. Natural rivers are never in a state of continuous equilibrium
but a variable, shifting equilibrium, or quasi-equilibrium (Langbein and Leopold
1964). Subsequently, channel forms constantly adjust around a medium to long-
term average condition. In some cases insensitive river engineering can act as a
catalyst within dynamic, metastable equilibrium (Schumm 1975) to carry the
system over a geomorphic threshold into a new equilibrium regime. The
enhanced channel design procedure presented here uses confidence bands to
derive stable channel dimensions within the range found in nature. Using this
principle, the river is "directed" towards a stable channel configuration, suitable
for the target "type" of channel, but is encouraged to participate in its own
recovery. Designing outside the stable limits may impose an unstable condition
that can lead to evolution towards a new metastable equilibrium regime.
A case study is presented illustrating calculation of the effective discharge

and also demonstrating a technique adopted to validate restored channel
geometry in terms of medium to long-term equilibrium sediment conveyance by
comparing the range of sediment-transporting flows in the supply reach with that
of the restored channel.
Restoration of Willful Creek
The headwaters of Willful Creek are located in the Piedmont zone of the
eastern United States and the majority of the system is found in the Coastal Plain.
The drainage area upstream of a gauging station located immediately
downstream from the project reach, is 12.25 km2. There is only one significant
tributary upstream from the project reach, which is gauged near the confluence
and contributes considerably to the sediment budget. The drainage system is
characterized by a high sediment load of sandy-gravel material pulsed through
the system by a flashy flow regime and a relatively steep energy gradient.
Although field reconnaissance revealed that many streams in this coastal

zone appear to be approaching a new state of equilibrium following the period of
urbanization, a number of sites have been targeted for restoration to advance this
recovery process and recreate stable channel configurations without
compromising the designated level of flood protection. A 1.5 km reach of
Willful Creek was restored in September 1996. The prerestoration channel
morphology was characterized by low sinuosity and poor depth diversity as a
result of previous channelization works.
The objectives of the channel design were to: recreate the diverse structure
and function of a meandering channel to a river system of relatively low
sinuosity; protect the bank lines from erosion; improve the aesthetic quality and
amenity value of the stream within an urbanized watershed and maintain the
present level of flood protection with embankments. The restored channel was
designed to be static-stable, that is minimizing aggradation and degradation,
while inhibiting the migrating tendency of a natural meandering river by
protecting the bank lines from erosion. The restoration design, included:
a. A low flow channel with wide, shallow point bars within rock-lined
embankments designed to contain the 50-year recurrence interval flood.
b. Increased sinuosity from approximately 1.05-1.1 to 1.5-1.7. This led to a

decline in slope from approximately 0.0038 to 0.0025.
c. Decreasing meander wavelength from approximately 370 to 90 metres.
d. Restructuring in-stream morphology with asymmetrical cross sections

around bendways and uniform cross sections at meander crossings.
e. Low stage rock vortex weirs at meander crossings to control the grade.
/ Root wads and riprap around bendways to prevent bank erosion and
lateral shift in planform.
g. Willow planting to stabilize the wide, shallow point bars.
Observations both upstream and downstream from the restored reach

revealed a considerably lower degree of sinuosity than that restored with an
almost straight alignment and occasional alternate bar features. Therefore, the
restoration involved meander creation or enhancement rather than reinstatement.
Post-project channel change
Sediment impacts. Post-project reconnaissance revealed significant

planform and cross-sectional channel change. In particular the stream has
reduced its sinuosity from that constructed and it has experienced considerable
sedimentation in the bendways. Cross-sectional surveys were undertaken at both
a representative bend apex and crossing within the restored channel. Consider-
able aggradation, at both the bend apex and crossing sections, has occurred
during the post-project period, between September 1996 and November 1998
(Figure 13). The morphological change is summarized in Table 1. The low flow
channel has straightened by eroding through the unprotected sandy point bars.
Post-project channel change can be summarized as: significant aggradation at
bends and crossing; reduced sinuosity and channel steepening, via redistribution
of sediments, and; isolation of the designed meander bendways as pools are filled
and the thalweg adopts a straighter alignment.
Table 1
Post-Project Channel Change at a Representative Bend Apex and
Crossing (specific gravity of bed-material is 2.65, density of water
is 1,000 kg m3)
Cross Section Bend Apex Crossing
Volume change (mV1 per unit length) 5.01 2.50
Aggradation (tonnes per unit length) 13.29 6.63

1
Aggradation rate (tonnes yr" per unit length) 6.05 3.01
Recovery mechanisms. In summary, the energy of the restored channel

was not sufficient to transmit the magnitude of sediment transported from
upstream. The restored reach was not designed as a component within the
broader river system and as a result the reach has acted as a sediment bottleneck
which prompted a complex series of channel changes following project
implementation. The degree of post project channel change may also have been
amplified by the fact that bankside vegetation probably did not have time to
establish before widening and straightening began. Although the restored
channel was intended to be static-stable, the shallow, wide point bars were not
stabilized sufficiently to withstand the erosive force of high flow events.
I12]
03 10 ■' ^^^.
"O
above
CD
r—
CO
c v,>
atio
f Aggradation -*- 11 November 1998

Y Migration
ä> 6
: ► '■ -«-20 March 1995 (restored design)
LJJ
C) 5 10 15 20 25 30 35 40 45 50
Distance (from left bank) (m)
I12]
11
E i
To 10 ■'
(above
CD
CO
"*s»^-^-*—"—~3*T^
c 7
: •—*—• •
.2 : Aggradation -•-11 November 1998
To
5 6: -•-20 March 1995 (restored design)
Lu :
3 5 10 15 20 25 30 35 40 45 50
Distance (from left bank) (m)
Figure 13. Design cross sections and post-project channel change: i) above: bend apex;
ii) below: meander crossing (200 percent vertical exaggeration)
Channel restoration design procedure
Channel restoration design is the reconstruction of a river channel to a geo-

metric configuration which is self-sustaining and in balance with imposed flow
and sediment regimes and the character of the catchment landscape (Soar et al.
1998). Techniques to design restored channels are often based on a combination
of field experience, reference reach observations and basinwide regime-type
curves. While these approaches can yield appropriate target channel dimensions
in some cases, they do not explicitly account for upstream sediment supply into
the restored channel and the capacity of the restored reach to transmit sediment to
the channel downstream. If sediment transport continuity through the restored
reach is not achieved, it is unlikely that restored channel dimensions will be
stable.
An enhanced channel design procedure currently in development (Soar et al.

1998; Copeland and Hall 1998) assesses sediment transport using a one-
dimensional analytical method (Copeland 1994) that simultaneously solves flow
resistance, sediment transport and hydraulic geometry equations at the effective
discharge within designated confidence limits. Two important stages of the
design procedure are the calculation of the effective discharge, which is derived
from magnitude-frequency analysis in the supply reach, and a sediment budget
analysis to derive a Capacity-Supply Ratio (CSR) and verify whether the restored
channel dimensions will minimize aggradation and degradation over the medium
to long-term. The SCR may be used as a basis for refining the design slope to
ensure reach-average dynamic stability and can help direct post-project
maintenance of short-term channel change.
Supply reach magnitude-frequency analysis
A supply reach analysis is undertaken to describe the distribution of

sediment-transporting flows and calculate the effective discharge, which are
input to the restored channel from upstream. The effective discharge is used as
the channel-forming discharge in the design procedure. Assessment of the sedi-
ment load entering the restored channel is necessary to ensure that channel stabil-
ity is maintained within the restored reach, in terms of near zero net aggradation
or degradation over the medium to long-term. Magnitude-frequency analysis
requires a record of flow events and a sediment-rating relationship to generate a
histogram of sediment-transporting flows for the reach of interest. If sediment
transport data are unavailable, a sediment frequency histogram can be derived
using a representative cross section and appropriate sediment transport equations.
Although Willful Creek is gauged immediately downstream from the

restored channel, reconnaissance upstream of the restored reach failed to identify
a suitable stable site in the immediate supply reach. Hence, a modified procedure
was used to synthesize a sediment discharge histogram and effective discharge
for the restored reach from available flow data and surveyed cross sections
upstream of the tributary confluence. In the absence of measured load, (Meyer-
Peter and Müller 1948) the bed-load equation was adopted to derive a sediment
frequency histogram and determine the effective discharge. The Limerinos
equation (Limerinos 1970) was used to characterize the roughness of the bed and
generally yielded a composite Manning n-value of approximately 0.030 for the
bed portion of the channel. The channel banks were assigned a Manning n-value
of 0.085, which is suitable for moderately dense scrub with some trees. Initially,
a flow-frequency histogram was derived using 25 class intervals ranging from the
critical discharge for sediment transport to the maximum-recorded discharge. As
the effective discharge could not be readily defined, the number of discharge
classes was incrementally increased until a definitive peak on the sediment
frequency histogram, giving an effective discharge 29.8 mV1 with 30 arithmetic
discharge classes (Figure 14).
Simulated channel restoration design
The enhanced channel restoration design procedure was applied to calculate

depth and slope at the effective discharge using the Meyer-Peter and Müller
(1948) bed load equation and the Limerinos (1970) flow resistance equation, both
suitable for gravel-bed rivers. Confidence bands for stable bankfull width were
derived from modified Hey and Thorne (1986) hydraulic geometry equations
suitable for streams with well developed riparian vegetation and greater than
20 5000
D Discharge Frequency
• 4500 v
OMain Channel Bed Load Yield
□ Tributary Bed Load Yield
12
2000 m
1000
2- 500
Hh R R n n n
\ ""- V to <oto<V<oc>><V<bO)nDto.Ä>n3^
fo K
S' -?' <£' # s riy ,?' ^'S1' .0)' A' to' V *V -' P)' A' Jo' .>' v'V
-9 -9 «P to 5 ^
<V <o °>
■V' lS>' t?'
V <nr v a/ cv o/ V ^y" V to" 9)" °D A Ä ^3
.-. .^ -^ KT -^ -
Mean Daily Discharge Range (m3s
Figure 14. Supply Reach Magnitude-Frequency Analysis derived from stable

reaches in the tributary and main channel upstream of the tributary
confluence
50 percent tree/shrub cover. The simulated design assumes moderately rough

banks with a Manning n-value of 0.085 and side slopes of 1.0.
Figure 15 presents the results of the analytical design in terms of a stable

width-slope curve (primary axis) and stable width-depth curve (secondary axis).
Suitable design solutions are located within the 95 percent confidence band. The
chart shows that the restored design slope of 0.0025 underestimates the simulated
range of slopes within the confidence band. This indicates that aggradation would
be expected in the restored channel. The valley slope through the restored reach
is estimated as 0.0043, giving the constructed restoration design a sinuosity of
approximately 1.7, which is significantly greater than that calculated using the
enhanced design procedure. This may explain the observed channel response
through straightening. Table 2 presents a summary of the restored and simulated
design parameters.
Sediment budget assessment: Capacity-Supply Ratio (CSR)
The potential success of a river project is often defined in terms of

performance based on a single flow event and the sediment load transported by
this event. This approach does not account for the potential for instability driven
by other flow events in the long-term record. The potential for restoring sediment
continuity through the restored reach requires an assessment of the sediment
budget, which may be determined by the magnitude and frequency of all
sediment-transporting flows. To attain geomorphic stability in the medium to
long-term, the mean annual sediment load for the restored channel (capacity)
must match the mean annual sediment load in the supply reach (supply).
0.0050 T—i 1 1
95% Confidence Limits

Stable Slope Curve
0.0045 • •2.5
Restored Design Slope
Stable Bankfiill Depth Curve ^^
0.0040 ■ ^^^ •2
Bankfiill Depth (m)

^^^
o 0.0035 -
en
CO
^__^^
0.0030 -
0.0020 -
10 20 30 40 50 60 70 80
Bankfiill Width (m)
Figure 15. Stable width-depth-slope diagram. The range of stable slope

(primary axis) and depth (secondary axis) are shown between
95 percent confidence intervals of the mean response
Table 2
Comparison of the Restored Channel Design with the Simulated
Design from the Enhanced Design Procedure
Design Width (m) Mean Depth (m) Slope Sinuosity
Restoration 24.5 0.90 0.0025 1.7
Simulation 20.43 1.01 0.00322 1.29
(19.17 to 21.82) (0.96 to 1.20) (0.00315 to 0.00350) (1.26 to 1.40)
Note: Width and depth refer to bankfull reference level. Values in parentheses are within
95 percent confidence limits of the mean response.
A Capacity-Supply Ratio (CSR), may be calculated as the bed-material load

transported through the restored reach by the natural sequence of flow events
over an extended time period divided by the bed-material load transported into
the restored reach by the same flow events over the same time period. These
loads are calculated by a numerical integration of a sediment-transport rating
curve and the flow duration curve. A successful project design has a CSR close
to unity. Values greater than 1.0 indicate potential degradation and values below
1.0 indicate potential aggradation. The CSR is used at the end of the design
procedure as a closure loop to: validate the efficacy of the restored channel
geometry; identify flows which may cause aggradation or degradation over the
short term, and; recommend minor adjustments to the channel design to ensure
that dynamic stability will be ensured over the medium to long-term. The CSR
can be used to refine the design width, depth and sinuosity (and hence slope) to
bring the CSR closer to unity and improve stability.
Achieving an optimum CSR, within 10 percent of unity, should ensure
dynamic stability while allowing the river itself to recover some of the fluvial
detail that cannot be designed by the engineer. It may be possible to adjust the
design parameters within the designated level of confidence to achieve an
optimum CSR, but if this cannot be achieved, the slope should be delicately
increased or decreased appropriately until the CSR is in the optimum range.
Refining the design slope will not affect the effective discharge in the restored
channel. The CSR was used in this study to determine the success of the actual
restored channel and that simulated from the enhanced design procedure.
Comparison of sediment capacity and supply for the actual restored design
revealed a CSR of 0.64: that is, the restored channel has the potential for
approximately 36 percent of the input load to be deposited in the restored reach
over the medium to long-term. This result is consistent with the observed
aggradation in the restored meander bends. The CSR for the initial simulated
design is 0.90, using mean values of width, depth and slope within the 95 percent
confidence band. By slightly increasing the slope from 0.00307 to 0.00324, the
CSR can be increased to unity.
The sediment capacity-supply analysis reveals that the restored alignment

was too sinuous for the prevailing flow regime and sediment supply. A dynami-
cally stable channel requires a lower sinuosity (approximately 1.33), which is
similar to that observed upstream. Notably, the CSR is derived from a simple
one-dimensional technique and is based on the total bed-material load trans-
ported. Disparities between capacity and supply within the sediment-frequency
histogram of Figure 16 show a potential for some minor channel change in the
short term.
6000
D Supply Bed Load
5000 - CD Capacity Bed Load
4000
o 3000
2000
2 1000
mfflmflllillU m rfl rfl rFl
£>■ 5»- £• $ & & & <§>• A' & & g g & £• S ^ g » <§>' £• ^ ^' & g & & #• *' &
V ^K > , K (T* ^(? *■? -S ^ ■? * "S V« ^* ^ * ,0 V« * .» * * -?,■?
<V <¥ # ® & £■
*£>&&£>&&>& A <g> ^ £,• £■ o, • 5, # ^ ■ £? & g- A- g- #■ A- - A' A' A • >•
<ff ^ 1 J
*Zr t-. K. K7 K^ »O tvT 1^ (y C\.
rtT (\ C\.
<v C\7
<V C\T
V (YT
'v
Mean Daily Discharge Range (m s )
Figure 16. Comparison of sediment supply and capacity for the enhanced
design based on 30 arithmetic discharge classes and increased slope
of 0.00324, CSR is 1.00. The minimum discharge is the critical
discharge for sediment transport in the supply reach
References
Andrews, E. D. (1980). "Effective and bankfull discharge of streams in the

Yampa basin, western Wyoming," Journal ofHydrology 46, 311-330.
Andrews, E. D. (1984). "Bed-material entrainment and hydraulic geometry of

gravel-bed rivers in Colorado," Geological Society ofAmerica Bulletin 95,
371-378.
Bagnold, R. A. (1966). "An approach to the sediment transport problem from

general physics," Professional Paper 422-1, U.S. Geological Survey,
Washington, DC.
Biedenharn, D. S., Little, C. D., and Thorne, C. R. (1987). "Magnitude and

frequency analysis of large rivers," Hydraulic Engineering, Proceedings of
the National Conference on Hydraulic Engineering, American Society of
Civil Engineers, R. M. Ragan, ed., Williamsburg, 782-787.
Biedenharn, D. S., and Thorne, C. R. (1994). "Magnitude-frequency analysis of

sediment transport in the Lower Mississippi River," Regulated Rivers 9, 237-
251.
Biedenharn, D. S., Derrick, D.L., Elliott, C, and Watson, C. C. (1997). "The

WES Stream Investigation and Streambank Stabilization Handbook,"
U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
Blench, T. (1957). Regime Behaviour of'Canals and Rivers. Butterworths

Scientific Publications, London.
Brownlie, W. R. (1981). "Prediction of flow depth and sediment discharge in

open channels," Report No. KH-R-43A, W. M. Keck Laboratory, California
Institute of Technology, Pasadena, CA.
Brush, L. M. (1961). "Drainage basins, channels and flow characteristics of

selected streams in central Pennsylvania," Professional Paper 282F,
United States Geological Survey, Washington, DC, 145-181.
Burkham, D. E. (1972). "Channel changes of the Gila river in Safford Valley,

Arizona, 1846-1970," Professional Paper 655G, U.S. Geological Survey,
Washington, DC.
References 39
Carling, P. A. (1988). "The concept of dominant discharge applied to two
gravel-bed streams in relation to channel stability thresholds/'Zsc/r//? Surface
Processes andLandforms 13, 355-367.
Carlston, C. W. (1965). "The relation of free meander geometry to stream

discharge and its geomorphic implications,"^wer/cfl« Journal of Science
263, 864-885.
Carlston, C. W. (1969). "Downstream variations in the hydraulic geometry of

streams: special emphasis on mean velocity," American Journal ofScience,
267, 409-509.
Chang, H. H. (1980). "Stable alluvial canal design," Journal of the Hydraulics

Division, ASCE 106(HY5), 873-891.
Charlton, F. W., Brown, P. M., and Benson, R. W. (1978). "The hydraulic

geometry of some gravel rivers in Britain," Report, IT 180, Hydraulic
Research Station, Wallingford.
Copeland, R. R. (1994). "Application of channel stability methods - Case

studies," Technical Report HL-94-11, Sept, 1994, U.S. Army Engineer
Waterways Experiment Station, Vicksburg, MS.
Copeland, R. R., and Hall, B. R. (1998). "Channel restoration hydraulic design

procedure," Proceedings of the Wetlands Engineering and River Restoration
Conference 1998. Denver, CO, March 22-26, Session G2, ASCE.
Copeland, R. R., McComas, D. N., Raphelt, N. K., and Thomas, W. A. (1997).

"User's manual for the hydraulic design package for channels (SAM),"
U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
Downs, P. W., and Thorne, C. R. (1996). "The utility and justification of river
reconnaissance surveys," Transactions of the Institute of British
Geographers. New Series, 21, 455-468. ISSN 0020-2754.
Dury, G. H. (1964). "Principles of underfit streams," Professional Paper 452A,

U.S. Geological Survey, Washington, DC.
Dury, G. H. (1973). "Magnitude-frequency analysis and channel morphology."

Fluvial geomorphology: A proceedings volume of the fourth annual
geomorphology symposia series. M. Morisawa, ed., Allen and Unwin,
Binghampton, New York, 91-121.
Emmett, W. W. (1975). "The channels and waters of the upper Salmon River,
Idaho," Professional Paper 870A, U.S. Geological Survey, Washington, DC.
Ferguson, R. I. (1973). "Channel pattern and sediment type," Area, 5, 523-537.
Gardiner, J. L. (1991). River projects and conservation: A manual for holistic

appraisal. Wiley, Chichester, UK.
40 References
Gupta, A., and Fox, H. (1974). "Effects of high magnitude floods on channel
form: a case study in Maryland Piedmont," Water Resources Research 10(3),
499-509.
H. R. Wallingford. (1992). "Standards of service for flood defence: reach

specification methodology," Report to Thames Region, National Rivers
Authority. D. Ramsbottom Principal Investigator, Hydraulics Research,
Wallingford, UK.
Hack, J. T., and Goodlett, J. C. (1960). "Geomorphology and forest ecology of a

mountain region in the Central Appalachian," Professional Paper 347,
Hey, R. D. (1975). "Design discharge for natural channels," Science,

Technology and Environmental Management. R. D. Hey and T. D. Davies,
eds., Saxon House, 73-88.
Hey, R. D. (1994). "Channel response and channel forming discharge: literature

review and interpretation," First Interim Report, U.S. Army Contract Number
R&D6871-EN-01.
Hey, R. D. (1997). "Channel response and channel forming discharge: literature

review and interpretation," Final Report, U.S. Army Contract Number R&D
6871-EN-01.
Hey, R. D., and Thome, C. R. (1986). "Stable channels with mobile gravel
beds," Journal ofHydraulic Engineering 112(8), 671 -689.
Inglis, C. C. (1941). "Meandering of rivers," Central Board of Irrigation (India)

Publication 24, 98-99.
Inglis, C. C. (1947). "Meanders and their bearing on river training," Institution

of Civil Engineers, Maritime and Waterways, Paper No. 7.
Inglis, C. C. (1949). "The behaviour and control of rivers and canals," Research
Publication 13, Central Waterpower Irrigation and Navigation Research
Station, Poona, 79-91.
Kellerhals, R. (1967). "Stable channels with gravel-paved beds," Journal of the

Waterways and Harbours Division, Proceedings of the American Society of
Civil Engineers, 93, 63-84.
Kennedy, R G. (1895). "The prevention of silting in irrigation canals,"

Proceedings of the Institution of Civil Engineers 119, 281-290.
Langbein, W. B., and Leopold, L. B. (1964). "Quasi-equilibrium states in

channel morphology," American Journal of Science 262, 782-794.
References 41
Langbein, W. B., and Leopold, L. B. (1966). "River meanders - theory of
minimum variance," Professional Paper 422-H, U.S. Geological Survey,
Washington DC.
Lane, E. W. (1955). "Design of stable channels," Transactions of the American

Society of Civil Engineers 120, 1234-1279.
Leopold, L. B. (1994). A view of the river. Harvard University Press,

Cambridge.
Leopold, L. B., and Maddock, T. (1953). "The hydraulic geometry of stream

channels and some physiographic implications," Professional Paper 252,
Leopold, L. B., Wolman, M. G., and Miller, J. P. (1964). Fluvial processes in

geomorphology, W. H. Freeman & Company, San Francisco.
Leopold, L. B., and Wolman, M. G. (1957). "River channel patterns - braided,

meandering and straight," Professional Paper 282B, U.S. Geological Survey,
Washington, DC.
Limerinos, J. T. (1970). "Determination of the Manning coefficient from

measured bed roughness in natural channels," Water Supply Paper, 1898B,
Lindley, E. S. (1919). "Regime channels," Proceedings of the Pimjub

Engineering Congress 7, 63-74.
Meyer-Peter, E., and Müller, R. (1948). "Formulas for bed-load transport,"

Proceedings of the International Association ofHydraulic Research, 3rd
Annual Conference, Stockholm, 39-64.
Millar, R. G., and Quick, M. C. (1993). "Effect of bank stability on geometry of

gravel rivers," Journal ofHydraulic Engineering, ASCE 119(12), 1343-
1363.
Nixon, M. (1959). "A study of bankfull discharges of rivers in England and

Wales," Proceedings of the Institution of Civil Engineers 12, 157-175.
Osterkamp, W. R., and Hedman, E. R. (1982). "Perennial streamflow

characteristics related to channel geometry and sediment in Missouri River
Basin," Professional Paper 1242, U.S. Geological Survey, Washington, DC.
Pickup, G., and Warner R. F. (1976). "Effects of hydrologic regime on

magnitude and frequency of dominant discharge," Journal of Hydrology 29,
51-75.
42 References
Raphelt, N. K. (1990). "Guidance on the selection and use of sediment
discharge formulas," Hydraulic Engineering, Proceedings of the 1990
National Conference, H. H. Chang and J. C. Hill, eds., 1, San Diego, CA,
July 30 to August 3, 1990, ASCE New York.
Richards, K. S. (1982). Rivers: form and process in alluvial channels.

Methuen, London, 361.
Robbins, L. G. (1977). "Suspended sediment and bed-material studies on the

Lower Mississippi River," Potamology Investigations, Rep. 300-1,
U.S. Army Corps of Engineers, Vicksburg District, Vicksburg, MS.
Schick, A. P. (1974). "Formation and obliteration of desert stream terraces - a

conceptual analysis," Zeitschrift für Geomorphologie, Supplement 21, 88-
105.
Schumm, S. A. (1971). "Fluvial geomorphology; channel adjustments and river

metamorphosis," River Mechanics, H. W. Shen, ed., 1, Fort Collins,
Colorado, 5.1-5.22.
Schumm, S. A. (1975). "Episodic erosion: A modification of the geomorphic

cycle," W. N. Melhorn and R. C. Flemal, eds., Theories of Landform
Development, SUNY Binghampton, Publications in Geomorphology, pp. 69-
85.
Schumm, S. A., and Lichty, R. W. (1963). "Channel widening and floodplain

construction along Cimarron River in south-western Kansas." Professional
Paper 352-D, U.S. Geological Survey, Washington DC, 71-88.
Simons, D. B., and Albertson, M. L. (1960). "Uniform water conveyance

channels in alluvial material," Journal of the Hydraulics Division,
Proceedings of the American Society of Civil Engineers 96(HY5), 33-71.
Soar, P. J., Thorne, C. R, Downs, P. W., and Copeland, R. R. (1998).

"Geomorpic engineering for river restoration design," Proceedings of the
Wetlands Engineering and River Restoration Conference 1998, Denver, CO,
March 22-26, Session G2, ASCE.
Thomas, W. A. (1996). "HEC-6T user's manual," Mobile Boundary Hydraulics,

Clinton, MS.
Thorne, C. R. (1998). Stream Reconnaissance Guidebook: Geomorphological

Investigation and Analysis ofRiver Channels, J. Wiley and Sons, Chichester,
UK, ISBN 0-471-968560, 127.
Thorne, C. R., Biedenharn, D. S., Hey, R. D., Soar, P. J., and Watson, C. C.
(1999). "Effective discharge for channel restoration design: Calculation,"
Proceedings of the International Water Resources Engineering Conference,
Seattle, WA, August 8-11, ASCE.
References 43
Thorne, C .R., Simon, A., and Allen, R. (1996). "Geomorphological river
channel reconnaissance for river analysis, engineering and management,"
Transactions of the Institute of British Geographers, New Series, 21, 469-
483.
Toffaleti, F. B. (1968). "A procedure for computation of the total river sand
discharge and detailed distribution, bed to surface," Technical Report No. 5,
Committee on Channel Stabilization, U.S. Army Corps of Engineers,
Vicksburg, MS.
U.S. Army Corps of Engineers. (1994). Channel Stability Assessment for Flood
Control Projects, EM 1110-2-1418, Washington DC.
U.S. Department of Agriculture. (1977). Design of Open Channels, Technical

Release No. 25, Soil Conservation Service, Washington DC.
United States Geological Survey. (1993). "Nationwide summary of

U.S. Geological Survey regional regression equations for estimating
magnitude and frequency of floods for ungauged sites, 1993," Water-
Resources Investigations Report 94-4002.
Watson, C. C, Gessler, D., Abt, S. R., Thornton, C. I., and Kozinski, P. (1996).
"Demonstration erosion control monitoring sites, 1995 evaluation," Annual
report DACW39-92-K-0003. Colorado State University.
Watson, C. C, Dubler, D., and Abt, S. R. (1997). "Demonstration erosion

control project report," Design Hydrology Investigations, Fort Collins, CO,
submitted to U.S. Army Engineer Waterways Experiment Station,
Vicksburg, MS, August, 1997.
White, W. R., Bettess, R., and Paris, E. (1982). "Analytical approach to river
regime," Journal of the Hydraulics Division, ASCE, 108(HY10), 1179-1193.
Williams, G. P. (1978). "Bankfull discharge of rivers," Water Resources

Research 14(6), 1141-1154.
Wolman, M. G, and Gerson, R. (1978). "Relative scale of time and

effectiveness," Earth Surface Processes and Landforms 3, 189-208.
Wolman, M. G., and Miller, J. P. (1960). "Magnitude and frequency of forces in

geomorphic processes," Journal of Geology 68, 54-74.
Woodyer, K. D. (1968). "Bankfull frequency in rivers," Journal ofHydrology 6,

114-142.
Yevjevich, V. (1972). Probability and Statistics in Hydrology, Water Resources

Publications, Fort Collins, CO, USA.
44 References
Glossary
1. Channel Forming Discharge Concepts

1.58-Year Return Period Discharge (ßi 5)
The discharge with a return period of 1.58 years, derived from the observed
annual maximum flow series. Also called the most probable annual flood.
Results of research on the use of <2i.s in dominant discharge analysis are reported
by Hey (1975) for the UK and Leopold, Wolman, and Miller (1964) for the USA.
2-Year Return Period Discharge (Q2)

The discharge with a return period of 2 years, derived from the observed annual
maximum flow series. Results of research on the use of Q2 in American rivers are
reported by Biedenharn et al. (1987).
Bankfull Discharge (Qh)

The maximum discharge which can be contained within the channel without
over-topping the banks. Leopold, Wolman, and Miller (1964) proposed that this
flow is responsible for forming and maintaining the morphology of the channel.
Bankfull stage refers to the water surface elevation during bankfull flow and can
be identified from various criteria (Williams, 1978). Research papers reporting
the use of Ob include: Leopold and Wolman (1957), Andrews (1980), Charlton,
Brown, and Benson (1978), and Hey and Thorne (1986).
Channel-Forming Discharge
The discharge that most efficiently drives the fluvial processes responsible for
forming and maintaining the main morphological features and dimensions of the
channel. Synonymous with dominant discharge.
Design Discharge
The steady discharge used in the engineering design of a stable channel or flood
defense scheme to define the upper boundary of the operating range of discharges
for the project.
Discharge
The volume of water passing through a cross section in a stream per unit time.
Usually expressed in cubic metres or cubic feet per second.
Glossary 45
Dominant Discharge (<2dom)
The single, steady discharge which would produce the same cross-sectional
morphology, alluvial features, planform geometry and dimensions as those
generated by the actual flow regime (Inglis 1949).
Effective Discharge (Qe)

Discharge responsible for transporting the largest fraction of the bed-material
load in a stable channel over a period of years (Andrews 1980). Defined by the
peak in a histogram of bed-material load versus discharge developed using the
principles of magnitude and frequency analysis (Wolman and Miller 1960).
Flow doing most work

The steady discharge which performs the most geomorphic work, where work is
defined in terms of sediment transport (Leopold and Miller 1960).
Mean Annual Discharge (Qma)

The yearly-averaged discharge. Papers reporting research involving Qma include:
Carlston (1965, 1969); Dury (1964); Leopold and Maddock (1953); Schumm
(1971).
Mean Annual Flood {Q233)

The discharge corresponding to the probability of exceedance of the mean annual
flood event in a Gumbel extreme value type 1 probability distribution (EV1)
derived from the observed annual maximum flow series. This event has a
recurrence interval of 2.33 years. Papers reporting research involving Q2r,
include Brush (1961) and Ferguson (1973).
Most Probable Annual Flood (£>, 5S)

The discharge corresponding to the probability of exceedance of the modal
annual flood event in a Gumbel extreme value type 1 (EV1) derived from the
observed annual maximum flow series. This event has a recurrence interval of
1.58 years. Papers reporting research involving Q\ 58 include Woodyer (1968) and
Dury (1973).
2. Sediment load
Bed Load
Component of the total sediment loadm&Az up of sediment particles moving in
frequent, successive contact with the bed (Bagnold 1966). Transport occurs at or
near the bed, with the submerged weight of particles supported by the bed. Bed
load movement takes place by gravitational processes of rolling, sliding or
saltation.
Bed-material Load
Portion of the total sediment load composed of grain sizes found in appreciable
quantities in the stream bed. In gravel-bed rivers the bed-material load moves as
bed load, but in sand-bed streams significant quantities of bed-material load
move as suspended load.
46 Glossary
Fine Material Load
Portion of the total sediment load composed of particles finer than those found in
the stream bed, and frequently assumed to be the fraction finer than 0.062mm.
Often synonymous with wash load.
Measured Suspended Load

Portion of the total sediment load measured by conventional suspended load
samplers. Includes a large proportion of the suspended load but excludes that
portion of the suspended load moving very near the bed (that is, below the
sampler nozzle) and all of the bed load.
Sediment Concentration
The concentration of sediment in the stream represented by the ratio of sediment
discharge to the water discharge. Usually expressed in terms of milligrams per
litre or parts per million (ppm). It is normally assumed that the density of the
water-sediment mixture is approximately equivalent to the density of the water.
This assumption is acceptable if the concentration is less than 16,000 mg/1.
Sediment Discharge
The mass of sediment that passes through a cross-section in a stream per unit
time. Usually expressed in kilograms per second or tonnes per day.
Suspended Bed-material Load

That portion of the bed-material load that is transported in suspension within the
water column.
Suspended Load
Component of the total sediment load made up of sediment particles moving in
continuous suspension within the water column. Transport occurs above the bed,
with the submerged weight of particles supported by anisotropic turbulence
within the body of the flowing water.
Total Sediment Load
The total mass of granular sediment transported by the stream. Can be broken
down by source, transport mechanism or measurement status as:
1. Bed-material load + wash load, or

2. Bed load + suspended load, or
3. Measured load + unmeasured load.
Unmeasured Suspended Load

That portion of the total sediment load that passes beneath the nozzle of a
conventional suspended load sampler. It consists of near-bed suspension load
and bed load.
Glossary 47
Wash Load
Portion of the total sediment load composed of grain sizes finer than those found
in appreciable quantities in the stream bed. The sum of bed-material load and
wash load makes up the total sediment load.
3. Other terms
Ephemeral Stream
A water course in which channel processes and morphology are significantly
affected by the fact that the discharge of water is intermittent. To be comparable
with the definition of a perennial stream, this may be taken as a water course
which exhibits a measurable surface discharge less than 80 percent of the time
(Osterkamp and Hedman 1982).
Hydraulic Geometry
A geomorphological expression introduced by Leopold and Maddock (1953) to
describe the morphology of an alluvial river as a function of dominant discharge.
The concept is similar to regime theory, but differs in the way that the dominant
discharge is expressed. With respect to the hydraulic geometry of an alluvial
river, the dominant discharge is the single flow event which is representative of
the natural sequence of events which actually occur. Regime theory was
developed for canals, which do not experience a range of flows. Hence, the
dominant discharge for regime theory is the steady, operating discharge.
Flow Duration Curve

A graphical representation of the percent of time (;c-axis) that a specific discharge
(y-axis) is equalled or exceeded during the period of record for which the curve
was developed.
Perennial Stream
A stream which exhibits a measurable surface discharge more than 80 percent of
the time (Osterkamp and Hedman 1982).
Regime Theory
A self-formed alluvial channel is in regime if there are no net changes in
discharge capacity or morphology over a period of years. The concept was
originally developed by engineers designing canals to convey a steady discharge
with neither erosion or siltation in India and Pakistan (Kennedy 1895; Lindley
1919) and, later, in North America (Blench 1957).
Aeaiment Kating
Sediment Rating curve
Curve
A graphical representation of the non-linear relationship between discharge
(x-axis) and sediment discharge (y-axis).
48 Glossary
Form Approved
REPORT DOCUMENTATION PAGE OMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the
data needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing
this burden to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-
4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently
valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.
1. REPORT DATE (DD-MM-YYYY) 2. REPORT TYPE 3. DATES COVERED (From - To)
August 2000 Final Report
A. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
Effective Discharge Calculation: A Practical Guide
5b. GRANT NUMBER
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER
David S. Biedenham, Ronald R. Copeland, Colin R. Thorne,
Philip J. Soar, Richard D. Hey, Chester C. Watson 5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION REPORT

NUMBER
U.S. Army Engineer Research and University of East Anglia
Development Center Norwich NR4 7TJ UK; ERDC/CHLTR-00-15
Coastal and Hydraulics Laboratory
3909 Halls Ferry Road Colorado State University
Vicksburg,MS 39180-6199; Department of Civil Engineering
Fort Collins, CO 80523
University of Nottingham
Nottingham NG72RD UK;
9. SPONSORING / MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR'S ACRONYM(S)
Flood Damage Reduction Research Program
Headquarters, U.S. Army Corps of Engineers
Washington, DC 20314-1000 11. SPONSOR/MONITOR'S REPORT
NUMBER(S)
12. DISTRIBUTION / AVAILABILITY STATEMENT

Approved for Public Release; distribution is unlimited.
13. SUPPLEMENTARY NOTES
14. ABSTRACT
This report presents a procedure for calculating the effective discharge for alluvial rivers. An alluvial river adjusts the shape of its
channel to the wide range of flows that mobilize the boundary sediments. However, in many rivers it has been shown that a single
representative discharge can be used to determine a stable channel geometry. One method to determine this "channel-forming" discharge is
based on the hypothesis that the discharge that transports the most sediment, over time, is the discharge that forms the channel. This is
termed the effective discharge. Two other methods commonly used are the bankfull discharge and a discharge with a specific recurrence
interval. While it may, under some circumstances, be possible to estimate the channel-forming discharge from the bankfull discharge, in
practice, identification of bankfull stage in the field is often problematic. Even if bankfull stage can be identified, channel roughness and
slope typically must be assumed to determine the bankfull discharge. Another method for determining channel-forming discharge is to
assume a specific frequency from the annual flood peak frequency curve. However, there does not seem to be a recurrence interval that is
generally applicable to alluvial rivers. The procedure for effective discharge calculation presented in this report is designed to have general
applicability, have the capability to be applied consistently and to integrate the effects of physical processes responsible for determining
channel dimensions. An example of the calculations necessary and applications of the effective discharge concept are presented.
15. SUBJECT TERMS

Bankfull discharge Effective discharge River engineering River restoration
Channel-forming discharge Hydraulic geometry River management Stable channel design
16. SECURITY CLASSIFICATION OF: 17. LIMITATION 18. NUMBER 19a. NAME OF RESPONSIBLE PERSON
OF ABSTRACT OF PAGES
a. REPORT b. ABSTRACT c. THIS PAGE 19b. TELEPHONE NUMBER (include area
code)
UNCLASSIFIED UNCLASSIFIED 63
Standard Form 298 (Rev. 8-98)
Prescribed by ANSI Std. 239.18

Tardir Tiffs A383261.tiff

Uploaded by

Copyright:

Available Formats

Tardir Tiffs A383261.tiff

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tardir Tiffs A383261.tiff

Uploaded by

Copyright:

Available Formats

US Army Corps

Effective Discharge Calculation:

Approved for public release; distribution is unlimited.

The findings of this report are not to be construed as an

® PRINTED ON RECYCLED PAPER

Effective Discharge Calculation:

Colin R. Thome, Philip J. Soar

Prepared for U.S. Army Corps of Engineers

Conversion Factors, SI to Non-SI Units of Measurement vii

1—Channel Forming Discharge Concept 1

2—Effective Discharge Calculation 4

3—Evaluation and Problem Solving 19

Problems with Lowest Discharge Class 19

Effective Discharge Calculation for the Mississippi River at Vicksburg 22

Figure 1. Derivation of total sediment load-discharge histogram (iii) from

Figure 2. Daily mean flow-duration curve: Sevier River, Hatch UK

Figure 3. Downstream daily flow-duration curves: River Wye, UK

Figure 4. Comparison of sediment relationships for Abiaca Creek,

Figure 5. Computational procedure for generating a flow-frequency

Figure 6. Computational procedure for generating a bed-material load

Figure 7. Computational procedure for generating a bed-material load

Figure 8. Effective discharge duration versus drainage area 21

Figure 11. Bed-material load histogram: Lower Mississippi River at

Figure 12. Long-channel variation in bank top elevations: Lower Mississippi

Figure 14. Supply Reach Magnitude-Frequency Analysis derived from

Figure 15. Stable width-depth-slope diagram. The range of stable slope

SI units of measurement used in this report can be converted to Non-SI units

meters 3.281 feet

cubic meters 35.32 cubic feet

square kilometers 0.39 square miles

tonne 1.1 US ton

Channel-forming discharge can be estimated using one of three prescribed

All three methodologies for estimating channel-forming discharge provide

Applications of the effective discharge concept outlined in the report include

Based on field observations, Inglis (1947) determined that the channel-

To explain this phenomenon it is necessary to recognize that any local

In humid environments, perennial rivers usually recover their long-term,

Chapter 1 Channel-Forming Discharge Concept

Equivalence between bankfull and effective discharges for natural alluvial

In practice, major problems are likely to arise when attempting to use

In many studies channel-forming discharge is estimated from the recurrence

In practice, bankfull discharge is often assumed to have a recurrence interval

Chapter 1 Channel-Forming Discharge Concept

Although the channel-forming discharge concept is not universally accepted,

Ideally, the method used to determine the channel-forming discharge should

Chapter 1 Channel-Forming Discharge Concept

The first step in an effective discharge calculation is to group the discharge

Chapter 2 Effective Discharge Calculation

Figure 1. Derivation of total sediment load-discharge histogram (iii) from flow

m-1 !IIIIIII ! 11mil! mi mi;:iirn'iiiiiijiiiüinijmiimMhiiüiiiiiiiiiuihiii linn I i IIIIIII i i um in

Chapter 2 Effective Discharge Calculation

Chapter 2 Effective Discharge Calculation

In the absence of 1 hour or 15-minute data, recorded hydrographs from

Period of record. The period of record must be sufficiently long to include

A reasonable minimum period of record for an effective discharge calcula-

At locations where gauging records are either unavailable or are found to be

Chapter 2 Effective Discharge Calculation

Drainage area, flow-duration curve method. This method relies on the

Regionalized duration curve method. An alternative to the use of

"'''"" '" '