Hydraulic Design of Stream Restoration Projects - [PDF Document] (2024)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (1)






Hydraulic Design of Stream RestorationProjectsRonald R. Copeland, Dinah N. McComas, Colin R. Thorne,Philip J. Soar, Meg M. Jonas, and Jon B. Fripp

September 2001










Approved for public release; distribution is unlimited.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (2)


The contents of this report are not to be used for advertising,publication, or promotional purposes. Citation of trade namesdoes not constitute an official endorsem*nt or approval of the useof such commercial products.

The findings of this report are not to be construed as an officialDepartment of the Army position, unless so designated by otherauthorized documents.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (3)

ERDC/CHL TR-01-28September 2001

Hydraulic Design of Stream Restoration Projectsby Ronald R. Copeland, Dinah N. McComas

Coastal and Hydraulics LaboratoryU.S. Army Engineer Research and Development Center3909 Halls Ferry RoadVicksburg, MS 39180-6199

Colin R. Thorne, Philip J. SoarUniversity of NottinghamNottingham, NG7 2RD, UK

Meg M. Jonas, Jon B. Fripp

U.S. Army Engineer District, Baltimore10 South Howard StreetBaltimore, MD 21201

Final report

Approved for public release; distribution is unlimited

Prepared for U.S. Army Corps of EngineersWashington, DC 20314-1000

Hydraulic Design of Stream Restoration Projects - [PDF Document] (4)





Purpose ...............................................................................................................1Definition of a Stream Restoration Project .......................................................3Design Philosophy .......................................................................................... 11Project Study Teams ....................................................................................... 13

2�Project Objectives and Constraints ................................................................. 15

Project Objectives ........................................................................................... 15Flood Damage Reduction Techniques ........................................................... 18Project Constraints .......................................................................................... 19

3�Hydrology ........................................................................................................ 20

General ............................................................................................................ 20Frequency Analysis......................................................................................... 20

Peak discharge analysis............................................................................ 20Regional regression .................................................................................. 21Errors in regression analysis .................................................................... 22

Flow-Duration Curves .................................................................................... 23Hydrologic Models ......................................................................................... 23Channel-Forming Discharge Concept ............................................................ 25

Bankfull discharge.................................................................................... 27Discharge with a specific recurrence interval.......................................... 33Effective discharge ................................................................................... 33Examples of channel-forming discharge representations........................ 36Channel-forming discharge related to drainage area............................... 39

Channel-Forming Discharge Summary.......................................................... 39Stormwater Management................................................................................ 40

General hydrologic effects of urbanization and stormwatermanagement.............................................................................................. 40Conventional dry ponds ........................................................................... 42Extended detention ponds ........................................................................ 44Wetland-pond systems ............................................................................. 44Infiltration basins and bioretention .......................................................... 44Stormwater management guidance.......................................................... 45

Hydraulic Design of Stream Restoration Projects - [PDF Document] (5)


4�Stability Analysis............................................................................................. 46

Geomorphic Assessment................................................................................. 47Data assembly........................................................................................... 48Field investigations .................................................................................. 48Identification of geomorphologically similar reaches............................. 50Assessment of reach condition................................................................. 51Channel typing and classification ............................................................ 52Methods for assessing historical channel stability .................................. 53Specific gage analysis .............................................................................. 54Comparative surveys and mapping.......................................................... 54

Hydraulic Geometry Assessment ................................................................... 55Background............................................................................................... 55Developing hydraulic geometry relations................................................ 56Choice of independent variables.............................................................. 57Use of stream typing systems to refine hydraulic geometryrelations .................................................................................................... 57Transfer of hydraulic geometry relations from one watershedto another .................................................................................................. 57Special problems of urbanized streams ................................................... 57Uncertainty in hydraulic geometry relations ........................................... 58Application of hydraulic geometry relations to assess channelstability ..................................................................................................... 58

Analytical Stability Assessment ..................................................................... 59Hydraulic calculations.............................................................................. 59Bed stability.............................................................................................. 61Sediment rating curve analogy analysis .................................................. 61Sediment budget analysis......................................................................... 62Nonequilibrium sediment transport ......................................................... 63Integration and application of results....................................................... 64

5�Hydraulic Design Methodology...................................................................... 66

Design Discharges........................................................................................... 66Design discharge for low flows ............................................................... 66Main channel discharge............................................................................ 67Habitat and hydraulic structure design discharge.................................... 67

Threshold Channels ........................................................................................ 68Alluvial Channels............................................................................................ 69

Alluvial channel design variables ............................................................ 70Analogy methods ............................................................................... 70Hydraulic geometry methods ............................................................ 71Extremal hypotheses.......................................................................... 80Constrained dependent variables....................................................... 80Calculation of the remaining unknown design variables.................. 80

Planform ................................................................................................... 81Natural variability in morphology around meander bendways............... 87Practical channel design equations for meander bend geometry ............ 93

Sediment Impact Assessment ......................................................................... 95Topics to report......................................................................................... 95Sediment budget analysis......................................................................... 96Sediment rating curve analysis ................................................................ 96

Hydraulic Design of Stream Restoration Projects - [PDF Document] (6)


Numerical Sedimentation Modeling............................................................... 97Operation and Maintenance Requirements .................................................... 98Quality Management....................................................................................... 98

6�Conclusions ..................................................................................................... 99

References............................................................................................................ 101

Appendix A�Flow-Duration Curves for Effective Discharge Calculation ........A1

Appendix B�Stream Reconnaissance Data Sheets by Thorne (1993) ...............B1

Appendix C�Stream Reconnaissance Data Sheets by Baltimore District .........C1

Appendix D�Guidelines for Sampling Bed Material .........................................D1

Appendix E�Computation of Average Annual Sediment YieldUsing Weighted Events ......................................................................................... E1

Appendix F�Example Scope of Work for Stability Assessment ....................... F1

Appendix G�Example Sediment Impact Assessment and StableChannel Design......................................................................................................G1

SF 298

List of Figures

Figure 1. Flow chart for systematic approach to hydraulic designof stream restoration projects.........................................................2

Figure 2. Bank erosion of vegetated bank, Johnson Creek, MS...................4

Figure 3. Bank stabilization with stone toe protection, JohnsonCreek, MS.......................................................................................4

Figure 4. Bank stabilized and revegetated bank, Johnson Creek,MS ..................................................................................................5

Figure 5. Channel bed and bank erosion due to degradation,Hotophia Creek, MS ......................................................................5

Figure 6. Channel stabilization with grade control, HotophiaCreek, MS.......................................................................................6

Figure 7. Stable channel upstream from grade control, HotophiaCreek, MS.......................................................................................6

Figure 8. Habitat enhancement with the use of low-flow stoneweirs, Paint Branch, MD................................................................7

Hydraulic Design of Stream Restoration Projects - [PDF Document] (7)


Figure 9. Bank protection and habitat enhancement using stonelunkers, Rapid Creek, Rapid City, SD...........................................7

Figure 10. Habitat enhancement features added to flood controlproject, Sammamish River, WA....................................................8

Figure 11. Incised urban stream upstream from restoration reach,San Antonio River, San Antonio, TX............................................8

Figure 12. River restoration to provide aesthetics and recreation,San Antonio River, San Antonio, TX............................................9

Figure 13. Project designed for flood protection with habitatenhancement features boulder clusters, and low-flowdeflectors, Zumbro River, Rochester, MN....................................9

Figure 14. Project designed to provide flood protection, aesthetics,and habitat enhancement, South Platte River, Denver,CO ............................................................................................... 10

Figure 15. Project designed to provide flood protection, aesthetics,and recreation, Cherry Creek, Denver, CO ................................ 10

Figure 16. Idealized stream cross section .................................................... 12

Figure 17. Channel velocity trends .............................................................. 13

Figure 18. Submodels of HEC-HMS modeling process.............................. 25

Figure 19. Bankfull depth using width-depth ratio...................................... 28

Figure 20. Stage-discharge rating curve Bogue Chitto River nearBush, LA ..................................................................................... 30

Figure 21. Stage-discharge rating curve Mississippi River at TarbertLanding, MS................................................................................ 31

Figure 22. Long-channel variation in bank top elevations: LowerMississippi River......................................................................... 31

Figure 23. Derivation of total sediment load-discharge histogramfrom flow frequency and sediment load rating curves............... 34

Figure 24. Relationship between effective discharge, Qe, andbankfull discharge, Qb, for 57 U.S. sand bed riverchannels....................................................................................... 36

Figure 25. Relationship between the 2-year return period flow, Q2,and bankfull discharge, Qb, for 57 U.S. sand bed riverchannels....................................................................................... 37

Figure 26. Relationship between the discharge marking the upperlimit of the range of discharges which transport 50percent of the average annual bed material load, Qe50,and bankfull discharge, Qb, for 57 U.S. sand bed riverchannels....................................................................................... 38

Hydraulic Design of Stream Restoration Projects - [PDF Document] (8)


Figure 27. Relationship between the discharge marking the upperlimit of the range of discharges which transport75 percent of the average annual bed material load, Qe75,and bankfull discharge, Qb, for 57 U.S. sand bed riverchannels....................................................................................... 38

Figure 28. Bankfull discharge as a function of drainage area ..................... 39

Figure 29. Schematic of a watershed relative to project location................ 41

Figure 30. An idealized effect of stormwater pond on channelvelocities...................................................................................... 43

Figure 31. Schematic of a watershed relative to multiple smallponds ........................................................................................... 43

Figure 32. Levels of stability assessment..................................................... 47

Figure 33. Sediment rating curve analogy analysis of existingconditions .................................................................................... 62

Figure 34. Hydraulic geometry relationship for width for the UpperSalmon River Basin, ID .............................................................. 72

Figure 35. Equilibrium channel slope versus drainage area forHickahala Creek, Batupan Bogue, and Hotopha Creek,MS ............................................................................................... 73

Figure 36. Best-fit hydraulic geometry relationships for width forU.S. sand bed rivers with banks typed according todensity of tree cover.................................................................... 74

Figure 37. Confidence intervals applied to the hydraulic geometryequation for width based on 32 sand bed streams withless than 50 percent tree cover on the banks .............................. 74

Figure 38. Confidence intervals applied to the width hydraulicgeometry equation based on 26 sand bed rivers with atleast 50 percent tree cover on the banks..................................... 75

Figure 39. Downstream width hydraulic geometry for NorthAmerican gravel bed rivers and U.K. gravel bed rivers............. 75

Figure 40. Downstream width hydraulic geometry for NorthAmerican gravel bed rivers with confidence bands................... 76

Figure 41. Downstream width hydraulic geometry for UnitedKingdom gravel bed rivers with confidence bands.................... 76

Figure 42. Downstream width hydraulic geometry for UnitedKingdom gravel bed rivers with confidence bands, basedon 36 sites in the United Kingdom with erodible banks............ 77

Hydraulic Design of Stream Restoration Projects - [PDF Document] (9)


Figure 43. Downstream width hydraulic geometry for UnitedKingdom gravel bed rivers with confidence bands, basedon 43 sites in the United Kingdom with resistant banks............ 77

Figure 44. Stability curve from stable channel analytical method .............. 81

Figure 45. Meander parameters.................................................................... 82

Figure 46. Hydraulic geometry relationship for meanderwavelength with confidence intervals, based on acomposite data set of 438 sites ................................................... 83

Figure 47. Definition of sine-generated curve ............................................. 85

Figure 48. Effect of the shape factor on channel sinuosity with thesine-generated curve ................................................................... 86

Figure 49. Cumulative distribution of radius of curvature-to-widthratio derived from a composite data set of 263 sites.................. 86

Figure 50. Meander cross-section dimensions for restoration design ......... 89

Figure 51. Ratio of bend apex width to inflexion point width as afunction of meander bend type only, for sinuosities of atleast 1.2........................................................................................ 90

Figure 52. Ratio of pool width to inflexion point width as a functionof meander bend type only, for sinuosities of at least 1.2.......... 91

Figure 53. Cumulative distribution of the pool-offset ratio for alltypes of meander bend studied ................................................... 92

Figure 54. Dimensionless maximum scour depth in meander poolsas a function of radius of curvature-to-width ratio..................... 93

Figure 55. Analogy sediment analysis of proposed projectconditions .................................................................................... 97

List of Tables

Table 1. Possible Field Indicators of River Stability/Instability.............. 50

Table 2. Reach Condition Assessment ..................................................... 51

Table 3. Hydraulic Geometry Width Predictors for Sand BedChannels ...................................................................................... 78

Table 4. Hydraulic Geometry Width Predictors for Gravel BedChannels ...................................................................................... 79

Hydraulic Design of Stream Restoration Projects - [PDF Document] (10)


Table 5. Hydraulic Geometry Relationships for MeanderWavelength ................................................................................. 84

Table 6. Ranges of Physical Characteristics Found in DifferentMeander Bend Types Identified from the 1981 RedRiver Hydrographic Survey Between Index, AR, andShreveport, LA............................................................................ 88

Table 7. Constant Values Used to Estimate the Mean Ratio ofBend Apex Width to Inflexion Point Width WithinConfidence Bands for Different Types of MeanderBends and for Sites with Sinuosity of at Least 1.2 .................... 90

Table 8. Constant Values Used to Estimate the Mean Ratio ofPool Width to Inflexion Point Width Within ConfidenceBands for Different Types of Meander Bends and forSites with Sinuosity of at Least 1.2 ............................................ 91

Table 9. Uncertainty in Estimates of Width Variability AroundMeander Bends and Location of Pools....................................... 94

Hydraulic Design of Stream Restoration Projects - [PDF Document] (11)



The work described in this report was authorized by Headquarters, U.S.Army Corps of Engineers (HQUSACE) as part of the Flood Damage ReductionResearch Program. The guidelines developed herein are products of Work Unit32878 �Channel Restoration Design.� The Program Monitor was Mr. Richard J.DiBuono, HQUSACE. The Program Manager was Ms. Carolyn Holmes, Coastaland Hydraulics Laboratory (CHL), U.S. Army Engineer Research andDevelopment Center (ERDC). The principal investigator for the work unit wasDr. Ronald R. Copeland, CHL.

The report was prepared by Dr. Ronald R. Copeland, CHL, Ms. Dinah N.McComas, CHL, Drs. Colin R. Thorne and Philip Soar, University ofNottingham, and Ms. Meg Jonas and Mr. Jon Fripp, U.S. Army EngineerDistrict, Baltimore. The U.S. Army Corps of Engineer Committee on ChannelStabilization and Dr. David Biedenharn, CHL, made valuable contributions andprovided review and comment. Committee on Channel Stabilization membersinclude: Dr. Ronald R. Copeland, CHL, Chairman; Ms. Dinah N. McComas,CHL, Secretary; Mr. Larry E. Banks, U.S. Army Engineer District, Vicksburg;Mr. William E. Branch, U.S. Army Engineer Division, Northwest; Dr. Craig J.Fischenich, Environmental Laboratory, ERDC; Ms. Meg M. Jonas, U.S. ArmyEngineer District, Baltimore; Mr. Thomas J. Pokrefke, CHL; Mr. John I. Remus,U.S. Army Engineer District, Omaha; Mr. Edward F. Sing, U.S. Army EngineerDivision, South Pacific; Mr. Michael F. Spoor, U.S. Army Engineer District,Huntington; Mr. Scott E. Stonestreet, U.S. Army Engineer District, Sacramento;and Mr. Howard M. Whittington, U.S. Army Engineer District, Mobile.Technical Review was also provided by Mr. Doug Otto, U.S. Army EngineerDistrict, Mobile; Mr. Pat Foley, U.S. Army Engineer District, St. Paul; andMr. Rene Vermeeren, Mr. Dave Cozakos and Mr. Nick Adelmeyer, U.S. ArmyEngineer District, Los Angeles.

The study was performed under the supervision of Dr. Yen-Hsi Chu, Chief ofthe River Sedimentation Branch, CHL, Mr. Thomas J. Pokrefke, Acting DeputyDirector of CHL, and Mr. Thomas W. Richardson, Acting Director of CHL.

At the time of publication of this report, Dr. James R. Houston was Directorof ERDC, and COL John W. Morris III, EN, was Commander and ExecutiveDirector.

The contents of this report are not to be used for advertising, publication,or promotional purposes. Citation of trade names does not constitute anofficial endorsem*nt or approval of the use of such commercial products.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (12)

Chapter 1 Introduction 1

1 Introduction

PurposeThe purpose of this document is to provide a systematic hydraulic design

methodology to hydraulic engineers involved in stream restoration projects. Theobjective of the methodology is to fit the stream restoration project into thenatural system within the physical constraints imposed by other projectobjectives and constraints.

The hydraulic design of a stream restoration project should provide for achannel that is in dynamic equilibrium with its sediment load. A stable channelis defined as a channel where the planform, cross section, and longitudinal profileare sustainable over time. The channel may migrate and still be stable, althoughchannel migration may not always be acceptable due to project constraints. Themagnitude of long-term aggradation and/or degradation in a stable channelshould be small enough to allow for economical channel maintenance. Thedesign methodology presented herein is systematic, i.e., when used by differentengineers, with the same project objectives, design results should be similar. Themethodology is based on sound physical principles and is applicable to both fixedand mobile boundary streams.

The systematic approach to hydraulic design of stream restoration projectspresented herein has several components. Chapter 2 outlines defining projectobjectives and constraints. Chapter 3 provides an overview of how to determinehydrologic data that may be of importance in the hydraulic design process.Chapter 4 outlines the stability assessment methodologies that are important toestablish baseline geomorphological conditions and to evaluate the effectivenessand geomorphological impacts of project alternatives. Chapter 5 presents themethodology for hydraulic design of project features and for assessing hydraulicand sediment transport impacts of alternatives. This systematic approach isiterative, as shown in Figure 1.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (13)

2 Chapter 1 Introduction

Figure 1. Flow chart for systematic approach to hydraulic design of streamrestoration projects

Hydraulic Design of Stream Restoration Projects - [PDF Document] (14)

Chapter 1 Introduction 3

Definition of a Stream Restoration ProjectStream restoration projects typically are intended to improve or restore

environmental conditions in the stream and the adjacent stream corridor.Channel stability is often essential to the maintenance of favorable environmentalconditions. Restoration projects do not necessarily require returning a system tosome predisturbance condition, as this is seldom feasible. The objective of arestoration project is then a partial recovery of the natural geomorphic, hydraulic,and ecological functions of the stream. It follows that the project design teamrequires expertise in the fields of geomorphology, hydraulics, and ecology.

U.S. Army Corps of Engineers stream restoration projects are frequentlyassociated with or are part of a flood-control project. Thus, projects have morethan one objective and compromises may be required to meet essential portionsof each objective. For example, it may not be possible to construct a channel thatboth carries the design flood flow and provides optimum transport of theupstream sediment load.

The scales and purposes of stream restoration projects vary significantlydepending on project objectives. In general, stream restoration projects can haveone or more of the following three general goals:

a. Enhance channel stability and thus reduce channel maintenance.

b. Improve the environment.

c. Provide aesthetic or recreation benefits.

Channel stability is usually addressed by providing bank protection and/orgrade control. Typical examples that primarily address aspects of channelstability are provided in Figures 2 through 7. The projects in these figurescombine traditional engineering protection methods with vegetation to providefor a stable stream. Other projects have been designed primarily to providehabitat enhancement. Figures 8 through 10 show features that were added toexisting projects to provide better aquatic habitat. Figures 11 and 12 show a riverrestoration project in San Antonio, Texas where the primary objectives wereaesthetics and recreation. Many stream restoration projects are multiobjective.These projects can be technically challenging but have widespread support sincethey address many concerns. Several examples of multiobjective projects areprovided in Figures 13 through 15.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (15)

4 Chapter 1 Introduction

Figure 2. Bank erosion of vegetated bank, Johnson Creek, MS

Figure 3. Bank stabilization with stone toe protection, Johnson Creek, MS

Hydraulic Design of Stream Restoration Projects - [PDF Document] (16)

Chapter 1 Introduction 5

Figure 4. Bank stabilized and revegetated bank, Johnson Creek, MS

Figure 5. Channel bed and bank erosion due to degradation, Hotophia Creek,MS

Hydraulic Design of Stream Restoration Projects - [PDF Document] (17)

6 Chapter 1 Introduction

Figure 6. Channel stabilization with grade control, Hotophia Creek, MS

Figure 7. Stable channel upstream from grade control, Hotophia Creek, MS

Hydraulic Design of Stream Restoration Projects - [PDF Document] (18)

Chapter 1 Introduction 7

Figure 8. Habitat enhancement with the use of low-flow stone weirs, PaintBranch, MD

Figure 9. Bank protection and habitat enhancement using stone lunkers, RapidCreek, Rapid City, SD

Hydraulic Design of Stream Restoration Projects - [PDF Document] (19)

8 Chapter 1 Introduction

Figure 10. Habitat enhancement features (root wads) added to flood-controlproject, Sammamish River, WA

Figure 11. Incised urban stream upstream from restoration reach, San AntonioRiver, San Antonio, TX

Hydraulic Design of Stream Restoration Projects - [PDF Document] (20)

Chapter 1 Introduction 9

Figure 12. River restoration to provide aesthetics and recreation, San AntonioRiver, San Antonio, TX

Figure 13. Project designed for flood protection with habitat enhancementfeatures boulder clusters and low-flow deflectors, Zumbro River,Rochester, MN

Hydraulic Design of Stream Restoration Projects - [PDF Document] (21)

10 Chapter 1 Introduction

Figure 14. Project designed to provide flood protection, aesthetics, and habitatenhancement, South Platte River, Denver, CO

Figure 15. Project designed to provide flood protection, aesthetics, andrecreation, Cherry Creek, Denver, CO

Hydraulic Design of Stream Restoration Projects - [PDF Document] (22)

Chapter 1 Introduction 11

Design PhilosophyA wide variety of analysis techniques can be applied to channel design and

stream restoration. A sound stream restoration design incorporates techniquesfrom fluvial geomorphology, engineering and stream ecology. The study area towhich these techniques are applied must extend beyond the limits of the projectsite to the extent that both the project�s effect on the stream system and thestream system�s effect on the project reach can be determined. This may requireanalysis of the entire watershed. Many of the principles and techniques found inEM 1110-2-1418, �Channel Stability Assessment for Flood Control Projects,�are equally applicable to hydraulic design of stream restoration projects.

Fluvial geomorphology techniques provide insight relative to generalresponses of a river system to a variety of imposed changes. These techniquesare important in analyzing the stability of the existing stream system and inidentifying the source of instabilities. Fluvial geomorphology techniques alsoprovide generalized guidance related to appropriate cross-section geometry andchannel planform. It is important to recognize that the science of fluvialgeomorphology utilizes both qualitative and quantitative analyses of observedfeatures and channel forms. As a result, trends and changes in the fluvial systemmay be inferred from the history of channel evolution and its past response tohuman interventions. However, quantitative assessment and design, for aspecific project area, requires use of physically-based calculations. Thesecalculations are based on principles and techniques in hydrology, hydraulics, andsediment transport. These considerations demonstrate that the contributions ofgeomorphology, hydrology, hydraulics, and sediment transport arecomplementary, not alternatives.

Stream restoration design should consider a variety of flow conditions.Rarely does the behavior of a channel under a single discharge adequately reflectthe range of design conditions required of a stream restoration project. So,during the design process, a range of discharges must be considered even thoughinitial stream restoration design focuses on a single design discharge. Figure 16illustrates a sketch of an idealized channel cross section and the various flowconditions that should be examined. In Figure 16, Qlfc indicates the low-flowcondition. The cross-sectional area at this discharge often defines the limitingbiologic condition for aquatic organisms. Minimum depths are often specified asa design goal at this discharge. In gravel bed streams, it is often desirable todesign the channel section at this flow level so that fine sediment does notdeposit at this discharge, as this risks smothering the coarse bed and benthicorganisms in a layer of sand and silt. Analysis and design for this area mayinvolve incipient motion and threshold techniques. Qolw indicates the ordinarylow water or base flow condition. In many situations, low flow and base flow canbe assumed to be the same. The cross-sectional perimeter at a flow of Qolw maybe further defined into a benthic zone (where species are attached to or buried inthe substrate) and an aquatic zone (which includes organisms in the water).Habitat features are often designed for the cross-sectional area below Qolw.While this area can be very active in a natural stream, it is often desirable toprovide some stabilization if a large investment in habitat structures is to bemade. Although the geometry of these habitat structures may be designed forstream conditions at a discharge of Qolw, hydraulic parameters at a higher

Hydraulic Design of Stream Restoration Projects - [PDF Document] (23)

12 Chapter 1 Introduction

discharge should be used to determine the stability of the structure and/orrequirement for bank stabilization. It is important to note that watershed land usechanges, such as urbanization, can have a strong effect on the flows at this level.

Figure 16. Idealized stream cross section

Qohw represents ordinary high water in Figure 16. Riparian vegetationtypically begins between the ordinary low water and the ordinary high water. Ifneeded, and if design conditions are met, the channel banks above the ordinarylow-water elevation may be suitable for vegetative bank protection. Sedimenttransport typically becomes an issue at this flow, especially in alluvial channels.

Bankfull stage is typically defined at a point where the width to depth ratio isa minimum. In Figure 16, this is shown at Qbf. In many cases, ordinary highwater and bankfull stage are synonymous.

There are many geomorphic regime relationships that provide relationshipsbetween drainage area, discharge, and the cross-section geometry at bankfullstage. In many situations, the channel velocity begins to asymptotically approacha maximum at this stage. It has also been observed that, in some cases, lateralmomentum losses can result in a decrease in channel velocity during rising stagesas flow spills onto the floodplain. However, when the floodplain is narrow orheavily obstructed, channel velocities may continue to increase with rising stage.This phenomenon is illustrated in the three idealized cross sections in Figure 17.In Figure 17a, there is a large increase in available conveyance once the flowsexceed the top of bank elevation. As a result, a large increase in flow producesonly little or no increase in channel velocity. In this situation it may beappropriate to use the bankfull hydraulic conditions to design bank protection orto assess the stability of in-stream habitat structures. However, Figures 17b and17c illustrate conditions where the overbank conveyance is limited bytopography or vegetation. In these situations, the channel velocity can continueto increase with increased flow. As a result, it may be appropriate to use asignificantly larger discharge to design channel features such as bank protectionand habitat structures. While the difference between the different cross sectionspresented in Figure 17 is very obvious in the idealized sketches, it is oftendifficult to assess the hydraulic behavior of overbank flows in the field.Therefore, it is necessary that hydraulic and sediment transport calculations beused to assess channel and overbank conditions under a wide range of inbank andoverbank flows.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (24)

Chapter 1 Introduction 13

Figure 17. Channel velocity trends

The riparian zone is typically found above the bankfull stage. This area isfavorable to plants and animals that inhabit land that is rarely submerged.Aquatic wetlands can also occur in this area. In Figure 16, Qlf represents a largeflow that floods the riparian zone. Flows at this stage and below can be adjustedthrough the use of stormwater management techniques. Sediment transport andcontinuity analysis should address these flows as well as lower discharges. Ifchannel avulsions are a concern in the study reach, incipient motion analysistechniques and threshold methods may also be required for flows in this zone. Ifa project involves riparian plantings, it is advisable to assess the expected depthsand velocities in the terrestrial zone with reference to the tolerances of theproposed species.

Some consideration should be given to extreme flood flows for streamrestoration projects. In Figure 16, Qrf represents a regulatory floodplain such asthe 100-year flood. Some analysis should be done to assess the impact of thestream restoration project on flood elevations in the floodplain. Conventionalstormwater management techniques are typically insufficient to affect flows atthis level. The stability requirements for the features used in the lower parts ofthe channel are typically defined somewhere between Qlf and Qrf. In addition,sediment continuity continues to be an issue at these flows.

Project Study TeamsSince few people possess all skills necessary to conduct a successful stream

restoration study and design, an interdisciplinary approach is required. This isespecially the case for the establishment of objectives. While the exact makeupof the team can vary, it should include engineering, geomorphological, andecological expertise.

Stream restoration projects differ from traditional hydraulic engineeringprojects in several ways. First, there is more potential variety in the goals.Second, there is less consensus on how the study should be conducted. Questionssuch as - what surveys and data collection are necessary, what analyses should beperformed, what design techniques should be used, and what measures should beimplemented, are frequently debated. Third, the success of the project may bejudged as much by appearance as by function. And fourth, an interdisciplinary

Hydraulic Design of Stream Restoration Projects - [PDF Document] (25)

14 Chapter 1 Introduction

approach using geomorphologists, engineers, and biologists is essential to thesuccess of the project. Since the study team will likely includegeomorphologists, biologists, ecologists, landscape architects, andrepresentatives of resource agencies as well as hydraulic engineers, it isimportant to define areas of responsibility clearly so that the team can functioneffectively. The hydraulic engineer performs all analyses and stable channeldesign calculations related to hydrology, hydraulics, water quality, andsedimentation. This may include geomorphic assessment, stream classification,assessment of watershed and channel stability, when the engineer has beentrained in the relevant techniques in geomorphology.

This report provides guidelines for determining the types of hydraulicanalyses that should be employed for a variety of stream restoration projects.Information is provided that describes setting objectives, data requirements,analytical techniques, limitations, and interpretations of possible results. Thecoverage provided is sufficient to the task of designing a restoration project witha stable channel capable of conveying the water and sediment supplied fromupstream. However, no attempt has been made to include or evaluate all theapproaches to channel analyses and design that are currently in use.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (26)

Chapter 2 Project Objectives and Constraints 15

2 Project Objectives andConstraints

Project ObjectivesAs a result of increased public concern for the environment, many Federal,

state, and local governments along with grassroots organizations are activelyengaged in stream restoration. Engineers are being asked for analyses anddesigns that focus on restoring, establishing, or maintaining natural streamenvironments. The perceived success or failure of many stream restorationprojects can be as much a function of the success criteria selected at the outset asof the design. Therefore, the importance of establishing achievable studyobjectives is critical. Once established, these objectives will define the type andamount of data collection, methodologies for assessments and designs, andcondition of the design itself. Project objectives should be clearly stated in thescope of work or project management plan.

The first step in a stream restoration project, as with any engineering project,is to clearly define project objectives in cooperation with stakeholders. Is theobjective to create an aesthetic setting, a natural setting, to enhance fish orwildlife habitat, to prevent bank erosion or channel degradation, or somethingelse? All these objectives could conceivably be considered legitimate goals of astream restoration project, though in practice it may be necessary to compromiseon one objective in order to meet others.

In establishing objectives for a stream restoration project, it is advisable toassess at least the following six issues:

a. The existing condition of the stream and watershed.

b. The scale and severity of the resource loss or degradation due to streaminstability.

c. Causal factors and controls that have resulted in the current streamcondition. In this context it is useful to establish whether currentinstability in the channel is being driven by the current flow regime or isa product of past conditions.

d. The condition into which the channel is likely to evolve without aproject. This often involves a strong reliance on geomorphic prediction

Hydraulic Design of Stream Restoration Projects - [PDF Document] (27)

16 Chapter 2 Project Objectives and Constraints

coupled with engineering judgment. This question incorporates the issueof what keeps the stream reach from restoring itself.

e. The physical constraints on possible restoration measures such as waterquality, available rights-of-way or construction area, as well as budgetconstraints.

f. The range of alternative solutions that are both feasible and acceptable tothe stakeholders.

These issues are best addressed through a preliminary watershedreconnaissance, which is discussed in more detail in a later section.

Many studies initiate with the objective defined as simply �fixing� thestream. The generality of such an objective can lead to problems. Clearlydefining the objectives offers a clear approach as well as reduces ambiguity forthe study team members. Objectives should not only be specific but also berealistic and achievable.

Restoring a stream to a �natural� condition is also often used as an objectivein many studies. There can be an attraction to defining this natural condition byaesthetic guidelines, but care must be taken to assure that it is appropriate forboth the watershed type as well as the study constraints. It is also important torealize that a stream that is behaving �naturally� within the context of watershedconditions may still be detrimental to riparian land use and may not possess adesirable habitat. For example, a stream reach just above the confluence with amajor river could be braided, aggrading, lacking pools, and exhibiting frequentout-of-bank flooding yet be behaving naturally. This reach could be altered toreduce flooding to force the stream into a single thread to enhance in-stream fishhabitat for a particular fish type. However, since these changes would be alteringthe natural condition, it is important to recognize that considerably moreengineering effort would be required over what would be required if the targetconditions were more in keeping with the existing morphology of the stream.

Restoring streams to some historical condition or to a reference conditionmay be stated as an objective. If this is the approach, care must be taken toassure that physical or biological changes or differences in the watershed orwatersheds do not prohibit a return to a historic or reference condition. The riskis that the historical channel morphology would be inappropriate to the modernwatershed context, and thus would require heavy maintenance to sustain it. Forexample, the objective for an incised and widening stream in an urban watershedcould be to restore it to provide habitat for a sensitive fish species that waspresent before development. Changes in runoff patterns, sediment load andwater quality could make this objective impossible to achieve if the project focusis only on the study reach. Stormwater management, sediment trapping, andwater quality enhancement may be required to restore the water quality andsediment regimes to predevelopment conditions. The use of historic geometricconditions as an objective can also be problematic. Using the same hypotheticalurban stream as an example, the watershed conditions that caused the currentstream condition may cause the channel to alter again unless additionalstabilization measures are included in the restoration design.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (28)

Chapter 2 Project Objectives and Constraints 17

Some stream restoration studies have objectives defined in terms ofimproving the habitat for a particular species. This can provide general designobjectives such as target depths and velocities. One problem with this approachis that the needs of the target species may not be consistent with a channelmorphology appropriate to supporting a wide biodiversity in a varied ecosystem.A better approach may be to make biodiversity optimization an objective if it isachievable for the project area.

When a stream restoration project is part of a local flood damage reductionproject, design objectives are stated in ER 1110-2-1405. The hydraulic design ofa local flood protection project must result in a safe, efficient, reliable, and cost-effective project with appropriate consideration of environmental and socialaspects. Issues related to safety include potential hazards to humans andproperty, creation of a false sense of security, and consequences of flowsexceeding the design channel capacity. Issues related to efficiency include bothhydraulic conveyance and operation and maintenance. Issues related toreliability include achieving project purposes throughout the project economiclife, and the proper functioning of appurtenances such as gates, weirs, deflectors,and bank stabilization. Cost-effectiveness includes both the initial project costsand operation, maintenance, and replacement costs. Channel restoration projectsare often perceived to have significantly higher maintenance costs thantraditional single objective flood-control projects. This is generally attributed tocosts related to maintaining appropriate vegetation density on the overbanks.This cost is offset, however, in a properly designed restored channel that is self-sustaining in terms of sediment transport. Simple flood control channels thathave widths and depths that are much larger than the stable regime channeldimensions suffer chronic sedimentation problems and tend to lose conveyancecapacity due to invasive vegetation. Both problems require heavy maintenancecosts. Environmental and social aspects include fish and wildlife habitats,aesthetics, recreational opportunities, handicap access, and mitigation of adverseimpacts. ER 1110-2-1405 lays out the project hydraulic design process and theformat for hydraulic design reporting.

In general, the engineering means to achieve the objectives of streamrestoration can be divided into three general categories based on the focus of theproposed solution: (a) hydrologic work, (b) habitat work, and (c) hydraulic work.Hydrologic work can be accomplished through the use of stormwater ponds orthrough the modification of reservoir release schedules to modify the runoffregime as necessary to meet project objectives. Habitat work includes theconstruction of structures or features on the bed, bank, and/or riparian area tomodify the biologic function of the stream. This can include measures thatprovide in-stream cover, low-flow channels, scour holes, riparian plantings, andsubstrate modification. Hydraulic work includes a variety of techniques thatcenter on measures that affect the geomorphic characteristics of the channel.They can include measures to provide the channel dimensions and geometriesrequired to produce a stable or regime condition, local works essential to supplythe morphological diversity necessary to support a wide range of habitats, and thestructures needed to hold the channel in its new alignment by preventing bankerosion.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (29)

18 Chapter 2 Project Objectives and Constraints

Flood Damage Reduction TechniquesThere is growing public interest in modifying existing flood damage

reduction projects to restore natural environmental functions while still providingflood protection benefits. Careful planning, analysis, and design are required forthe successful implementation of these changes. Some of the most commonmodifications are listed here. They are listed in decreasing order of preferencefrom the standpoint of achieving a naturally behaving system.

a. Flood setbacks. This technique involves removing structures from thefloodplain and restoring the channel to its historic configuration. Thechannel dimensions are designed so that the inflowing sediment load canbe transported through the project reach without significant aggradationor degradation. Original planform can be restored. The stream is left tofreely meander and flood its overbanks. Overbank flooding should occurevery one or two years on the average. In urban areas, this option istypically infeasible due to real estate costs.

b. Levee setbacks. This technique is essentially the same as flood setbacks,except that the overbank floodplain is limited by levees. The leveesshould not encroach upon the meander belt so that the channel may stillmigrate within this morphologically active corridor.

c. Two stage channels. This type of project involves an upper channelsection to provide flood conveyance, with a natural low-flow channelwithin it to provide habitat enhancement and improved sedimenttransport capacity. The upper area can be designed to provide publicrecreation or wetland habitat. The low-flow channel can be designed in ameandering pattern within the upper channel. However, both channelsare essentially held in a static but geometrically stable condition. Naturalstability in terms of sediment transport may not be maintainable becausethe top bank elevation of the low-flow channel will be lower than thenatural bankfull elevation. Aggradation in the upper channel should beexpected if the stream is an alluvial stream, but not as much as wouldoccur in a large single-stage trapezoidal channel. Careful assessment ofsediment transport capacity is needed to design the low-flow channel.Modification of an existing flood-control channel to this type can involvealteration to bridge piers and abutments, alteration to utilities and realestate acquisition since the construction of the upper channel willtypically involve an expanded area in order to maintain floodconveyance.

d. Relief channels. This technique typically involves restoring the channelto its original configuration and constructing a high-flow channel orrelief culvert to provide flood conveyance. The restored channelprovides habitat benefits while the high-flow channel can be designed toprovide wetland or lowland habitat or for recreational benefits. Thehigh-flow channel is functioning as a detached floodplain. Real estatecosts for the high-flow channel can be an issue. Careful consideration ofthe sediment transport conditions of the stream is required. In streamswith high bedload, the loss of transport capacity at the entrance to the

Hydraulic Design of Stream Restoration Projects - [PDF Document] (30)

Chapter 2 Project Objectives and Constraints 19

bypass channel can result in sediment deposition in the restored channel.Where the relief channel reenters the restored channel, the increase insediment transport capacity can result in bed degradation.

e. Addition of in-stream habitat features. This can include the addition ofboulders, wing deflectors, stone weirs, and lunker-type habitats withinthe existing flood damage prevention project. A low-flow meanderingchannel may be established within the flood channel. This low-flowchannel is not the same as a natural regime channel and maintenanceafter flood events may be required. These features should be designed towithstand the forces of the flood flows. The effects of adding in-streamhabitat features on channel conveyance and sediment transport must beconsidered. The reliability of the flood control project should not becompromised.

f. Addition of bank vegetation. Trees and shrubs can provide lowlandhabitat, channel shading, and aesthetic benefits. This type of project isoften the easiest to implement since it involves minimal modifications tothe existing project. However, vegetation may increase the channelroughness and a careful hydraulic analysis is required to assess suchimpacts. A hydraulic analysis can be used to aid in the selection of plantspecies. In addition, the possible impacts of debris clogging oninfrastructure and channel stability should be assessed.

Additional information on ranking flood damage reduction alternatives isavailable in U.S. Army Corps of Engineers� Engineering Manuals. EM 1110-2-1418 provides a ranking based on channel stability and EM 1110-2-4000provides a ranking based on sedimentation issues.

Project ConstraintsThe process of determining constraints is just as important as establishing

objectives. An ideal stream restoration design might include a natural channelfree to migrate laterally and longitudinally down the valley, connectedhydraulically to its floodplain and with natural vegetation along the banks. Sucha design would preclude most types of development in the floodplain. This maynot be feasible, so a less than ideal solution may be required. Constraints areparticularly common in urban floodplains and include rights-of-way, highwaysand bridges, utility crossings, buildings, archeological and historical sites, andcemeteries. Another common concern is the erosion of polluted sediment in thestreambed or in the banks. To maintain water quality standards it may bebeneficial to make sure these polluted sediments stay in place. These constraintsmay make it necessary to stabilize the banks, preventing the natural channelmigration process. Another constraint common to Corps projects is flooddamage reduction. Overbank flows, which may be beneficial for habitatdevelopment, are often not acceptable for economic, political, and/or socialreasons.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (31)

20 Chapter 3 Hydrology

3 Hydrology

GeneralHydrologic computations are an integral part of stream restoration projects.

A wide variety of techniques are available to the designer. The level of accuracyrequired for a specific hydrologic analysis generally depends on the specificcharacteristics of each individual project. The selection of the appropriatemethodology should be done with a firm understanding of the assumptions,accuracy, data requirements, and limitations of the approach chosen. Thischapter outlines some of the most common techniques and offers generalguidelines regarding selection criteria. For more complete information on thedetails regarding the assumptions and limitations of specific models, the originaldocumentation associated with each of the models should be reviewed. Finaldecisions regarding the suitability of a particular model for a particular projectmust be determined using engineering judgment on a case-by-case basis.

In order to design a stream restoration project with long-term stability, it isnecessary to evaluate the full range of flows that will affect the channel. Thisrequires the development of a historical hydrograph and/or a flow-duration curve.Estimates and calculations of existing base flows, channel-forming flows, andflood flows are often required. Estimates of future flow conditions are oftenrequired to properly assess future project performance. Base flows often definecritical habitat conditions. Estimates of channel-forming discharges are used todetermine channel dimensions. Flood flow estimates are used to determinestability of structures and natural channel features, as well as for scour depthprediction. The choice of a maximum design flow for stability analysis should bebased on project objectives and consequences of failure. For example, the 100-year discharge might be used to design bank protection in a densely populatedarea while a 10-year discharge might be appropriate in a rural stream.

Frequency Analysis

Peak discharge analysis

The objective of hydrologic peak frequency analysis is to relate themagnitude of a given flow event with the frequency of that event�s occurrence.This is accomplished using stream flow gage records. Data can be stratified by

Hydraulic Design of Stream Restoration Projects - [PDF Document] (32)

Chapter 3 Hydrology 21

seasons depending on study goals. Gage records should contain at least 10 yearsof consecutive peak flow data, and they should span at least one wet year and onedry year. The frequency analysis requires that the flow data consist ofindependent events. A variety of hydrologic techniques are available for theprediction of the frequency of flow events. In general, the hydrologic analysisfor the gage should follow the recommendations of Guidelines for DeterminingFlood Flow Frequency, Bulletin 17B (USIACWD 1982). Typically, threedifferent asymptotic forms of extreme value distributions (Types 1, 2, and 3) areused in the frequency analysis.

Most estimates of infrequent flow events are made with the annual durationseries where the number of values in the data subset is equal to the number ofyears of record. An annual maximum (or minimum) duration series is composedof the largest (or smallest) value in each year. Since these events occur on anannual basis, it is usually safe to assume that each observation is independent.

When the desired event has a frequency of occurrence of less than 2 to 5years, a partial duration series is recommended. This is a subset of the completerecord where the values are above a preselected base value. The base value istypically chosen so that there are no more than three events in a given year. Inthis manner, the magnitude of events that are equaled or exceeded three times ayear can be estimated. Care must be taken to assure that small peaks associatedwith large events are not included in the analysis to ensure that independence ispreserved. The return period for events estimated with the use of a partialduration series is typically 0.5 years less than what is estimated by an annualseries (Linsley, Kohler, and Paulhus 1975). While this difference is fairly smallat large events (100 years for a partial vs. 100.5 years for an annual series), it canbe significant at more frequent events (1 year for a partial vs. 1.5 year for anannual series). It should also be noted that there is more subjectivity at the endsof both the annual and partial duration series.

Gage records provide an actual representation of the hydrologic behavior of awatershed. However, when a gage record is of short duration, or of poor quality,or the results are judged to be inconsistent with field observations or soundengineering judgment, then the analysis of the gage record should besupplemented with other methods. It is important to assess the applicability ofthe historic gage data to current conditions. For example, rapid increases ofimperviousness in an urban watershed may have increased peak flows, renderinghistoric gage data obsolete. Correction of gage data is possible but can beproblematic. If an invalid portion of a record is used, the results will be biased.It is also important to note that many stream restoration sites do not have anappropriate gage in the area.

Regional regression

Federal, state and local agencies have developed regional regression relationsto estimate peak discharges at ungaged sites. Regional regression equations areeasy-to-use and provide relatively reliable and consistent findings when appliedby hydraulic engineers. They are typically statistical models that quantify generalregional relationships between flow of a specific recurrence interval and a

Hydraulic Design of Stream Restoration Projects - [PDF Document] (33)

22 Chapter 3 Hydrology

watershed's physiographic, hydrologic, and meteorological characteristics. Themost simple regression relation is of the form:

= + ci iQ a b A (1)

where: Qi is the dependent variable, such as flow; Ai is an independent variable,such as drainage area; and a, b, and c are regression coefficients derived from thedatabase. The U.S. Geological Survey (USGS) has developed regionalregression relations for many states (USGS 1993). Multiple regression analysiscontains more than one independent variable and can thus account for morevariations between watersheds.

Errors in regression analysis

Regional regression relations should include relevant parameters to accountfor unique characteristics of the study watershed. It is also important to considerthe confidence limits of the regression relationships and how they relate topredicted discharges for the study reach. One of the most commonly usedmeasures of goodness of fit is the correlation coefficient, expressed typically asr2. This is a measure that describes how well a regression equation explains therelationship between the variables. It is the dimensionless ratio of the explainedvariation in the dependent variable over the total variation in the dependentvariable. A correlation coefficient of 1.00 indicates that the values of thedependent variable can be calculated exactly using the independent variable inthe given data set. Since this value is dimensionless, it can be used to comparegoodness of fit of different regression relations. It does not provide a quantifiedexpected variation. If the correlation is linear, it does not matter which variableis considered independent. However, if the relationship is nonlinear, theregression coefficients will be dependent on the choice of independent variableand curve fit. It should also be noted that a high degree of correlation (r2 close to1) does not necessarily imply causation or direct dependence between thevariables. In all cases, the reasonableness of the causation between theindependent and dependent variables should be examined. Additionally,variation is expected in natural systems. Data collection techniques should beexamined if the calculated r2 implies near perfect correlation (very close to 1) offield data.

Another measure of the robustness of a regression relation is the standarderror of estimate, expressed typically as SY,X. This is the root mean square of theestimates. It is a measure of the scatter about the regression line of theindependent variable. In general, the standard error of estimate is not reflexive,i.e., SY,X is not a measure of how well the independent variable correlates to thedependent variable. The standard of estimate has similar properties to thestandard deviation. Since the standard error of estimate has dimensions, itprovides a measure of possible variation.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (34)

Chapter 3 Hydrology 23

Flow-Duration CurvesDetermination of the quantity of sediment transport is dependent on flow

duration estimates. Flow-duration curves are typically calculated using the entireperiod of record. Procedures for this are described in �Hydrologic FrequencyAnalysis,� EM 1110-2-1415. Data are typically sorted by magnitude, and thepercent of the time that each value is exceeded is calculated. Data can bestratified by seasons depending on study goals. Regional flow-duration curvescan be developed using drainage area. Drainage area can be used to transferflow-duration information from gaged sites to ungaged areas; however, the ratioof gaged to ungaged drainage area should be between 0.5 and 2 for reliableresults. Typically, there is more error in transferring or estimating the ends of aflow-duration curve. Flow duration is dependent on watershed conditions. Ifregional flow duration relations are to be developed, it is recommended that ameasure of watershed development be included as an independent variable.

Two methods for estimating a flow-duration curve for ungaged sites aredescribed in Appendix A. They are the drainage area � flow-duration curvemethod and the regionalized duration curve method. With the drainage area �flow-duration curve method graphs of a specified recurrence interval dischargeversus drainage area are developed for a number of sites on the same stream orwithin hydrologically similar portions of the same drainage basin. If data arereasonably hom*ogenous, power functions may be fit using regression and used togenerate a flow-duration curve for the ungaged location. With the regionalizedduration curve method a nondimensional flow-duration curve is developed for ahydrologically similar gaged site by dividing discharge by bankfull discharge ora specified recurrence interval discharge. Then a specified recurrence intervaldischarge is computed for the ungaged site using the aforementioned USGSregression equations. Finally, the flow-duration curve for the ungaged site isderived by multiplying the dimensionless flows (Q/Q2) from the nondimensionalcurve by the site Q2. It should be noted that both methods simply provide anapproximation to the true flow-duration curve for the site because perfecthydrologic similarity never occurs.

It is often important to determine how the proposed restoration project willperform with low or normal flows. In addition, seasonal flow variations can havecritical habitat importance. For example, a project goal may include a minimumflow depth during a critical spawning period for salmonoid species and a lowerminimum depth for resident fish species. The same techniques used to developflow-duration curves for sediment analysis can also be used to assess and designfor habitat conditions.

Hydrologic ModelsHydrologic models have long been used to determine discharges for various

recurrence intervals. Models are particularly applicable where gages arenonexistent, limited, or do not reflect current conditions. Models provide theability to estimate existing as well as future rainfall runoff patterns for a varietyof conditions. The accuracy of models is dependent on calibration data, which

Hydraulic Design of Stream Restoration Projects - [PDF Document] (35)

24 Chapter 3 Hydrology

may not be available. However, if the issues that are to be addressed arecomparative in nature rather than absolute, the importance of calibration isdiminished. Brief statements on the use of the models are provided in thefollowing paragraphs.

The rational method (rational formula) is one of the easiest models toimplement. It is can be used for drainage areas up to 80 ha (200 acres). Use ofthe rational formula on larger drainage areas requires sound engineeringjudgment to ensure reasonable results. The hydrologic assumptions underlyingthe rational formula include constant and uniform rainfall over the entire basinand a rainfall duration equal to the time of concentration. If a basin has morethan one main drainage channel, if the basin is divided so that hydrologicproperties are significantly different in one section versus another, if the time ofconcentration is greater than 60 min, or if storage is an important factor, then therational method is not appropriate.

The National Resource Conservation Service (NRCS) (formally SCS) TR-55method (SCS 1986) provides a graphical method for computing peak dischargesof drainage basins with areas ranging from 4.0 ha (10 acres) up to 800 ha(2,000 acres, 3.1 square miles). The TR-55 method is segmental (i.e., flow time iscomputed by adding the travel times for the overland, shallow concentrated (rill),and channel segments). TR-55 considers hydrologic parameters such as slope,roughness, losses, rainfall intensity, soil type, land use, and time. Somehydrologists have stated that TR-55 tends to produce conservatively highestimates of peak flows. TR-55 should be used with caution when the design ishighly sensitive to the computed peak flow values. Although TR-55 has fewerassumptions than the rational formula, it also assumes that rainfall is uniformover the entire basin. Additional assumptions include:

a. Basin drained by a single main channel or by multiple channels withtimes of concentration within 10 percent of each other.

b. Time of concentration between 0.1 and 10 hr.

c. Storage in the drainage area is less than 5 percent of the runoff volumeand does not affect the time of concentration.

d. A single composite curve number can accurately represent the watershedrunoff characteristics.

The HEC-1 model is a rainfall-runoff model developed by the U.S. ArmyCorps of Engineers, Hydrologic Engineering Center (USAEHEC 1981). It can beused with basins of almost any size and complexity. HEC-1 is designed tosimulate the surface runoff resulting from precipitation over a watershed byrepresenting that watershed as an interconnected system of components. Thesecomponents consist of surface runoff, stream channels, and reservoirs. Eachcomponent is represented by a set of parameters, which specify itscharacteristics, and the mathematical relations, which describe its physicalprocesses. The end result of the HEC-1 modeling process is the computation ofrunoff hydrographs for the subbasins and stream channels. The program iscomposed of five basic submodels as illustrated in Figure 18.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (36)

Chapter 3 Hydrology 25

Figure 18. Submodels of HEC-HMS modeling process

HEC-1 assumes that the rainfall is spatially uniform over each subbasinmodeled. NRCS rainfall time distributions, loss methods, dimensionless unithydrographs, and the lag equations often are used; however, careful considerationmust be given to the assumptions and limitations underlying these methods. Forexample, the NRCS has published an upper limit on basin size for the NRCS lagequation of 800 ha (2,000 acres, 3.1 square miles) (SCS 1985). The upper limiton basin area for the NRCS loss method (i.e., runoff curve number) is not wellestablished; however, a limit of 52 km2 (20 square miles) has been suggested.These limitations may be overcome by subdivision of the watershed andappropriate routing. Various GIS packages can be used as an interface to HEC-1.These GIS techniques systematize the computation of the physiographic andhydrologic parameters required by HEC-1. Similar hydrologic models includeTR-20 and HEC-HMS.

Channel-Forming Discharge ConceptNatural alluvial streams experience a wide range of discharges and adjust

their shape and size during flow events that have sufficient energy to mobilizeeither the stream�s bed or banks. However, hydraulic design has been attemptedusing only a single representative discharge for many stream restoration projects.Using a representative or channel-forming discharge may be appropriate fordetermining initial or preliminary design channel dimensions, but the difficulty inthe determination of the channel-forming discharge and the uncertainty related tothe concept itself makes its sole use untenable for reliable and effective hydraulicdesign.

The channel-forming discharge concept is based on the idea that for a givenalluvial channel geometry there exists a single steady discharge that givenenough time would produce channel dimensions equivalent to those produced bythe natural long-term hydrograph. This discharge therefore dominates channelform and process and may be used to make morphological inferences. Althoughconceptually attractive, this definition is not necessarily physically feasiblebecause bank line vegetation, bank stability and even the bed configurationwould be different in a natural stream than in a stream with a constant discharge.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (37)

26 Chapter 3 Hydrology

The channel-forming discharge concept is not universally accepted. However,most river engineers and scientists agree that the concept has merit, at least forperennial and ephemeral streams in humid environments and perennial streams insemiarid environments. For channels in arid environments where runoff isgenerated by localized high-intensity storms and the absence of vegetationensures that the channel will adjust to each major flood event, the channel-forming discharge concept is generally not applicable.

Care must be exercised in applying the channel-forming discharge procedure,particularly in unstable channels and those that have experienced catastrophicevents during the period of record because flow-frequency and sediment-transport relations may have changed or be changing with time as the channeladjusts. Results may, therefore, represent a condition that does not accuratelydepict present flow and sediment-transport conditions.

Until the 1960s it was widely assumed that floods of great magnitude but lowfrequency controlled channel form because of the nonlinear relationship betweendischarge and sediment transport capacity. Sediment transport increasesexponentially with discharge. This view was challenged by Wolman and Miller(1960) who demonstrated that in most streams over an extended period of timethe total amount of sediment transported by a discharge of a given magnitudedepends not only on its transport capacity, but also its frequency of occurrence.Thus, although extremely large events can produce spectacularly high sedimentloads, they happen so infrequently and last such a short time that their overallcontribution to the total sediment movement during a long period is relativelysmall. Small events also make a small contribution to the total sediment movedbecause their high frequency of occurrence is offset by their very low sedimenttransport capacity. It follows from this logic that flows of both moderatemagnitude and moderate frequency are responsible for the greatest amount ofsediment movement. Wolman and Miller defined moderate frequency as eventsoccurring at least once each year or two and in many cases several or more timesper year.

Assigning a single value to this theoretical channel-forming discharge isproblematic. The following deterministic approximations for channel-formingdischarge have been suggested as follows:

a. The natural bankfull channel discharge.

b. A discharge based on statistical return intervals.

c. The effective discharge or that discharge which, over time, does the mostwork and transports the most sediment.

Systematic methodologies for determining each of these approximations forthe channel-forming discharge are presented in the following sections.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (38)

Chapter 3 Hydrology 27

Bankfull discharge

The bankfull discharge is the maximum discharge that the channel canconvey without overflowing onto the floodplain. This discharge is considered tohave morphological significance because it represents the breakpoint between theprocesses of channel formation and floodplain formation. Based on boththeoretical and empirical arguments, bankfull discharge is generally recognizedas being the moderate flow that best fits Wolman and Miller�s channel-formingdischarge concept for streams in dynamic equilibrium. Leopold, Wolman, andMiller (1964) proposed that bankfull discharge was responsible for maintainingchannel shape in natural alluvial channels and therefore was equivalent tochannel-forming discharge. However, in an unstable channel that is adjusting itsmorphology to changes in the hydrologic or sediment regime, bankfull dischargecan vary markedly from channel-forming discharge. Therefore, the expression�bankfull discharge� should not be used to refer to �channel-forming discharge,�but should be reserved to refer only to the maximum discharge that the channelcan convey without overflow onto the floodplain.

Bankfull discharge is determined first by identifying bankfull stage and thendetermining the discharge associated with that stage. Identifying the relevantfield features that define the bankfull stage can be problematic. Many fieldindicators have been proposed, but none appear to be generally applicable or freefrom subjectivity (Williams 1978). The most common definition of bankfullstage is the elevation of the active floodplain (Wolman and Leopold 1957; Nixon1959). Another common definition of bankfull stage is the elevation where thewidth to depth ratio is a minimum (Wolman 1955; Pickup and Warner 1976).This definition, diagramed in Figure 19, is systematic and relies only on accuratefield surveys. In some cases the highest elevation of channel bars may be used asan indicator of bankfull stage (Wolman and Leopold 1957). Woodyer (1968)defines the bankfull stage of streams having several overflow surfaces as theelevation of the middle bench. Wolman (1955) combines the width to depth ratiocriterion with identifying a discontinuity in the channel boundary such as achange in its sedimentary or vegetative characteristics. Schumm (1960) definedbankfull stage as the height of the lower limit of perennial vegetation, primarilytrees. Similarly, Leopold (1994) states that bankfull stage is indicated by achange in vegetation, such as herbs, grasses, and shrubs. Given the number ofcriteria in common use to define bankfull stage and the considerable experiencerequired to apply them, it is not surprising that there can be wide variability infield determination of bankfull stage.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (39)

28 Chapter 3 Hydrology

Figure 19. Bankfull depth using width-depthratio (after Knighton 1984)

The field identification of bankfull indicators is often difficult and subjective(Knighton 1984). The stream reach should be identified as stable and alluvialbefore field personnel attempt to identify bankfull stage indicators. If the projectreach is unstable or nonalluvial, it may be possible to find indicators of bankfullstage in stable alluvial reaches upstream or downstream on the same stream. Theprocess of identifying bankfull indicators is often an iterative and subjectiveprocess that involves a great deal of judgment.

If a reach is not stable and not alluvial, indicators of bankfull stage will beunreliable. Some examples are given as follows:

a. If a reach is nonalluvial, then sediment transport capacity normallyexceeds sediment supply, and deposits would be missing orunderdeveloped. Using underdeveloped deposits as bankfull indicatorswould result in too low a channel-forming discharge. Deposits couldalso be relics of extreme flood events, in which case they wouldnormally give too high a channel-forming discharge.

b. If the channel is degrading, then sediment transport capacity exceedssediment supply, and the observations above for the nonalluvial channelhold true. In addition, since the bed of the channel is lowering, formerfloodplain deposits are being abandoned (they are in the process of

Hydraulic Design of Stream Restoration Projects - [PDF Document] (40)

Chapter 3 Hydrology 29

becoming terraces.) Using these deposits as indicators would give toohigh a channel-forming discharge.

c. If the channel is aggrading, the in-channel deposits could be incorrectlymistaken for bankfull stage indicators. Since the bed of the stream isrising, using the existing floodplain as an indicator would give too low adischarge. (The floodplain will aggrade as well, but usually at a slowerrate than the channel.)

Confusion often occurs when criteria suggest a bankfull stage at an elevationthat is not close to the top of either bank. This condition suggests that thechannel is not in equilibrium, that the existing channel geometry may not bestable, and that the channel-forming discharge would be poorly approximated bythe bankfull discharge. Since stream restoration is most often practiced inunstable channels and watersheds (instability is often the reason for restoration),field determination of bankfull stage may be impractical or impossible. In fact,attempting to determine a channel-forming discharge from an unstable stream isin conflict with the theoretical premise that is the basis for the channel-formingdischarge concept.

Once bankfull stages are estimated for a reach of the stream, then bankfulldischarge can be estimated. Ideally, the discharge associated with bankfull stagecan be determined from a stage-discharge rating curve based on measured data atthe project site. When floodplain conveyance is significant with respect tochannel conveyance, there will be a distinct break in the stage-discharge ratingcurve at bankfull stage as shown in Figure 20. The data scatter in Figure 20occurs because stage is not a unique function of discharge in alluvial streams. Itis therefore necessary to estimate a rating curve through the data scatter. It isbest to consider that the bankfull discharge will have a range rather than a singlediscrete value. Uncertainty associated with the stage-discharge relationship isaddressed in EM 1110-2-1619. In cases where floodplain conveyance is notsignificant with respect to channel conveyance, there may not be a distinct breakin the stage-discharge rating curve (Figure 21). In this case the bankfulldischarge may not have as much morphological significance as when floodplainflow is significant. Lacking gage data at the site, a stage-discharge rating curvecan be determined from a backwater analysis. Ideally, the downstream startingwater-surface elevation will be based on data from a gaging station. Theaccuracy of this rating curve will depend on the uncertainties associated withassigned hydraulic roughness coefficients and the cross-section geometry.Uncertainty is greatest when the stage-discharge rating curve is estimated from asingle cross section. In this case both hydraulic roughness and energy slope mustbe assigned. It is best if the determination of bankfull stage occurs over a reachlength of at least one wavelength or 10 channel widths. An example of a

Hydraulic Design of Stream Restoration Projects - [PDF Document] (41)

30 Chapter 3 Hydrology

comparison of bankfull stage and a computed water-surface elevation is shown inFigure 22. Note in Figure 22 that bankfull stage is taken to be at the bottom ofthe top-of-bank data scatter because this represents the elevation at which flowonto the floodplain begins. Also note that considerable variability in bankfullstage could be estimated if only a single top-of-bank point were used in theanalysis. The hydraulic engineer determines what method is best suited tocompute the bankfull discharge from the bankfull stage indicators. For example,backwater computations may be required in some cases, while normal depthcomputations will be sufficient in others.

Figure 20. Stage-discharge rating curve Bogue Chitto River near Bush, LA

Hydraulic Design of Stream Restoration Projects - [PDF Document] (42)

Chapter 3 Hydrology 31

Figure 21. Stage-discharge rating curve Mississippi River at Tarbert Landing,MS

Figure 22. Long-channel variation in bank top elevations: Lower MississippiRiver (Biedenharn and Thorne 1994)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (43)

32 Chapter 3 Hydrology

The following guidelines are provided relative to field determination ofbankfull discharge and use of bankfull discharge as the channel-formingdischarge:

a. Bankfull discharge is geomorphologically significant only in stablealluvial channels. Therefore, the reach where bankfull stages aredetermined should be stable and the streambed should be mobile atbankfull flow.

b. The estimates of bankfull discharge most appropriately used to determinechannel dimensions for the main channel are those based upon top-of-bank indicators. A stage identified by the edge of the active channel, thebeginning of woody vegetation, or the top of channel bars may havevalue for designing those particular features in a restored channel, butshould not be used for establishing the bank height of a stable channel.Only bankfull discharges that are top-of-bank discharges aremorphologically significant in establishing the channel-formingdischarge.

c. An exception to the preceding rule is in a stable and alluvial incisedstream that has formed a new floodplain within the incised channel. Inthis case, the top of the high bank is now an abandoned floodplain orterrace, and there should be newly formed top-of-bank features withinthe older incised channel. However, it is important to remember that thenew floodplain may not yet be fully formed; that is, the channel may notbe stable (it may still be aggrading). This would give misleading valuesfor the bankfull discharge.

d. Assuming that the bankfull discharge for one reach of a stream is thesame as the bankfull discharge in another reach may not be appropriate.The location of the break between the channel and the floodplain isinfluenced by many factors, including (but not limited to) the following:

(1) confinement of the floodplain

(2) hydrologic regime

(3) sediment supply

(4) bed and bank sediment size and cohesiveness

(5) size and type of vegetation on the floodplain and within the channel

(6) controls on channel width, slope and alignment

For example, the bankfull discharge taken from a reach with a narrowfloodplain may be inappropriate for use on another reach on the samestream, which has a wide floodplain.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (44)

Chapter 3 Hydrology 33

Discharge for a specific recurrence interval

Due to difficulties in the identification of bankfull stage and discharge, manyresearchers have related the channel-forming discharge to a specific recurrenceinterval discharge. In these studies the researchers have typically studied stablestreams where bankfull stage could readily be determined and where streamgages were located nearby. Under these conditions, bankfull discharge andchannel-forming discharge were assumed equivalent and most of the literatureaddressing specific return interval discharges uses the two terms interchangeably.This can be confusing, and it should be remembered that these studies areactually comparing two methods for approximating the channel-formingdischarge, and not actually comparing an approximation method to the truevalue.

In general, bankfull discharge in stable channels is often assumed tocorrespond to an annual flood recurrence interval of approximately 1 to 2.5 yearsand the 1.5-year recurrence flood has been shown to be a representative mean ofmany streams (Leopold 1994). Wolman and Leopold (1957) suggested that thebankfull discharge had a recurrence interval of 1 to 2 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 forthree British gravel bed rivers, the 1.5-year discharge provided a water-surfaceelevation that passed through the scatter of bankfull discharges measured alongthe course of the rivers. Richards (1982) suggests that in a partial duration series,bankfull discharge equals the most probable annual flood, which has a 1-yearreturn period. However, there are many instances where the channel-formingdischarge does not fall within the 1 to 2.5-year range. Recurrence intervalrelations are intrinsically different for channels with flashy hydrology than forthose with less variable flows. For instance, Williams (1978) clearly showed thatout of 35 floodplains he studied in the United States, the bankfull dischargevaried between the 1.01- and 32-year recurrence interval, and that only about athird of those streams had a bankfull discharge recurrence interval between 1 and5 years. In a similar study, Pickup and Warner (1976) determined that bankfullrecurrence intervals ranged from 4 to 10 years. Because of such discrepancies,many have concluded that recurrence interval approaches tend to generate poorestimates of bankfull discharge. Hence, field verification is recommended toinsure that the selected discharge reflects morphologically significant features.

Effective discharge

The effective discharge is defined as the mean of the discharge incrementthat transports the largest fraction of the annual sediment load over a period ofyears (Andrews 1980). The effective discharge incorporates the principleprescribed by Wolman and Miller (1960) that the channel-forming discharge is afunction of both the magnitude of the event and its frequency of occurrence. Anadvantage of using the effective discharge is that it is a calculated value notsubject to the problems associated with determining field indicators. It iscalculated by integrating the flow-duration curve and a bed-material-sedimentrating curve. A graphical representation of the relationship between sedimenttransport, frequency of the transport, and the effective discharge is shown in

Hydraulic Design of Stream Restoration Projects - [PDF Document] (45)

34 Chapter 3 Hydrology

Figure 23. The peak of curve C from Figure 23 marks the discharge that is mosteffective in transporting sediment, and therefore it is hypothesized that it does themost work in forming the channel.

Figure 23. Derivation of total sediment load-discharge histogram (c) from flowfrequency (a) and sediment load rating curves (b)

In various types of stable alluvial streams researchers have demonstrated theequivalence between bankfull and effective discharges (Andrews 1980; Carling1988; Hey 1997). However, the effective and bankfull discharges are not alwaysequivalent as reported by Benson and Thomas (1966); Pickup and Warner(1976); Webb and Walling (1982); Nolan, Lisle, and Kelsey (1987); and Lyons,Purcherelli, and Clark (1992). This suggests that the effective discharge may notalways be an adequate surrogate for the channel-forming discharge. However, itmay simply reflect uncertainties in (a) the determination of the bankfullelevation, (b) the calculation of the bankfull discharge corresponding to thatelevation, (c) the calculated sediment transport rating curve and/or (d) theinherent uncertainties in the effective discharge calculation.

The recommended procedure to determine the effective discharge is executedin three steps. They are as follows:

a. The flow-duration curve is derived from available stream gage data.

b. A bed-material-sediment rating curve is constructed from sediment dataor calculated using a bed-material sediment transport equation.

c. The flow-duration curve and the bed-material-sediment rating curve areintegrated to produce a sediment load histogram that displays sedimentload as a function of discharge for the period of record. The histogrampeak is the effective discharge increment.

Calculating a flow-duration curve involves selecting the type of dischargedata to be used and a method for subdividing the observed range of discharge

Hydraulic Design of Stream Restoration Projects - [PDF Document] (46)

Chapter 3 Hydrology 35

into classes. The period of record should be at least 10 to 15 years. In manycases, mean-daily discharges are used because these data are readily availablefrom the U.S. Geological Survey (USGS) and others. However, except for largerivers, mean-daily flows tend to be low-flow biased because they mask theeffects of short-duration peak flows on sediment transport. Dischargesrepresenting shorter time periods than a day, such as the 15-min data collected bythe USGS, provide a more accurate means of establishing a sediment transportrating relation. These data, although superior for a broader size range of streamsand rivers, are not readily available, but may sometimes be obtained via specialrequest. The discharge data is typically divided into about 25 class intervals ofequal size, although the appropriate number of intervals may vary. Effectivedischarge calculations require the use of arithmetic intervals for the dischargeclass. This is different from calculating a flow duration curve where logarithmicintervals are frequently employed. One indicator that additional intervals may beneeded is when the discharge mode occurs in the lowest discharge class, whichfrequently occurs in small, flashy streams.

A bed-material-sediment rating curve, showing sediment concentration as afunction of water discharge, can be calculated or determined from measured data.If measured suspended data are used, the wash load component should beremoved from the total concentration. If sediment data are not available, bed-material sediment transport rates can be calculated from a variety of sedimenttransport functions. Frequently, the logarithms of sediment concentrations areplotted against the logarithms of discharge and regressed to create a simple ratingrelation. However, power functions derived this way are often inadequate todefine the transport relation because they overestimate transport at high flows. Inaddition, depending on how the bed-material varies with discharge, and whenand how the bed-material gradation data used in the sediment transport equationwere determined, the predicted transport can also be overestimated at lowdischarges. This necessitates using two or three linear segments or a curvedrating.

The sediment load histogram is developed by multiplying the frequency (inpercent) in each discharge class by the bed-material sediment transport ratecorresponding to the mean of the class interval. The mean of the class intervalwith the greatest transport load is the effective discharge. In some cases,however, there may not be a single class interval representing a maximum.Instead, the peak average transport rate may spread across a range of classes,indicating that there is no single effective discharge but that significantgeomorphic work is performed by a range of flows.

Since channel instability is the result of an imbalance in sediment supply andtransport capacity, the greatest advantage of using effective discharge inrestoration design lies in the fact that it requires quantification of the sedimenttransport capacity of a channel for a given hydrologic regime. Various channelgeometries can be examined for their competence to transport the incomingsediment load, facilitating comparison of permutations of channel dimensions inorder to optimize sediment transport efficiency within logistical constraints. Thisinformation is also useful when predicting the impact of alteration of watershedconditions with respect to sediment loads (e.g., upstream dam removal) orhydrology (e.g., urbanization) on channel stability.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (47)

36 Chapter 3 Hydrology

Examples of channel-forming discharge representations

Using data collected from 57 stable sand bed rivers in the United States,Thorne and Soar (2000) compared the bankfull discharge with otherrepresentations of channel-forming discharge. The bankfull discharge was takenas the best representation of channel-forming discharge because the rivers werestable. Of course, the bankfull discharge, in this case, was calculated frommeasured hydraulic dimensions, slope estimated from topographic maps and fieldestimates of hydraulic roughness and must therefore also be considered anapproximation of the channel-forming discharge.

The effective discharge is compared to the calculated bankfull discharge inFigure 24. This figure confirms the results of earlier research that the effectivedischarge is significantly lower than the bankfull discharge in most cases,particularly at low discharges, and only approximates the bankfull discharge athigh discharges in a small proportion of streams. In general, the effectivedischarge provides an adequate lower bound for the range of bankfull discharges.

Figure 24. Relationship between effective discharge, Qe, and bankfull discharge,Qb, for 57 U.S. sand bed rivers. Solid line is the best-fit powerrelationship. Dotted line is equality

In contrast, as shown in Figure 25, the 2-year flow event is greater than thebankfull discharge in most cases and provides an adequate upper bound to therange of bankfull discharges. The best-fit line in Figure 25 is linear at the 95percent significance level and represents bankfull discharges at approximately 60percent of 2-year flow over the range of the data.






1 10 100 1000 10000

Effective Discharge, Q e (m3s-1)





ge, Q

b (m

3 s-1)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (48)

Chapter 3 Hydrology 37

Figure 25. Relationship between the 2-year return period flow, Q2, and bankfulldischarge, Qb, for 57 U.S. sand bed rivers. Solid line is the best-fitpower relationship. Dotted line is equality

The effective discharge is statistically defined as the steepest gradient on thecumulative sediment frequency curve. Thorne and Soar (2000) hypothesized thatthe median discharge on this curve might have greater morphologicalsignificance than the effective discharge. The median discharge is defined as theupper limit of the range of discharges that transport 50 percent of the averageannual bed material load (Qe50). Although providing a closer association withbankfull discharge, the median flow with respect to sediment transport, Qe50,underestimates the bankfull discharge in most cases (Figure 26). However, the75 percent flow with respect to sediment transport, Qe75, provides a betterrelationship with bankfull discharge with an r2 value of 0.82 for the best-fit line(Figure 27). Furthermore, there is no statistical difference between the best-fitline in Figure 27 and the line of perfect agreement at the 95 percent significancelevel. The Qe75 discharge corresponds in many cases to either a high in-bank flowor a flow that just overtops the bank, and could provide engineers with a usefulalternative to the effective discharge as a surrogate for the channel-formingdischarge in stable sand bed rivers.





10 100 1000 10000

2-yr Return Period Flow, Q 2 (m3s-1)





ge, Q

b (m

3 s-1)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (49)

38 Chapter 3 Hydrology

Figure 26. Relationship between the discharge marking the upper limit of therange of discharges that transport 50 percent of the average annualbed material load, Qe50, and bankfull discharge, Qb, for 57 U.S. sandbed rivers. Solid line is the best-fit power relationship. Dotted line isequality

Figure 27. Relationship between the discharge marking the upper limit of therange of discharges that transport 75 percent of the average annualbed material load, Qe75, and bankfull discharge, Qb, for 57 U.S. sandbed rivers. Solid line is the best-fit power relationship. Dotted line isequality





10 100 1000 10000

Q e50 (m3s-1)





ge, Q

b (m

3 s-1)





10 100 1000 10000

Q e75 (m3s-1)





ge, Q

b (m

3 s-1)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (50)

Chapter 3 Hydrology 39

Channel-forming discharge related to drainage area

Use of regional regression curves for determining channel-forming dischargeas a sole function of the drainage area is not recommended. Drainage area isonly one of many parameters affecting runoff.

Within physiographically similar watersheds it may be useful to develop achannel-forming discharge versus drainage area curve for use in that watershed.Emmett (1975) developed such a curve for the Salmon River in Idaho (Figure28). Emmett chose stable channel reaches for his study and assumed thatbankfull discharge was equivalent to channel-forming discharge. Although theregression line fits the data in a visually satisfactory fashion, it should be notedthat for a drainage area of about 181 km2 (70 square miles), the bankfulldischarge varied between about 8.5 m3/s and 25.5 m3/s (300ft3/s and 900 ft3/s).This large range should not necessarily be attributed to errors in fieldmeasurements, but rather to the natural variation in bankfull discharge withdrainage area.

Figure 28. Bankfull discharge as a function of drainage area (Emmett 1975)

Channel-Forming Discharge SummaryChannel forming discharge can be estimated using a prescribed

methodology. One such deterministic discharge is the bankfull discharge.Another deterministic discharge used to represent the channel-forming dischargeis a specified recurrence interval discharge, typically between 1 and 3 years. Thethird is effective discharge.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (51)

40 Chapter 3 Hydrology

All three methodologies for estimating channel-forming discharge presentchallenges. In practice, problems often arise when attempting to identifybankfull stage in the field. Although several criteria have been identified to assistin field identification of bankfull stage, ranging from vegetation boundaries tomorphological breaks in bank profiles, considerable expertise is required to applythese in practice, especially on streams which have in the past undergoneaggradation and degradation. Recurrence intervals for channel-formingdischarge are generally in the range of 1 to 3 years, but have been shown to varywidely (4 to 10 years) for different types of streams. Calculation of effectivedischarge requires hydrologic and sediment data. Without nearby gage data,effective discharge calculations require use of an assumed hydraulic roughnessand selection of a reliable sediment transport equation. In light of thesechallenges, it is recommended that all three methods be used and cross-checkedagainst each other to reduce the uncertainty in the final estimate of the channelforming flow.

Stormwater ManagementStormwater management can be an important component of a broad

restoration program in urban and suburban watersheds. Stormwater managementcan provide for channel stability as well as habitat benefits. There are a variety ofstormwater management techniques that are in use today. An appropriatetechnique should be selected based on a firm understanding of the technique�slimitations and capabilities and the hydrologic effects of urbanization. Thehydrologic implications of various management practices need to be taken intoconsideration. This section outlines some of the more common techniques andoffers general guidelines regarding selection criteria.

General hydrologic effects of urbanization and stormwatermanagement

The primary hydrologic result of urbanization is an increase of runoff from arainfall event of a given recurrence interval. Urbanization results in an increase inthe impervious area of the watershed. Impervious areas such as parking lots,roads, and roofs increase the rainfall excess by reducing the volume of rainfallthat can be absorbed through infiltration. Gutters, culverts, and storm sewers alsoreduce the travel time across the watershed, which increases the rate of rise of therunoff hydrograph. In addition, as an area is developed, natural ponding areasare reduced which further increases the rate of rise and total volume of the urbanrunoff hydrograph. Only in rare cases, such as the development of poorly tilledcropland into large lot residential areas, would runoff volume decrease as thewatershed developed.

A large watershed can typically be broken into three areas relative to aproject location as shown in Figure 29.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (52)

Chapter 3 Hydrology 41

Figure 29. Schematic of a watershed relative toproject location

The upper portion of the watershed is typically very sensitive todevelopment. Stormwater management should focus on long detention times inthis area to prevent the upper watershed peak flows from coinciding with thepeaks in the lower portion of the watershed. Measures that decrease travel timewithin reaches such as piping and channelization can have detrimental affectsdownstream for similar reasons. Since the base flows are naturally lower in thisupper portion, infiltration measures are especially important to maintain baseflow for habitat purposes.

In the middle portion of the watershed, the detention should be intermediatein nature. The goal for stormwater management in this area is to delay flowslong enough to allow runoff from the lower portion to clear the watershed.However, the delay should not be significant enough to cause an overlappingwith the peak flows from the upper portion of the watershed.

In the lower area, stormwater management that relies on long detention timesshould be avoided since it may result in peak runoff from this lower area beingdelayed to coincide with the peak runoff from the upper watershed. As a result,this lower area is typically less sensitive to development.

Watersheds and development patterns are unique. It is recommended that abasin hydrologic model be developed and used in conjunction with stormwaterplanning to avoid adverse interactions between stormwater managementmeasures.

The four basic types of stormwater management ponds are listed in thefollowing sections. New structures often make use of features from all four.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (53)

42 Chapter 3 Hydrology

Conventional dry ponds

Since the 1970s, many state governments have required that the hydrologiceffects of urbanization be mitigated through the implementation of stormwatermanagement practices primarily of the form of stormwater ponds. Stormwaterponds are designed to reduce the effects of development on nuisance levelflooding. The ponds are directed towards maintaining the post development peakdischarge of the 2-year and 10-year storm events and have been locally veryeffective for this design goal. While stormwater ponds have met with success inreducing the peaks of storm events, conventional dry ponds do little to reduce theoverall quantity of runoff. It has only been in recent years that the use ofstormwater ponds for water quality enhancement and channel-forming flows hasbeen implemented. In addition, research has suggested that if stormwater pondsare designed without consideration to their relationship in the watershed, theirinteraction may result in an increase of peak discharge over what would haveoccurred if they had not been constructed at all (Ferguson 1991). As a result, itis advisable to develop a watershed hydrologic model if stormwater managementis a significant portion of a watershed study so that the impacts of existing andproposed stormwater ponds on the watershed can be determined.

A conventional flood reduction stormwater management pond can adverselyaffect stability of the downstream channel. Figure 30 illustrates an idealizedeffect of a stormwater pond. In this example, the pond maintains the peakvelocities of post development conditions at the predevelopment level. While thismight indicate that flooding would not be exacerbated, a geomorphic or stabilityassessment is required to determine if the channel stability is adversely affected.For example, if the erosion threshold velocity is V2, then postdevelopmentconditions with the pond should have minimal effect on the channel morphology.However, if the erosion threshold velocity is V1, then the pond can increasechannel erosion. Since channel erosion is time dependent, a pond can makedownstream channel erosion worse over conditions that existed without a pond.

Many stormwater management plans recommend a series of stormwaterponds throughout the watershed. This is schematically illustrated in Figure 31.The use of a number of ponds at the upper portions of the watershed will allowfor a more uniform control of the entire watershed hydrology. It is typicallyeasier to achieve multiple objectives (flood reduction, stability, and ecological)with multiple ponds. However, it is recommended that a watershed model beused for the planning and permitting of these features in order to avoid adverseinteractions between the ponds. A drawback to the use of multiple ponds is thetypical increase for maintenance over what is typically required for a singlelarger structure. A benefit is that since these are typically smaller structures, thedam safety requirements are typically less than for larger dams and thus they aresimpler to design and construct.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (54)

Chapter 3 Hydrology 43

Figure 30. An idealized effect of stormwater pond on channel velocities

Figure 31. Schematic of a watershed relative tomultiple small ponds

Retrofitting older storm water ponds to provide additional features such asthose documented herein can often provide significant benefits. Existingstormwater ponds throughout the Northeast are being retrofitted to provide suchbenefits. Utilizing the existing stormwater drainage system and increasing thestorage volume of an existing pond can be a cost-effective approach to streamrestoration in urban and suburban watersheds.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (55)

44 Chapter 3 Hydrology

Extended detention ponds

Extended detention ponds can provide both water quality benefits and reduceerosive flows. The most common design storms are the 1-year rainfall event orthe event that generates 0.5 in. of runoff. The first 0.5 in. of runoff is consideredto provide a �first flush� of the watershed and contains a significantconcentration of pollutants. The 1-year event is also considered for erosioncontrol. The design storms are detained for 12 to 24 hr as measured between thecentroid of the inflow to the centroid of the outflow hydrograph. This results in alonger detention time and a decrease in the peak discharge over what would haveoccurred without the pond. The water quality benefits are provided by detainingwater for enough time to allow sediments (and their attached pollutants) to settleto the bottom of the pond. The stream stability benefits are based on the premisethat the increased volume of runoff from the developed watershed is offset by areduced peak discharge.

As noted to be the case with dry ponds, if stormwater ponds are designedwithout consideration to their relationship in the watershed, their interaction mayresult in an increase of peak discharge over what would have occurred if they hadnot been constructed at all. Maintenance requirements should be considered dueto the settling out of suspended sediments. To facilitate maintenance, a sedimentforebay is recommended for these systems. Effects of the structure on fishpassage as well as thermal loading to the stream should also be considered.Temperature of water stored in detention ponds typically increases with time andmay adversely affect cold water fisheries downstream.

Wetland-pond systems

Wetland-pond systems are used to provide aesthetic, habitat, and waterquality benefits. Often, large systems include nature and fitness trails. Habitatbenefits can be provided with high and low marshes, nesting islands, and plantingdiversity. It is important to note that since these features are a sink for a varietyof pollutants, the choice of planted species is more limited than in conventionalwetland creation sites. Water quality enhancement is a result primarily of thesettlement of pollutant-laden sediment, and physical filtration of particulatematter as well as nutrient uptake.

As with any shallow impoundment, a drawback for the use of wetland-pondsystems is primarily thermal loading to downstream reaches. Effects of thestructure on fish passage as well as public safety should also be considered.Maintenance can be more extensive due to the settling out of suspendedsediments. If they are not maintained, they can become a source of pollutantsduring dry weather. To facilitate maintenance, a sediment forebay isrecommended for these systems.

Infiltration basins and bioretention

Infiltration designs are often a preferred first choice since they seek to mimicpredevelopment hydrology. They provide quasihabitat benefits through

Hydraulic Design of Stream Restoration Projects - [PDF Document] (56)

Chapter 3 Hydrology 45

increased base flow and water quality benefits though filtration. However, theyare limited to areas that have well drained soils and often require large areas. Asediment forebay is recommended since many of the infiltration designs that arecurrently in use are prone to failure by clogging.

Bioretention projects typically involve the use of shallow ponding areas andinfiltration. The use of mulching and vegetation reduces the possibility ofclogging and failure of the infiltration components of the bioretention systems.Because they are relatively small, they can be incorporated into the landscapingplans of almost any site. The primary benefit of this type of project is improvedwater quality and the maintenance of base flow. Bioretention and infiltrationdesigns typically do not affect runoff during larger events.

Stormwater management guidance

There are a wide variety of reports, technical papers, and manuals thataddress different aspects of stormwater management design and usage (AmericanSociety of Civil Engineers 1993; McCuen and Moglen 1988; Moglen andMcCuen 1988; Ferguson 1994; and Schueler 1994). State and local governmentsare often in the forefront of the development of design guidance (MarylandDepartment of the Environment, 1984 and 1987; and Prince Georges County,MD 1997).

Hydraulic Design of Stream Restoration Projects - [PDF Document] (57)

46 Chapter 4 Stability Analysis

4 Stability Analysis

Stability analyses are necessary for stream restoration projects. The purposeof the stability analysis is to identify the dominant fluvial processes in the streamsystem. Knowledge of dominant channel processes allows prediction of theproposed project�s impact on channel morphology and channel stability and theeffect the natural processes will have on the functionality of the project. Anaccurate assessment of the stability of various stream reaches and the types ofinstability occurring in the stream system (aggradation, degradation, planforminstability, etc.) is the foundation for the designer�s understanding of thewatershed�s dominant physical processes.

The most basic form of stability analysis is the assessment of bed stability -the determination of whether the channel bed is aggrading, degrading, or stable.Other aspects of stability assessment are bank stability, planform stability,historic or future changes in hydrology or sediment inflow, and changes inchannel width or cross section. This chapter will discuss the methods availablefor assessing channel stability. An example scope of work for a stability analysisis given in Appendix F.

A channel is considered stable when the prevailing flow and sedimentregimes do not lead to aggradation or degradation or to changes in the channelcross-sectional geometry over the medium to long term. It is important torecognize that short-term changes in sediment storage, channel shape, andplanform are both inevitable and acceptable in natural channels with unprotectedbank lines. Evaluation of stability can be undertaken at various levels, rangingfrom geomorphic assessments based on qualitative methods to quantitativetechniques using numerical data and analytical techniques. There are three levelsof stability assessment ranging from empirical reconnaissance-level methods tomore process-based analytical techniques (Figure 32).

Hydraulic Design of Stream Restoration Projects - [PDF Document] (58)

Chapter 4 Stability Analysis 47

Figure 32. Levels of stability assessment

The appropriate level of detail for a particular evaluation depends on thestatus of the study, the perceived seriousness of potential problems, the scale ofthe project and the resources available (EM 1110-2-1418). This study procedureis consistent with the staged sediment study approach outlined in EM 1110-2-4000. In the Corps of Engineers, all three levels of the stability assessment arethe functional responsibility of the Hydraulic and Hydrologic EngineeringSections and should be performed by an engineer with experience in riverengineering and geomorphology.

Geomorphic AssessmentThe geomorphic assessment provides the process-based framework to define

past and present watershed dynamics, develop integrated solutions, and assess theconsequences of remedial actions. This is an essential part of the design processwhether planning bank protection for a single stream bank or attempting todevelop a comprehensive plan for an entire watershed. A geomorphicassessment may be divided into the following three components: (a) dataassembly; (b) field investigation; and (c) identification of geomorphologicallysimilar reaches. Channel classification is also discussed under this heading.

Hydrology AssessmentsHydrology Assessments

Geomorphic AssessmentGeomorphic Assessment

Analytical StabilityAssessment

Analytical StabilityAssessment

Hydraulic GeometryAssessment

Hydraulic GeometryAssessment

Level 1Data assemblyField investigationIdentification of similar reachesHistorical channel stability

Regime curvesConfidence bands

Bed stabilitySediment budgetNon-equilibriumSediment transport

Level 2

Level 3

Hydraulic DesignHydraulic Design

Hydraulic Design of Stream Restoration Projects - [PDF Document] (59)

48 Chapter 4 Stability Analysis

Data assembly

The first step in the geomorphic assessment is to gather and compile existingdata. Historical data are used to identify trends, to provide information on ratesof landform change in the watershed, and to help the engineer determine land useimpacts upon current conditions. Data requirements depend upon projectobjectives and watershed characteristics. Guidelines for data collection areprovided in EM 1110-2-4000, EM 1110-2-1418, and Biedenharn et al. (1997).

Field investigations

Field reconnaissance is undertaken to gather data and make observationsleading to an understanding of the active processes and condition of the stream.It is critical that experienced personnel conduct this effort. Field reconnaissanceis used to describe the geomorphological landform of study reaches and toidentify potentially destabilizing phenomena based on reach-scale evidence oferosion, sediment storage, and deposition. Basic information on how to conductfield investigations to collect data for channel stability assessment is contained inthe following publications: EM 1110-2-4000; EM 1110-2-1418; Biedenharn et al.(1997), which contains detailed discussion on field equipment and a descriptionof features to look for in the field; and Thorne (1993).

Collection of field data can be aided with the use of appropriate fieldassessment data sheets. Example data sheets are provided in Appendixes Band C. These sheets are comprehensive and should be adapted to specific studyneeds. Guidance for carrying out detailed reconnaissance surveys is given byDowns and Thorne (1996); Thorne, Simon, and Allen (1996); and Thorne (1998).The level of effort required to conduct a field reconnaissance varies depending onconditions. It is recommended that a consistent technique be utilized and that itbe tailored to the watershed conditions and the study goals. It is alsorecommended that a trial run be conducted with a formulated field sheet to assesstime requirements and assessment coverage before initiation of a largewatershed-level field effort.

Field assessments are best made during low-water conditions and during thedormant season when the banks can be more readily examined. However, it isimportant to recognize that conditions may be different at high flows. For safetyand logistical reasons, field work is best accomplished by teams of at least twopeople. Field work (particularly in urban areas) may raise significant health andsafety issues. Potential hazards include crime, needles, and exposure to rawsewage and waterborne pathogens such as hepatitis. It is recommended that aminimal team consist of a biologist who is familiar with characteristics of aquaticand riparian habitat of the study area and an engineer who is experienced inhydraulics, hydrology, geomorphology, and sediment transport. Field work goesmuch better if at least one member of the team is familiar with the area.Inspections at bridge crossings should be treated with caution since bridges arefrequently placed at constrictions and/or at bedrock outcrops - locations that maynot be characteristic of the stream as a whole. However, valuable indicators ofstream stability can be observed at bridges and other points where infrastructurecrosses the stream. In assessing streams in the field, it is important to keep in

Hydraulic Design of Stream Restoration Projects - [PDF Document] (60)

Chapter 4 Stability Analysis 49

mind that a channel typically has four degrees of freedom: width, depth, slope,and planform.

During the field reconnaissance at least the following basic informationshould be collected:

a. Descriptions of watershed development and land use, floodplaincharacteristics, channel planform, and stream gradient.

b. Assessment of historical conditions. This can be obtained via interviewswith knowledgeable landowners.

c. Measurements of low-flow and bankfull channel dimensions and channelslope in critical reaches. Identification of terraces and active floodplains.

d. Characterization of the channel bed. Determine if it is bedrock, erodiblecohesive material, armored or alluvial. Determine the gradation of anyarmor layer and collect bed-material samples of the substrate layer.Guidelines for collection of bed material samples are given inAppendix D.

e. Descriptions of riverbank profiles, bank materials, and evidence of bankinstability.

f. Descriptions of point bars, pools, riffles, bed instability, and evidence ofsedimentation processes.

g. Observations of impacts due to channel alterations and evidence ofstream recovery.

h. Descriptions of channel debris and bed and bank vegetation.

i. Preliminary stream restoration alternatives should be identified so thatinformation can be gathered on possible constraints such as access,utilities, staging areas.

j. Photographic records of critical stream and watershed characteristics.

There are many possible indicators of stream stability. A range of fieldindicators within a watershed is given in Table 1. It is important to recognize thatthese are not absolutes and that items listed as possible indicators of instabilitymay occur in natural and/or stable streams and vice versa. Therefore it isimportant that those conducting the field assessment be experienced in theaccurate interpretation of the results of stream reconnaissance.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (61)

50 Chapter 4 Stability Analysis

Table 1Possible Field Indicators of River Stability/Instability

Evidence ofDegradation

Terraces (abandoned floodplains)Perched channels or tributariesHeadcuts and knickpointsExposed pipe crossingsSuspended culvert outfalls and ditchesUndercut bridge piersExposed or �air� tree rootsLeaning treesNarrow/deep channelBanks undercut, both sidesArmored bedHydrophytic vegetation located high on bank

Evidence ofa*ggradation

Buried structures such as culverts and outfallsReduced bridge clearancePresence of midchannel barsOutlet of tributaries buried in sedimentSediment deposition in floodplainBuried vegetationPerched main channelSignificant backwater in tributariesUniform sediment deposition across the channelHydrophobic vegetation located low on bank or dead in floodplain

Evidence of Stability

Vegetated bars and banksLimited bank erosionOlder bridges, culverts and outfalls with bottom elevations at or near gradeMouth of tributaries at or near existing main stem stream gradeNo exposed pipeline crossings

It is important to recognize the possible pitfalls of field assessments. Theseinclude observer bias, temporal limitations, and spatial limitations. Issues relatedto observer bias can be partially overcome with the consistent use of trainedpersonnel. This practice will minimize relative differences betweenobservations. Temporal bias can be minimized by examination of historicalrecords, but these may be incomplete. Having the field team walk a continuousreach of stream can reduce spatial bias. Field investigation should extend bothupstream and downstream of the project reach, and ideally should be conductedat several different periods of the year.

During the stream reconnaissance, it is important to locate and observe bothstable and unstable areas within the particular study reach. By observing theareas that have the worst problems, one will be able to establish the upper limitsof erosion, sedimentation, and/or flooding. It is equally important to visit reachesof the system where these problems are either not as apparent or absent. This willallow the engineer to define a total envelope of values associated with the studyarea and to understand the variability of the physical characteristics of the variousreaches in the stream.

Identification of geomorphologically similar reaches

The information gathered in the data assembly and field investigation shouldbe used to divide the channel into geomorphologically similar reaches. Whenestablishing reach limits, consideration should be given to: differences in channelslope, tributary locations, presence of geologic controls, planform changes,location of channel control structures (grade control structures, dams, culverts,etc.), changes in bed material size, major sediment sources (gravel mines,

Hydraulic Design of Stream Restoration Projects - [PDF Document] (62)

Chapter 4 Stability Analysis 51

sediment laden tributaries, etc.), changes in channel evolution type, or othersignificant hydrologic or geomorphic changes. Initial reach limits may be madeearly during the field investigation, but may be refined following more detailedanalyses.

Assessment of reach condition

At the conclusion of a field investigation, a summary of channel stability ineach reach is assessed. This summary may include the use of general typing andscoring techniques related to the existing condition of individual reaches. Themany techniques available range in complexity and required effort. The choiceof an assessment technique should be made with consideration of the study goalsafter the field investigations have been performed. An example of basic typing isas shown in Table 2.

Table 2Reach Condition AssessmentCondition Bed Bank

Stable The channel bed is as close to a stable condition ascan be expected in a natural stream. If the reachexhibits signs of local bed scour or deposition with alow rate of change, it would fall into this category.

The channel banks are as close to a stable conditionas can be expected in a natural stream and appear tohave a low potential to erode. Banks are predominantlycovered with extensive vegetation, boulders, orbedrock formations. If the reach exhibits signs of localbank erosion within an allowable rate of change, itwould fall into this category.


The channel bed in the reach is in a moderately stablecondition. However, the reach may be in transition.Reaches where the bed is experiencing bedaggradation or degradation at a low rate of changewould fall into this category. In addition, moderate tohigh local bed scour or deposition would fall into thiscategory. For example, rapid aggradation immediatelyabove and scour immediately below a minor debrisblockage (such as a single tree blocking the channel).

The channel banks in the reach are in a moderatelystable condition and exhibit medium erodibility. Banksare partially vegetated with moderately erodible soils.Typically, parallel flows would not result in bankerosion. The reach may be in transition. Reaches withbanks that exhibit moderate local bank erosion thatdoes not appear to be spreading would fall into thiscategory. For example, in an otherwise stable reach, asingle section of the bank could fall into the stream andresult in local, moderate bank erosion.

Unstable The channel bed in the reach is in an unstablecondition. Reaches where the bed is undergoingwidespread bed aggradation or degradation at amoderate rate would fall into this category. Moderatelyscoured reaches or reaches where many of the poolsare filled with loose sediment would fall into thiscategory.

The channel banks in the reach are predominantlyunstable. Reaches where the banks are experiencingwidespread erosion at a moderate rate would fall intothis category. Reaches where the channel banks areundergoing local bank erosion at a high rate of changeand where the erosion is not likely to be self healingwould also fall into this category.


The channel bed in the reach is in a very unstablecondition. Typically the channel shows no signs ofapproaching equilibrium with the current shape andplanform. Reaches where the bed is undergoingwidespread aggradation or degradation at a high ratewould fall into this category. Severely scoured reacheswould fall into this category. Reaches where all of thepools are filled with loose sediment would also fall intothis category.

The channel banks in the reach exhibit high erodibilityand do not have any controls that would restrictextensive changes in planform or shape. Riparian rootmasses are not present to slow rapid bank retreat. Anyparallel or impinging flows would cause extensive bankerosion. Reaches with near vertical to overhangingbanks.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (63)

52 Chapter 4 Stability Analysis

Channel typing and classification

Channel typing or classification is a useful though not essential step inchannel assessment. A channel can be described in detail without selecting aclassification system and assigning the stream reach to a certain class. Typing orclassification is useful if one is developing or using hydraulic geometry relationswith separate regression equations for different types of streams. Suchrelationships should result in regression equations with better accuracy and lessuncertainty.

Determining a channel type relies on developing a channel description basedprimarily on observation. The channel description includes parameters such aschannel and floodplain geometry, bed and bank material, planform, vegetation,bed forms, evidence of aggradation or degradation, grade control, alluvial orthreshold conditions, etc. Channel typing is an elementary level of streamclassification, using generic terms. For instance, a stream may be typed as ameandering sand bed channel.

Channel classification involves the selection of a classification system,normally developed by a specific person (e.g., Brice 1984 or Schumm 1977), andthe categorization of a channel into a specific class based on factors andmeasurements such as planform and planform features, dominant mode ofsediment transport, entrenchment ratio, sinuosity, etc. There are numerous streamclassification systems. Some of the most widely used are described in EM 1110-2-1418 and in the Federal Interagency Stream Restoration Working Group(1998). Some limitations of stream classification systems include the following:

a. The classification is a �snapshot� of the existing condition of the streamand does not give any information about trends, such as whether a streamis stable, aggrading, degrading, or approaching a critical geomorphicthreshold.

b. Water quality or the biological health of a stream cannot be determinedfrom a geomorphic classification system.

c. The classification is a generalization of stream behavior, which theindividual stream may conform to well or poorly.

Channel evolution models differ from classification systems in that they areused to predict sequential stages in channel response. For example, the incised-channel evolution model developed by Schumm, Harvey, and Watson (1984)predicts the sequence of changes which will occur in a channel as a headcutmoves upstream. The model stages are shown in EM 1110-2-1418, Figure 2-23.Simon (1989) has developed a similar model. Channel evolution models can beused to predict trends (aggradation, degradation, and channel widening) at aproject site, and to prioritize restoration work along a stream channel.

Regime-type relationships that express bankfull width as a function ofbankfull discharge can be used to provide initial predictions for stabledimensions of restored channels. However, the equations are valid only for thestream type and range of parameters from which they were derived. Hence,

Hydraulic Design of Stream Restoration Projects - [PDF Document] (64)

Chapter 4 Stability Analysis 53

when designing a stable channel it is essential to apply only the morphologicalequations appropriate to the stream type of the target restored channel. In thegeomorphological stability assessment the existing channel stream type should bedetermined and an appropriate target stream type should be recommended basedon characteristics from stable reference-reach sites. Classification of rivers mightbe used as a basis for typing the stream. There are many different methods ofclassifying alluvial rivers at the reach scale, ranging from simple descriptions tomore comprehensive systems (see Federal Interagency Stream RestorationWorking Group 1998). According to Thorne (1997), �the action of the drivingvariables of water and sediment inputs on the boundary conditions presented bythe floodplain topography, bed sediments, bank materials and riparian vegetationproduces the characteristic channel morphology of an unconfined alluvialstream.� More comprehensive typologies are limited in practice because theyrequire strong geomorphic insight and understanding to apply consistently andusefully (Thorne 1997) and in many cases there are insufficient morphologicalequations to match the number of subcategories. On this basis, it is recommendedthat channels should be typed according to the nature of bed sediments and bankcharacteristics and the typing used to guide engineers in choosing appropriatehydraulic geometry equations for use in stability assessments and channelrestoration design.

In summary, data obtained during the field investigation and historical datacollection can be used to determine the target stream type, in terms of boundarysediments, riparian vegetation and meander pattern. In many cases, the type anddensity of bank vegetation will be different from that present in the referencereaches due to ecological, aesthetic, and recreational objectives. It is importantthat target vegetation is identified prior to channel design as it influences flowresistance. Otherwise the stability status of the restored channel could beaffected.

Methods for assessing historical channel stability

The analysis of historical data from stream gages, surveys, and mapping cangive useful information about channel stability, any aggradation /degradationtrends, rates of lateral movement, and planform changes. The review of aerialphotographs taken at different time periods is a useful starting point. These arenormally available for any site, even when gage data or historic surveys areabsent. The use of historic data has some potential pitfalls, however, especiallywhen comparing surveys performed several years apart, or gage data with gaps inthe record. For example, the fact that the existing thalweg is lower than thehistoric thalweg normally indicates that degradation is the dominant process, butit does not always indicate that the stream is currently degrading. The streammay have degraded to a point below the existing streambed, reversed its trend ofinstability and then aggraded so that the existing dominant process is aggradation(Schumm, Harvey and Watson 1984). Results of the historical data analysisshould be compared to both the results of the field investigation and theanalytical stability assessment before reaching final conclusions.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (65)

54 Chapter 4 Stability Analysis

Specific gage analysis

If gage data are available, one of the most useful tools available to theengineer and geomorphologist for assessing the historical stability of a riversystem is the specific gage record. A specific gage record is a graph of stage fora specific discharge at a particular gaging location plotted against time. Achannel is considered to be in equilibrium if the specific gage record shows noconsistent increasing or decreasing trends over time, while an increasing ordecreasing trend is indicative of aggradation or degradation, respectively.

The first step in a specific gage analysis is to establish the stage-dischargerelationship at the gage for the period of record being analyzed. A rating curve isdeveloped for each year in the period of record. A regression curve is then fittedto the data and plotted on the scatter plot. Once the rating curves have beendeveloped, the discharges to be used in the specific gage record must be selected.This selection will depend largely on the objectives of the study. It is usuallyadvisable to select discharges that encompass the entire range of observed flows.A plot is then developed showing the stage for the given flow plotted againsttime.

Specific gage records are an excellent tool for assessing the historicalstability at a specific location. However, specific gage records indicate only theconditions in the vicinity of the particular gaging station and do not necessarilyreflect river response upstream or downstream of the gage. Therefore, thespecific gage record should be coupled with other assessment techniques in orderto assess reach conditions, or to make predictions about the ultimate response ofa river.

Comparative surveys and mapping

One of the best methods for directly assessing channel changes is to compareboth channel thalweg and cross sections. Thalweg surveys are taken along thechannel at the lowest point in the cross section. Comparison of several thalwegsurveys taken at different points in time allows the engineer and geomorphologistto chart the change in the bed elevation through time and track the migration ofheadcuts or aggradation zones through the system. Cross-section surveysprovide information about channel widening or narrowing.

There are certain limitations that should be considered when comparingsurveys on a river system. When comparing thalweg profiles it is often difficult,especially on larger streams, to determine any distinct trends of aggradation ordegradation if there are deep scour holes, particularly in bendways. Theexistence of very deep local scour holes may completely obscure temporalvariations in the thalweg. This problem can sometimes be overcome byeliminating the pool sections and focusing only on the crossing locations, therebyallowing aggradation or degradation trends to be more easily observed. Reliablesurvey comparisons can be made only if the surveys are hom*ologous in riversand streams that have significant bed form movement and/or seasonal variationsin sediment transport.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (66)

Chapter 4 Stability Analysis 55

While thalweg profiles are a useful tool it must be recognized that theyreflect only the behavior of the channel bed and do not provide information aboutthe channel as a whole. For this reason it is usually advisable to study changes inthe cross-sectional geometry. Cross-sectional geometry refers to width, depth,area, wetted perimeter, hydraulic radius, and channel conveyance at a specificcross section.

If channel cross sections are surveyed at permanent monumented rangelocations, then the cross-sectional geometry can be compared directly fordifferent time periods. At each range, the cross section plots for the various timeperiods can be overlaid and compared. When available cross sections are notlocated by permanent monuments it is often advisable to compare reach averagevalues of the geometric parameters. This requires the study area to be dividedinto distinct reaches based on geomorphological characteristics. Next the cross-sectional parameters are calculated at each cross section and averaged for theentire reach. Then the reach average values can be compared for each surveyperiod. Cross-sectional variability between bends (pools) and crossing (riffles)can obscure temporal trends, so it is often preferable to use only cross sectionsfrom crossing reaches when analyzing long-term trends of channel change.

Comparison of time sequential maps or aerial photographs can provideinsight into planform evolution, and change or instability of the channel. Ratesand magnitude of channel migration (bank caving), locations of natural andmanmade cutoffs, and spatial and temporal changes in channel width andplanform geometry can be determined from analysis of historical information.With this type of data, channel response to imposed conditions can bedocumented and used to substantiate predictions of future channel response to aproposed alteration. Contemporary planform data can be obtained from aerialphotos, maps, or from field investigations.

Hydraulic Geometry Assessment


A common component of empirical approaches to stable channel design restson downstream hydraulic geometry analysis. This approach employs a statisticaltreatment of data sets linking flow regime, sediment characteristics and resultingchannel forms under dynamically stable conditions. Hydraulic geometry theory isbased on the concept that a river system tends to develop in a way that producesan approximate equilibrium between the channel and the inflowing water andsediment load (Leopold and Maddock 1953). The stable channel does notchange significantly in profile, cross section, or planform characteristics over thelong term. Stable does not mean static: a stable channel may be activelymeandering. Since many natural channels are stable over a wide range of flows,the empirical hydraulic geometry relations used to describe them have been ofgreat interest to river engineers. Hydraulic geometry relations typically correlatean independent or driving variable, such as discharge or drainage area, todependent variables such as width, depth, slope, and velocity. These relations areempirically derived, and their development requires a relatively large amount of

Hydraulic Design of Stream Restoration Projects - [PDF Document] (67)

56 Chapter 4 Stability Analysis

data. EM 1110-2-1418 discusses the historic development, limitations, andapplication of hydraulic geometry relations. The development of hydraulicgeometry relations and their use for stability assessment will be discussed in thefollowing section.

Developing hydraulic geometry relations

The development of hydraulic geometry relations for a watershed or region isnot a trivial task. It is best performed by engineers and geomorphologists withextensive experience in the region. Some excellent examples of regionalhydraulic geometry studies are Emmett (1975); Charlton, Brown, and Benson(1978); Bray (1982); and Hey and Thorne (1986). Hydraulic geometry datashould be collected in stable, alluvial reaches. Channel dimensions are typicallytreated as dependent variables and are best determined from field surveys.Discharge is typically the independent variable.

Hydraulic geometry relations can be developed for a project reach, a stream,a watershed, or a physiographic region. The various sources of data are listedand described as follows in order of preference:

a. Given the natural variation of stream and watershed characteristics, thepreferred source of data for a project reach is the reach itself. Thischoice may not be feasible, either because the reach is not stable andalluvial, or because the reach has been altered.

b. The second preferred choice is data from the same stream at stable,alluvial reaches.

c. The third choice is data from other streams in the project watershed,although care must be taken to ensure that data are acquired fromportions of the watershed with physiographic conditions similar to thoseof the project reach.

d. The fourth choice is relations developed for a different watershed in asimilar physiographic region.

e. Generalized relations, or relations developed for other parts of thecountry, are a last choice, and should be evaluated on a case-by-casebasis. For example, relations developed for the Piedmont region ofVirginia may be transferable to the Piedmont region of Maryland. Butrelations developed for the glaciated areas of northern Pennsylvaniawould probably not be transferable to the nonglaciated areas of the samestate. The use of hydraulic geometry relations outside the area in whichthey were developed is discussed in more detail in the followingparagraphs.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (68)

Chapter 4 Stability Analysis 57

Choice of independent variables

The fundamental assumption of hydraulic geometry theory is that the shapeof a channel can be related to measurable or predictable hydraulic parameters.Therefore, cross-sectional form is inherited from the imposed natural sequence ofwater and sediment inflows and boundary conditions. As the discharge usuallyexplains most of the variance in geometry, bankfull width, depth, and velocity arenormally plotted as dependent variables against discharge. Although drainage areais often used as an independent variable (due to its ease of measurement), it ismerely a surrogate for discharge, and may be poorly correlated with bankfulldischarge within a watershed. While these relationships may be used to providerough estimates of channel dimensions at ungaged sites, they should be appliedwith caution if used to design stable channels. The choice between drainage areaand discharge as the independent variable also depends on the processes occurringin the watershed. For instance, in an urbanizing watershed, the relationship betweendischarge and drainage area will vary both spatially and over time, making drainagearea a poor choice for an independent variable.

Use of stream typing systems to refine hydraulic geometry relations

In general, data sets used in hydraulic geometry analysis are regionally-basedand apply to a particular locality rather than a stream type. Consequently,applying the resultant morphological equations beyond the parent region must beexercised with extreme caution. Alternatively, hydraulic geometry relationsdeveloped for various subsets of streams within a classification system based on bedand bank sediment and vegetation characteristics could reasonably be expected tohave less scatter since some of the secondary factors are taken into account.

Transfer of hydraulic geometry relations from one watershed toanother

The transfer of hydraulic geometry relations developed for one watershed toanother watershed should be performed with care. The two watersheds should besimilar in historical land use, physiography, hydrologic regime, precipitation,vegetation, etc. For example, relations developed for pristine watersheds should notbe transferred to urban watersheds. Relations developed for areas with snowmelthydrology should not be transferred to areas dominated by convective storms. Sincedischarge is the variable that shapes the channel, relations based on discharge can betransferred with more confidence than those based on drainage area.

Special problems of urbanized streams

Urbanized streams present particular problems in both the development and theapplication of hydraulic geometry relations. Land use and runoff characteristicsusually vary greatly, even within a single watershed. The multiplicity of manmadestructures, such as storm sewers, bridge openings, culverts, and stormwatermanagement facilities, changes the amount, duration, and timing of flows. Thiswould be expected to greatly increase data variability. (These factors make

Hydraulic Design of Stream Restoration Projects - [PDF Document] (69)

58 Chapter 4 Stability Analysis

discharge more poorly correlated with drainage area, and, hence, would makedischarge the better choice than drainage area as an independent variable.) Locatingstable, alluvial reaches may be difficult.

Uncertainty in hydraulic geometry relations

A sufficient number of data points must be measured to ensure that theresults from hydraulic geometry analysis are statistically valid. For example, ifany three or four random data points were used, a different relation could easilybe derived. The fewer and more widely scattered the data points, the lessconfidence one has in any derived trend. Even with quite a few data points in arelatively hom*ogeneous watershed, there is a great deal of scatter in the data dueto natural variability.

Natural rivers which are in regime have stable morphologies that broadlyconform to regime or hydraulic geometry relationships, linking the dependentparameters of channel form to independent controls of flow regime, boundarymaterials, and riparian vegetation. However, rivers do not follow regime lawsprecisely. In fact, every river displays local departures from the expected channelform described by morphological equations and possesses inherent variability inspace and time. While it is true that natural channel forms are in generalpredictable, it is also true that each river is in detail unique. Regime dimensionsin the natural domain should be interpreted only as representative reach-average,ideal or target conditions about which channel morphology fluctuates in time andspace.

The coefficient of determination, r2, in hydraulic geometry analysisnumerically represents the amount of variation that can be explained by theselected independent variable. The lower the r2 value, the less useful the relationis (and the wider the scatter in the data). The natural variability of data in arelatively hom*ogeneous watershed such as the upper Salmon River watershed(Emmett 1975) underlines the importance of viewing the data used to develop thecurve (not just the curve itself), along with statistical parameters such as r2 valuesand confidence limits. Equations given without plotted data points or statisticalparameters should be verified for applicability.

Statistical confidence bands can be used effectively to introducenonuniformity into restored channel designs and have been applied for thispurpose by Soar and Thorne (2001). Advanced texts on probability and statisticsdescribe standard methods of computing correlation coefficients and settingconfidence limits on data (e.g., Myers 1990; Graybill and Iyer 1994).

Application of hydraulic geometry relations to assess channelstability

Hydraulic geometry analysis can be used in a geomorphic assessment of thestudy reaches to provide semiquantitative information on channel stability andsensitivity to change. The hydraulic geometry observed for the existing channelmay be compared to that predicted for a stable channel in a reference reach using

Hydraulic Design of Stream Restoration Projects - [PDF Document] (70)

Chapter 4 Stability Analysis 59

new or existing equations and associated bands of uncertainty. If the data for theproject reach fall outside the 95 percent confidence band applied to the referencehydraulic geometry data, then there is reason to believe that the project reach inquestion may be unnatural or unstable. However, this method should be usedonly to provide an indication of stability because data points that lie far from thebest-fit regression line could be influenced by other factors such as geology, landuse, or vegetation that are not common to the rest of the data set.

The use of hydraulic geometry relations and confidence bands to assess thestability of a given channel reach requires that the watershed and stream channelcharacteristics of the reach in question are not dissimilar to the reference data setused to develop the hydraulic geometry relations and that the data scatter isknown, so that confidence bands can be derived. When applying this stabilityassessment, the two most reliable hydraulic geometry equations are thoseexpressing bankfull width as a function of bankfull discharge, for different typesof bed and bank characteristics, and meander wavelength as a function ofbankfull width. These relationships exhibit the least variability as opposed toother combinations of the dependent and independent variables (for examples seeHey and Thorne 1986 and Williams 1986).

In summary, the application of downstream hydraulic geometry relationshipsrequires that the actual data be plotted and the statistical coefficients calculated.Hydraulic geometry relations and associated confidence bands can be used as apreliminary guide to indicate potential stability or instability in stream reaches,but these indications should be checked using other techniques due to the widenatural variability of the data.

Analytical Stability AssessmentObservations and hydraulic geometry relations may be used to identify

possible stability problems, but analytical methods are required to determine themagnitude of a stability problem. An analytical stability analysis requirescalculation of hydraulic parameters such as velocity and shear stress for the rangeof natural discharges. The hydraulic resistance of the channel boundary isdetermined from field observations and measurements. Sufficient field samplingof the streambed should be conducted to determine the spatial variability, size,and gradation of the bed material. Sediment inflow is estimated from measureddata or by calculation.

Hydraulic calculations

Hydraulic parameters can be determined using normal depth assumptions orby a more rigorous backwater analysis. The SAM hydraulic design package(Thomas et al. 2000) can be used to average hydraulic parameters if normal depthassumptions are adequate. There are several available computer programs,including HEC-2, HEC-RAS, WSPRO and HY-22 if a gradually varied, steadyflow assumption is more appropriate.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (71)

60 Chapter 4 Stability Analysis

Normal depth. The SAM hydraulic design package is available to calculatenormal depth in a compound channel with variable hydraulic roughness. Severalcomplex channel compositing schemes are available for the calculation. Channelhydraulic parameters are calculated separate from overbank hydraulic parametersand effective channel hydraulic parameters are calculated for use in sedimenttransport relationships. Hydraulic roughness can be varied across the crosssection and different roughness equations can be used for different portions of thecross section.

Reliability of the normal depth calculation is directly related to the reliabilityof the input data. Sound engineering judgment is required in the selection of arepresentative cross section. The cross section should be located in a uniformreach where flow is essentially parallel to the bankline (no reverse flow oreddies). This typically occurs at a crossing or riffle. Determination of theaverage energy slope can be difficult. Thalweg slopes and low-flow water-surface slopes may not be representative of the energy slope at morphologicallysignificant flows. Slope estimates should be made over a significant length of thestream (a meander wavelength or 20-channel widths). Hydraulic roughnessmust be estimated based on field observations and measurements. Severaltechniques are recommended in EM 1110-2-1601 and the SAM Users Manual(Thomas et al. 2000).

Backwater analysis. Hydraulic models are used to calculate water-surfaceprofiles, flow and lateral velocity distributions, flow regimes, and scour potential.For stream restoration projects that are likely to involve revisions to the FederalEmergency Management Agency�s (FEMA) flood insurance rate maps, selectionof the hydraulic model should be coordinated carefully with FEMA. Somestandard hydraulic models are discussed in the following paragraphs.

HEC-RAS. HEC-RAS (USACE, HEC 2001) is the recommended model forperforming hydraulic calculations for steady, gradually varied (over distance),one-dimensional, open channel flow. HEC-RAS includes a culvert module that isconsistent with HDS-5 and HY-8. The bridge hydraulics algorithms now includethe WSPRO models. HEC-RAS applies conservation of momentum, as well asenergy and mass, in its hydraulic analysis. HEC-RAS includes all the featuresinherent to HEC-2 and WSPRO plus several friction slope methods, mixed flowregime support, automatic n value calibration, ice cover, quasi 2-D velocitydistribution, and superelevation around bends.

HEC-2. HEC-2 (USACE, HEC 1990) performs hydraulic calculations forsteady, gradually varied (over distance), one-dimensional, open channel flow.One of HEC-2�s technical limitations is the normal bridge routines and standard-step backwater computations use energy conservation only. Conservation ofmomentum is used only in the special bridge routines when there are bridgepiers.

WSPRO. The WSPRO computer program was developed by the U.S.Geological Survey (USGS) and is comparable to HEC-2, except for the fact thatWSPRO had special subroutines for analysis of water-surface profiles at bridgelocations. All of these WSPRO subroutines have been incorporated into

Hydraulic Design of Stream Restoration Projects - [PDF Document] (72)

Chapter 4 Stability Analysis 61

HEC-RAS. The current version of WSPRO must be used with caution since ithas known bugs and is no longer being supported.

HY-22. HY-22 is a small tool kit of relatively simple computer programs forperforming the hydraulic analyses described in the �Urban Drainage DesignManual,� Hydraulic Engineering Circular No. 22, FHWA (Brown 1996). HY-22includes pavement drainage, open channel hydraulics, critical depth computation,computation of storage volume, and simple reservoir routing.

Bed stability

After hydraulic parameters have been calculated for a range of discharges, itis important to determine the discharge at which the streambed begins to move.This can be accomplished using the threshold criteria described in EM 1110-2-1418. This step is especially important in a channel with an armor layer.Sediment transport capacity dramatically increases when the armor layer isdisrupted or destroyed and the coarse material becomes thoroughly mixed withthe substrate material. Stability of vegetated or gravel banks can be determinedusing allowable velocity methods or shear stress methods. A mobile streambedis not necessarily unstable, but mobile beds require a higher level of analysis todetermine stability.

Sediment rating curve analogy analysis

The sediment-rating curve analogy analysis is a relatively simple techniquethat can be used to assess the sediment transport characteristics of an existing orproposed stream project. The basic approach is to assess the sediment transportcharacter of a study reach by comparing its sediment transport capacity to that ofits supply reach. If the supply reach is not fully alluvial, a reference reach maybe used as a surrogate for the supply reach. The sediment rating curve analogyanalysis is suitable for streams where the sediment supply is not limited, that iswhere the stream is alluvial. It is generally not suitable for threshold streams.

This qualitative technique does not require stream gage data or sedimentgage data. It does require an estimate of the supply reach grain size distribution,an estimated range of peak flows, and a description of hydraulic characteristics ofboth the study and supply reaches. Hydraulic information can be based on normaldepth calculations or hydraulic modeling. Peak flows can be estimated usingregional regression curves or hydrologic modeling. Sediment transport capacityis calculated for a range of discharges in both the existing and supply reaches. Bycomparing the sediment rating curves of the two reaches, an estimate can bemade of the sediment transport capacity of the study reach relative to the capacityof the sediment supply reach. This is illustrated in Figure 33.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (73)

62 Chapter 4 Stability Analysis

Figure 33. Sediment rating curve analogy analysis of existing conditions

The comparison of the supply reach and study reach sediment rating curvesshown in Figure 33 indicates that there is a strong possibility that the existingstudy reach is depositional for flows above Q1. This estimated condition shouldbe checked by field observations to detect evidence of an aggradational trend. Toimprove channel stability the sediment rating curve for the project channelshould be as close as possible to the supply reach sediment rating curve.

Sediment budget analysis

Channel stability is ultimately determined by the ability of the channel topass the incoming sediment load while not scouring its bed. If sediment transportcapacity is less than sediment supply then the channel will aggrade. On the otherhand if the capacity is greater than the supply and the bed is alluvial then thechannel will degrade. A determination of the potential for aggradation ordegradation in a channel reach requires an assessment of the reach-scalesediment budget. The sediment budget compares the quantity of sedimenttransported into the reach with the sediment transport capacity of the reach. Thisis accomplished using the magnitude and frequency of all sediment-transportingflows. The following steps are recommended for conducting a sediment budgetanalysis.

a. Assemble information about the stream. This includes geometric,sedimentation, and hydrologic information. Missing data may be filledin from detailed site reconnaissance completed during a geomorphicassessment.

b. Calculate hydraulic parameters for a typical or average reach for a rangeof discharges. This range should extend from the average annual lowflow to the peak of the design flood. Average hydraulic parameters canbe determined from HEC-2 results using SAM.m95; or from normaldepth calculations for a designated typical cross-section geometry usingSAM.hyd.

c. Select an appropriate sediment transport function for the study reach.This can be achieved by comparing calculated sediment transport tomeasured data, taking care to ensure that bed-material load is beingcompared. When no data are available, one may rely on experience with

Hydraulic Design of Stream Restoration Projects - [PDF Document] (74)

Chapter 4 Stability Analysis 63

similar streams in the region. SAM.aid will designate the best sedimenttransport equation for rivers with similar hydraulic and sedimentcharacteristics.

d. Calculate three sediment transport rating curves for the existing channelin the assessment reach, upstream of the assessment reach (the supplyreach), and downstream from the assessment reach. Sediment transportrating curves should also be determined for any tributaries that might beaffected by the assessment reach.

e. Calculate sediment yield for the supply reach and the assessment reachand the downstream reach using the flow-duration sediment dischargerating curve method. This should be done using a flow-duration curve toobtain average annual sediment yield and a flood hydrograph to obtainyield during a flood event.

f. Calculate trap efficiency by comparing the supply reach and assessmentreach sediment yields. Also calculate trap efficiency for the assessmentreach compared to the downstream receiving channel. A positive trapefficiency indicates deposition and a negative value indicates erosion. Ifthe assessment reach is stable the trap efficiency is near zero.

The preferred method for calculation of average annual sediment yield is theflow-duration sediment-discharge method described in EM 1110-2-4000,Chapter 3. This method requires sufficient gage data to develop the flow-duration curve and requires either measured bed-material load data or calculationof a sediment-discharge rating curve using an appropriate sediment transportrelationship.

Often sufficient gage data are not available to calculate a flow-duration curvefor the project reach. In these cases, there are two approaches that can be used tocompute average annual sediment yield. The first is to synthesize a flow-duration curve using the drainage-area flow-duration curve method or theregionalized duration method (Appendix A), and then use standard methods tocompute sediment yield. The second approach is to compute sediment yields forhydrographs of various frequencies and then weight them according to theirprobability of occurrence. This is not a frequently used method but is discussedin Appendix E because of its usefulness for certain applications.

Nonequilibrium sediment transport

HEC-6 (USACE, HEC 1993) is a one-dimensional moveable boundary openchannel flow numerical model designed to simulate and predict changes in riverprofiles resulting from scour and deposition over moderate time periods,typically years, although applications to single flood events are possible. Acontinuous discharge record is partitioned into a series of steady flows of variabledischarge and duration. For each discharge, a water-surface profile is calculated,providing energy slope, velocity, depth, and other variables at each cross section.Potential sediment transport rates are then computed at each section. These rates,combined with the duration of the flow, permit a volumetric accounting of

Hydraulic Design of Stream Restoration Projects - [PDF Document] (75)

64 Chapter 4 Stability Analysis

sediment within each reach. The amount of scour or deposition at each section isthen computed and the cross-section geometry is adjusted for the changingsediment volume. Computations then proceed to the next flow in the sequenceand the cycle is repeated using the updated cross-section geometry. Sedimentcalculations are performed by grain size fractions, allowing the simulation ofhydraulic sorting and armoring.

HEC-6 is a powerful tool that allows the designer to estimate long-termresponse of the channel to a predicted series of water and sediment supply. Theprimary limitation is that HEC-6 is one-dimensional, i.e., geometry is adjustedonly in the vertical direction and average hydraulic parameters are assumed in thecomputations. Changes in channel width or planform cannot be simulated.

Integration and application of results

The final part of a stability assessment of a channel system is accomplishedby integrating the information from all the available analyses. Analysis usingeach of the geomorphic tools discussed previously may yield a verdict ofa*ggradation, degradation, or dynamic equilibrium with respect to the channelbed, and stable or unstable with respect to the banks. Often the individualassessments produce contradictory results. For instance, the field investigationsmight indicate that a channel reach is vertically stable, but the empiricalrelationships and SAM results indicate that the channel should be degrading. Inthis case one would have to assign a level of confidence to the variouscomponents based on the reliability and availability of the data, and one�s ownexperience with each tool in order to reconcile these contradictory results. Onceagain, it is obvious that there is no cookbook answer, and that sound judgementbased on insight and experience must always be incorporated when making astability assessment.

The information gained from the channel stability assessment can be appliedto determine potential evolutionary trends in the stream system. This isdependent on having a clear understanding of the dominant geomorphicprocesses at work in the watershed, and a conceptual model of how the streamsystem reacts to imposed changes. For example, the incised channel evolutionmodel (Schumm, Harvey, and Watson 1984) is a conceptual model of thereaction of a stream system to a base-level lowering without changes in theupstream land use or sediment supply. In the channel evolution model, theevolution of the stream (at any one point) follows a predictable series of stages.In watersheds where this model applies, the engineer can predict the futureevolution of various channel reaches, based on an assessment of the currentchannel condition gained from the channel stability assessment. However, inmany watersheds, the effects of base level lowering may be combined with otherperturbations, such as increased runoff or decreased sediment supply and this willcause a more complicated response in the stream system than is described in theincised channel evolution model. The engineer should attempt, as much aspossible, to develop a conceptual understanding which explains the historic andfuture evolution of the streams within the study watershed.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (76)

Chapter 4 Stability Analysis 65

The scope of a hydraulic analysis of a stream restoration project will varydepending on the stage of the planning process and the magnitude of channelinstability problems. Appendix F is an example scope of work for a stabilityanalysis that might be conducted early in the planning process to define dominantgeomorphological processes.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (77)

66 Chapter 5 Hydraulic Design Methodology

5 Hydraulic DesignMethodology

Design DischargesIn order to design a stream restoration project with long-term stability which

is sustainable without the need for maintenance dredging or grade control, it isnecessary to evaluate the full range of flows that will affect the channel. Astream restoration project usually has several design flows selected to meetvarious objectives. A narrow deep channel may be designed for lower dry-seasonflows (base flow) to meet habitat requirements during biologically criticalperiods. Channel dimensions for the main channel may be selected to convey aflow crucial to channel stability (channel-forming discharge), while projectfeatures, such as bank protection and habitat enhancement structures, may bedesigned to withstand a significant flood event, normally a 10 percent chanceexceedance discharge or larger. The appropriate types of design discharges fordifferent project elements are discussed in the following sections. Although aparticular project may not require the use of all of these flows for design, theengineer should still consider how the project will perform during low,intermediate, and high flows.

Design discharge for low flows

Normally, biological objectives drive project design for low flows. Forinstance, under many hydrologic regimes, summer low flows are often a criticalperiod for fish, and project goals may include narrowing the low-flow channel toprovide the increased depths necessary to support the fish population. Designflows may also be necessary to provide the depths and velocities essential for fishspawning or fish passage during other critical times of the year. Coordinationwith study team biologists is essential to make sure an appropriate flow (or rangeof flows) is selected. Design of a low-flow channel may be required as part of achannel modification. The 7-day annual low flow is often used for critical low-flow design. Guidance for flow depths and velocities required or tolerated by awide variety of fish species can be found in Bell (1986) and Morrow andFischenich (2000).

Hydraulic Design of Stream Restoration Projects - [PDF Document] (78)

Chapter 5 Hydraulic Design Methodology 67

Main channel discharge

If the stream channel is realigned or reconstructed, a suitable designdischarge must be selected for an initial estimate of reach-averaged stablechannel dimensions. This is normally larger than the one-year frequency event.Issues related to the selection of a channel-forming discharge for stable channeldesign are covered in Chapter 3. However, project constraints may not allow fora channel that carries only the channel-forming discharge. For instance, achannel that is larger than the regime channel may be required for floodconveyance. In such cases, a compound channel may be designed with a mainchannel that has other than ideal dimensions. Constraints that influence thedesign discharge for the main channel may also include the capacity of theupstream and downstream channels, utilities or rights-of-way that limit width,slope or alignment, and flooding concerns. It should be noted that stable channeldesign includes the evaluation of sediment transport capacity for a range of flows(not just the design discharge) to determine long-term maintenance requirementsand whether the project is likely to aggrade or degrade significantly in the future.

A single channel-forming discharge can be estimated by determining thebankfull flow, calculating the effective discharge or selecting a specificrecurrence interval discharge. However, inspection of a natural channel revealsthat variability is inherent to natural fluvial systems. Hence, when designingchannels that are intended to replicate natural channel features, but also remainstable over long periods of time, it is important to establish both the degree oflocal morphological diversity expected for the channel and its stability over arange of discharges.

After a preliminary design has been prepared, channel stability checks mayinclude simulation of sediment transport under either selected hydrologic eventsor the entire flow-duration curve for the available period of record. This type ofanalysis will indicate whether the channel will experience unacceptable levels ofscour or deposition during discharges above and below the design flow, andwhether aggradational or degradational trends will be significant within the lifespan of the project.

Habitat and hydraulic structure design discharge

Constraints such as floodplain development or flood damage reductionrequirements mean that successful stream restoration often includes bankprotection, grade control, and in-stream construction of habitat features. Livingplant materials are often used in association with inert materials, such as timberor rock, and manufactured products. To accomplish a reasonably self-sustainingholistic ecosystem, a combined technology approach is required. Sound physicalprinciples and well established engineering formulae are used in the analysis anddesign of both soft and hard features.

A significant flood event (normally no smaller than the 10-year frequencydischarge) is used to size stone and compute scour depths. The goal is that thehard project features will withstand a flood of this magnitude without majordamage, movement, or flanking. Impacts that might reasonably be expected

Hydraulic Design of Stream Restoration Projects - [PDF Document] (79)

68 Chapter 5 Hydraulic Design Methodology

during a flood event would be deposition of sediment and debris; combined withlocal and general scour, erosion and stone movement; and destruction ofvegetation. Often in urban settings and flood damage reduction channels theone-percent chance exceedance discharge is used to size stone and compute scourdepths. In addition, the impact of the project on flood elevations and conveyancemust be evaluated. Often the impact on the water-surface profile for the one-percent chance exceedance discharge must be submitted as part of the project�spermitting requirements. It may also be necessary to compute the impact of theproject on more frequent flood events, or for a larger event.

Threshold ChannelsAs defined herein, a threshold channel is a channel in which channel

boundary material movement is not a stability issue during the design flow. Theterm threshold is used because the channel geometry is designed such thatapplied forces from the flow are below the threshold for movement of theboundary material. This class of stream includes cases where the streambed iscomposed of very coarse material or erosion resistant bedrock. Streams wherethe boundary materials are remnants of processes no longer active in the streamsystem may be threshold streams. Examples are streambeds formed by highrunoff during the recession of glaciers or dam breaks and streams armored due toreduction in the upstream sediment supply and degradation. Fine sediment maypass through threshold streams as throughput or wash load. In general, thissediment should not be considered part of the stream boundary for stabilitydesign purposes even if there are temporary deposits on the streambed at lowflow. However, throughput or wash load may be an environmental issue.Threshold channels do not have the ability to adjust their geometry, as do alluvialchannels, because the material forming the channel boundary is unerodible underthe normal range of flows, and there is no significant exchange of materialbetween the sediment carried by the stream and the bed. At flows larger than thedesign flow or during extreme events, threshold channels may becomedestabilized for short periods, with harmful morphological impacts and thispossibility must be borne in mind.

There is not always a clear distinction between threshold and alluvialchannels. One reach of the stream may be alluvial while another has thecharacteristics of a threshold channel. A threshold stream reach can be changedto an alluvial reach by flattening the slope. A stream may be alluvial at lowdischarges when there is an adequate sediment supply and then act like athreshold channel at high discharges. If an armor layer is present, a stream maybe a threshold channel at low flows and on the rising limb of a flood hydrograph,but an alluvial channel at high flows when the armor layer is mobilized, and onthe falling limb of the flood hydrograph when sediment is being deposited. It istherefore important to evaluate channels through their entire flow range todetermine how they will react to natural inflow conditions and how their stabilitystatus may change as a function of discharge.

Hydraulic design methods for threshold channels are well established andavailable from several sources. Maximum permissible velocity methods areapplicable for a variety of boundary materials and guidance can be found in

Hydraulic Design of Stream Restoration Projects - [PDF Document] (80)

Chapter 5 Hydraulic Design Methodology 69

EM 1110-2-1601 and EM 1110-2-1418. The U.S. National ResourceConservation Service (NRCS) has developed allowable velocity designprocedures for drainage channels (USDA 1977 and EM 1110-2-1418). The U.S.Bureau of Reclamation developed the tractive force method (Lane 1955; and EM1110-2-1418) for design of irrigation canals, primarily with gravel beds.Threshold design methods do not provide unique solutions for channeldimensions of width, depth, and slope. However, this limitation is not critical tothe hydraulic design in terms of stability because the boundary is immobile.

The concepts of channel-forming discharge and hydraulic geometry aregenerally not applicable to threshold channels because these channels do notadjust their dimensions to the natural runoff hydrograph.

Theoretical threshold-channel design methods have been developed thatconsider the lateral turbulent diffusion of downstream momentum in a crosssection with a laterally-uniform bed gradation (Parker 1978; Parker 1979; Ikeda,Parker, and Kimura 1988; Ikeda and Izumi 1990; and Diplas and Vigilar 1992).These methods are discussed by the ASCE Task Committee on Hydraulics, BankMechanics, and Modeling of River Width Adjustment (1998).

Threshold methods are also used to design stream features such as bank toeprotection, riffles, and deflector dikes. The Corps� riprap design procedure (EM1110-2-1601) is appropriate for design of these features. This procedure allowsfor use of rounded stone as well as angular stone more commonly used in flood-control projects.

Alluvial ChannelsAlluvial streams have bed and banks formed of material transported by the

stream under present flow conditions. There is an exchange of material betweenthe inflowing sediment load and the bed and banks of the stream. Alluvialchannels adjust their width, depth, slope and planform in response to changes inwater or sediment discharge.

The hydraulic design methodology described herein is intended for caseswhere an historically stable channel has been realigned creating instability, orwhere hydrologic and/or sediment inflow conditions have changed so much thatthe channel is currently unstable. A stream is defined as stable when it has theability to pass the incoming sediment load without significant degradation oraggradation and when its width, depth, and slope are fairly consistent over time.Bank erosion and bankline migration are natural processes and may continue in astable channel. When bankline migration is deemed unacceptable, thenengineering solutions must be employed to prevent bank erosion. Bankprotection technology is not addressed in this report, but a review of issues anddesign considerations can be found in Biedenharn et al. (1997).

Hydraulic Design of Stream Restoration Projects - [PDF Document] (81)

70 Chapter 5 Hydraulic Design Methodology

Alluvial channel design variables

The hydraulic design variables of width, depth, slope, and planform aredependent variables in an alluvial channel. Their magnitudes are determined bythe independent variables of sediment inflow, water inflow, and bankcomposition. The downstream water-surface elevation is an independent variablethat could have a significant effect on the dependent variables in some cases.Boundary resistance along the channel banks and sometimes along the bed can beboth dependent and/or independent depending on local circ*mstances. Thehydraulic design methodology provides a method for determining the magnitudeof the dependent variables given the magnitudes of the independent variables.

The design philosophy is to employ the best available physically basedmethodologies to determine the design variables. Average magnitudes for width,depth, and slope are determined first. The initial or preliminary average channelgeometry is initially determined using a single channel-forming discharge. Latera full range of discharges is used to evaluate the channel design, and the initialdesign may be adjusted. Analytical techniques are employed to ensure that thecombination of design variables are compatible. With three unknowns, threeequations are required to determine the magnitude of each design variable. Ahydraulic resistance equation, such as Manning�s equation, can be one designequation. A sediment transport equation, such as Brownlie�s equation, can be thesecond design equation. Resistance and sediment transport equations are wellestablished and can be used with a reasonable level of confidence in the designprocess. One additional equation is needed. Four alternatives are consideredherein for this third equation: (a) analogy methods, (b) hydraulic geometryrelationships, (c) constraint of one of the variables, or (d) adopting an extremalhypothesis.

When channel width is not constrained by rights-of-way limitations, thepreferred method for determining one of the unknown dependent variables is toapply geomorphic principles such as an analogy method or a hydraulic geometryrelationship. Several techniques are available.

Analogy methods. If the existing channel is stable in the project reach anattempt should be made to retain the same channel geometry in the restoredchannel.

If the channel is unstable in the project reach, a design top width for thestable project channel can be determined by assigning a measured average topwidth from a reference reach. The reference reach must be stable and alluvialand have the same channel-forming discharge and boundary conditions as theproject reach. The reference reach may be upstream and/or downstream from theproject reach, or in a different but physiographically similar watershed. The bedand banks in the project and reference reaches must be composed of similarmaterial, and there should be no significant hydrologic, hydraulic, or sedimentdifferences between the reaches. This technique is inappropriate for streamswhere the entire fluvial system, or a significant part of it, is in disequilibrium.

An alternative to the reference reach approach is to reconstruct the channel toa stable predisturbance width and planform. This is feasible if historical width

Hydraulic Design of Stream Restoration Projects - [PDF Document] (82)

Chapter 5 Hydraulic Design Methodology 71

and planform information can be determined from mapping, aerial photos, and/orsoil borings. However, this technique is inapplicable if the watershed water andsediment runoff characteristics have changed over time, as the historically stablechannel form will no longer be stable in the current watershed context.

Hydraulic geometry methods. Hydraulic geometry theory is based on theconcept that a river system tends to develop in a predictable way, producing anapproximate equilibrium between the channel and the inflowing water andsediment (Leopold and Maddock 1953). The theory typically relates a dependentvariable, such as width or slope, to an independent or driving variable, such asdischarge or drainage area. Hydraulic geometry relations are sometimes stratifiedaccording to bed material size, bank vegetation, or bank material type. Hydraulicgeometry relationships are developed from field observations at stable andalluvial cross sections. These relationships were originally used as descriptors ofgeomorphologically adjusted channel forms. As design tools, hydraulicgeometry relationships may be useful for preliminary or trial selection of thestable channel width.

A hydraulic geometry relation for width can be developed for a specific river,watershed, or for streams with similar physiographic characteristics. Data scatteris expected about the developed curve even in the same river reach. An exampleof a hydraulic geometry relationship between bankfull discharge and bankfullwater surface width developed for a mountainous watershed can be found inEmmett (1975). He collected data at 39 gaging stations in the Salmon RiverDrainage Basin, ID. The relationship between bankfull discharge and bankfullwidth is shown in Figure 34. Emmett�s mean regression line had a regressioncoefficient (r2) of 0.92. Nevertheless a wide range of bankfull widths were foundfor any specific bankfull discharge. This range does not necessarily indicateinstability or different physiographic conditions (Emmett gave no indication inhis report that any of his sites were unstable) but rather the wide range ofpossible stable widths for a given channel-forming discharge. The data scatter inFigure 34 also demonstrates the importance of using confidence bands withhydraulic geometry relationships in geomorphologic stability assessment.

It follows that the more dissimilar the stream and watershed characteristicsare, the greater the expected data scatter is. It is important to recognize that thisscatter represents a valid range of stable channel configurations due to variablessuch as geology, vegetation, land use, sediment load and gradation, and runoffcharacteristics. The composition of the bank is very important in thedetermination of a stable channel width. It has been shown that the presence andpercentage of cohesive sediment in the bank and the amount of vegetation on thebank may significantly affect the stable alluvial channel width.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (83)

72 Chapter 5 Hydraulic Design Methodology

Figure 34. Hydraulic geometry relationship for width for the Upper Salmon RiverBasin, ID (Emmett 1975)

A regional slope-drainage area hydraulic geometry relationship can bedeveloped for physiographically similar watersheds. An empirical regionalstability relationship that defines the stable channel slope (equilibrium slope) as afunction of drainage area (a surrogate for discharge) has been developed forseveral watersheds in north Mississippi (Figure 35). Channel slopes weremeasured in the field at several locations where stable reaches could beidentified. Drainage area was determined from topographic maps. Theequilibrium slope was used to set the slope between grade control structures inunstable reaches. The slope-drainage area curve can be a valuable relationshipfor initial understanding of stream morphology in an unstable watershed.However, the relationship is empirical and extrapolation to other watersheds, orthe same watershed during a different time period, is risky. Constant fieldverification is necessary for continued value.

When a hydraulic geometry relationship is to be used for a channelrestoration design it is best to use one developed from stable alluvial reaches ofthe project stream. It is required that the stable reaches used to develop therelationship have similar physiographic conditions to each other and the projectreach. If there are no stable reaches or if the range of discharges is insufficient,other streams or tributaries in the same watershed may be used to develop thehydraulic geometry relationship. The third choice is to use regional relationshipsdeveloped for other watersheds in the same physiographic region. In all cases itmust be remembered that data used to develop hydraulic geometry relationshipsshould come from stable reaches and that the watersheds and channel boundaryconditions should be similar in the project channel.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (84)

Chapter 5 Hydraulic Design Methodology 73

Figure 35. Equilibrium channel slope versus drainage area for Hickahala Creek,Batupan Bogue, and Hotopha Creek, MS

Lacking data to develop more reliable hydraulic geometry relationships,generalized width predictors for various stream types with different bankcharacteristics have been developed and are presented in Figures 36 through 43.These predictors include confidence limits and may be used for general guidancewhen stream or watershed specific data cannot be obtained.

a. Hydraulic geometry for meandering sand bed rivers

Hydraulic geometry width predictors (Figure 36) were developed from datacollected from 58 meandering sand bed rivers in the United States (Soar andThorne 2001). Sufficient data were collected to determine both bankfulldischarge and effective discharge. Data were collected from stable reaches, sobankfull discharge should be the most reliable approximator for the channel-forming discharge. In many of these meandering sand bed rivers, the effectivedischarge was significantly less that the bankfull discharge. For design purposes,the bankfull discharge was used to define the width predictor. The data weredivided into two sets: type T1 where there was less that 50 percent tree cover onthe banks (Figure 37) and type T2 where there was greater than 50 percent treecover on the banks (Figure 38). All sites were treelined to some degree, therefore

Hydraulic Design of Stream Restoration Projects - [PDF Document] (85)

74 Chapter 5 Hydraulic Design Methodology

the predictors should not be used for grasslined or thinly vegetated banks. Thepercentage of silt and clay in the banks was not found to be significant inaffecting width for these rivers, possibly because the root-binding properties ofthe trees were more significant in stabilizing the bank than cohesive forces.

Figure 36. Best-fit hydraulic geometry relationships for width for U.S. sand bedrivers with banks typed according to density of tree cover

Figure 37. Confidence intervals applied to the hydraulic geometry equation forwidth based on 32 sand bed streams with less than 50 percent treecover on the banks (T1). S.I. units, m and m3/sec (English units, ftand ft3/sec)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (86)

Chapter 5 Hydraulic Design Methodology 75

Figure 38. Confidence intervals applied to the width hydraulic geometry equationbased on 26 sand bed rivers with at least 50 percent tree cover onthe banks (T2). S.I. units, m and m3/sec (English units, ft and ft3/sec)

Figure 39. Downstream width hydraulic geometry for North American gravel bedrivers, W = 3.68 Qb

0.5, and U.K. gravel bed rivers, W = 2.99 Qb0.5





1 10 100 1000

Bankfull Discharge, Q b (m3s-1)




th, W


Wolman (1955) Nixon (1959)Emmett (1972) Charlton et al. (1978)Emmett (1975) Hey and Thorne (1986)Williams (1978) Griffiths (1981)Andrews (1984) North America DataAnable (1996) U.K. Data

Hydraulic Design of Stream Restoration Projects - [PDF Document] (87)

76 Chapter 5 Hydraulic Design Methodology

Figure 40. Downstream width hydraulic geometry for North American gravel bedrivers, W = a Qb

0.5 with confidence bands. Based on 94 sites in NorthAmerica. S.I. units, m and m3/sec (English units, ft and ft3/sec)

Figure 41. Downstream width hydraulic geometry for United Kingdom gravel bedrivers, W = a Qb

0.5 with confidence bands. Based on 86 sites in theUnited Kingdom. S.I. units, m and m3/sec (English units, ft and ft3/sec)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (88)

Chapter 5 Hydraulic Design Methodology 77

Figure 42. Downstream width hydraulic geometry for United Kingdom gravel bedrivers, W = a Qb

0.5 with confidence bands. Based on 36 sites in theUnited Kingdom with erodible banks. S.I. units, m and m3/sec (Englishunits, ft and ft3/sec)

Figure 43. Downstream width hydraulic geometry for United Kingdom gravel bedrivers, W = a Qb

0.5 with confidence bands. Based on 43 sites in theUnited Kingdom with resistant banks. S.I. units, m and m3/sec(English units, ft and ft3/sec)

The hydraulic geometry width predictor is expressed by the general equation:

bW a Q=

Hydraulic Design of Stream Restoration Projects - [PDF Document] (89)

78 Chapter 5 Hydraulic Design Methodology

Where W is the channel top width, Q is the channel-forming discharge, andvalues for the coefficient a and the exponent b are given in Table 3. Thehydraulic geometry width predictors each include two sets of confidence bands.The 95 percent mean response limit provides the band in which one can be 95percent confident that the mean value of the width will occur. This is theconfidence interval for the regression line. This provides the range of averagevalues of width that can be expected for a given discharge. The 90 percent singleresponse limit provides the envelope curves that contain 90 percent of the datapoints. This is the confidence interval for an individual predicted value. Thisprovides the engineer with the range of possible widths that have been observedto correspond to a given discharge. The confidence interval on an individualpredicted value is wider than the confidence interval of the regression line sinceit includes both the variance of the regression line plus the squared standarddeviation of the data set. While the equations given in Table 3 may be used forpreliminary design purposes, they are subject to several limitations. In theabsence of stage-discharge relationships at each site, the equations are based onflow resistance considerations. As cross-sectional geometry was used to calculatedischarge, discharge is not truly independent of width in this analysis.Furthermore, only one cross section was measured at each site in order tomaximize the size of the data set and identification of the bankfull referencelevel, although based on field experience and geomorphic criteria, is alwayssubject to a degree of uncertainty. These factors contribute to the observedvariability in the width relationships. Finally, small rivers are not wellrepresented in the data set and should not be applied when discharge is less than17 m3s-1 in type T1 channels and less than 38 m3s-1 in type T2 channels.

Table 3Hydraulic Geometry Width Predictors For Sand Bed Channels

Data Source Sample size a

90% singleresponselimit for a

95% meanresponselimit for a b r2

All sand-bed streams 58 4.24(2.34)


3.90-4.60(2.15-2.54) 0.5 0.76

Type T1:<50% tree cover 32 5.19


4.78-5.63(2.64-3.11) 0.5 0.87

Type T2:>%50% tree cover 26 3.31


3.04-3.60(1.68-1.99) 0.5 0.85

Note: r2 refers to linear regression equations (not given) where b was variable. Exponent b was found not to be statistically differentfrom 0.5 which was chosen for convenience. S.I. units m and m3/sec (English units ft and ft3/sec) W = a Qb

b. Hydraulic geometry for gravel bed rivers.

A review of the published gravel bed stream data and hydraulic geometrywidth predictors for North American and British streams (Soar and Thorne 2001)revealed that North American gravel bed rivers are generally wider than thosefound in the U.K., assuming discharge and other conditions are equal. NorthAmerican data used to develop the hydraulic geometry relationship included datafrom Brandywine Creek in Pennsylvania (Wolman 1955), Alaskan streams

Hydraulic Design of Stream Restoration Projects - [PDF Document] (90)

Chapter 5 Hydraulic Design Methodology 79

(Emmett 1972), Upper Salmon River in Idaho (Emmett 1975), Colorado, NewMexico, Oregon, Pennsylvania, Tennessee, Utah, West Virginia, and Wyoming(Williams 1978), Alberta, Canada (Annable 1996), and the Rocky Mountainregion of Colorado (Andrews 1984). United Kingdom data included data fromNixon (1959), Charlton, Brown, and Benson (1978), and Hey and Thorne (1986).The hydraulic geometry relationships are shown in Figure 39. The difference inthese regression curves cannot satisfactorily be explained using the sitedescriptions given in original publications. A possible explanation is that theU.K. sites have on the average more resistant banks than the North Americansites. Another plausible explanation is that width in mobile-gravel bed streamsvaries with flow variability and the North American sites on the average may bemore flashy. Still another possibility is that the North American sites may bemore active, that is have a higher concentration of sediment transport. Furtherresearch is required to validate these hypotheses.

The hydraulic geometry width predictors for United Kingdom and NorthAmerican gravel bed streams are presented with confidence bands in Figures 40and 41, respectively. Exponents and coefficients for the hydraulic geometryequation are given in Table 4. The gravel bed river data comprise a wide rangeof bank material types (e.g., cohesive, sand, gravel, and composite banks ofvarious strata). However, different width-discharge relationships based ondifferent types of bank material could not be derived for the North Americanriver data from the limited information available. There were sufficient dataavailable from the UK gravel bed rivers to develop distinct width predictorsbased on erodible banks (low density of trees) and resistant banks (high densityof trees). These are presented in Figures 42 and 43. These hydraulic geometryrelations may be used for preliminary design purposes, recognizing thatconsiderable variability may occur for areas different from the streams used inthe development of the equations.

Table 4Hydraulic Geometry Width Predictors for Gravel Bed Channels

Data Source Sample size a

90% singleresponselimit for a

95% meanresponselimit for a b r2

All North American gravelbed streams 94 3.68


3.45-3.94(1.90-2.18) 0.5 0.80

All United Kingdom gravelbed streams 86 2.99


2.83-3.16(1.56-1.74) 0.5 0.80

<5% tree or shrub cover,or grass-lined banks (UKstreams)

36 3.70(2.04)


3.49-3.92(1.93-2.16) 0.5 0.92

≥5% tree or shrub cover(UK streams) 43 2.46


2.36-2.57(1.30-1.42) 0.5 0.92

Note: r2 refers to linear regression equations (not given) where b was variable. Exponent b was found not to be statisticallydifferent from 0.5, which was chosen for convenience.S.I. units m and m3/sec (English units ft and ft3/sec) W = a Qb

Hydraulic Design of Stream Restoration Projects - [PDF Document] (91)

80 Chapter 5 Hydraulic Design Methodology

Extremal hypotheses. If a reliable hydraulic geometry relationship cannotbe determined from field data or when sediment transport is significant,analytical methods may be employed to obtain a range of feasible solutions.Analytical methods employ an extremal hypothesis as a third equation. Oneextremal hypothesis assumes that a channel will adjust its geometry so that thetime rate of energy expenditure is minimized (Chang 1980; Copeland 1994).Another assumes that sediment transport is maximized within the constraints onthe system (White, Bettess, and Paris 1982; Millar and Quick 1993). These areequivalent assumptions. Computer programs or look-up charts are required tosolve the resistance, sediment transport, and extremal equations simultaneously.The SAM hydraulic design package contains a program to solve these equationsusing either the Brownlie (1981) resistance and sediment transport equations forsandbed streams, or the Limerinos (1970) resistance equation and the Meyer-Peter and Muller (1948) sediment transport equation for gravel-bed streams.

The advantage of using an extremal hypothesis is that a unique solution canbe obtained for the dependent variables of width, depth, and slope. However,extensive field experience demonstrates that channels can be stable with widths,depths, and slopes different from those found at the extremal condition. Also thesensitivity of energy minima or sediment transport maxima to changes in drivingvariables may be low, so that the channel dimensions corresponding to theextremal value are poorly defined.

Constrained dependent variables. In many cases, project constraints limitthe theoretical variability in channel geometry. For example, the channel slopecannot be greater than the valley slope for a long reach. The channel width maybe limited by available rights-of-way, or flood risks, and damages may limitallowable depth. For these and many other reasons, the selection of one of thedependent design variables may be based on established project constraints.

Calculation of the remaining unknown design variables. Once one of thedependent design variables is determined, the other two should be calculatedusing one of several resistance and sediment transport equations available in theliterature. Appropriate equations can be chosen from those described in EM1110-2-1601, EM 1110-2-1418, or the SAM Users Manual (Thomas et al. 2000).

In coarse-bed streams where bed-material sediment transport is small, or instreams with bedrock outcrops or with cohesive beds, threshold design methodsmay be used to calculate depth and slope. However, in sand bed streams, bed-material sediment transport is typically significant and an analytical procedurethat considers both sediment transport and bed form roughness is required.

The stable-channel analytical method in the Corps hydraulic design packageSAM may be used to determine the unknown dependent design variables. Thismethod is based on a typical trapezoidal cross section and assumes steady,uniform flow. The method is especially applicable to small streams because itaccounts for sediment transport, bed form and grain roughness, and bankroughness. This procedure assumes a fully mobile bed. Details are available inthe SAM users manual.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (92)

Chapter 5 Hydraulic Design Methodology 81

The stable channel analytical method in SAM produces a family of solutionsfor slope and depth for specified widths for a selected discharge (Figure 44).These curves represent combinations of width, depth, and slope that satisfy thesediment transport and roughness equations. The wide range of possible solutionscan be narrowed by the assigned project constraints. For example, a maximumwidth constraint might be imposed by the available rights-of-way, a maximumdepth constraint might be imposed by flood-control considerations, and/or amaximum slope constraint would be imposed by the valley slope. Lackingproject constraints a hydraulic geometry relationship with confidence limits forwidth could be used to select a range of stable slopes and depths, or the extremalassumption can be applied and the unique solution occurs at the minimum slopeon the stable channel design curve.

Figure 44. Stability curve from stable channel analytical method


This step involves determining a meander wavelength, an appropriatechannel length for one meander wavelength, and then laying out a planform.Existing methods often rely on the user locating a reference or control reach oneither the study stream or another suitable stream from which to develop atemplate for the meander planform. This may often be problematic due to thenonavailability of a reference reach, subtle but important fluvial, sedimentary ormorphological differences between it and the study reach, or restrictions on therights-of-way, which preclude the introduction of meanders with the sameamplitudes observed in the reference reach. Alternatively, meander wavelengthcan be determined using hydraulic geometry techniques. The most reliablehydraulic geometry relationship is wavelength vs. width. As with thedetermination of channel width, preference is given to wavelength predictorsfrom stable reaches of the existing stream either in the project reach or inreference reaches. Lacking data from the existing stream, general guidance is

Hydraulic Design of Stream Restoration Projects - [PDF Document] (93)

82 Chapter 5 Hydraulic Design Methodology

available from several literature sources. A composite relationship has beendeveloped by Soar and Thorne (2001) combining nine data sets and 438 sites.Their mean linear regression predictor for wavelength is


where: λ is meander wavelength and W is channel width. Definitions ofplanform descriptive variables are shown in Figure 45. Confidence bands aboutthis equation are shown in Figure 46. The r2 for the wavelength equation was0.88 for a linear regression equation with a variable exponent on W. Thisexponent was found not to be significantly different from 1.0 so the exponentwas fixed at 1.0 for convenience. Only sites with sinuosities of at least 1.2 andbankfull widths between 1 m and 1,000 m were used in development of thisregression equation. Within these constraints, meander wavelengths rangebetween 10.4 m and 19,368 m and sinuosities range between 1.2 and 5.3. Theequation corrected for bias is:


An unbiased hydrologic equation for meander wavelength within 95 percentconfidence limits on the mean response suitable for engineering design is:

Wto )47.1226.11(=λ

Figure 45. Meander parameters

Hydraulic Design of Stream Restoration Projects - [PDF Document] (94)

Chapter 5 Hydraulic Design Methodology 83

Figure 46. Hydraulic geometry relationship for meander wavelength withconfidence intervals, λ = 10.23 W, based on a composite data set of438 sites

According to Hey (1976) and Thorne (1997), twice the distance betweensuccessive riffles (or pools) in a straight channel equals 4πW (12.57 W). This isbased on the assumption that the average size of the largest macro-turbulenteddies (or helical flow cell) is half the channel width. The preceding equationshows that the upper range of stable meander wavelengths is numerically veryclose to this value and similar to the coefficient of 12.34 given by Richards(1982). This corroborates the assertion by Leopold and Wolman (1957, 1960)that the matching of waveforms in bed topography and planform is related to themechanics of the flow, and in particular to the turbulent flow structuresresponsible for shaping the forms and features of meandering channels.

The following data sources were used in the development of these equations:Leopold and Wolman (1957) data from U.S. rivers (21 sites); Leopold andWolman (1960) data compiled from various sources and including rivers inFrance (1 site), U.S. (34 sites) and one model river (total of 36 sites); Carlston(1965) data from U.S. rivers (29 sites); Schumm (1968) data from midwesternU.S. rivers (25 sites); Chitale (1970) data from large alluvial rivers in Africa (1site), Canada (1 site), India (16 sites), Pakistan (2 sites) and U.S. (1 site) (total of21 sites); Williams (1986) data compiled from various sources and includingrivers in Australia (2 sites), Canada (7 sites), Sweden (17 sites), Russia (1 site),U.S. (16 sites) and one model river (total of 44 sites); Thorne and Abt (1993)data from various sources including measurements from the Red River 1966 (35sites) and 1981 (39 sites) hydrographic surveys between Index, Arkansas, andShreveport, LA, and rivers in India (12 sites), The Netherlands (1 site), U.K. (48sites) and U.S. (18 sites) (total of 153 sites); Annable (1996) data from streams inAlberta, Canada (30 sites); and Cherry, Wilco*ck, and Wolman (1996) data fromU.S. rivers, predominantly sand bed (79 sites).

Hydraulic Design of Stream Restoration Projects - [PDF Document] (95)

84 Chapter 5 Hydraulic Design Methodology

Other hydraulic geometry relationships for meander wavelength from theliterature are given in Table 5.

Table 5Hydraulic Geometry Relationships for Meander WavelengthAuthor Equation UnitsLeopold and Wolman (1960) λ= 10.9 W1.01 feetInglis (1941) λ = 6.06 W0.99 feetYalin (1992) λ = 6 W lengthDury (1965) λ = 30 Qbf

0.5 feet, cfsCarlston (1965) λ = 8.2 Qbf

0.62 feet, cfsCarlston (1965) λ = 106.1 Qma

0.46 cfsSchumm (1967) λ = 1890 Qma

0.34 M-0.74 feet, cfsNotes: λ = meander wavelength

W = widthQbf = bankfull dischargeQma = mean annual dischargeM = silt-clay factor

The channel meander length is simply the meander wavelength times thevalley slope divided by the channel slope.

wavelength valley slopechannel meander lengthchannel slope


Once meander wavelength is determined, one way to lay out the planform isto cut a string to the appropriate channel length and lay it out on a map. Another,more analytical approach, is to assume a sine-generated curve for the planformshape as suggested by Langbein and Leopold (1966) and calculate x-ycoordinates for the planform. Their theory of minimum variance is based on thehypothesis that the river will seek the most probable path (the path that providesthe minimum variance of bed shear stress and friction) between two fixed points.The sine-generated curve is defined in Figure 47 and by the following equation:




Φ = angle of meander path with the mean longitudinal axis

ω = maximum angle a path makes with the mean longitudinal axis in radians

s = the curvilinear coordinate along the meander path

M = the meander arc length

The shape parameter, ω, is a function of the channel sinuosity, P, which canapproximated by the following equation (Langbein and Leopold 1966):

PP 12.2 −

Hydraulic Design of Stream Restoration Projects - [PDF Document] (96)

Chapter 5 Hydraulic Design Methodology 85

Figure 47. Definition of sine-generated curve

Figure 48 shows how the shape parameter of a sine-generated curve definesthe shape of the stream.

Calculation of the points on a sine-generated curve is a rather tediousnumeric integration for Φ. However, it can be accomplished using a computerprogram such as the one in the SAM hydraulic design package. The sine-generated curve produces a very uniform meander pattern. A combination of thestring layout method and the analytical approach would produce a more naturallooking planform.

The radius of planform curvature is not constant in the sine-generated curvebut ranges from a maximum value at the inflexion point to a minimum curvaturearound the bend apex. The average radius of curvature is centered at the bendapex for a distance of approximately one sixth of the channel meander length.

Most reaches of stable meandering rivers have radius of curvature-to-widthratios between 1.5 and 4.5. Of the 438 sites used to derive the wavelength-widthrelationship in Figure 44, radius of curvature is recorded for 263 of the sites.This subset was used to develop a cumulative distribution curve of radius ofcurvature-to width ratios (Figure 49). This figure shows that 33.5 percent, 52.9percent, and 71.2 percent of the sites have radius of curvature-to-width ratiosbetween 2 and 3, 2 and 4, and 1.5 and 4.5 respectively. The final planformlayout should have ratios within the normal range.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (97)

86 Chapter 5 Hydraulic Design Methodology

Figure 48. Effect of the shape factor on channel sinuosity with the sine-generated curve (Langbien and Leopold 1966)

Figure 49. Cumulative distribution of radius of curvature-to-width ratio derivedfrom a composite data set of 263 sites

If the calculated meander length is too great, or if the required meander beltwidth is unavailable, grade control structures may be required to reduce thechannel slope and stabilize the bed elevations.






0.1 1 10 100

Radius of curvature to width ratio, R c / W






y of


s (%


R c / W = 2

R c / W = 3

R c / W = 4

Hydraulic Design of Stream Restoration Projects - [PDF Document] (98)

Chapter 5 Hydraulic Design Methodology 87

In streams that are essentially straight (sinuosity less than 1.2) riffle and poolspacing may be set as a function of channel width. The empirical guide of 5 to 7channel widths applies here (Knighton 1984). Two times this riffle spacing givesthe total channel length through one meander pattern.

Natural variability around meander bendways

Thorne (1988) and Soar and Thorne (2001) compiled empirical data sets ofcross section and planform dimensions from meander bends in the Red Riverbetween Index, AR, and Shreveport, LA. The Red River in this reach is typical oflarge meandering rivers, having a wide variety of both bend geometries and bankmaterials. These studies provided a useful baseline database for examining thevariability of width around meander bends. Of course, while the Red River in thestudy reach is representative of meandering rivers in general, if applied elsewherethese equations should be used with caution. In the data set, each bend wasclassified as one of three types based on the Brice (1975) classification system:equiwidth meanders - denoted as Type-e (Te) meanders, meanders with point bars- denoted as Type-b (Tb) meanders, and meanders with point bars and chutechannels - denoted as Type-c (Tc) meanders.

a. Equiwidth meandering. Equiwidth indicates that there is only minorvariability in channel width around meander bends. These channels aregenerally characterized by: low width/depth ratios; erosion resistantbanks; fine-grain bed material (sand or silt); low bed material load; lowvelocities; and low stream power. Channel migration rates are relativelylow because the banks are naturally stable.

b. Meandering with point bars. Meandering with point bars refers tochannels that are significantly wider at bendways than crossings, withwell-developed point bars but few chute channels. These channels aregenerally characterized by: intermediate width/depth ratios; moderatelyerosion resistant banks; medium grained bed material (sand or gravel);medium bed material load; medium velocities; and medium streampower. Channel migration rates are likely to be moderate unless banksare stabilized.

c. Meandering with point bars and chute channels. Meandering with pointbars and chute channels refers to channels that are very much wider atbendways than crossings, with well-developed point bars and frequentchute channels. These channels are generally characterized by: moderateto high width/depth ratios; highly erodible banks; medium to coarsegrained bed material (sand, gravel, and/or cobbles); heavy bed materialload; moderate to high velocities; and moderate to high stream power.Channel migration rates are likely to be moderate to high unless banksare stabilized.

Ranges of physical characteristics pertaining to each of the meander bendtypes are given in Table 6. Figure 50 provides a definition sketch for channelcross-section geometries and dimensions through a meander.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (99)

88 Chapter 5 Hydraulic Design Methodology

Table 6Ranges of Physical Characteristics Found in Different Meander Bend Types Identifiedfrom the 1981 Red River Hydrographic Survey Between Index, AR, and Shreveport, LA

n S (106) P Wi / Dm Dmax / Di Rc / Wi

Type-e 20(8)

65 to 268(133 to 268)

1.0 to 2.1(1.2 to 2.1)

34.2 to 74.1(38.3 to 74.1)

1.6 to 2.4(1.7 to 2.4)

0.9 to 9.3(0.9 to 5.2)

Type-b 34(19)

76 to 294(105 to 294)

1.0 to 2.0(1.1 to 2.0)

36.8 to 121.0(36.8 to 102.4)

1.5 to 2.6(1.7 to 2.6)

1.5 to 9.1(1.5 to 6.1)

Type-c 13(10)

91 to 201(91 to 201)

1.1 to 2.3(1.2 to 2.3)

33.5 to 88.2(33.5 to 88.2)

1.6 to 2.4(1.6 to 2.4)

2.2 to 6.8(2.2 to 5.2)

Note: n = number of meander bends studied; S = water-surface slope; P = sinuosity; Wi / Dm = inflexion point width-to-mean depthratio; Dmax / Di = maximum scour depth in pool-to-mean depth at inflexion point; Rc / Wi = radius of curvature-to-inflexion point widthratio. Values in parentheses refer to meander bends with sinuosity 1.2 or greater.

Two dimensionless parameters can be used to describe the width variabilityaround meander bends based on the enhanced Red River data set. These are theratio of bend apex width to inflexion point width, Wa/Wi, and the ratio of width atthe location of maximum bend pool scour to inflexion point width, Wp/Wi.Theoretically, these parameters adjust according to the degree of curvature andthe type of meander bend. To derive new morphological relationships, sinuosity,P, was preferred as the independent variable rather than the radius of curvature-to-width ratio, which would have resulted in width appearing on both sides of theregression equations.

Morphologic relationships for the width ratios as a function of meander typewere developed for channels with sinuosities greater than 1.2. This is acommonly accepted threshold between straight channels with only slightsinuosity and meandering channels with moderate to high sinuosity. The bedapex width to inflexion point width ratio, Wa/Wi, was found to be independent ofsinuosity. Data are plotted with confidence limits in Figure 51. Values for theratios for each type of meander bend can be determined from Table 7 and thefollowing equation, where p denotes the level of significance and corresponds tothe 100(1-p)% confidence level.


W uai

a +=


Hydraulic Design of Stream Restoration Projects - [PDF Document] (100)

Chapter 5 Hydraulic Design Methodology 89

Figure 50. Meander cross-section dimensions for restoration design. Note: Pointbars defined by shaded regions; Lm = meander wavelength, Z =meander arc length (riffle spacing); Am = meander belt width, Rc =radius of curvature; Θ = meander arc angle; W = reach averagebankfull width; D = depth of trapezoidal cross section; Dm = meandepth (cross-sectional area / W); Dmax = maximum scour depth inbendway pool; Wi = width at meander inflexion point; Wp = width atmaximum scour location; Wa = width at meander bend apex








A�A Am







A A�

B B�

C C�


Bend Apex

Inflexion Point

Maximum PoolScour







A A�

B B�

C C�








A A�

B B�

C C�








A A�

B B�

C C�


Bend Apex

Inflexion Point

Maximum PoolScour

Hydraulic Design of Stream Restoration Projects - [PDF Document] (101)

90 Chapter 5 Hydraulic Design Methodology

Table 7Constant Values Used to Estimate the Mean Ratio of Bend ApexWidth to Inflexion Point Width, Wa/Wi, Within Confidence Bandsfor Different Types of Meander Bends and for Sites with Sinuosityof at Least 1.2 (Coefficients Pertaining to the 99, 95 and 90 percentConfidence Limits are Given)

a u0.01 u0.05 u0.1

Type-e 1.05 0.08(0.29)



Type-b 1.35 0.05(0.27)



Type-c 1.79 0.09(0.36)



Note: Values given refer to mean response confidence limits. Value in parentheses is used tocalculate single response confidence limits.

Morphologic relationships for the width ratios as a function of meander typewere developed for the ratio of pool width at the location of maximum scour toinflexion point width (Wp/Wi ) for channels with sinuosities greater than 1.2.This ratio was also found to be independent of sinuosity. Data and confidencelimits are plotted in Figure 52. Values for the ratios for each type of meanderingriver can be determined from the following equation and Table 8.




p +=

Table 8Constant Values Used to Estimate the Mean Ratio of Pool Width(at Maximum Scour Location) to Inflexion Point Width, Wp/Wi,Within Confidence Bands for Different Types of Meander Bendsand for Sites with Sinuosity of at least 1.2. Coefficients Pertainingto the 99, 95, and 90 Percent Confidence Limits are Given.

a u0.01 u0.05 u0.1

Type-e 0.95 0.15(0.56)



Type-b 1.15 0.12(0.64)



Type-c 1.29 0.26(1.07)



Note: Values given refer to mean response confidence limits. Value in parentheses is used tocalculate single response confidence limits.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (102)

Chapter 5 Hydraulic Design Methodology 91

Figure 51. Ratio of bend apex width to inflexion point width, Wa/Wi as a functionof meander bend type only, for sinuosities of at least 1.2. Confidencelimits of a mean response are shown at the 95 percent level. Sourcedata: 1981 Red River hydrographic survey. Note: Filled symbols =sinuosity of at least 1.2; empty symbols = sinuosity less than 1.2

Figure 52. Ratio of pool width (at maximum scour location) to inflexion pointwidth, Wp/Wi as a function of meander bend type only, for sinuositiesof at least 1.2. Confidence limits of a mean response are shown atthe 95 percent level. Source data: 1981 Red River hydrographicsurvey. Note: Filled symbols = sinuosity of at least 1.2; emptysymbols = sinuosity less than 1.2

Hydraulic Design of Stream Restoration Projects - [PDF Document] (103)

92 Chapter 5 Hydraulic Design Methodology

While the location of meander inflexion points and bend apices aregeometrically defined, the location of pools, defined by the position of maximumbend scour, is not only controlled by the meander configuration but by thecomplex velocity distribution and large-scale coherent flow structures whichpulse sediment along the channel to form alternate zones of scour and fill. Innatural meanders, the deepest pool is usually located downstream from the bendapex and restoration design should mimic this natural attribute in constructedmeanders. The pool location in a meander bend can be represented empiricallyby a pool-offset ratio, defined as the ratio of the channel distance between bendapex and maximum scour location to the channel distance between bend apexand downstream inflexion point, Za-p / Za-i . The pool�offset ratio was found tobe independent of sinuosity. Neither was a distinct relationship found for thedifferent meander types. The range and cumulative distribution function for thepool-offset ratio is shown in Figure 53. The mean value for the ratio was 0.36and the range was �0.4 to 1.08.

Figure 53. Cumulative distribution of the pool-offset ratio, Za-p/Za-i, for all types ofmeander bend studied. Confidence limits on the mean response areshown. Source data: 1981 Red River hydrographic survey

Data from a wide range of rivers (Thorne and Abt 1993; Maynord 1996)were used to develop morphological equations for the maximum scour depth inpools. The data were divided into two subsets using a width-to-depth thresholdvalue of 60, which is an approximate modal value. The best-fit morphologicalrelationships are given by Soar and Thorne (2001) as:


i <DW





max ln190142WR




i ≥DW

, 10i

c <WR





max ln540982WR



Hydraulic Design of Stream Restoration Projects - [PDF Document] (104)

Chapter 5 Hydraulic Design Methodology 93

A practical safe design curve may then be defined by considering both equationsas





max 5451−



DD (5)

This equation is an asymptotic relationship with a theoretical minimumDmax/Dm of 1.5 representing pool scour depths expected in a straight channel witha pool-riffle bed topography. From this upper-bound relationship, Dmax/Dm rangesfrom 4 to 3 for Rc/Wi between 1.8 and 3. For channels with an Rc/Wi of less than1.8 pool depth is independent of bend curvature and it is recommended that thedimensionless scour depth should be fixed at 4. All three relationships areportrayed in Figure 54, which shows that this equation is a safe curve for bothclasses of Wi/Dm.

Figure 54. Dimensionless maximum scour depth in meander pools as a functionof radius of curvature-to-width ratio. Source data: Thorne and Abt(1993); Maynord (1996)

Practical channel design equations for meander bend geometry

Assuming that confidence is primarily a function of sample size in theanalysis of planform width variability, it is possible to derive a mean band ofuncertainty, u, suitable for all three types of meander bends to provide a set ofpractical design equations. The cumulative effects of e-type, b-type and c-typebends are represented by the binary parameters, Te, Tb and Tc, respectively. Thevalue of Te has a value of 1 for all three types of bend and represents the smallestplanform width ratio. If point bars are present but chute channels are rare, then Tbis assigned a value of 1 and Tc is assigned a value of 0. If point bars are presentand chute channels are common, then both Tb and Tc are assigned values of 1.Obviously Tc can only be given a value of 1 when Tb has a value of 1.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (105)

94 Chapter 5 Hydraulic Design Methodology

Bend Apex(P>=1.2)

05 0 30 0 44= ⋅ + ⋅ + ⋅ ±ae b c


W 1 T T T uW (6)

Pool Width(P>=1.2):

95 0 20 0 14= ⋅ + ⋅ + ⋅ ±pe b c


W0 T T T u

W (7)

For all three bend types and sinuosities greater than 1, the pool offset ratio isgiven by

Pool-Offset(P>1.0) 36= ⋅ ±a-p


Z0 u


Values of u refer to confidence limits on the mean response as given in Table 9.

Table 9Uncertainty, u, in Estimates of Width Variability Around MeanderBends and Location of Pools (Values Refer to Confidence Limitson the Mean Response)

Confidence Limits (%) Wa / Wi Wp / Wi Za-p / Za-i

99 0.07 0.17 0.11

95 0.05 0.12 0.08

90 0.04 0.10 0.07

A practical design equation for predicting or constructing maximum scourdepths at bends is the upper-bound curve in Figure 54, given by the followingequation





max 5.45.1−



DD (9)

For sites where active meandering is not permitted, bank protection will berequired along the outer bank to prevent erosion. In addition, this equationshould be used together with bank stability charts to establish whether bankstabilization against mass failure is also necessary.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (106)

Chapter 5 Hydraulic Design Methodology 95

Sediment Impact AssessmentThe potential success of a river project is often defined in terms of

performance based on a single flow event and the sediment load transported bythis event. This approach does not account for the potential for instability drivenby other flow events in the long-term record. The potential for restoring sedimentcontinuity through the restored reach requires an assessment of the sedimentbudget, which is determined by the magnitude and frequency of all sediment-transporting flows. To attain geomorphic stability through sediment continuity inthe medium- to long-term, the mean annual sediment load for the restoredchannel (capacity) must match the mean annual sediment load in the supply reach(supply).

On this basis, the sediment impact assessment is a closure loop at the end ofthe design procedure to: (a) validate the efficacy of the restored channelgeometry; (b) identify flows which may cause aggradation or degradation overthe short term (these changes are inevitable and acceptable in a dynamicchannel); and (c) recommend minor adjustments to the channel design to ensurethat dynamic stability will be continued over the medium- to long-term. This canbe accomplished using a sediment budget approach for relatively simple projectsor by using a numerical model that incorporates solution of the sedimentcontinuity equation for more complex projects. An example of a sedimentimpact assessment is given in Appendix G.

Adopting this approach should result in a low maintenance channel, withenvironmental and economic benefits that are sustainable in the long-term. Thisstep is especially important if the restored reach is part of a flood damagereduction project. In such cases it may be necessary to design a channel that isless than ideal in terms of channel stability in order to achieve flood-controlbenefits. Typically, a compound channel design provides the best combinationof benefits.

Topics to report

The following subject areas should be included in a sediment impactassessment report. Some of these items should have been completed early in thestudy process as part of the geomorphological assessment.

a. The project boundaries and study area boundaries should be identified.The study area should include the area affected by the project. Theproject's effect on water-surface elevations and sediment transportcapacity upstream and downstream of the proposed improvements shouldbe determined. This includes effects the project may have on tributaries,such as headcutting or induced deposition.

b. Available data sources should be identified and the need for additionaldata collection determined.

c. A site reconnaissance should be conducted to identify the stability of theexisting channel as well as existing problems upstream and downstream

Hydraulic Design of Stream Restoration Projects - [PDF Document] (107)

96 Chapter 5 Hydraulic Design Methodology

from the proposed project area. The type of bed-material sediment loadshould be determined, and bed-material samples collected. Aggradationand/or degradation in the project reach should be noted. The land use inthe basin should be noted, especially if there has been any significantchanges recently or historically.

d. A brief history of stream behavior in the study reach should bedeveloped. This history should describe aggrading and/or degradingtrends, land use changes, behavior of the system during flood events, andhistorical changes to and by the river system.

e. A sediment budget analysis should be conducted. This is therecommended approach for determining the severity of long-termaggradation or degradation trends, maintenance requirements, reliabilityduring passage of a design flood hydrograph, the need for upstream andtributary control measures to allow for changes in the water-surfaceelevations due to the project, and the need to make certain the tailwaterrating curve is stable.

Sediment budget analysis

The sediment budget analysis is the analytical backbone of the sedimentimpact assessment. This analysis provides relative stability comparisons forvarious alternatives, and provides an assessment of the general stability ofproposed plans. The level of confidence that can be assigned to the sedimentbudget approach is a function of the reliability of the available data about thestream and the project. The recommended steps for conducting a sedimentbudget analysis during the design phase of the study are essentially the same asthose used during the geomorphic assessment, as discussed in chapter 4. In thefinal design phase of the study, sediment yield from the supply reach is comparedto sediment yield through the designed channel reach.

Sediment rating curve analysis

The sediment-rating curve analysis discussed as part of the geomorphicassessment in chapter 4 can also be used to evaluate the project design. Thisqualitative technique does not require stream gage data or sediment gage data. Asediment rating curve is calculated for the proposed and project reachesfollowing the same procedure as is used in the sediment budget analysis.However, instead of gage data, peak flows are used which can be estimated usingregional regression or hydrologic modeling. The basic approach is to assess thesediment transport character of a study reach by comparing its sediment transportcapacity to that of its supply reach. This approach is illustrated in Figure 55.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (108)

Chapter 5 Hydraulic Design Methodology 97

Figure 55. Analogy sediment analysis of proposed project conditions

A comparison of the two sediment-rating curves in Figure 55 indicates thatthe project reach should be able to transport the incoming sediment load througha discharge of Q2. Above this discharge, deposition is a possibility with a strongpossibility of aggradation occurring above Q3. These discharges can becompared to the peak discharges of estimated storm frequencies to provide aqualitative estimate of project life. However, since there is no calibration, theactual quantity of deposition cannot be estimated. In addition, this approach doesnot account for changes in sediment transport capacity, which may occur assediment deposits in the section and changes its geometry. This technique can beused in conjunction with the sediment budget analysis to assess possible impactsof extreme storm events.

Numerical Sedimentation ModelingThe 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, wherechanging one significant geometric feature or boundary condition affects othergeometric features and flow characteristics both temporally and spatially.Changes at any given location in a stream system are directly related to theinflow of sediment from upstream. This makes the application of the sedimentcontinuity equation essential to any detailed analysis. The most significant ofthese relationships and the continuity of sediment mass are accounted for in thenumerical model approach. The fact that application of a numerical sedimentmodel requires knowledge of sediment transport and river mechanics should notbe a deterrent to its use; that knowledge is required for any responsible designwork in a river system. It should be expected that an analysis of system responsein a complicated system, such as a mobile-bed river system, would require someengineering effort. That effort should be based on analysis of the physical lawsthat govern the system. The system cannot be expected to adhere to constraintsplaced on it in violation of natural physical laws, no matter how well intentionedor frugally those constraints were developed. The critical decision with respect tousing a numerical model should be based on whether or not significant changesare expected to occur in the system as a result of the proposed design work. In

Hydraulic Design of Stream Restoration Projects - [PDF Document] (109)

98 Chapter 5 Hydraulic Design Methodology

the Corps, this decision typically is reached in the reconnaissance level planningstudy using the sediment impact assessment approach.

Operation and Maintenance RequirementsHydraulic design input for assessing operation and maintenance requirements

for a stream restoration project are the same as requirements for hydraulic designinput for local flood protection projects found in ER 1110-2-1405. Operationand maintenance requirements for a project should ensure the functionality of theproject and should be clearly outlined in the project design document and clearlydefined in the project Operation and Maintenance manual. These requirementsshould stand the tests of safety, reliability, functionality, cost-effectiveness andenvironmental consciousness. Environmental considerations in the conduct ofoperation and maintenance should be clearly defined in the Operation andMaintenance manual and coordinated in advance with applicable resourceagencies to ensure that these requirements can be implemented in a timelymanner and also within environmental resource constraints.

Operation should include a monitoring program to ensure that the projectbehaves as designed. It must be recognized that there are design uncertaintiesassociated with stream restoration design. It may be necessary to make projectdesign adjustments due to unexpected response. It is also possible that aninfrequent high flow event may impart severe damage to the project beforestabilization measures such as vegetation have had time to become established.The monitoring program should include data collection quantifying changes inaverage channel dimensions, bank erosion, aggradation or degradation, erosion inthe vicinity of structures, and vitality of vegetation.

Quality ManagementStream restoration projects should undergo established quality management

processes to ensure that the products being developed meet or exceedexpectations.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (110)

Chapter 6 Conclusions 99

6 Conclusions

This report presents a new systematic methodology for hydraulic andmorphological design of stream restoration projects. The methodology employsboth geomorphological principles and analytical engineering formulae. Theobjective of the methodology is to fit the stream restoration project into thenatural system within physical constraints imposed by past development of thefloodplain and other project objectives. It is critical that the design process hasparticipation from all the project stakeholders and from a range of scientificdisciplines to ensure that the project will meet expectations and objectives.

The basic steps in the hydraulic design methodology are as follows:

a. Define project objectives and constraints in cooperation withstakeholders.

b. Determine the hydrologic regime, including discharge frequencies, flow-duration curves, and channel-forming discharge.

c. Conduct a geomorphological analysis to assess historical channelstability and to determine the dominant geomorphological trendscurrently active in the watershed.

d. Determine average hydraulic dimensions for a stable main channel usingboth geomorphological principles and hydraulic formulae. The requireddependent design dimensions are width, depth, slope, and planform.Techniques presented in this report include analogy methods, hydraulicgeometry methods, and analytical methods that can be facilitated usingthe SAM hydraulic design package.

e. Conduct a sediment impact assessment to determine the impact that thefull range of natural flows will have on project stability. The primaryfocus of the sediment impact assessment is the sediment budget thatcompares sediment inflow to the project to sediment transport capacitythrough the project. The initial design should be refined and modifieduntil input and capacity are closely matched.

The scope of a hydraulic analysis of a stream restoration project will varydepending on the stage of the planning process and the magnitude of channelinstability problems. Two examples of stream restoration projects are describedin Appendixes F and G. Appendix F is a scope of work for a stability analysis

Hydraulic Design of Stream Restoration Projects - [PDF Document] (111)

100 Chapter 6 Conclusions

that might be conducted early in the planning process to define dominantgeomorphological processes. Appendix G is an example sediment impactassessment using methods described in this report.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (112)

References 101


American Society of Civil Engineers (ASCE). (1993). �Design and constructionof urban stormwater management systems - manuals and reports ofengineering practice, No. 77,� Prepared by the Task Committee of the UrbanWater Resources Research Council of the American Society of CivilEngineers and the Water Environment Federation, New York.

American Society of Civil Engineers (ASCE) Task Committee on Hydraulics,Bank Mechanics, and Modeling of River Width Adjustment. (1998). �Riverwidth adjustment, Part I: Processes and mechanics, and Part II: Modeling,�Journal of Hydraulic Engineering ASCE, 124(9), 881-917.

Andrews, E. D. (1980). �Effective and bankfull discharge of streams in theYampa Basin, western Wyoming,� Journal of Hydrology 46, 311-330.

Andrews, E. D. (1984). �Bed-material entrainment and hydraulic geometry ofgravel-bed rivers in Colorado,� Bulletin of the Geological Society ofAmerica, 95, 371-378.

Annable, W. K. (1996). Database of morphological characteristics ofwatercourses in southern Ontario. Ministry of Natural Resources, Ontario,Canada.

Bell, M. C. (1986). Fisheries handbook of engineering requirements andbiological criteria. Fish Passage Development and Evaluation Program, U.S.Army Engineer Division, North Pacific, Portland, OR.

Benson, M. A., and Thomas, D. M. (1966). �A definition of dominantdischarge,� Bulletin of the International Association of Scientific HydrologyXI, 76-80.

Biedenharn, D. S.; Derrick, D. L.; Elliott, C., and Watson, C. C. (1997). The WESstream investigation and streambank stabilization handbook. U.S. ArmyEngineer Waterways Experiment Station, Vicksburg, MS.

Biedenharn, D. S., Copeland, R. R., Thorne, C. R., Soar, P. J., Hey, R. D., andWatson, C. C. (2000). �Effective discharge calculation: A practical guide,�ERDC/CHL TR-00-15, U.S. Army Engineer Research and DevelopmentCenter, Vicksburg, MS.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (113)

102 References

Biedenharn, D. S., and Thorne, C. R. (1994). �Magnitude-frequency analysis ofsediment transport in the Lower Mississippi River,� Regulated Rivers 9, 237-251.

Bray, D. I. (1972). �Generalized regime-type analysis of Alberta rivers,� Ph.D.diss., Department of Civil Engineering, University of Alberta Publication.

Bray, D. I. (1982). �Regime equations for gravel-bed rivers,� Gravel bed rivers.R. D. Hey, J. C. Bathurst, and C. R. Thorne, ed., Wiley, Chichester, U.K.517-542.

Brice, J. C. (1975). �Airphoto interpretation of the form and behavior of alluvialrivers,� Final report to the United States Army Research Office, Durham,under Grant Number DA-ARD-D-31-124-70-G89, Washington University,St. Louis, MO.

Brice, J. C. (1984). �Planform properties of meandering rivers,� RiverMeandering, Proceedings of the �83 Rivers Conference. ASCE, NewOrleans, LA. October 24-26, 1983, C. M. Elliott, ed., 1-15.

Brown, S. A. (1996). �Urban drainage design manual,� Hydraulic EngineeringCircular No. 22. FHWA-SA-96-078. Federal Highway Administration,Washington, DC.

Brownlie, W. R. (1981). �Prediction of flow depth and sediment discharge inopen channels,� Report No. KH-R-43A, California Institute of Technology,W.M. Keck Laboratory of Hydraulics and Water Resources, Pasadena, CA.

Carling, P. A. (1988). �The concept of dominant discharge applied to two gravel-bed streams in relation to channel stability thresholds,� Earth SurfaceProcesses and Landforms 13, 355-367.

Carlston, C. W. (1965). �The relation of free meander geometry to streamdischarge and its geomorphic implications,� American Journal of Science263, 863-865.

Chang, H. H. (1980). �Stable alluvial canal design,� Journal of the HydraulicsDivision ASCE, 106 (HY5), 873-891.

Charlton, F. G., Brown, P. M., and Benson, R. W. (1978). �The hydraulicgeometry of some gravel rivers in Britain,� Report IT 180, HydraulicsResearch Station, Wallingford, U.K.

Cherry, D. S., Wilco*ck, P. R., and Wolman, M. G. (1996). �Evaluation ofmethods for forecasting planform change and bankline migration in flood-control channels,� Final report to U.S. Army Engineer WaterwaysExperiment Station, Vicksburg, MS, Johns Hopkins University, Baltimore,MD.

Chitale, S. V. (1970). �River channel patterns,� Journal of the HydraulicsDivision, ASCE, 96, 201-221.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (114)

References 103

Church, M. A., and Kellerhals, R. (1978). �On the statistics of grain sizevariation along a gravel river,� Canadian Journal of Earth Sciences 15,1151-1160.

Church, M. A., McLean, D. G., and Wolcott, J. F. (1987). �River bed gravels:sampling and analysis,� Sediment Transport in Gravel-Bed Rivers, C. R.Thorne, J. C. Bathurst, and R. D. Hey, ed., John Wiley and Sons Ltd.,Chichester, Great Britain, 43-88.

Copeland, R. R. (1994). �Application of channel stability methods � Casestudies,� TR HL-94-11, U.S. Army Engineer Waterways Experiment Station,Vicksburg, MS.

Copeland, R. R. (1995). �Sediment impact assessment, Carlsbad, New Mexico,�Miscellaneous Paper HL-95-7, U.S. Army Engineer Waterways ExperimentStation, Vicksburg, MS.

Diplas, P., and Vigilar, G. G. (1992). �Hydraulic geometry of thresholdchannels,� Journal of Hydraulic Engineering ASCE, 118(4), 597-614.

Downs, P. W., and Thorne, C. R. (1996). �The utility and justification of riverreconnaissance surveys,� Transactions of the Institute of BritishGeographers, New Series, 21, 455-468. ISSN 0020-2754.

Dury, G. H. (1965). �Theoretical implications of underfit streams,� U.S.Geological Survey, Professional Paper, 452C.

Dury, G. H. (1973). �Magnitude-frequency analysis and channel morphology,�Fluvial geomorphology: A proceedings volume of the fourth annualgeomorphology symposia series. M. Morisawa, ed., Binghampton, NY, Allenand Unwin, 91-121.

Einstein, H. A. (1950). �The bed-load function for sediment transportation inopen channel flows,� Technical Bulletin No. 1026, U.S. Department ofa*griculture, Soil Conservation Service, Washington, DC.

Emmett, W. W. (1972). �The hydraulic geometry of some Alaskan streams southof the Yukon River,� U.S. Geological Survey, Open File Report 72-108.

Emmett, W. W. (1975). �The channels and waters of the upper Salmon River,Idaho,� Geological Survey Professional Paper 870A, Washington, DC.

Einstein, H. A. (1950). �The bed-load function for sediment transport in openchannel flows,� Technical Bulletin No. 1026, U.S. Department ofa*griculture, Soil Conservation Service, Washington DC.

Federal Interagency Stream Restoration Working Group. (1998). Stream corridorrestoration: Principles, processes, and practices NRCS, Washington, DC.http://www.usda.gov/stream_restoration.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (115)

104 References

Ferguson, B. K. (1991). �The failure of detention and the future of stormwaterdesign,� Landscape Architecture Techniques and Practice � ConstructionDecember.

Ferguson, B. K. (1994). Stormwater infiltration. Lewis Publishers, CRC Press,Boca Raton, FL.

Geological Survey Professional Paper 422-H, Washington, DC.

Graybill, F. A., and Iyer, H. K. (1994). Regression analysis: Concepts andapplications. Duxbury Press, Belmont, CA.

Hey, R. D. (1975). �Design discharge for natural channels,� Science, technologyand environmental management, R. D. Hey and T. D. Davies, ed., SaxonHouse, 73-88.

Hey, R. D. (1976). �Geometry of river meanders,� Nature 262, 482-484.

Hey, R. D. (1997). �Channel response and channel forming discharge: literaturereview and interpretation,� Final Report for the U.S. Army Contract NumberR&D 6871-EN-01.

Hey, R. D., and Thorne, C. R. (1983). �Accuracy of surface samples from gravel-bed material,� Journal of Hydraulic Engineering 109, 842-851.

Hey, R. D., and Thorne, C. R. (1986). �Stable channels with mobile gravel-beds,� Journal of Hydraulic Engineering ASCE, 112, 6, 671-689.

Ikeda, S., Parker, G., and Kimura, Y. (1988). �Stable width and depth of straightgravel rivers with heterogeneous bed materials,� Water Resources Research,24(5), 713-722.

Ikeda, S., and Izumi, N. (1990). �Width and depth of self-formed straight gravelrivers with bank vegetation,� Water Resources Research 26(10), 2353-2364.

Inglis, C. C. (1941). �Digest of answers to the Central Board of Irrigationquestionnaire on meandering of rivers with comments on factors controllingmeandering and suggestions for future actions,� Central board of irrigationannual technical report, 1939-1940 session. Publication 24, A. R.Edgecombe, ed., Poona, India, 100-114.

Knighton, A. D. (1984). Fluvial forms and processes. Arnold, London.

Langbein, W. B., and Leopold, L. B. (1966). �River meanders - Theory ofminimum variance,� Geological Survey Professional Paper 422-H,Washington DC.

Lane, E. W. (1955). �Design of stable channels,� Transactions of the AmericanSociety of Civil Engineers. 120, 1234-1279.

Leopold, L. B. (1994). A view of the river. Harvard University Press, Cambridge.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (116)

References 105

Leopold, L. B., and Maddock, T. (1953). �The hydraulic geometry of streamchannels and some physiographic implications,� Geological SurveyProfessional Paper 252, Washington, DC.

Leopold, L. B., and Wolman, M. G. (1957). �River channel patterns - braided,meandering and straight,� Geological Survey Professional Paper 282B,Washington, DC.

Leopold, L. B., and Wolman, M. G. (1960). �River meanders,� Bulletin of theGeological Society of America 71, 769-794.

Leopold, L. B., Wolman, M. G., and Miller, J. P. (1964). Fluvial processes ingeomorphology. W.H. Freeman & Company, San Francisco.

Limerinos, J. T. (1970). �Determination of the Manning coefficient frommeasured bed roughness in natural channels,� U.S. Geological SurveyWater-Supply Paper 1898-B, Washington DC.

Linsley, R. K., Kohler, M. A., and Paulhus, J. L. H. (1975). Hydrology forengineers. McGraw-Hill Series in Water Resources and EnvironmentalEngineering, New York.

Maryland Department of the Environment. (1984). Infiltration design andpractice. Baltimore MD.

Maryland Department of the Environment. (1987). Design procedures forstormwater management detention structures. Baltimore, MD.

Maynord, S. T. (1996). �Toe scour estimation in stabilized bendways,� Journalof Hydraulic Engineering ASCE, 122(8), 460-464.

McCuen, R. H., and Moglen, G. E. (1988). �Multicriterion stormwatermanagement methods,� Journal of Water Resources Management ASCE,114 ,4.

Meyer-Peter, E., and Muller, R. (1948). �Formulas for bed-load transport,� Proc.Second Meeting of the International Association for Hydraulic Research,Stockholm, Sweden, 39-64.

Millar, R. G., and Quick, M. C. (1993). �Effect of bank stability on geometry ofgravel rivers,� Journal of Hydraulic Engineering ASCE, 119 (12), 1343-1363.

Moglen, G. E., and McCuen, R. H. (1988). �Effects of detention basins on in-stream sediment movement,� Journal of Hydrology 104.

Myers, R. H. (1990). Classical and modern regression with applications. Seconded., Duxburg Press, Belmont, CA.

Nixon, M. (1959). �A study of bankfull discharges of rivers in England andWales,� Proceedings of the Institution of Civil Engineers 12, 157-175.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (117)

106 References

Nolan, K. M., Lisle, T. E., and Kelsey, H. M. (1987). �Bankfull discharge andsediment transport in northwestern California,� Proceedings of the CorvallisSymposium, Erosion and Sedimentation in the Pacific Rim, InternationalAssociation of Hydrological Sciences, Publication No. 165, WallingfordU.K. 439-449.

Parker, G. (1978). �Self-formed straight rivers with equilibrium banks andmobile bed, Part I � The sand-silt river,� Journal of Fluid Mechanics 89,109-125.

Parker, G. (1979). �Hydraulic geometry of active gravel rivers,� Journal of theHydraulics Division, ASCE, 105(9), 1185-1201.

Parker, G. (1990). �The ACRONYM series of Pascal programs for computingbedload transport in gravel rivers,� External Memorandum No. M-220, St.Anthony Falls Hydraulic Laboratory, University of Minnesota.

Pickup, G., and Warner, R. F. (1976). �Effects of hydrologic regime onmagnitude and frequency of dominant discharge,� Journal of Hydrology 29,51-75.

Prince Georges County, Maryland. (1997). Low impact development designmanual. Department of Environmental Resources, Marlboro, MD.

Richards, K. S. (1982). Rivers: Form and process in alluvial channels. Methuen,London.

Scharffenberg, W. A. (2001). Hydrologic Modeling System HEC-HMS User�sManual, USACE Hydrologic Engineering Center, Davis, CA.

Schoklitsch, A. (1930). �Handbuch des wasserbaues,� Springer, Vienna (2nd ed.)English translation by S. Shulits., �The Schoklitsch bed-load formula,�Engineering. London, England, June 21, 1935, 644-646; and June 28, 1935,687.

Schueler, T. (1994). �The importance of impervious,� Watershed ProtectionTechniques 1(3).

Schumm, S. A. (1960). �The shape of alluvial channels in relation to sedimenttype,� Geological Survey Professional Paper 352-B, Washington DC.

Schumm, S. A. (1967). �Meander wavelength of alluvial rivers,� Science, 157,1549-1550.

Schumm, S. A. (1968). �River adjustment to altered hydrologic regime �Murrumbidgee River and Palaeochannels, Australia,� U.S. GeologicalSurvey Professional Paper, 598.

Schumm, S. A. (1977). The fluvial system. Wiley, New York.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (118)

References 107

Schumm, S. A., Harvey, M. D., and Watson, C. C. (1984). Incised channels,morphology, dynamics and control. Water Resources Publications, Littleton,CO.

Shulitis, S. (1935). �The Schoklitsch bed-load formula,� Engineering London,England, June 21, 644-646 and June 28, 687.

Simon, A. (1989). �A model of channel response in disturbed alluvial channels,�Earth Surface Processes and Landforms 14, 11-26.

Soar, P. J., and Thorne, C. R. (2001). �Channel restoration design for meanderingrivers,� ERDC/CHL CR-01-1, U.S. Army Engineer Research andDevelopment Center, Vicksburg, MS.

Thomas, W. A., Copeland, R. R., McComas, D. N., and Raphelt, N. K. (2000).�User�s manual for the SAM hydraulic design package for channels,� U.S.Army Engineer Research and Development Center, Vicksburg, MS.

Thorne, C. R. (1988). �Bank processes on the Red River between Index,Arkansas and Shreveport, Louisiana,� Final Report to U.S. Army EuropeanResearch Office, under contract DAJA45-88-C-0018, Department ofGeography, Queen Mary College, U.K.

Thorne, C. R. (1997). �Channel types and morphological classification,� Appliedfluvial geomorphology for river engineering and management. C. R. Thorne,R. D. Hey, and M. D. Newson, ed., Wiley, Chichester, U.K., 175-222.

Thorne, C. R. (1998). Stream reconnaissance handbook. Wiley, Chichester, U.K.

Thorne, C. R., and Abt, S. R. (1993). �Velocity and scour prediction in riverbends,� Contract Report HL-93-1, U.S. Army Engineer WaterwaysExperiment Station, Vicksburg, MS.

Thorne, C. R., Simon, A., and Allen, R. (1996). �Geomorphological riverchannel reconnaissance for river analysis, engineering and management,�Transactions of the Institute of British Geographers, New Series, 21, 469-483.

Thorne, C. R., and Soar, P. J. (2000). �Analysis of channels with compoundchannels for channel restoration design,� Final Report submitted to U.S.Army, Research Development and Standardization Group � U.K., London,under contract No. N68171-00-M-5506, Project No. W90C2K-8913-EN01,School of Geography, University of Nottingham, Nottingham, U.K.

U.S. Army Corps of Engineers (USACE). (1982). �Hydraulic Design for LocalFlood Protection Projects,� ER 1110-2-1405, Washington, DC.

U.S. Army Corps of Engineers (USACE). (1993). �Hydrologic frequencyanalysis,� EM 1110-2-1415, Washington, DC.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (119)

108 References

U.S. Army Corps of Engineers (USACE). (1994a). �Channel stability assessmentfor flood control projects,� EM 1110-2-1418, Washington, DC.

U.S. Army Corps of Engineers (USACE). (1994b). �Hydraulic design of floodcontrol channels,� EM 1110-2-1601, Washington, DC.

U.S. Army Corps of Engineers (USACE). (1995). �Sedimentation investigationsin rivers and reservoirs,� EM 1110-2-4000, Washington, DC.

U.S. Army Corps of Engineers (USACE). (1996). �Risk-based analysis for flooddamage reduction studies,� EM 1110-2-1619, Washington, DC.

U.S. Army Corps of Engineers Hydrologic Engineering Center (USACE, HEC).(1981). �HEC-1 flood hydrograph package, Users manual,� Davis, CA.

U.S. Army Corps of Engineers Hydrologic Engineering Center (USACE, HEC).(1990). �HEC-2 water surface profiles, users manual,� Davis, CA.

U.S. Army Corps of Engineers Hydrologic Engineering Center (USACE, HEC).(1993). �HEC-6 scour and deposition in rivers and reservoirs, users manual,�Davis, CA.

U.S. Army Corps of Engineers Hydrologic Engineering Center (USACE, HEC).(2001). �HEC-RAS river analysis system, users manual,� Davis, CA.

U.S. Army Engineers (USAE). (1991). Tidal Hydraulics, EM 1110-2-1607,Office of the Chief of Engineers, Washington, DC.

U.S. Army Engineers (USAE). (1989). �Sediment investigations of river andreservoirs,� EM 1110-2-4000, Office of the Chief of Engineers, Washington,DC, Change 1.

U.S. Department of Agriculture (USDA). (1977). Design of open channels.Technical Release No. 25. Soil Conservation Service, Washington, DC.

U.S. Interagency Advisory Committee on Water Data (USIACWD). (1982).�Guidelines for determining flood flow frequency, Bulletin 17-B of theHydrology Subcommittee,� Office of Water Data Coordination, U.S.Geological Survey, Reston, VA.

U.S. Geological Survey (USGS). (1993). �Nationwide summary of U. S.Geological Survey regional regression equations for estimating magnitudeand frequency of floods for ungaged sites,� Water-Resources InvestigationsReport 94-4002.

U.S. Soil Conservation Service (SCS). (1985). National engineering handbook,Section 4, �Hydrology.� U.S. Department of Agriculture, Washington, DC.

U.S. Soil Conservation Service (SCS). (1986). �Urban hydrology for smallwatersheds,� Technical Release 55 (TR-55).

Hydraulic Design of Stream Restoration Projects - [PDF Document] (120)

References 109

U.S. Soil Conservation Service (SCS). (1992). �Computer program for projectformulation hydrology,� Technical Release 20 (TR-20).

Watson, C. C., Dubler, D., and Abt, S. R. (1997). �Demonstration erosion controlproject report, design hydrology investigations,� Fort Collins, Colorado,submitted to U.S. Army Engineer Waterways Experiment Station,Vicksburg, MS.

Webb, B. W., and Walling, D. E. (1982). �The magnitude and frequencycharacteristics of fluvial transport in the Devon drainage basin and somegeomorphological implications,� Cantena, 9, 9-23.

White, W. R., Bettess, R., and Paris, E. (1982). �Analytical approach to riverregime,� Journal of the Hydraulics Division ASCE, 108 (HY10), 1179-1193.

Williams, G. P. (1978). �Bankfull discharge of rivers,� Water ResourcesResearch 14(6), 1141-1154.

Williams, G. P. (1986). �River meanders and channel size,� Journal ofHydrology 88, 147-164.

Wolman, M. G. (1954). �A method for sampling coarse river-bed material,�American Geophysical Union Transactions. 35(6), 951-956.

Wolman, M. G. (1955). �The natural channel of Brandywine CreekPennsylvania,� Geological Survey Professional Paper 271, Washington, DC.

Wolman, M. G., and Leopold, L. B. (1957). �River flood plains: Someobservations on their formation,� Geological Survey Professional Paper 282-C, Washington, DC.

Wolman, M. G., and Miller, J. P. (1960). �Magnitude and frequency of forces ingeomorphic processes,� Journal of Geology 68, 54-74.

Woodyer, K. D. (1968). �Bankfull frequency in rivers,� Journal of Hydrology 6,114-142.

Yalin, M. S. (1992). River mechanics. Pergamon Press, New York.

Yang, C. T. (1973). �Incipient motion and sediment transport,� Journal of theHydraulics Division, American Society of Civil Engineers, 99, (HY10),1679-1704.

Yang, C. T. (1984). �Unit stream power equation for gravel,� Journal of theHydraulics Division, American Society of Engineers, 110 (12) 1783-1797.

Yevjevich, V. (1972). Probability and statistics in hydrology. Water ResourcesPublications, Fort Collins, CO.

Yuzyk, T. R. (1986). �Bed material sampling in gravel-bed streams, sedimentsurvey section, environment Canada,� IWD-HQ-WRB-SS-86-8, Ottawa,Ontario.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (121)

Appendix A Flow Duration Curves for Effective Discharge Calculation A1

Appendix AFlow Duration Curves forEffective DischargeCalculation1

A standardized procedure is required to ensure that effective dischargecalculations are accurate and that results from different sites can be compared. Tobe practical, the procedure must use only data that are readily available fromgaging stations, or that can be synthesized using limited additional computations.

The basic approach is to divide the range of river flows during the period ofrecord into a number of arithmetic classes and then calculate the total sedimentquantity transported by each class. This is achieved by multiplying the frequencyof occurrence of each flow class by the median sediment load for that flow class(Figure A1). The initial data required are flow duration data and a sedimenttransport rating curve.

The calculated value of the effective discharge depends to some extent on thesteps used to manipulate the input data to define the flow regime and sedimenttransport function. The procedure described here represents the best practice inthis regard, based on extensive first-hand experience in using flow and sedimenttransport data to determine the effective discharge.

Gaged SitesThe first step in an effective discharge calculation is to group the discharge

data into equal arithmetic flow classes and determine the number of eventsoccurring in each class during the period of record. Logarithmic or nonequalwidth arithmetic classes introduce systematic bias into the calculation of effectivedischarge and should not be used. Grouping the discharge data is usuallyaccomplished using a flow-duration curve, which is a cumulative distributionfunction of observed discharges at the gaging station. Figure A2 is an example ofa flow-duration curve calculated for the Sevier River, UT. The flow-duration

1 Extracted from Biedenharn et al. (2000). All references cited in this appendix are listedin the References section following the main text.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (122)

A2 Appendix A Flow Duration Curves for Effective Discharge Calculation

curve defines the percentage of time a particular discharge is equaled orexceeded. The frequency of occurrence of each discharge class is calculated fromthis curve. Three critical components must be considered when developing aflow-duration curve: the number of discharge classes; the time base for dischargeaveraging; and the length of the period of record. It is important that thehistorical record is hom*ogeneous, i.e., watershed conditions are unchanged.

Class interval and number of classes

The selection of class interval can influence the effective dischargecalculation. Intuitively, it might be expected that the smaller the class intervaland, therefore, the greater the number of classes, the more accurate would be theoutcome. However, when too small an interval is used, discontinuities appear inthe discharge frequency distribution. These in turn produce a rather irregularsediment load histogram having multiple peaks. Therefore, the selected classinterval should be small enough to accurately represent the frequency distributionof flows but large enough to produce a continuous distribution.

There are no definite rules for selecting the most appropriate interval andnumber of classes, but Yevjevich (1972) stated that the class interval should notbe larger than s/4, where s is an estimate of the standard deviation of the sample.For hydrological applications he suggested that the number of classes should bebetween 10 and 25, depending on the sample size.

Hey (1997) found that 25 classes with equal, arithmetic intervals produced arelatively continuous flow frequency distribution and a smooth sediment-loadhistogram with a well defined peak, indicating an effective discharge whichcorresponded exactly with bankfull flow. A smaller interval, and correspondinglylarger number of classes, produced anomalous results. Experience has shown thatin some cases, 25 classes produce unsatisfactory results and that up to 250 classesmay be required. Particular care has to be exercised on rivers where there is ahigh incidence of very low flows. Under these circ*mstances, the effectivedischarge may be biased towards the lowest discharge class.

Time base

Mean daily discharges are conventionally used to construct the flow-durationcurve. Although this is convenient, given the ready availability of mean dailydischarge data from the United States Geological Survey (USGS), it can, in somecases, introduce error into the calculations. This arises because mean daily valuescan underrepresent the occurrence of short-duration, high magnitude flow eventsthat occur within the averaging period, while overrepresenting effects of lowflows.

On large rivers, such as the Mississippi, the use of the mean daily values isacceptable because the difference between the mean and peak daily discharges isnegligible. However, on smaller streams, flood events may last only a few hoursand the peak daily discharge can be much greater than the corresponding meandaily discharge. Excluding the flood peaks and the associated high sediment

Hydraulic Design of Stream Restoration Projects - [PDF Document] (123)

Appendix A Flow Duration Curves for Effective Discharge Calculation A3

loads can result in underestimation of the effective discharge. Rivers with a highflashiness index, defined as the ratio of the instantaneous peak flow to theassociated daily mean flow, are most likely to be affected. To avoid this problemit may be necessary to reduce the time base for discharge averaging from 24 hr(mean daily) to 1 hr, or even 15 min on flashy streams. For example, aninvestigation of discharge data for 11 USGS gaging stations in the Yazoo RiverBasin, MS, revealed that the annual yields of bed material calculated using meandaily discharge data were approximately 50 percent less than the yieldscalculated using 15-min data (Watson, Dubler, and Abt 1997). These arerelatively small basins (drainage areas less than 1,000 km2) with high rainfallintensities and runoff characteristics that have been severely affected by land-usechange and channel incision. Consequently, hydrographs are characterized bysteep rising and falling limbs, with events peaking and returning to base flow inmuch less than 24 hr.

In practice, mean daily discharge data may be all that are available for themajority of gaging stations and these data may be perfectly adequate. However,caution must be exercised when using mean daily data for watersheds with flashyrunoff regimes and short-duration hydrographs. The use of 15-min data toimprove the temporal resolution of the calculations should be consideredwhenever the available flow records allow it.

In the absence of 1-hr or 15-min data, recorded hydrographs from USGSgaging records can be used to refine the high discharge portion of the flow-duration curve. Actual instantaneously recorded hydrographs can be used todetermine 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 ofmorphologically significant flows, but not so long that changes in the climate,land-use or runoff characteristics of the watershed produce significant changes inthe data. If the period of record is too short, there is a significant risk that theeffective discharge will be inaccurate due to the occurrence of unrepresentativeflow events. Conversely, if the period is too long, there is a risk that the flow andsediment regimes of the stream at the beginning of the record may besignificantly different to current conditions.

A reasonable minimum period of record for an effective dischargecalculation is about 10 years, with 20 years of record providing more certaintythat the range of morphologically significant flows is fully represented in thedata. Records longer than 30 years should be examined carefully for evidence oftemporal changes in flow and/or sediment regimes. If the period of record at agaging station is inadequate, consideration should be given to developing aneffective discharge based on regional estimates of the flow duration.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (124)

A4 Appendix A Flow Duration Curves for Effective Discharge Calculation

Ungaged SitesAt locations where gaging 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 isby using records from nearby gaging 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 ofthe flow-duration characteristics and that there can be considerable uncertainty inthe results. The reliability of these methods is a function of the quality of theexisting gage data, and the morphologic similarity between the gaged andungaged locations. Caution is advised whenever the existing gage data arelimited, or the site in question has a significantly different morphologic characterthan the gaged site.

Drainage area - flow duration curve method

This method relies on the availability of gaging station data at a number ofsites on the same river as the ungaged location. First, flow duration curves foreach gaging station are derived for the longest possible common period of record.This guarantees comparability between the data, as all the gaging stations haveexperienced the same flow conditions, and ensures that the curves represent thelonger period. Provided there is a regular downstream decrease in the dischargeper unit watershed area, then a graph of discharge for a given exceedanceduration against upstream drainage area will produce a power function withvirtually no scatter about the best fit regression line. Figure A3 shows thisrelationship for the River Wye, UK (Hey 1975). The equations generated by thismethod enable the flow-duration curve at an ungaged site on that river to bedetermined as a function of its upstream watershed area.

For sites on streams where there is only one gaging station, flow-durationcurves can be estimated at ungaged locations provided the streams are tributariesto rivers where the relation between discharge and drainage area conforms to aknown power function. Estimates of the contributing flow to the main stem canbe obtained from the difference between discharges on the main stem above andbelow the tributary junction. Discharge - drainage area relations can then bederived for the tributary given the flow-duration curve at the gaging station andthe predicted curve at its confluence with the main stem. However, this techniqueshould not be used if there are distinct and abrupt downstream changes in thedischarge per unit area for the watershed. This could occur if portions of thedrainage area consisted of different hydrological regions. In this case it wouldbe preferable to use the regionalized duration curve method described in the nextparagraph.

Regionalized duration curve method

An alternative to the use of watershed area to generate a flow-duration curvefor an ungaged site is to use a regional-scaling method based on data from

Hydraulic Design of Stream Restoration Projects - [PDF Document] (125)

Appendix A Flow Duration Curves for Effective Discharge Calculation A5

watersheds with similar characteristics. For example, Emmett (1975) andLeopold (1994) suggest using the ratio of discharge to bankfull discharge (Q/Qb)as a nondimensional index to transfer flow-duration relationships between basinswith similar characteristics. However, bankfull discharge does not necessarilyhave either a consistent duration or return period (Williams 1978).

To avoid this problem, a nondimensional discharge index was proposed byWatson, Dubler, and Abt (1997) using the regionalized 2-year discharge tonormalize discharges (Q/Q2). For ungaged sites the 2-year discharge may beestimated from regionalized discharge frequency relationships developed by theUSGS (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 (Q/Q2) can be used to transfer a flow-duration relationship to an ungaged site from a nearby gaged site. The gaged sitemay be within the same basin, or an adjacent watershed.

To transfer a flow duration relationship within a watershed use the followingsteps:

a. Develop the regionalized flow-duration curve. Using a flow-durationcurve from a gaged site in a physiographically similar watershed, dividethe discharges in the flow-duration relationship by the Q2 for the gagedsite. This creates a dimensionless flow-duration curve. If more than onegage site is available an average dimensionless flow-duration curve forall the sites can be developed.

b. Compute the Q2 for the ungaged site.

c. Calculate the flow-duration curve for the ungaged site. Multiply thedimensionless ratios from the regionalized flow-duration curve by theungaged Q2.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (126)

A6 Appendix A Flow Duration Curves for Effective Discharge Calculation

Figure A1. Derivation of total sediment load-discharge histogram (c) from flowfrequency (a) and sediment load rating curves (b)

Figure A2. Daily mean flow-duration curve: Sevier River, Hatch, UT (from Hey1997)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (127)

Appendix A Flow Duration Curves for Effective Discharge Calculation A7

Figure A3. Downstream daily flow-duration curves: River Wye, UK 1937-1962(Hey 1975)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (128)

Appendix B Stream Reconnaissance Sheets, Thorne (1993) B1

Appendix BStream ReconnaissanceSheets, Thorne (1993)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (129)

B2 Appendix B Stream Reconnaissance Sheets, Thorne (1993)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (130)

Appendix B Stream Reconnaissance Sheets, Thorne (1993) B3

Hydraulic Design of Stream Restoration Projects - [PDF Document] (131)

B4 Appendix B Stream Reconnaissance Sheets, Thorne (1993)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (132)

Appendix B Stream Reconnaissance Sheets, Thorne (1993) B5

Hydraulic Design of Stream Restoration Projects - [PDF Document] (133)

B6 Appendix B Stream Reconnaissance Sheets, Thorne (1993)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (134)

Appendix B Stream Reconnaissance Sheets, Thorne (1993) B7

Hydraulic Design of Stream Restoration Projects - [PDF Document] (135)

B8 Appendix B Stream Reconnaissance Sheets, Thorne (1993)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (136)

Appendix B Stream Reconnaissance Sheets, Thorne (1993) B9

Hydraulic Design of Stream Restoration Projects - [PDF Document] (137)

Appendix C Stream Reconnaissance Data Sheets by Baltimore District C1

Appendix CStream Reconnaissance DataSheets by Baltimore District

Hydraulic Design of Stream Restoration Projects - [PDF Document] (138)

C2 Appendix C Stream Reconnaissance Data Sheets by Baltimore District

Hydraulic Design of Stream Restoration Projects - [PDF Document] (139)

Appendix C Stream Reconnaissance Data Sheets by Baltimore District C3

Hydraulic Design of Stream Restoration Projects - [PDF Document] (140)

C4 Appendix C Stream Reconnaissance Data Sheets by Baltimore District

Hydraulic Design of Stream Restoration Projects - [PDF Document] (141)

Appendix C Stream Reconnaissance Data Sheets by Baltimore District C5

Hydraulic Design of Stream Restoration Projects - [PDF Document] (142)

Appendix D Guidelines for Sampling Bed Material D1

Appendix DGuidelines for Sampling BedMaterial

Purpose of Bed Material SamplingKnowledge of streambed characteristics is necessary for a variety of

engineering and environmental purposes related to stream restoration projects.Bed material sampling programs must be carefully designed to meet the particularneeds of a specific study. Stream restoration studies may include objectivesrelated to: the source, transport and fate of pollutants; fish habitat; resourcemanagement; morphological trends and/or river engineering works.Contaminates typically attach to cohesive sediment and therefore are distributedover a wide area, especially in areas where flow velocity is low. Sampling for acontaminate study should concentrate on depositional zones in the stream andoverbank. Fish habitat studies may be concerned with the suitability of thestreambed for spawning. Sampling for this type of study should be relativelyextensive, identifying lateral, longitudinal, and temporal variations in the surfacelayer over a wide area of the stream. Resource management studies arefrequently concerned with the need or feasibility of sand and gravel mining. Coreor substrate sampling that identifies vertical variation of the streambed isessential for this type study. Morphologic and engineering studies are typicallyconcerned with changes in the character of the river over time. These studiesfrequently require knowledge of the grain size distribution of both the bed surfacematerial and subsurface material for sediment transport calculations, critical shearstress determinations, determining potential for particle sorting and armoring, andfor determining hydraulic roughness.

Bed material sampling is frequently conducted in order to make sedimenttransport calculations. For this purpose the sampling program should identify a�representative� bed material gradation, but it is also necessary to identify anylateral, longitudinal, vertical, and/or temporal variation in bed materialcomposition. Lateral variations in bed material gradation can be much moresignificant than longitudinal variations. In sand bed streams the sample istypically taken from the upper five centimeters of the bed surface. In gravel bedstreams with coarse surface layers, samples of both the surface and subsurface

Hydraulic Design of Stream Restoration Projects - [PDF Document] (143)

D2 Appendix D Guidelines for Sampling Bed Material

layers are required. Ideally, bed material samples should be taken at differenttimes during the year to account for seasonal variations.

Table D-1 provides guidance relative to where a bed-material sample mightbe taken as a function of the type of geomorphologic or engineering analysis to beconducted.

Table D1Bed Material Sampling SitesPurpose of analysis Sample locationTo estimate the maximum permissible velocity in athreshold stream Riffle

To estimate the minimum permissible velocity in athreshold stream Areas of local deposition

To estimate sediment yield for an alluvial stream Crossing or middle bar

To quantify general physical habitat substrate condition Bars, riffles, and pools

Bed Material CharacteristicsDeposited sediment is sampled to provide information on the individual

sediment particles, the sediment mixture, and the bulk sediment deposit. Particlecharacteristics include grain size, shape, specific gravity, lithology, andmineralogy. The quantity and type of contaminates attached to particles arefrequently of interest. Data that describes the distribution of the various particlesizes and of specific contaminates are frequently required. Characteristics of thesediment deposit itself include: stratigraphy, density, and compaction. For someof these purposes a sample can be disturbed, others require undisturbed sampling.Different samplers and sampling procedures are appropriate for differentenvironments. Water depth and velocity and bed material size are the mostimportant factors used to identify appropriate samplers and sampling procedures.

When the sediment particles are noncohesive, mechanical forces dominatethe behavior of the sediment in water. The three most important properties thatgovern the hydrodynamics of noncohesive sediments are particle size, shape, andspecific gravity. A discussion of these properties is found in �SedimentationInvestigations in Rivers and Reservoirs,� EM 1110-2-4000.

The boundary between cohesive and noncohesive sediments is not clearlydefined. It can be stated, however, that cohesion increases with decreasingparticle size for the same type of material. Clays are much more cohesive thansilts. Electrochemical forces dominate cohesive sediment behavior. The threemost common minerals that have electrochemical forces causing individualparticles to stick together are illite, kaolinate, and montmorillonite. The dispersedparticle fall velocity, flocculated fall velocity of the suspension, clay and nonclaymineralogy, organic content, and cation exchange capacity characterize cohesivesediment. The fluid is characterized by the concentration of important cations,anions, salt, pH, and temperature. More detailed information is presented in�Tidal Hydraulics,� EM 1110-2-1607.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (144)

Appendix D Guidelines for Sampling Bed Material D3

Sampling ProceduresSeveral factors influence both sampling site selection and sampling

procedure. The most significant factor is the data necessary to meet theobjectives of the study at hand. The objective of a bed material samplingprogram may be to determine a representative bed gradation for a particular reachof a stream, or it may be to determine the variability and diversity of the sedimentbed. Data needs should be clearly defined before the sampling program isplanned. The second factor to consider is field conditions. Will the bed of thestream be wet or dry? Is the site accessible by road, boat, trail, or only byhelicopter? Field conditions will determine both the practicality and type ofsampling equipment to be used in the sampling program. Another factor thatinfluences the type of sampling equipment and the appropriate samplingprocedure is the character of the streambed itself. Sand bed streams typicallyhave a more uniform bed gradation and therefore require a smaller volumesample than gravel bed streams. Typically, equipment appropriate for samplingsand bed streams is inappropriate for gravel bed streams. Thus, it is necessary toknow the general streambed characteristics before the sampling program isestablished. Finally, available resources must be considered as a limiting factorwhen establishing a bed-sampling program. Equipment, manpower, and fundsare frequently limited and therefore priorities must be established.

It is helpful if the bed material sampling location is near a stream gagingstation in order to relate sediment data to measured hydrologic and hydraulicdata.

Site Selection for Representative SamplingThere is no simple rule for locating representative sampling sites or reaches.

The general rule is as follows: Carefully select sampling locations and avoidanomalies that would bias either the calculated sediment discharge or thecalculated bed stability. The location must be representative of the hydraulic andsedimentation processes that occur in that reach of the river. The site should bemorphologically stable (constant slope and width upstream and downstream). Toensure data reflects reach-averaged river conditions there should be no tributaryinflow in the proximity of the site as it may interfere with the hom*ogeneity of thesection by supplying sediment for deposition. The site should not be locatedadjacent to a zone of active bank erosion as the material deposited in the channelnear the area may not be representative of the reach. Although bridges providegood access, bridge crossings are typically not appropriate sampling sites becausethey are located at natural river constrictions or their abutments and piers createconstrictions and local scour. Dead water areas behind sand bars or otherobstructions should be avoided, as these are not representative of average flowconditions.

Sand Bed StreamsSand bed streams are characterized by a relatively hom*ogeneous bed material

gradation. Vertical and temporal variability is typically insignificant in stable

Hydraulic Design of Stream Restoration Projects - [PDF Document] (145)

D4 Appendix D Guidelines for Sampling Bed Material

streams. Longitudinal variability typically occurs over distances of manykilometers. However, lateral variability, especially in bends, can be significant.In sand bed rivers, sampling of material is most frequently carried out in the lowflow channel. The equipment and methodology depends on the river depth andvelocity. The task can be accomplished in flowing streams either by wading orfrom a boat, or in ephemeral and intermittent streams in the dry. Verticalvariations in the bed material are usually insignificant in flowing water andsamples are collected from the surface. However, in standing water or on drybeds, a layer of fine material deposited on the recession of a flood hydrograph issometimes found on the surface. It is standard practice to remove this finesurface layer before collecting a sample from this kind of area.

Einstein (1950) recommended using only the coarsest 90 percent of thesampled bed gradation for computations of bed material load. He reasoned thatthe finest 10 percent of sediment on the bed was either material trapped in theinterstices of the deposit or a lag deposit from the recession of the hydrograph andshould not be included in bed material load computations.

Representative bed material sampling in sand bed streams may beaccomplished by one of two methods. Employing the cross-section approachrequires selecting a site and time for sampling where and when the bedcharacteristics are typical. This method requires considerable experience, andunanimity of opinion about where and when the typical condition occurs cannotbe expected even among experienced river scientists. Frequently, judgement isinfluenced by the type of streams the sampler has experienced and by theintended use of the data. Employing the reach approach where samples fromseveral systematically selected cross sections are averaged to obtain arepresentative sample may eliminate some uncertainty associated with the cross-section approach.

Cross-section approach. This approach requires the selection of arepresentative cross section. In streams with relatively uniform depths, betweenfive and three samples should be taken across the section to account for lateralvariations. In streams with variable depths more samples are required. Twentyverticals are commonly taken in braided streams. Taking bed material samples atcrossings where flow distribution is typically more uniform reduces the lateralvariation in the samples. However, at low flow, crossings may develop a surfacelayer gradation that reflects sediment transport conditions at the lower discharge,which may be coarser or finer then the bed gradation at bankfull discharge. Also,crossings are typically submerged and more elaborate sampling equipment isrequired than at exposed bars where a shovel is frequently sufficient. However,samples collected on a point or alternate bar may exhibit considerable variation.Figure D-1 illustrates typical bed material gradation patterns on a point bar. Notethat although the typical grain sizes found on the bar surface form a pattern fromcoarse to fine, there is no one location which always captures the precisedistribution that will represent the entire range of sedimentation processes.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (146)

Appendix D Guidelines for Sampling Bed Material D5

Figure D1. Gradation pattern on a point bar

Reach approach. An alternative to the cross-section approach is the reachapproach.1 A reach is defined as a portion of the stream with similar morphology(identified by its hom*ogeneity). Generally, five cross sections are laid out in thehom*ogeneous reach. If there is a gage in the reach, locating the center crosssection near the gage is preferred. If the stream reach is straight, the spacingshould be approximately two to five stream widths, and if the reach ismeandering, the spacing should occur within one meander length as shown inFigure D-2. The same criteria used in the cross-section approach to determine thenumber of verticals are applied here. The reach approach is especially applicableto rivers with meanders of different wavelengths and amplitudes.

1 Zrymiak, P. (1997). �Field procedures for sediment date collection: Vol. 2, bedmaterial,� (draft) Environment Canada, Ottawa, Ontario.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (147)

D6 Appendix D Guidelines for Sampling Bed Material

Figure D2. Bed sampling locations for sand bed streams (Zrymiak 1997, op. cit.)

Gravel Bed StreamsCoarse beds (gravel, cobble, and boulder) are characterized by significant

vertical, spatial, and temporal bed material variability. The most distinctivecharacteristic is a coarse surface layer that may form in both the low flow channeland on bars. Frequently the low flow channels of coarse bed streams are armoredwith large cobbles and boulders while bars consist primarily of sand and gravel.

Since the spatial variability in most coarse bed streams is high, it is verydifficult to perform representative sampling. River bars are frequently chosen assampling sites and specific bar types have been determined to be more

Hydraulic Design of Stream Restoration Projects - [PDF Document] (148)

Appendix D Guidelines for Sampling Bed Material D7

representative than others. A bar type hierarchy established to aid site selection(Bray 1972; Yuzyk 1986) is shown in Figure D-3. Midchannel and diagonal barsare selected as most ideal sites because they are exposed to the highest velocities,which transport the largest materials. Point bars are not as ideal becausevelocities are highly variable, decreasing toward the inside bank. Channel side orlateral bars are least desirable because they exist in zones of low velocities due toboundary and bank effects. In small streams with no bars and a pool-rifflesequence the riffles may be sampled to characterize bed-material size. However,the riffle can be expected to be much coarser at low flow when sediment transportis typically negligible than at bankfull flow when sediment transport is active.

Based on the assumption that the coarsest materials in the bed exert thepredominant effect on channel behavior and flow resistance, some recommendthat samples be collected at the upstream end of bars (Bray 1972; Church andKellerhals 1978; Yuzyk 1986). Sediments at this location are indicative of thesediments in the main channel, readily identifiable and generally exposed. Theupstream end of bars usually consists of the coarsest material in the channel andnot the average size in the reach because the upstream end of the bars is thelocation most frequently exposed to the highest stream velocities.

In coarse bed streams it is necessary to determine the characteristics of boththe surface and subsurface bed layers. Bulk sampling is employed to characterizethe subsurface layer. Both bulk and areal sampling are employed to characterizethe surface layer. Bulk surface sampling is preferred if it is possible to identifyand collect only the surface layer material. This is difficult when the surface layerhas a wide range of size classes. Bulk surface sampling provides informationabout the finer grain sizes trapped in the interstices of the surface layer, which isuseful for permeability studies for fish habitat and for sediment transport studies.Areal surface sampling is used to characterize the coarse surface layer and is usedto determine hydraulic roughness, critical shear stresses, armoring, and sedimenttransport.

A common methodology for areal sampling is a pebble count (Wolman 1954)where individual particles are collected at random by hand and the intermediateaxis is measured. This method requires that the stream be wadeable, althoughdivers may be employed. At least 100 particles should be included in the sample.However, to be very precise or to accurately measure small percentiles, thenumber of sampled particles should be increased. One method for choosing theparticles is a random walk laterally across the stream or longitudinally along apoint bar; another is to set up a longitudinal or square grid and measure particlesat the intersection of grid points. The gradation curve developed from these datais based on the number of particles in each size class not their weights orprojected surface areas. Studies have shown that particles smaller than 8 mm aretypically missed with areal sampling, especially if the bed surface is submerged,and thus the pebble count may be biased toward the larger sizes. This problemcan be overcome by truncating pebble-count samples at 8 mm and using a bulksurface sample to define the percentage and distribution of material finer than 8mm.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (149)

D8 Appendix D Guidelines for Sampling Bed Material

Figure D3. Coarse bed stream sampling hierarchy (Bray 1972)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (150)

Appendix D Guidelines for Sampling Bed Material D9

In addition to determining bed-material gradations, it is often important todetermine the characteristics of the stream bank. The bank material can helpdefine the stability of the channel section and may be responsible for a significantpercentage of the total sediment load. It is advisable to assess the soil type andgradation of each of the bank strata. Therefore several bulk samples should betaken at a given location. When sampling the bank, it is also advisable to assesssuch features as layering and lensing and to look for evidence of piping andseepage and related features.

Step-by-Step Field Sampling ProceduresMaintain detailed records of all data collected

Step 1. Select and mark out the required cross sections and the samplinglocations. Use as many of the site selection criteria previouslyoutlined as possible. The fixed permanent initial point should beon the left bank (looking downstream). Establish the control(horizontal and vertical) and reference all points.

Step 2 Sketch the site on data forms and reference the control points. Ifthe streambed contains a mixture of sand and gravel depositsthen map areas and record deposits of different size material.Develop a sampling strategy that will sample each zone.

Step 3 Collect a photographic record of the reach, controls, crosssections, sample locations (if possible), bed material (use a scalefor reference) and bank conditions.

Step 4 Select appropriate sampler for the task (based on depth, velocity,and sample requirements). Verify that the sampler is operational.

Step 5 Surface bulk sample - sand bed. Move to a sampling location. Inshallow streams use a tape to measure from the permanentlyfixed initial point (IP), and wade to a sampling vertical on thesection. Approach the sampling verticals from the downstreamside to prevent disturbing the bed at the sampling section. Indeep streams using a boat and some type of positioning system(tag line in narrow streams, electronic distance measurement(EDM) in wide streams), hold the boat steady over the samplinglocation. Obtain a sample of about 250 g at each chosen locationusing the selected sampler.

Surface areal sample - coarse bed. To obtain a surface arealsample in a coarse bed stream, several techniques are employed.These include random walks, setting up square or linear grids,and removing all the surface particles within a specified area.The spacing of the sampling points must be at least two times thediameter of the largest particle in the sampling area. This

Hydraulic Design of Stream Restoration Projects - [PDF Document] (151)

D10 Appendix D Guidelines for Sampling Bed Material

reduces the influence of nearby particles. Use 100 sample points(Wolman 1954; Hey and Thorne 1983; and Yuzyk 1986) for easeof data reduction. The random walk method devised by Wolman(1954) can easily be employed on a dry bed or in wadeable flow,and with more difficultly by divers. To obtain a sample, a teammember paces along a selected path, collecting a pebble witheach step. With closed or averted eyes the first pebble touched isselected. This method generally produces a sample biasedtoward coarse size classes. Other forms of grid sampling includelaying out a linear tape and selecting the pebble at a designatedinterval and laying out a preconstructed rectangular grid andselecting the pebble at grid point intersections. Collecting theentire surface layer within a specified area generally requires aspecialized sampler. The process may be aided by spray paintingthe surface if the bed is dry, although this technique is rathertedious. Regardless of the approach chosen, the measuringprocess may be streamlined in the field by using a gravelometerto measure the sieve diameter of each particle immediately afterthe particle is selected.

Surface bulk sample - coarse bed. To obtain a surface bulksample, carefully remove and collect all sediment in the surfacelayer to a thickness of the intermediate axis of the largest particlein the area. Care should be taken to insure that fine sediment isnot washed out of the sample. The required sample mass is afunction of the largest particle on the surface and can bedetermined from Figure D-4.

Subsurface bulk sample - coarse bed. If the surface layer has notalready been removed then scrape away the surface layer ofcoarse material to the thickness of the intermediate axis of thelargest particle in the area. The required sample mass is afunction of the largest particle in the subsurface and can bedetermined from Figure D-4.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (152)

Appendix D Guidelines for Sampling Bed Material D11

Figure D4. Bulk sampling standards for gravel and cobble streams (Yuzyk 1986;Church, McLean, and Wolcott 1987)

Step 6 Field sieving. This step is an alternative to transporting largebulk samples to a laboratory. Set up a weighing station. Thismay consist of a tripod with a scale suspended for weighing pailsof material. Assemble field sieve sets and insert correct sieves.Collect pails, spades, template, labels, field note forms, sturdyplastic bags, and tarpaulins. Spread out two tarpaulins. Obtaintare weights for the pails. Shovel subsurface material into pails,weigh and record. Pour material into top of the field sieves (8-,16-, 32-, 64-, 128-mm sieves). Rock and shake the sieve set untilmaterial has moved to its retained size sieve. Weigh materialretained on each sieve and on the pan. Record in field notes.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (153)

D12 Appendix D Guidelines for Sampling Bed Material

Save the material passing the finest sieve size for laboratoryanalysis. Save the 10 largest particles. Repeat the process untilthe required mass has been sieved. Measure the threeperpendicular axes of the 10 largest particles. Retain up to 10 kgof the combined material from the pan and discard the rest of thesample.

Step 7 Transfer the sample to a clean heavy-gage plastic bag.

Step 8 Complete and attach a label and sediment field note form foreach sample. Specify the stream, station, cross section, verticallocation, date, time, bed form and flow conditions, personnel oncrew, type of sampler, sample number, and sample depth.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (154)

Appendix E Computation of Average Annual Sediment Yield Using Weighted Events E1

Appendix EComputation of AverageAnnual Sediment Yield UsingWeighted Events

Average annual sediment yield can be approximated using calculated floodhydrographs and an adjustment factor. This calculation requires the followingdata:

a. discharge hydrographs for the project reach for a range of flood events

b. sediment-rating curve for the project reach

c. flow-duration data for a gage on the same stream or on a similar stream

d. discharge hydrographs at the gage location

e. sediment rating curve at the gage location.

The theory behind this methodology is that the annual sediment load can becomputed by adding up the contribution from hydrograph events, each weightedby their annual probability of occurrence. However, this summation gives onlythe contribution of the annual peak events, and not the smaller events that occurduring the year. A correction factor must be computed to account for theseevents. It is computed using both flow-duration data and event hydrograph dataat a gaged site. The gage should be as close to the project reach as possible:ideally on the same stream, but certainly in the same hydrophysiographic region.

Step 1. Compute event sediment yield at the project reach for a range offlood events, using the flood discharge hydrographs. In this example, the 2-, 5-,10-, 25-, 50-, and 100-year flood events are used. The SAM program can beused to perform this computation.

Step 2. Plot the sediment yield for each event against its annual frequency ofoccurrence to give a single event sediment yield vs. annual recurrence intervalcurve.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (155)

E2 Appendix E Computation of Average Annual Sediment Yield Using Weighted Events

Step 3. Determine the area under the curve to obtain the average annualsediment yield for the annual series of storm events (this is not the entire averageannual sediment yield). This is mathematically given as:


= ∫s iY Y dP

where Ys is the sediment yield calculated from the sum of the frequency events, Yiis the sediment yield associated with a given frequency of occurrence and P is theprobability. This equation may be approximated numerically as:

100 50 25 10 5 20 015 0 015 0 04 0 08 0 20 0 40= + + + + +sY . Y . Y . Y . Y . Y . Y (E1)

Note that there are other possible numerical approximations for the sedimentyield. The preceding equation is based on trapezoidal segments. The curveshould be examined to make sure the expressions for the first and last segmentsare reasonable. At the high end of the curve (above Y100), an additionalcomputation for Y200 may improve the estimate. At the lower end of the curve(below Y2), the recurrence interval where the yield goes to zero should beestimated to give the proper coefficient for that segment of the area. Acomputation for an event below the 2-year might improve the estimate.

Step 4. Compute the correction factor, J. The calculated area under thecurve represents the contribution of the series of annual storm events to theaverage annual sediment yield, but it does not include the contribution of lesserstorm events. The area under the curve must be multiplied by a correction factorJ to account for the difference between the annual peak series and the partialduration series and to account for other errors that may be associated with thenumerical integration and with the difference between synthesized eventhydrographs and a natural series of hydrographs. The correction factor is theratio of the sediment yield computed using measured flow data to the sedimentyield computed using the numerical approximation. The correction factor, J, iscomputed at a gage site, which should be as similar as possible to the project sitein sediment and hydrologic characteristics.

So the basic equation is:

100 50 25 10 5 2(0.015 0.015 0.04 0.08 0.20 0.40 )= + + + + +aY J Y Y Y Y Y Y (E2)

where Ya is the average annual sediment yield.

Step 4a. Compute sediment yield at the gage site using weighted eventhydrographs. This is a repeat of Steps 1, 2, and 3, except performed at the gagesite.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (156)

Appendix E Computation of Average Annual Sediment Yield Using Weighted Events E3

Step 4b. Compute sediment yield at the gage site using mean daily flowsand a site-specific sediment-rating curve. The SAM program can be used. If thestream is flashy, 15-min or 30-min data should be used if available.

Step 4c. Calculate the correction factor J, which is the ratio of the result ofStep 4b to the result of Step 4a:

sediment yield at the gage site computed using measured flowssediment yield at the gage site computed using weighted event hydrographs


Once the correction factor is computed, it may be reasonable to use itthroughout a watershed. This could be particularly useful on studies where ahydrologic model of the watershed has been set up (so that dischargehydrographs are available). Average annual sediment yields could be computedfairly easily for a number of sites without gage data.

Step 4d. Calculate Ya.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (157)

Appendix F Example Scope of Work for Stability Assessment F1

Appendix FExample Scope of Work forStability Assessment

Scope of Work

Preliminary Stream Restoration Assessment of Upper StudebakerRiver Watershed and Project Reach



Upper Studebaker River Section 206 Aquatic Ecosystem Restoration.


Upper Studebaker River, South Lake, CA.

Project Description

The U.S. Army Corps of Engineers (the Corps), and the City of South Lakeare undertaking an aquatic ecosystem restoration project on the UpperStudebaker River (USR) as authorized by Section 206 of the Water ResourcesDevelopment Act of 1996 (WRDA 96). Anthropogenic activities within the USRwatershed such as logging, grazing, and commercial and residential developmenthave impaired the natural functioning of its ecosystem. Additionally, theseactivities have also contributed to the ecosystem degradation of the USR�sterminus, South Lake. The Upper Studebaker River Section 206 AquaticEcosystem Restoration (USR 206) Project seeks to remedy anthropogenicimpacts to the USR watershed and South Lake aquatic ecosystems byimplementing measures to restore the conditions of, and relationships between,its channel, riparian, and wetland habitats.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (158)

F2 Appendix F Example Scope of Work for Stability Assessment

Study area, project reach, and subreaches

The study area for the USR 206 geomorphic assessment is the entire USRwatershed. The watershed of the Upper Studebaker River covers more than 50square miles. The upper end of the drainage basin begins in the mountains atabout elevation 3,048 m (10,000 ft). From its headwaters, the USR flowswesterly about 1.60934 km (1 mile), and then northerly about 8.04672 km(5 miles) through a steep, narrow canyon. Upon leaving the canyon the streamflows through a gently sloping valley that ends in South Lake. The upper basin ischaracterized by steeply rising, heavily timbered mountain slopes that terminatein large granitic outcrops at their crests. The upper basin is relatively pristine andhas experienced minimal anthropogenic impacts. The lower basin has beenimpacted by logging, agriculture and urban development. Figure F1 displays theUSR 206 study area location.

Figure F1. Study area and project reach

The project reach for the USR 206 project is located in the lower basin. It isdefined as the USR channel as well as the current and historic floodplain betweenSouth Lake and the Highway 10 bridge crossing upstream and includes thecommunity of Smallwood. The lower portion of the project reach has beenchannelized.

The USR 206 project reach has been subdivided into hydraulically relevantsubreaches. Table F1 defines the hydraulic subreaches.

Table F1USR 206 Project Hydraulic SubreachesHydraulicSubreach


EndStation Boundary Descriptions (D/S � U/S)

1 0+00 6+00 South Lake Blvd. to Rt 5 bridge crossing2 6+00 25+00 Rt 5 bridge crossing to upstream town limits of Smallwood. Channelized reach.3 25+00 90+00 Upstream town limits of Smallwood to abandoned dam. Agricultural overbank.

4 90+00 160+00 Abandoned dam to downstream main channel/cutoff channel confluence. Agriculturaloverbank.

5 160+00 200+00L Downstream main channel/cutoff channel confluence to upstream main channel/cutoffchannel confluence. Left descending channel. Gravel mining in reach.

6 160+00 200+00R Downstream main channel/cutoff channel confluence to upstream main channel/cutoffchannel confluence. Right descending channel. Gravel mining in reach.

7 200+00 250+00 Upstream main channel/cutoff channel confluence to Rt 10 Bridge. Logging in overbankarea.

Rt 10 Rt 5

Project ReachSmallwood

South Lake


Pristine areaminimal impacts

Hydraulic Design of Stream Restoration Projects - [PDF Document] (159)

Appendix F Example Scope of Work for Stability Assessment F3

Description of Services Required


The geomorphic processes of sediment generation and fluvial transport arefundamental and determining factors in the condition of the USR watershed�saquatic habitats. Therefore, a geomorphic assessment (GA) is to be performed insupport of the USR 206 project in order to characterize how geomorphicconditions within the USR watershed influence its ecosystem. Habitats ofprimary interest to the restoration effort are aquatic channel, riparian, andwetland habitats. Parameters of primary importance to the quality of aquatichabitat are diversity and stability.1 Therefore, particular emphasis shall be placedon characterizing how geomorphic conditions within the USR watershed impactthe ecology and relative stability of morphometric channel and floodplainconditions within the project reach.

The geomorphic processes of sediment generation and fluvial transportwithin the USR watershed are also relevant to the issue of South Lake�sdiminishing clarity, its most visible and noteworthy symptom of ecosystemdegradation. Therefore, another goal of the GA is to characterize howgeomorphic conditions within the USR watershed influence South Lake�sdiminishing clarity.

The GA shall consist of a reconnaissance-level analysis of the entire USRwatershed and a more detailed level analysis of the project reach. Watershedanalysis shall focus on characterizing sediment sources and contributing land-usepractices, particularly where sediment generation rates appear to be inordinatelyhigh. Project reach analysis shall focus on assessing the channel�s ecology,geomorphic stability, and behavior.

Review of available material

Familiarization of the study area shall be achieved through discussions withpertinent individuals and agencies and review of prior reports of the study area,including Federal Emergency Management Agency (FEMA) studies, hydrologicmodels, and aquatic surveys.

Field investigations

Generally, field investigations shall be performed as necessary tocharacterize geomorphic conditions in the USR watershed and support ageomorphic stability assessment of the project reach. It is expected that the fieldefforts will be conducted with an engineer who is familiar with hydraulics,sedimentation, and geomorphology as well as a biologist familiar with theecological area. Specific field investigations required are described in thefollowing sections. 1 Stability in this context is relative to channel conditions of dynamic equilibrium, i.e., thebalancing of sediment inflow and sediment transport capacity by relatively modest adjustments inchannel dimensions.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (160)

F4 Appendix F Example Scope of Work for Stability Assessment

Watershed assessment

Available material shall be researched, compiled, and reviewed. Fieldinvestigations shall be conducted as necessary to perform the following tasks. Itis expected that field examinations and verifications will be necessary forcompletion of these tasks.

a. Define the relevant geologic characteristics of the USR watershed andthe USR 206 project reach. Acquire and review available topographicdata including U.S. Geological Survey (USGS) topo maps, surveyedtopography, aerial photos, etc. Identify general subsurface, soil types,cover conditions, and relevant properties within the project reach.

b. Prepare a summary and a time line of the history of land-use activitiesand associated geomorphic conditions in the USR watershed. Include allidentifiable events of geomorphic significance.

c. Identify USR sediment sources and contributing land-use practices. Notewhere sediment generation rates appear to be inordinately high. Providea general characterization of sediment sources based on their relativecontribution to the project reach�s bed-load, suspended-load and wash-load supply.

d. Identify dominant geomorphic processes within the USR watershed.Assess whether each process is natural or anthropogenic.

e. Identify any apparent geologic and/or anthropogenic structural controlson the geomorphic conditions of the watershed.

f. Identify aquatic species that have been impacted in the study reach.

g. Assess point and nonpoint source water quality impacts on the watershedwith particular emphasis on the limitations that it may place on targetaquatic species.

h. Contact the state fish and wildlife agencies with regard to their stockingprogram and locate stocking sites on the maps (frequency, species, etc).Determine if the state has historic records of declining fisheries on anyparticular streams and/or reaches and locate these on a map. Assessrecords and studies to determine if specific blockages to fish passagehave adversely impacted fisheries.

Project reach assessment

Available material shall be researched, compiled, and reviewed. Fieldinvestigations shall be conducted as necessary to perform the following tasks. Itis expected that an experienced field crew will need to walk the entire reach.

a. Acquire and review available topographic data including USGS topomaps, surveyed topography, aerial photos, etc. Identify current andhistoric channel types, locations, planform characteristics (to include

Hydraulic Design of Stream Restoration Projects - [PDF Document] (161)

Appendix F Example Scope of Work for Stability Assessment F5

sinuosity, wavelength, and meander belt width). Prepare a figuredisplaying historic channel locations.

b. Compare current and historic topographic (planform and vertical) data.Identify indications of historic channel behavior and current channelcondition. When feasible, identify the effects of relevant anthropogenicactivities on channel morphology. Estimate historic lateral migrationrates. Prepare a figure displaying locations.

c. Identify current geomorphic subreaches. Prepare a table displaying thebeginning and ending stations of each subreach. Table F1 may be usedas a reference. Subsequent tasks shall reference geomorphic subreachesidentified, where appropriate.

d. Assess point and nonpoint source water quality impacts to the reach.Collect water quality samples from high flow and low flow conditionsfrom each of the subreaches. Prepare a figure displaying locations.

e. Characterize the longitudinal location of the project reach relative to thewatershed in terms of its sediment transport/geneneration behavior (i.e.,zone of erosion, transportation, or deposition). Prepare a figuredisplaying locations.

f. From field measurements, identify average channel morphometry foreach subreach to include bankfull channel dimensions (defined asminimum width to depth ratio), overbank characteristics, and base flowchannel dimensions.

g. Identify active and remnant floodplain surfaces (terraces). Prepare afigure displaying locations.

h. Characterize the bed material and bed forms. For each subreach, samplea representative reach for bed-material load calculations and onerepresentative area for low flow habitat conditions. Prepare a figure ortable identifying sample locations and characteristic hydraulic conditionsrelative to project stationing. Perform standard laboratory sizedistribution (sieve) analysis. Prepare a standard plot(s) of bed samplegradation curves and a table(s) of bed sample grain size data. Provide ageneral characterization of the sources of project reach bed material.

i. Provide a general characterization the ecology of each subreach. Theinventory of aquatic habitat will utilize existing data on benthos andfinfish sampling as well as a rapid bioassessment of physical instreamhabitat to indicate current habitat conditions.

j. Characterize the bank material and stratigraphy for each of thegeomorphic subreaches. For each subreach, collect representative bankmaterial samples. Perform standard laboratory-size distribution (sieve)or hydrometer analysis, as appropriate. Identify bank material soil typesand properties. Provide a general characterization of the relativecohesiveness and erodability of the bank materials. Identify bank

Hydraulic Design of Stream Restoration Projects - [PDF Document] (162)

F6 Appendix F Example Scope of Work for Stability Assessment

erosion and failure mechanisms. Characterize the existing bankvegetation.

k. Identify significant sediment sources and sinks within the study reach.Assess sediment impacts from sources in the upper basin above the studyreach.

l. Observe tributary, distributary, and relict channels in the project reach,and identify indications of channel behavior and geomorphic conditions.

m. Observe anthropogenic features including bridge abutments and piers,grade control structures, low flow crossings, and bank protection.Identify impacts of features and indications of channel behavior andgeomorphic conditions. Identify significant geomorphic controls in theproject reach.

n. Acquire and review USGS gaging station records, including surveyedcross sections and rating tables. Perform specific gage analysis andidentify indications of channel behavior and geomorphic conditions.

o. Characterize the current geomorphic stability of the subreaches in projectreach channel, whether incising or aggrading. Identify the severity andextent of any existing vertical or horizontal instabilities via a qualitativeindex.

p. Characterize the grade conditions of the channel, whether incised,aggraded, or at-grade. Apply appropriate channel evolution models toidentify current channel stage, subsequent stages of evolution, and theevolved stable channel form. Qualitatively estimate the time scale ofchannel recovery. Characterize the impacts of anthropogenic features onchannel morphology and stability.

q. Identify potential problem areas in the project reach. Characterize thepotential for significant increase and/or decrease in the project reachsediment supply. Characterize the sensitivity of channel form to suchvariations, including expected channel form response, and the magnitudeand time scale of expected adjustments. Characterize the impacts ofanthropogenic features on expected channel behavior.

r. Characterize the relative uncertainty of the assessment performed.Identify any additional analyses required to develop a reasonably certaingeomorphic assessment.

Restoration recommendations

Based on the geomorphic assessment performed, recommend restorationmeasures and appropriate analysis and design methodologies. Specific items tobe addressed are described in the following sections.

a. Recommend measures to restore geomorphic stability and ecologicalhealth to the project reach. Include an assessment of the appropriateness

Hydraulic Design of Stream Restoration Projects - [PDF Document] (163)

Appendix F Example Scope of Work for Stability Assessment F7

of bank protection, grade protection, instream habitat enhancement,channel realignment, wetland creation, and fish blockage removal.

b. Recommend measures to reduce impacts to water quality from theproject reach area. Include an assessment of the appropriateness of bankprotection, grade protection, channel realignment, and riparianmodifications.

c. Document each proposed measure. The information will include anestimate of the potential project size, the general project type, a sketch ofthe site, the impact to and proximity of utilities, and an assessment ofconstruction area and access. Provide a general estimate of costs.

d. Address specifically the appropriateness and feasibility of possibleprojects. Particularly address impacts to existing floodplains. Addresssocial and biologic controls that may limit possible projects.

Meeting attendance

There will be four required meetings to review study progress. These shall beat the initiation of work, at the completion of field investigation activities, at themidpoint of the study process and finally at the end of the study to presentfindings.

Quality Control and Quality AssuranceQuality control (QC) of the technical products produced under this scope of

work shall consist of development and execution of a Quality Control Plan(QCP), independent technical review (ITR), and Quality Control Certification(QCC). The experience and background of personnel selected for the fieldassessments will be reviewed to ensure that the study is conducted by qualifiedpersonnel. Products shall be reviewed for compliance with standard engineeringand professional practices, adequacy of the scope of the associated document,appropriateness of data used, consistency, accuracy, comprehensiveness, andreasonableness of results.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (164)

Appendix G Example Sediment Impact Assessment and Stable Channel Design G1

Appendix GExample Sediment ImpactAssessment and StableChannel Design

A sediment impact assessment was conducted as part of the reconnaissancelevel planning study for a flood damage reduction project for the City ofCarlsbad, NM (Copeland 1995).1 The purpose of the sediment impact assessmentwas to identify the magnitude of possible sediment problems that might beassociated with the proposed project. One potential source of flooding was DarkCanyon Draw, a tributary of the Pecos River (Figure G1). One of the flooddamage reduction alternatives being considered was a bypass channel that woulddivert Dark Canyon Draw around the City of Carlsbad. The proposed diversionwould begin near the city airport and flow northeasterly to the Pecos River to alocation about 8.04 km (5 miles) downstream from the city.

Depending on the diversion channel design, several sedimentation andchannel stability problems could occur. If a threshold channel is constructed (achannel designed with little or no sediment transport potential), then bed materialdelivered from upstream would deposit at the diversion entrance. Sedimentdeposits would have to be removed periodically. If a channel is designed to carrythe incoming sediment load, there will be a period of adjustment for the channel,as the bed and banks become established. Bed armoring may progress quickly orslowly, with extensive degradation, depending on the consistency of the materialthrough which the diversion channel is cut and the sequence of annual runoff thatoccurs. And finally, if the diversion channel is too efficient in terms of sedimenttransport capacity, it could degrade and induce additional channel degradationupstream from the diversion location. The sediment impact assessment wasconducted to determine the magnitude of possible sediment degradation oraggradation problems and to obtain relatively stable dimensions for the diversionchannel.

1 All references cited in this appendix are listed in the References section following themain text.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (165)

G2 Appendix G Example Sediment Impact Assessment and Stable Channel Design

Figure G1. Carlsbad and surrounding areas (To convert miles to kilometers,multiply by 1.609347)

Field ReconnaissanceA preliminary assessment of channel stability and potential sediment impacts

were determined during a two-day field reconnaissance. This briefreconnaissance provided insight for general observations related to channelstability.

Dark Canyon Draw transitions from a wide shallow alluvial channel,characteristic of Southwestern United States alluvial fans, at its canyon mouth toan incised arroyo at its confluence with the Pecos River. Gravel mining iscurrently active in the lower reaches of Dark Canyon Draw between the PecosRiver and the city airport and appeared to have been occurring for many years.Due to the gravel mining, the channel had been both widened and deepened. Thechannel also showed signs of incision/degradation upstream from the airport.The bed and banks of the incised channel were capable of supplying significantquantities of sediment to the stream. The bed surface of Dark Canyon Drawconsisted primarily of coarse gravel and cobbles. Banks were generallycomposed of loose alluvial material ranging in size from clays and silts toboulders. The channel tended to migrate laterally, eroding banks and creatingremnant gravel bars in former channels. Armoring was generally observed in the

Hydraulic Design of Stream Restoration Projects - [PDF Document] (166)

Appendix G Example Sediment Impact Assessment and Stable Channel Design G3

existing low flow channel. However, at high flows the channel would migrate,mobilizing significant sediment from the gravel bars and from bank erosion.

Bed material samples were collected during the field reconnaissance.Sample size class distributions were determined using the Wolman (1954) pebblecount method. Due to the limited scope of the sediment impact assessment,samples were collected at only two sites. Both surface and subsurface sampleswere collected at the mouth of the canyon, which is several kilometers (miles)upstream from the proposed diversion channel. There was no coarse surfacelayer at the second site, which was located on a gravel bar, about 1.5 km (1 mile)downstream from the canyon mouth. The thoroughly mixed bed form was anindication that active-layer mixing had occurred during the last flow event at thissite. Median grain size ranged between 22 and 55 mm for all the samples. Thegradation determined at the downstream site was selected as the representativegradation for the sediment impact assessment because it was characteristic of afully mobile bed. Bed material gradations determined from these samples areshown in Figure G2.

Figure G2. Bed material gradations, Dark Canyon Draw

HydrologyHydrographs used in the sediment impact assessment were developed using

the HEC-1 hydrograph package (USACE, HEC 1981). These were used tocalculate sediment yield for flood events. The peak discharge for the 1-percent

Hydraulic Design of Stream Restoration Projects - [PDF Document] (167)

G4 Appendix G Example Sediment Impact Assessment and Stable Channel Design

exceedance flood was 2,000 m3/s (75,000 ft3/s). The 10-percent chanceexceedance hydrograph was assumed to have the same shape as the 1-percentchance exceedance flood; discharges on the hydrograph were calculated bymultiplying the 1-percent chance exceedance hydrograph by the ratio of thepeaks. The 10-percent chance exceedance peak discharge was 570 m3/s (20,000ft3/s). The 1-percent chance exceedance hydrograph for Dark Canyon Draw isshown in Figure G3.

Figure G3. One-percent chance exceedance hydrograph, Dark Canyon Draw (Toconvert cubic feet per second to cubic meters per second, multiply by0.02831. To convert acre-feet to cubic meters, multiply by 1,233.489)

A flow-duration curve was developed from 18 years of USGS mean dailyflow data from the Dark Canyon at Carlsbad gage. Durations of published peakflows greater than the maximum mean daily flow were added to the flow-duration data by assuming the historical flood hydrographs had shapes similar tothe 1-percent change exceedance hydrograph. The flow-duration curve is shownin Figure G4.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (168)

Appendix G Example Sediment Impact Assessment and Stable Channel Design G5

Figure G4. Dark Canyon Draw at Carlsbad, NM (To convert cubic feet persecond to cubic meters per second, multiply by 0.02831)

Average Hydraulic ParametersA typical reach in the existing Dark Canyon Draw channel was selected from

a HEC-2 (USACE, HEC 1990) backwater model. The typical reach chosen forthis analysis was about 3.21 km (2 miles) long and was located adjacent to theCarlsbad Airport. The reach was considered to be in a state of nonequilibriumdue to its proximity to gravel mining operations. A reach further upstream, lessinfluenced by gravel mining operations, would have been preferred fordetermining long-term sediment yield. However, the existing backwater modeldid not extend any further upstream. It was recommended that additional cross-section surveys upstream be obtained for more detailed sediment studies.

Water-surface elevations and hydraulic variables were calculated using theHEC-2 model for a range of discharges. Average values for hydraulic variableswere then determined using the reach-length weighted averaging procedure inSAM (Thomas et al. 2000).

Sediment Transport Rating CurveThe bed material sediment yield of Dark Canyon Draw is important when

considering sediment transport and channel stability questions. The bed materialsediment load consists of the sediment sizes that exchange with the streambed asthey are transported downstream. The bed-material yield is most likely to be

Hydraulic Design of Stream Restoration Projects - [PDF Document] (169)

G6 Appendix G Example Sediment Impact Assessment and Stable Channel Design

relatively small compared to the total sediment yield because the bed of DarkCanyon Draw consists primarily of gravels and cobbles. The wash loadcomponent of the total sediment yield will be transported through the system tothe Pecos River, unless it is trapped by a reservoir or introduced into a pondedarea.

Sediment transport was calculated using several sediment-transport equationsavailable in the SAM program. The equations chosen were equations thatincluded at least some data from gravel bed rivers in their development. As canbe seen from the sediment-discharge rating curves, shown in Figure G5, there is awide range in predicted sediment transport rates. There are no available data onDark Canyon Draw to aid in the selection of a transport equation. However, theguidance program in SAM identified the North Saskatchewan and Elbow Riversin Saskatchewan, Canada, as having similar median bed grain sizes, depths,velocities, and slopes as Dark Canyon Draw at high flow. The guidance programdetermined from the available set of equations in SAM that the Schoklitschequation (Shulits 1935) best reproduced measured data on the NorthSaskatchewan and Elbow Rivers. Based on the comparison of calculatedsediment transport rating curves using different sediment transport functionsshown in Figure G5, the Schoklitsch equation will produce a relatively lowsediment yield. In order to cover the uncertainty range in the calculated bedmaterial sediment yield, two additional sediment transport equations were chosento calculate yield. The Parker equation (Parker 1990) was used to represent ahigh sediment transport load, and the Einstein (1950) equation was chosen torepresent an intermediate sediment transport load.

Figure G5. Bed material sediment transport rating curves, Dark Canyon Draw (Toconvert cubic feet per second to cubic meters per second, multiply by0.02831. To convert tons to newtons, multiply by 8,896.443)

Hydraulic Design of Stream Restoration Projects - [PDF Document] (170)

Appendix G Example Sediment Impact Assessment and Stable Channel Design G7

Diversion Channel DesignThe stable-channel analytical design method in SAM was used to size the

low-flow channel. This method provides channel dimensions that will transportthe incoming bed-material sediment load for a specified discharge. The methoduses the Brownlie (1981) equation to calculate sediment transport and roughnesson the channel bed. This equation was not developed for gravel bed streams, andpredicts lower sediment transport rates at lower discharges than other testedequations (Figure G5). This apparent deficiency in the sediment-transportequation is accounted for later by testing the resultant cross-section geometryusing other transport equations.

The criteria chosen for the diversion channel design were: (a) a compositechannel geometry with a low-flow channel designed to carry the effectivedischarge, and (b) the overbank designed using threshold criteria for the one-percent chance exceedance flood. Assigned side slopes were 1V:3H, with a sideslope Manning's roughness coefficient of 0.05.

The effective discharge is the discharge that transports the largest percentageof the bed material sediment load. This was determined by integrating the flow-duration curve for Dark Canyon Draw and a sediment-transport rating curvedeveloped using the Einstein formula. A plot of percentage of bed-material loadversus discharge increment is shown in Figure G6; an effective discharge of 240m3/s (8,500 ft3/s) was indicated.

Figure G6. Effective discharge, Dark Canyon Draw (To convert cubic feet persecond to cubic meters per second, multiply by 0.02831)

The inflowing sediment concentration was determined for the effectivedischarge from the sediment-transport rating curve developed for the typicalreach of Dark Canyon Draw. The Brownlie sediment-transport equation wasused for typical reach to be compatible with the calculations in the design reach.The bed material gradation in the diversion channel was assumed to be the same

Hydraulic Design of Stream Restoration Projects - [PDF Document] (171)

G8 Appendix G Example Sediment Impact Assessment and Stable Channel Design

as in Dark Canyon Draw. This is a reasonable assumption for the long-termcondition in the diversion channel, but not for initial conditions. The transitionfrom initial to final conditions could be determined in future, more detailedstudies using a numerical model such as HEC-6 (USACE, HEC 1993).

Using the natural slope between the proposed Dark Canyon Draw diversionand the Pecos River, a unique solution for width and depth was determined forthe effective discharge channel. The average slope between Dark Canyon Drawat the airport and the Pecos River is 0.0047. The ground slope is steeper at theairport and becomes very mild as it crosses the Pecos River floodplain. A moredetailed analysis should include different channel geometries due to variation inslope.

The stable channel curve for 240 m3/s (8,500 ft3/s) is shown in Figure G7.This curve defines combinations of width and slope that would provide formovement of the inflowing sediment load through the diversion channel. Theaverage slope for the diversion channel, if no drop structures were employed,would be 0.0047. With this slope, the stable channel method suggests that a basewidth of about 122 m (400 ft) would be stable. The calculated depth was 1.1 m(3.5 ft).

Figure G7. Preliminary diversion channel design (To convert feet to meters,multiply by 0.3048)

The width of the overbank portion of the channel was determined by trial anderror using the threshold velocity criteria from EM 1110-2-1418. With a medianbed material size of about 30 mm and a water depth of 1.5 m (5 ft), a thresholdvelocity up to 1.8 m/s (6 ft/s) would be appropriate for channel stabilityconsiderations. Roughness on the overbank was calculated using the Brownlieroughness predictor. The total width of the overbank and channel wasdetermined to be 850 m (2,800 ft). The details of the final geometry is shown inFigure G8. If the threshold velocity is exceeded, degradation can be expected.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (172)

Appendix G Example Sediment Impact Assessment and Stable Channel Design G9

The extent of degradation can be estimated in a more detailed study using theHEC-6 numerical sedimentation model.

Figure G8. Cross section Dark Canyon Draw Diversion Channel (To convertcubic feet per second to cubic meters per second, multiply by0.02831. To convert feet to meters, multiply by 0.3048)

Sediment BudgetThe magnitude of potential aggradation or deposition problems in the Dark

Canyon channel can be determined by calculating bed material sediment yieldthrough a typical reach of the existing channel and comparing it to calculatedsediment yield in the project reach.

Bed material sediment yield was calculated for the existing channel using theflow-duration sediment transport curve method and SAM. Sediment yields werecalculated for the 1-percent and 10-percent chance exceedance floods usingsynthetic hydrographs, and for average annual conditions, using the flow-duration curve. Bed-material sediment yields calculated using three differentsediment transport equations are listed in the following tabulation.

Calculated Bed-Material Sediment Yield1

Dark Canyon Draw1-percent exceedanceflood

10-percent exceedanceflood Average Annual

m3 yds3 m3 yd3 m3 yd3

Schoklitsch 2,400 3,100 530 690 180 230Einstein 11,300 14,800 3,300 4,300 1,300 1,700Parker 27,700 36,200 4,100 5,400 1,100 1,5001 Sediment yield volume calculated assuming specific weight of deposit of 1,500 kg/m3 (93 lbs/ft3)

Sediment yield was determined in the diversion channel using the sameprocedure that was used to calculate sediment yield in the typical reach of theexisting channel. Trap efficiency was then determined for flood hydrographs andfor average annual conditions.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (173)

G10 Appendix G Example Sediment Impact Assessment and Stable Channel Design

The potential for aggradation/degradation in the diversion channel for a 10-and 1-percent chance exceedance floods and for average annual conditions wasdetermined using the sediment budget approach. Bed material sediment yieldwas calculated using three sediment transport equations and compared to thecalculated bed material sediment yield in the existing Dark Canyon Draw. Bedmaterial sediment transport was assumed to occur only in the low flow channel inthe diversion. Calculated bed material sediment yield and its percentage of thetotal bed material yield calculated for Dark Canyon Draw is shown in thefollowing tabulation. This tabulation indicates that deposition will occur in thediversion channel for all cases tested. For the 1-percent chance exceedanceflood, between 34 and 38 percent of the inflowing bed material sediment loadwill deposit in the diversion channel. For the 10-percent chance exceedanceflood, between 12 and 17 percent of the inflowing bed material load will deposit.For average annual conditions, between 6 and 18 percent of the inflowingsediment load will deposit. A range of the quantities of deposition can bedetermined from these calculations. Recall that the Schoklitsch equationproduced sediment transport quantities closest to the measured data from a riverwith similar characteristics.

Calculated Bed-Material Sediment Yield 1Diversion Channel

1-Percent Exceedance Flood 10-Percent Exceedance Flood Average AnnualSediment TransportFunction m3 yd3

Percentof Inflow m3 yd3

Percentof Inflow m3 yd3

Percentof Inflow

Schoklitsch 1,600 2,050 66 450 590 86 150 190 82

Einstein 7,500 9,800 66 2,900 3,800 88 1,200 1,600 94

Parker 17,100 22,400 62 3,400 4,500 83 1,000 1,300 87

1 Sediment yield volume calculated assuming specific weight of deposit of 1,500 kg/m3 (93 lbs/ft3)

At the next level planning, it would be necessary to evaluate the temporaldevelopment of the diversion channel using the HEC-6 numerical sedimentationmodel. In this sediment impact assessment, the bed material gradation wasassumed to be already developed. A more detailed study would requireknowledge of the existing soil profile through which the channel will be cut. Thearmoring process would have to be simulated with a numerical model. Inaddition, the slope of the diversion channel will vary between the diversion pointand the Pecos River. This requires a more detailed analysis of spacial variabilityin the sedimentation processes.

Hydraulic Design of Stream Restoration Projects - [PDF Document] (174)


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 maintainingthe 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 forreducing 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 notdisplay a currently valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.1. REPORT DATE (DD-MM-YYYY)

September 20012. REPORT TYPE

Final report3. DATES COVERED (From - To)




Hydraulic Design of Stream Restoration Projects





Ronald R. Copeland, Dinah N. McComas, Colin R. Thorne, Philip J. Soar,Meg M. Jonas, Jon B. Fripp



U.S. Army Engineer Research and Development CenterCoastal and Hydraulics Laboratory, 3909 Halls Ferry Rd., Vicksburg, MS 39180-6199;University of Nottingham, University Park, Nottingham, NG7 2RD, England;U.S. Army Engineer District, Baltimore, 10 South Howard Street, Baltimore, MD 21201





U.S. Army Corps of EngineersWashington, DC 20314-1000


Approved for public release; distribution is unlimited



This document provides a systematic hydraulic design methodology to aid hydraulic engineers in the design of stream restorationprojects. The objective is to achieve a channel design that fits into the natural system within the physical constraints imposed by otherproject objectives. In the Corps of Engineers, stream restoration projects are frequently associated with or part of a flood-control project.Thus, projects have more than one objective and compromises may be required to meet essential portions of each objective. Thehydraulic design of a stream restoration project should provide for a channel that is in dynamic equilibrium with its sediment load. Asound stream restoration design incorporates techniques from both fluvial geomorphology and physics. The study area to which thesetechniques are applied must extend beyond the limits of the project site to the extent that both the project�s effect on the stream systemand the stream system�s effect on the project reach can be determined. The iterative systematic approach presented includes definingproject objectives and constraints; determining appropriate hydrologic data; conducting a stability assessment of the existing streamsystem channel to establish baseline geomorphological conditions and to evaluate the effectiveness and geomorphological impacts ofproject alternatives; and a methodology for hydraulic design of project features and for assessing hydraulic and sediment transportimpacts of alternatives. Appendices provide useful tools and examples for use in this methodology.

15. SUBJECT TERMSAlluvial channelsBankfull discharge

Bed material samplingChannel-forming dischargeChannel stability

Effective discharge Hydraulic designFlood damage relief techniques Hydraulic geometry assessmentGeomorphic assessment SAM hydraulic design package (cont.)









UNCLASSIFIED 17319b. TELEPHONE NUMBER (includearea code)

Standard Form 298 (Rev. 8-98)Prescribed by ANSI Std. 239.18

Hydraulic Design of Stream Restoration Projects - [PDF Document] (175)


Sediment impact assessmentSediment transportStormwater managementStream restoration projects

Hydraulic Design of Stream Restoration Projects - [PDF Document] (2024)
Top Articles
Latest Posts
Article information

Author: Foster Heidenreich CPA

Last Updated:

Views: 6457

Rating: 4.6 / 5 (56 voted)

Reviews: 87% of readers found this page helpful

Author information

Name: Foster Heidenreich CPA

Birthday: 1995-01-14

Address: 55021 Usha Garden, North Larisa, DE 19209

Phone: +6812240846623

Job: Corporate Healthcare Strategist

Hobby: Singing, Listening to music, Rafting, LARPing, Gardening, Quilting, Rappelling

Introduction: My name is Foster Heidenreich CPA, I am a delightful, quaint, glorious, quaint, faithful, enchanting, fine person who loves writing and wants to share my knowledge and understanding with you.