Understanding River Restoration fluvial processes within river restoration design
Dr Jenny Mant Dr Philip J. Soar Overview
• accounting for sediment in river restoration • channel-forming discharge • what to think about for restoration design • class exercise
• FORM, PROCESSES AND DYNAMICS (change over time)
Context
Fluvial geomorphology is still an emerging science Interest in ‘applied’ fluvial geomorphology grew rapidly during the 1980s Key NRA/EA R&D reports in during the 1990s setting out a geomorphological approach to river management Consideration of river processes, sediment transfer and physical habitats recognised as vital for ‘sustainable’ river management and restoration In the UK there are few consultant geomorphologists (but this is changing!) Training courses are not widely available and there are no standard approaches to river restoration Accounting for Sediment in river restoration Starting point
Guidebook of Applied Fluvial Geomorphology Technical Report FD1914 (Thorne, Sear and Newson 2003)
Starting point
Applied Fluvial Geomorphology for River Engineering and Management (Thorne, Hey and Newson 1997)
Starting point
Fluvial forms and process A new perspective
( Knighton, D 1998) Sustainable restoration design
Why should we ‘account for sediment’? What do we mean by ‘continuity’ (or connectivity)?
Sediment Continuity… as a design ‘principle’ as a design ‘process’ in the design ‘procedure’
Define Fluvial Geomorphology...
Fluvial geomorphology is:
Fluvial is defined as: found in, or produced by a river or rivers Morphology is defined as: the scientific study of form and structure Geo relates to the surface of the earth
Fluvial geomorphology is the study of sediment sources, fluxes and storages within the river channel over short, medium and longer timescales and of the resultant floodplain morphology (Sear and Newson, 1993)
Why ‘account’ for sediment in restoration
“Consider the amount and type of sediment supplied to a stream channel. Why? Sediment is part of the balance (i.e. between energy and material load) that determines channel stability.”
“Lack of sediment relative to stream power, shear stress or amount of energy in the flow (discharge = m3/s) usually results in erosion of sediment from the channel boundary of an alluvial channel.” (indexes of transport capacity)
“Conversely, an oversupply of sediment relative to the transport capacity usually results in deposition of sediment in that reach”
Federal Interagency Stream Restoration Working Group 1998. Stream Corridor Restoration: Principles, Processes and Practices
Why ‘account’ for sediment in design
For sustainable channel restoration design...
“a reconstruction that modifies the size of a cross section and sinuosity should be analysed to ensure that upstream sediment loads can be transported through the reconstructed reach with minimal deposition or erosion.”
Federal Interagency Stream Restoration Working Group (FISRWG), 1998. Stream Corridor Restoration: Principles, Processes and Practices
River restoration objectives
What is the ‘primary’ objective of river restoration? Support a diverse biodiversity! Improve fisheries Improve conservation value of the river landscape Restore meanders!
Further objectives might include flood protection and recreation
Geomorphology and sediment transport are often given a lower priority or not considered at all!
‘Form without function’
Without accounting for sediment transport and river processes there is a risk of designing ‘form without function’
By imposing an unsustainable condition, the designed channel might not be able to support the targeted habitats over the long term.
The river can reject the imposed changes very quickly, especially if there is a high sediment load and sufficient energy, resulting in complex responses. River management NOT working with nature
‘Detroit riprap’
Geomorphology offers a better solution! Discharge
Velocity distribution in channel velocity profile bed roughness
Channel Bed material Bed shear geometry characteristics stress
availability selective transport of sediment competence & capacity erosion & Bedload deposition transport
Sediment supply after Ashworth and Ferguson (1986)
The ‘sediment system’
Schumm, 1977 Characterising the sediment system (relate to previous slide)
Hawkcombe Stream, Somerset Channel-slope coupling
a) Gorge Fully coupled: delivery of slope material to channel at all flows
b) Confined channel / floodplain Partially coupled: delivery of slope material to channel. Spatially discrete and sporadic activity of sources. Some limited storage.
c) Unconfined channel / floodplain Uncoupled /weakly coupled: delivery of slope material to channel at high flows or where tributaries join. Storage on floodplain. Sediment transfer
Transfer Reach
Flow Bank conditions Site constraints
Width Depth Slope Sediment transfer (Soar and Thorne, 2001)
Reference Transfer Downstream Reach Reach Reach (Supply) (Demand)
Stream type Local variability Bank conditions
Width Channel stability
Channel-forming discharge
Sediment load Depth Sediment and gradation Slope load Sedimentation
Braided channel?
Time
0 Channel Change Channel Unstable Channel Design Stable Channel Design Erosion Impact of channelisation on sediment transporting capacity
discharge sediment discharge
QwS+ ~ Qs+D50 bed slope median bed material size
A . Planform .B
.A .B Long profile Impact of channelisation on aggradation and degradation
Brookes, 1988 The channel is trying to ‘recover’ a new equilibrium through a combination of degradation and aggradation: negative feedback
Know your river!
Where in the fluvial system? ‘Type’ of river (meandering? mobile bed? hydrology? etc) Channel typing and characterisation: Reference reach(es) Natural morphologies to use as analogues? Supported sediment forms and features Supply reach Sediment transfer and connectivity Flow regime ( i.e. stream power) Bed and bank materials/vegetation/typology Project reach Bed and bank materials/vegetation Valley gradient and site constraints Where in the fluvial system? Stream Power
Stream Power: f(QS)
Q = discharge S = slope
FISRWG, 1998
Q = width x mean depth x mean v Independent and dependent variables controlling channel form Sediment transport classification
When discussing the sediment load it is vital to be clear about which component of the load is being dealt with. Predicting Sediment Transport Rate
In the UK measured sediment transport data are not available other than for research sites There are numerous equations to calculate the sediment (bed material) load in a stream. Accuracy can range between +/- 50% of actual transport rates unless calibrated against measured loads. The key to selecting an appropriate equation and improve accuracy is to find an equation developed for conditions that match those under which it is to be applied. Sediment Transport Rate
Bed material size Basis Sample Applications Comments Formula
Bagnold (1980) sand, gravel Stream Mimmshall Brook, Performed well in power R. Sence, R. Idle, tests against field Shelf Brook data using reach- average values. Both under and over predicts.
Bathurst, Graf gravel, cobble Discharge Shelf Brook, (Newson Performed well for and Cao and Bathurst 1991) steep, headwater (1987) streams (S > 0.1). Over-predicts and can produce negative loads. Ackers-White silt, sand Shear stress R. Sence, Usk, Performed well in tests (1973) updated gravel Colne, Stour, based on flumes and by HR Ecclesborne (HR rivers. Much better when Wallingford Wallingford 1992) calibrated against data (1990) from site in question. Over-predicts.
Newson (1986) silt, sand Catchment Shelf Brook, Sence, Provides estimate of gravel area Tawe, Idle Dunsop, annual sediment yield to Whitendale river. (Newson and Bathurst 1991) ‘Typing’ the bed and banks
Bed type: implications for slope, bed material composition, roughness and stable channel dimensions
Bathurst, 1997 Bank material and vegetation: implications for roughness and stable channel dimensions River profile
Riffle-pool sequence characteristic of both straight and meandering channels with heterogeneous bed material in the size range 2- 256mm (mostly gravel-cobble bed streams). Slope range generally 0.001 to 0.02. Important sediment storage and sorting for channel stability and high ecological value.
Step-pool sequence slopes are steep > 0.03 – 0.1, typically formed from accumulation of boulders and cobbles in confined valleys. Riffle-pool spacing measured in rivers
Diagram from Leopold and Wolman (1957). Riffle spacing in low sinuous channels has a spacing (on average) or 5-7 times the channel width. Pool-riffle characteristics
Regular spacing between successive units of 5-7 times the width (thus, the spacing is scale-related!) However, Keller and Melhorn (1978) showed that the spacing can range between 1.5 and 23.3 times the channel width with a mean of 5.9. Riffles generally absent from sand bed rivers and cobble- boulder bed streams, where they are replaced by step- pool units. Riffles have a coarser bed material than pools suggesting that a sorting mechanism is present. How is this maintained? Riffles
River Cole, nr Birmingham
River Gaunless, Co Durham WANDERING
POOL- RIFFLE PLANE BED
STEP-POOL
CASCADE
Transport Capacity 40 Know your river
Stream Reconnaissance (Thorne, 1998)
Fluvial Audit (Defra FD1914)
And … RIVERS MOVE!
Mississippi River, Mississippi
? and… RIVERS MOVE!
Gilwiskaw Brook, Leicestershire Stability and Equilibrium
‘Geomorphological’ stability:
A river that maintains the same average cross section dimensions and planform characteristics while adjusting its position on the floodplain is said to exhibit ‘dynamic equilibrium’. At this condition the width, depth and slope are adjusted so that there is neither net aggradation or degradation over time. Bank erosion and changes to bar forms are natural features of this dynamic equilibrium (and vital ecologically!).
Types of morphological ‘adjustment’
Downs, 1992 Types of morphological ‘adjustment’
Downs, 1992 Conclusions
Accounting for sediment continuity will provide more sustainable (and economic) solutions while meeting flood defence and biodiversity objectives
How? Problem is approached ‘holistically’ within the catchment context Designing ‘form with function’ Thus, ‘minimum maintenance’ requirement Geomorphological processes create dynamic and diverse habitats
Channel Forming Discharge representations and variability Overview
Concept and use in river restoration Representations of channel forming discharge Bankfull discharge Discharge of a fixed return period Effective discharge
Definition
‘At this discharge, equilibrium is most closely approached and the tendency to change is least. This condition may be regarded as the integrated effect of all varying conditions over a long period of time.’ (Inglis, 1947).
Important because…
Incomplete understanding of the ‘formative’ properties of all flows.
Flows (Q) at and near the bankfull stage have morphological significance. They are ‘effective’
In channel restoration design, a ‘design discharge’ is required that corresponds to the conveyance capacity of the restored channel
Purpose
In river restoration, the channel-forming discharge concept provides a relationship between the:
hydrologic characteristics of the catchment;
the hydraulic characteristics of the channel, and;
the geomorphic characteristics of the project reach.
Influence of Hydrology
HYDROLOGY (The natural distribution and sequence of flows) The ‘driving’ parameters
HYDRAULICS The mechanistic processes of open channel flow
SEDIMENT TRANSPORT Sediment erosion, transportation and deposition
MORPHOLOGY Channel configuration, forms and features
CHANNEL STABILITY Stable channel restoration designs Representations
In a natural river the most appropriate definition is the discharge conveyed at the elevation of the active floodplain (Wolman and Leopold, 1957).
However, there are three definable flows which have been taken to represent the 'dominant' flow based on the application of repeatable geomorphological and hydrological techniques
Bankfull discharge Discharge of a fixed return period Effective discharge
Bankfull discharge
The maximum flow that the channel can convey without overflowing onto the floodplain. Requires estimation of the bankfull stage and then determination of the discharge at that stage
A frequently applied method is to estimate the discharge corresponding to the channel width at the minimum width-to-depth ratio (Wolman, 1955) Bankfull discharge (Wolman 1955) Bankfull morphology Bankfull discharge
Bankfull Indicators: Elevation of the ‘active’ floodplain Highest elevation of active channel bars Lowest elevation of perennial vegetation Marked break in bank slope or materials
Requires stream reconnaissance techniques and knowledge of geomorphic stability criteria
BUT accurate location of bankfull indicators is not a routine procedure and is problematic and subjective Bankfull in an incised channel?
Long Creek, Mississippi Where is bankfull here? Bankfull discharge (main) methods
Direct gauging Use a stage-discharge rating curve from a nearby gauge Use an empirical relationship (eg. predict from catchment area) Use appropriate flow resistance (roughness) equations which can synthesize a stage-discharge curve
Bankfull discharge from catchment area
Emmett,1975 (in FISRWG, 1998)
Flow resistance equations There are many equations in the literature Manning equation: AR0.67S0.5 WD1.67S0.5 Q Q n n where: d1/ 6 n 50 21.1
after Strickler for static beds with no bedforms, or ‘n’ can be found from look-up charts or Barnes book (1967) for different conditions.
Q is bankfull discharge, A is bankfull section area, R is hydraulic radius (Area / Wetted Perimeter), S is slope, n is Manning roughness coefficient, W is width, D is depth, d50 is median particle size
See Bathurst (1997) in ‘Applied Fluvial Geomorphology’ book for other flow resistance equations suitable for mobile beds. Discharge of a fixed period
Rivers will remain dormant for long periods of time (low flow conditions) Most active in higher than normal flow. As the process of channel adjustment is purely mechanical the determining factors must persist for sufficient time for this adjustment to take place and stability to be reached. The ‘dominant’ discharge, or whatever discharge determines the river channel shape, must occur often enough to permit the channel to reach regime’.
(Nixon, 1959) Discharge of a fixed period
In general, bankfull discharge has a return period of between 1 and 2 years (Leopold et al., 1964)
However, there is a range of possible frequencies... 1.0 to 5.0 years (Wolman and Leopold, 1957) 4 to 10 years (Pickup and Warner, 1976) 50% of sites had return periods less than 1.25 years or greater than 1.75 years (Andrews, 1980) 0.56 to 3.44 years on Partial Duration Series (Hey and Heritage, 1988) 1 to 10 years (USACE, 1994) and others... Effective discharge
The dimensions of a dynamically stable river must be delicately adjusted to the sediment balance (Mackin, 1948)
Perennial rivers adjust their bankfull capacity to the flow that transports the greatest quantity of sediment load over a period of years (usually the period of flow record), or the flow which expends the greatest energy per unit time (Wolman and Miller, 1960).
Andrews (1980) termed this flow the Effective Discharge
Frequency of effective discharge
Effective Discharge corresponds to...
2 year flow event (Biedenharn et al., 1987) 2 year flow event (Watson et al., 1997) 1.46 year flow event (Orndorff and Whiting, 1999) and others…
All in the range of bankfull discharge return periods! Effective Discharge Calculation: A Practical Guide. ERDC/CHL TR-00-15 (August, 2000)
Requires a combination of empirical, statistical and mathematical methods
The method is designed to have general applicability, have the capability to be applied consistently and to integrate the effects of physical processes responsible for determining the channel dimensions Channel forming flow ?
Bankfull Discharge of Effective = = Discharge Fixed Return Discharge Period Channel Restoration Design Geomorphological approaches Candidates for river restoration?
River Idle, Nottinghamshire Candidates for river restoration?
West Branch, Patuxent River, Maryland Candidates for river restoration?
Spring Branch, Maryland What level of detail can we design to?
Frissell et al. (1986) Representative cross sections Restoration Approaches
Natural recovery: Do nothing
Prompted recovery: Rehabilitation of natural processes and forms
Managed recovery: Channel restoration design
Natural Recovery
Hawkcombe Stream, Somerset, UK ‘Prompted’ Recovery
Deflector 6A in the River Idle, Nottinghamshire, UK ‘Full restoration’ design methods
Historical reconstruction: the carbon copy solution
Reference reach geometry: the natural analogue solution
Regime analysis: the empirical design solution
Analytical channel design: the rational design solution
Channel Restoration Design i) Historical reconstruction: the carbon copy solution Historical reconstruction
Reinstatement of a previous channel configuration that possessed the type of channel configuration and range of forms and features required in the target restored channel Ideally, this ‘carbon copying’ approach involves replacing meanders exactly as found prior to disturbance One of the most widely practiced techniques in northern European countries. Two techniques are available: replicate meander planform from historical sources (e.g. air photographs, maps); excavate old river courses on the floodplain. Historical reconstruction
BUT Assumes rainfall-runoff/sediment discharge patterns in the catchment have not significantly varied over time (stationary drainage basin controls)
If land-use patterns and hydrology have changed - restored channel configuration likely to be unstable and could result in progressive aggradation or degradation. Channel Restoration Design ii) Reference reach geometry: the natural analogue solution Reference reach(es)
‘Pristine’ reaches are very rare to find in catchments targeted for river restoration but stable segments of river can usually be found. Undisturbed reaches close to the target restored channel may be used as channel restoration design blueprints If stable reaches can be identified in catchments with similar hydrological and physical characteristics/valley type, then channel geometry data could be scaled to the restored reach (eg. Adopting the same width-depth ratio)
Reference reach(es)
BUT Difficulty and subjectivity in locating ‘stable’ reaches and sites with similar boundary conditions as the target restored channel Difficulty identifying bankfull dimensions from field indicators.
Despite this, reference reaches should always be sought to examine the morphological characteristics that can be supported To find out more about reference sites…
Harrelson, C., Rawlins, C. L. and Potyondy, J. P. 1994 Reference Sites: An illustrated guide to field technique
Channel Restoration Design iii) Regime analysis: the empirical design solution Regime analysis (see handout)
‘A self-formed alluvial channel is ‘in regime’ if there are no net changes in discharge capacity or morphology over a period of years’
Cross-sectional form is inherited from the imposed natural sequence of flows and boundary sediments. Discharge usually explains most of the variance in geometry
Downstream Hydraulic Geometry equations (after Leopold and Maddock (1953): 0.5 0.4 W aQb D cQb
Qb is bankfull discharge, W is bankfull width and D is bankfull depth. ‘a’ and ‘c’ are constants for different ‘types’ of channel Meander wavelength 100000
Lm = (11.26 to 12.47) W within 95% confidence limits
10000
(m)
m L
1000
100
90% single response limit 10 Meander Wavelength,Meander 95% mean response limit Regression 1 1 10 100 1000 Bankfull Width, W (m) Regime equations
BUT Only limited equations available for different ‘types’ of channel (especially sand-bed rivers) Extreme caution must be exercised when applying regime equations to rivers with different characteristics to those used to derive the equations The equations are empirically derived and not based on flow/sediment mechanics (although some justification from analytical methods) Limited predictive capabilities (especially for depth and slope)
Channel Restoration Design iv) Threshold channel design: the engineered design solution Static Boundary Design
Some river beds are static and coarse sediment transport is negligible Considerations of shear stress and ‘initiation of motion’ can be used in the design, based on critical depths and maximum permissible velocities Can be used to size sediment for riffles in a restoration project. As a guide: d 11DS d is smallest particle size that can be moved, D is critical depth (assume bankfull) and S is the slope
Mechanics are beyond the scope of this workshop Channel Restoration Design v) Analytical channel design: the rational design solution Stable channel design procedure Supply Reach Phase 1 Assessment
Project Reach Phase 2 Reach-Average Assessment Channel Dimensions
Local Morphological Phase 3 Channel Design Variability
Sediment Impact Phase 4 Design Brief Assessment To find out more about the design approach and procedure…
Soar, P. J. and Thorne, C. R. 2001 Channel restoration design for meandering rivers
ERDC/CHL Report CR-01-1 United States Army Corps of Engineers, Engineer Research and Development Center, Vicksburg, Mississippi, September Meander features
Straight Channel Meandering Channel Thalweg
Pointbar
Riffle or Crossing
Pool Local morphological variability
Lm
q Rc
A’ A A m Z C’ B’
B
Apex C Local morphological variability
A A’
Wi W
D Inflexion Point m D
B B’
Wp
Maximum Pool Scour Dmax
C C’
Wa
Bend Apex Design equations (Soar and Thorne, 2001)
Width at bend apex, Wa / Width at inflexion point, Wi
Wa 105Te 030Tb 044Tc u Wi
Width at pool, Wp / Width at inflexion point, Wi
Wp 095Te 020Tb 014Tc u Wi
Apex to pool, Za-p / Apex to inflexion, Za-i Z a-p 036 u Za-i
Equiwidth meandering: Te=1 Tb=0 Tc=0
Meandering with point bars: Te=1 Tb=1 Tc=0
Meandering with point bars and chute channels: Te=1 Tb=1 Tc=1 Design conclusions
The procedure is not a ‘cook book’ for all types of river - has been developed for meandering rivers only Supply Reach Assessment and understanding the type of channel that can be supported by the range of boundary conditions are critical! Approaches that ‘account’ for sediment transfer in the design and with a sediment impact assessment as a closure loop ensure that restored river morphology and assemblage of physical habitats are sustainable However, full river restoration in the UK is still a practice in its infancy and further testing of approaches is required Case Study Whitemarsh Run. Maryland Case study: Whitemarsh Run
Before Restoration Whitemarsh Run
1 km downstream Whitemarsh Run: Restored design Response: straightening, sedimentation