Ben Plumb¹, Cal Jefferies^1,2, Matthew Iannetta¹, Jeff Hirvonen¹ and William K. Annable²

¹GeoProcess Research Associates, Ontario, Canada

²University of Waterloo, Waterloo, Canada

Estimating the hydraulic conditions that trigger erosion (incipient motion) is a common assessment/design consideration of many contemporary fluvial geomorphology, river engineering and aquatic habitat-related projects. Many methodologies and approaches exist, each with their respective applicability and limitations for a given stream/valley type. The inherent uncertainty of these methods continues to present challenges when characterizing and designing natural systems, especially when applied to the varied stream morphologies in Southern Ontario.

This study evaluates several common analytical and empirical erosion threshold methodologies against temporal hydrologic, morphologic and sediment transport data from a wide range of stream types in Southern Ontario. The field data include geomorphic monitoring data, hydrometric gauge records, bedload transport measurements (active and passive) and in-situ erodibility measurements of cohesive sediment spanning an overall period of record exceeding a decade from a range of semi-alluvial, bedrock, gravel-bed and cohesive streams.

Findings from each approach are discussed in the context of applicability bounds related to both stream type and project objectives. Field data collection is considered and evaluated to balance levels of field effort with specific erosion threshold-related problems. The study provides additional field-verification and benchmarking of common analytical and empirical tools with a specific focus on their application in Southern Ontario steams. Data limitations, inherent uncertainty, and applied research opportunities leveraging existing data are highlighted. Implications of various methods of erosion threshold characterization are discussed through the lens of formative river processes.

Patrick Padovan and Paul Villard

GEO Morphix Ltd., Campbellville, Ontario, Canada

Quantifying stage discharge relations and predicting instream velocities is important for natural channel design and provides insight on instream processes, estimating bankfull flows, defining erosion thresholds, calibrating hydrology models, and validating ecological targets. When building stage discharge rating curves, we often rely on the collection of discrete measurements of velocity and discharge in panel form over a range of flow conditions or water levels. Rating curves for smaller watercourses are generally based on discrete measurements collected during different storm events over multiple years. This approach assumes that there is no change in the form of the rating curve over the period of sampling. As a result, there is limited consideration of seasonal changes that occur within the watercourse and the associated seasonal variation in the stage discharge relation.

Given the discrete nature of traditional sampling protocols and the reasonable strength of most stage discharge relations that are built from them, we often miss the nuances associated with vegetation impacts on instream hydraulics in smaller streams. This presentation will examine a continuous data set of depth and velocity measurements collected in a small southern Ontario stream through three seasons (spring, summer, fall). Acoustic and pressure sensors were used to collect continuous data, equating to approximately 25,000 discrete data points. This is a significantly larger dataset than what is generally collected for the creation of discharge rating curves. The presentation will review and summarize the observed impacts of vegetation on roughness and channel velocity.

Scott Cowan

Stantec Consulting Ltd., Waterloo, Canada

A common objective in stream restoration design is to return a channel to its historical alignment, cross-sectional geometry, and riparian condition. However, restoration to a pre-disturbance condition is not always feasible, particularly in our urban environments where land-use changes have irreversibly changed the fundamental building blocks of our watercourses.

As a result of these irreversible changes, the management of our urban watercourses should not only consider maintaining or restoring a channel to a particular historical condition but should also explore rehabilitation alternatives that are in balance with a watershed’s likely future conditions of hydrology and sediment transport.

This presentation will discuss industry standard approaches to developing stream restoration design discharges. Additionally, an alternative approach to developing stream restoration design discharges will be explored that considers long-term trends in urban hydrological and sediment transport regimes. Lastly, this presentation will outline simple, tangible steps that project owners can take to improve the resilience of our urban stream restoration designs. It is expected that this presentation will stimulate discussion regarding the background data required to improve long-term stream restoration design outcomes.

Leif Burge and Megan Hendershot

Stantec Consulting, Kelowna, Canada

Stantec completed a Master Plan for Ellis Creek, that flows westward approximately 5 km from a reservoir, through industrial and urban areas, to the Okanagan River channel in Penticton, BC. Assessments for the master plan included: (1) a geomorphic field assessment, (2) a hydraulic assessment, (3) a bank erosion hazard assessment, (4) an aggradation or degradation assessment, and (5) sediment mobility and transport assessment. A one-dimensional hydraulic (HEC-RAS) model was developed to inform the shear stresses and inundation.  The sediment mobility was assessed using the mobility ratio (bed shear stress / critical shear stress) and the bags sediment transport model. Three historical dam breaches in 1921, 1941 and 1942 substantially impacted the channel geomorphology.  More recently, floods in 2017 and 2018 caused significant erosion of the channel bed and banks of Ellis Creek.  The channel was divided in to 13 reaches for analysis and grouped based on the assessments into natural, incised, deeply incised, transitional, aggraded, and channelized. Sediment supply within the study area is limited to the bed and the banks of the channel downstream of the reservoir.  Sediment is largely produced in an upstream section that is deeply incised, with banks up to 4 m tall. Sediment produced in the incised section was deposited in the downstream aggraded section. Aggradation decreased the capacity of the channel. Widespread flooding is predicted during the design flood flow event downstream. Conceptual designs were developed based on geomorphically stable bed sediment patterns, including riffle pools, step pools and plane beds with boulder clusters.  The restoration of the creek is estimated to take 20 years to complete.

Chris Cummings, CAN-CISEC

Matrix Solutions Inc., Mississauga, Canada

Natural channel designs in Southern Ontario have historically relied upon the application of a riffle-pool bed morphology to achieve various design objectives. This traditional approach utilizes riffle features as grade controls to provide long-term bed stability with the secondary benefit of introducing habitat diversity. Although there are characteristics that are common across most designs, designed riffle features will vary in slope and stone distribution and are commonly sized based on an estimated bankfull flow or greater discharges to ensure long-term stability under a range of high flow conditions. More frequent flow events are often considered during the design process from an ecological standpoint but may be neglected from a geomorphological perspective.

Through numerous years of design, implementation and monitoring cycles, a pattern of riffle failure has emerged, primarily in higher-gradient (typically ~1% or greater) watercourses. In many systems, major storm events have the most dramatic impact on riffle function and stability, however it has been observed that less significant flow events appear to also influence riffle stability. Channel instability can be exacerbated by riffle designs that fail to consider backwater effects and the protection of riffle tail areas that is provided from backwatering. Another commonly observed riffle ‘failure’ includes excessive piping and throughflow due to bi-modal substrate sizing.

Examples of designs that did not fully consider these factors are discussed from both the design process and construction perspectives. Lessons learned over the past 20 years in implementing natural channel designs with a riffle-pool morphology are presented including motivations for considering more frequent flows when evaluating channel function and stability.

Ben Miller¹, Fred Dobbs², Laura Wensink², and Paul Villard¹

¹GEO Morphix Ltd., Campbellville, Canada

²Nottawasaga Valley Conservation Authority, Utopia, Canada

Agricultural channels with alluvium dominated by sands and gravels can see substantial adjustments due to vegetation/land use changes, and increased volume and magnitude of flows from tile drainage. Significant increases in sediment supply also occur due to agricultural practices that diminish riparian vegetation cover. These changes can result in downstream floodplain sediment trapping, aggradation on the floodplain floor, downcutting of the bankfull channel and increasing entrenchment, more rapid bank erosion, and rapid development of higher channel sinuosity. Changes in riparian vegetation and rooting depth/density create positive feedback. In many cases, there may also be a disconnect between the floodplain and the bankfull channel due to floodplain aggradation and channel downcutting. In the Nottawasaga River watershed, restoration activities have incorporated terracing to provide greater floodplain connectivity and soft bioengineering of banks to increase resistance to bank erosion. This strategy has been effective for providing greater floodplain access, reducing channel shear stresses, and increasing bank strength. Decreasing channel sinuosity by bypassing meanders and creating oxbow channels provides greater floodplain access, sediment trapping, and brings the channel to a more natural, pre-agricultural form. The strategy of reducing channel length to restore habitat can be counter intuitive, but reducing channel length can be a critical step in restoring gradient in flat land streams to maintain riffles and provide important habitat for aquatic invertebrates and spawning fish. The logic behind this approach is discussed in the context of channel evolution models and sediment transport mechanisms. Several designed and constructed restoration projects are used to illustrate this approach and its potential advantages. Examining new potential techniques, particularly those that are counterintuitive, such as reducing channel length and sinuosity, provides an opportunity to re-evaluate traditional approaches to restoration.

Mark Hartley, B.Sc., M.Sc., P.Eng. and Fred Dobbs, B.Sc.

Nottawasaga Valley Conservation Authority, Utopia, Ontario

Working with watercourses that have cross-sections that are sensitive to erosion and scour within reaches that have a dynamic sediment load is challenging.  The current body of knowledge for open channel design that mimics natural fluvial processes recognizes that relatively straight alignments with uniform trapezoidal cross-sections are not sustainable and should be replaced with meandering planforms with variable cross-sections and a non-uniform profile.  However, trapesoidal cross-sections in both meandering and straight alignments are common in rural Ontario.  Rivers and streams that are experiencing adjustment for numerous reasons (incision) often take on a trapezoidal shape.  The most common cross-section used in the design of open drains under the Drainage Act is trapezoidal.  These situations often result in the main channel being disconnected from its floodplain and, accordingly, are considered unsustainable.  The floodplain, a natural attribute of rivers, is a large-scale depositional feature formed from a combination of within-channel and overbank depositional processes along with many intermediate sedimentary forms.  Recently, one of these intermediate sedimentary forms was identified in rural watercourses in Indiana and Ohio and has become known as a floodplain bench which in turn forms the main feature of the ‘two-stage channel design’.

The two-stage channel, which includes a low-flow channel as well as a floodplain area, improves bank stability, lowers the erosivity of flows, and increases sediment transport and storage capacity. This design lowers maintenance costs (dredging, bank stabilization) and is expected to lessen downstream yields of sediment.

A number of projects using this best management practice have been constructed in recent years by the NVCA Stewardship Program including sites on Lamont and Beeton Creeks.  The projects were completed in response to landowner concerns over bank stability and have been shown to not only address these concerns but to improve overall health of the reach.  Finally, the two-stage method was used in the recently completed design of the South Innisfil Creek Drain; a municipal drain in the Town of Innisfil.

Erin Kelly¹, Chase Konecny² and Hamish Trenam³

¹Stantec Consulting Ltd, Whitehorse, Canada

²Stantec Consulting Ltd, Waterloo, Canada

³Stantec Consulting Ltd, Markham, Canada

Toe protection structures are placed along the outside of meander bends to reduce the stream bank erosion potential and are often an important component of the long-term stability of an urban natural channel design (NCD) project. A more typical design utilizes wood-debris toe protection to protect stream banks where the increased roughness disrupts helical flow patterns and reduces near-bank shear stress. The construction of wood-debris toe protection can be time intensive and requires a ready source of appropriate wood (native, non-invasive species free of disease or infestation) that must be transported to site.

As an alternative to wood, can straw bale toe protection be used as an adaptation of wood-debris toe protection? Will strawbales provide the same toe protection, long term stability and vegetation growth when compared to a wood toe or a base condition (no protection)? To answer this question, both types of bank protection and a control group (no bank protection) was installed as part of a NCD realignment of an agricultural watercourse in Ontario. During construction, installation time for all types of structures was noted for comparison. Four years later, with monitoring data in hand, the effectiveness of the three toes based on stability, vegetation establishment and sediment accumulation within the pools can be determined.

This presentation will describe the risks and opportunities for each type of toe protection and will provide guidance on which systems are appropriate for their application.

Robin McKillop

Palmer, Toronto, Canada

One common cause of failure of a natural channel design (NCD) is the use of morphologies, substrates and/or bioengineering techniques that are incompatible with the local geomorphological setting. Three settings with unique implications for NCD are gullies, ravines and broad valleys. Case studies are presented to illustrate differences in the evolution of each of these three settings, considering energy expenditure and the conveyance of water, sediment and woody debris, highlighting the importance of setting in the development of NCD parameters.

Gullies are steep, V-shaped incisions into slopes or the edge of tableland. They evolve rapidly through interactions between fluvial and mass movement processes. In gullies, stability is achieved through structural aggregation of coarse substrates and/or woody debris, rather than by resistance of individual particles to flow-induced shear stress. An erosion mitigation design for an actively head-cutting gully in the East Oshawa Creek valley, in Oshawa, is presented to illustrate how flow energy can be accommodated and dissipated by mimicking the morphologies of steep mountain channels.

Ravines are transitional between gullies and broad valleys: they have steep sides and are locally affected by mass movements, but they also have semi-alluvial channels and isolated floodplains. Logjams are common but generally short-lived. Most floodwater is contained within the channel, with little opportunity for floodplain attenuation. Any stone placements must be sized to accommodate flows much higher than the “bankfull” or 2-year flows typically used for NCDs in unconfined settings. Partial failure of a newly reconstructed section of channel along the Mud Creek ravine, in Toronto, highlights the importance of grade control, “over-sized” stone placements, and separation from adjacent slope toes.

Broad valleys commonly contain a “classic” meandering channel that spills overbank onto a low, level floodplain every year or two, on average. Most morphological relationships that inform NCDs are based on such settings. A variety of sinuosities, riffle-pool sequences and cross-sectional geometries may be appropriate, and systems are resilient to periodic logjam formation. A recently implemented NCD along a tributary of Humber River, in Brampton, illustrates why there is greater flexibility in design parameters for realignment of a channel within a broad valley than for erosion control works in more confined settings.

Practitioners are encouraged to tailor their approaches to NCD to the particular setting in which their project is located for better performance in the long-term.

Chase Konecny, BASC, EIT and Andrew Doherty, P.Eng,

Stantec Consulting Ltd., Waterloo, Canada

Natural channel design balances driving and resisting forces to achieve project objectives and to create stable, self-sustaining systems. It is a continuously evolving practice that applies art and science, with new innovations and approaches being introduced, tested, and modified regularly. These innovations and approaches are part of a process called ‘adaptive management,’ which relies on post-construction monitoring to demonstrate performance and identify where changes or repairs are required.

However, there is no one-size-fits-all approach to post-construction monitoring, as projects vary greatly in their objectives, scale, and setting. It is necessary to consider a wide range of factors and potential approaches in developing comprehensive and successful monitoring programs.

The presentation will explore different approaches and considerations in developing a monitoring program, including technologies, performance metrics, and habitat and biological indicators. We will also discuss challenges encountered with different monitoring schemes and highlight feedback and perspectives from clients, regulators, and other project stakeholders.

This presentation will provide a lens on regional perspectives on monitoring by drawing on natural channel design case studies from across North America. We will also encourage audience discussion to gain a multi-disciplinary perspective. The overall intent of the presentation will be to spark a constructive dialogue on how practitioners can work together to develop monitoring programs that efficiently and effectively serve project objectives.