A Modeling Approach to Restoring Pool-Riffle Structure in an Incised, Straightened Channel of an Urban Stream

Urbanization impacts watershed hydrology resulting in hydromodification of in-stream hydraulics and rapid adjustment of channel morphology thereby disturbing natural geomorphic and ecological processes in stream systems. Pool-riffle structure, capable of supporting diverse biological ecosystems, is frequently degraded in urban streams because of channel incision and the loss of channel-scale helical flow patterns, which are responsible for initiating and maintaining pool-riffle sequences. The objective of the current study was to restore pool-riffle bedform and channel hydraulics to maintain pool-riffle structure in an incised, straightened 270 m reach of Beaver Creek (Knox County, Tennessee). Property boundaries laterally constrained the existing channel thereby limiting extensive planform alterations (i.e., macro-sinuosity) in the restoration design. ArcGIS® and River2D®, a two dimensional depth averaged finite element hydrodynamic model, were used to design bed, bank and in-stream features to support maintenance of velocity acceleration/deceleration sequencing, energy dissipation and complex flow patterns, and bed and bank stability. Design included spacing pool-riffle sequences by integrating into existing channel constrictions and expansions. Also at each riffle and run location, bank material (1–2 m by 10–20 m) was removed to dissipate energy and create hydraulic macro-eddies; and gravel substrate was specified for augmentation at riffles. Flow patterns at multiple flow conditions (from baseflow to bankfull) for both the current condition and the design channel were characterized, and aquatic habitat was evaluated in River2D® generating weightable usable area for various fish species. The design channel was manipulated in ArcGIS® and evaluated in River2D® to develop a sustainable design by altering bed and channel form to create self-regulating pool-riffle sequences, rather than relying solely on natural channel design structures. Boundary shear stresses were calculated from model output to size substrate, design stable in-stream features to augment flow structure, and evaluate bank and bed stability. This work illustrates the utility of computational fluid dynamic models for stream restoration in urban watersheds where channels are in disequilibrium and a reference reach is not relevant. Model development and a description of design will be presented, with a design focus on rehabilitating degraded stream habitat.