Evaluation of a geomorphic instream flow tool for conducting hydraulic‐habitat modelling

• Published in: River research and applications Vol. 37; no. 10; pp. 1520 - 1537
• Main Authors: Gronsdahl, Stefan, McParland, Dan, Eaton, Brett, Moore, R. Dan, Rosenfeld, Jordan
• Format: Journal Article
• Language: English
• Published: Chichester, UK John Wiley & Sons, Ltd 01.12.2021; Wiley Subscription Services, Inc
• Subjects: Benchmarks; Channel morphology; Depth; environmental flow needs; Fish; Fluid flow; fluid mechanics; fluvial geomorphology; Forest harvesting; forests; Geomorphology; Gravel; Habitats; hydraulic geometry; Hydraulics; hydraulic‐habitat modelling; Instream flow; instream‐flows; Methods; Modelling; North America; Performance evaluation; Polls & surveys; Rivers; Stream discharge; Stream flow; Streams; Surveys; Velocity; watersheds; North America
• ISSN: 1535-1459; 1535-1467
• DOI: 10.1002/rra.3847

Conventional hydraulic‐habitat modelling methods are time‐consuming to implement. In response to repeated calls for more efficient and practical approaches, researchers have developed a geomorphic instream‐flow tool (GIFT) that combines a method to simulate reach‐averaged hydraulics at flows less than bankfull and depth and velocity frequency distributions to develop streamflow‐fish habitat relationships. This approach requires fewer resources to implement than conventional methods, but it has not been widely adopted because it has been subject to minimal testing and validation. This study evaluates the performance of GIFT by comparing its outputs to empirical measurements and conventional model outputs from eight rivers in western North America. The results of this comparison indicate that the root mean square errors for average depth and velocity were 0.078 m and 0.047 m/s, respectively, and the fit of modelled depth and velocity frequency distributions was satisfactory (index of agreement >0.9) for 11 of 15 surveys for depth and 12 of 15 surveys for velocity. GIFT‐derived fish habitat‐streamflow relationships peaked at lower flows than benchmark relationships in smaller streams (mean annual discharge [MAD] < 0.15 m3/s) and are markedly differed from the benchmark in the largest river (MAD of 87 m3/s). GIFT was also paired with a geomorphic regime model to predict the direction of changes in channel morphology and fish habitat following forest harvesting in one watershed. GIFT provides an alternative to conventional modelling approaches for single‐thread, gravel‐bed rivers with a MAD of around 15 m3/s or less. Application of this technique outside of these bounds, or in other regions should proceed with caution, as these scenarios have not been tested.