This study is based on the Highland Water Research Catchment (HWRC), a national reference site for lowland floodplain forest streams. Forested floodplains represent the undisturbed land cover of most temperate and tropical river systems, but they are under threat from human resource management. In principle, the ability to simulate forested floodplain flows would underpin our understanding of a number of key environmental issues in these systems, including: flood dynamics; sediment, nutrient and pollutant fluxes, and; the evolution of physical habitat. However, in practice forested floodplain flows are relatively shallow and strongly modified by the complex floodplain topography and the presence of vegetation and organic debris on the woodland floor. In such instances flow blockage and diversions are common, and there is the possibility of frequent switches from sub-critical to locally super-critical flow. Such conditions may also favour intense turbulence generation, both by wakes and by shear. Unfortunately this complexity means that the present state of the art is such that it is not at all clear what terms dominate the governing flow momentum equations or what forms of turbulence model might be needed to close them. Even if numerical simulations could be undertaken, a lack of empirical data (the complexity of forested floodplain flows provides a high degree of spatial variability that inhibits field data collection) negates the possibility of model validation. To address these issues, this project will acquire the high-quality 3D flow velocity data needed to further our understanding of the mechanics of flow across forested floodplains. As noted previously, the high degree of spatial variability inhibits field data collection. The only feasible methodological approach, therefore, is to replicate natural floodplain flows within the controlled conditions of a laboratory flume. However, this raises the new challenge of replicating floodplain topography and associated hydraulics so that the flume experiment is a precise model of natural conditions. We will achieve this by employing Southampton's large-scale Chilworth flume facility to achieve 1:1 hydraulic scaling in tandem with the application of terrestrial laser scanning to obtain a natural floodplain Digital Terrain Model (DTM) at very high resolution. The use of laser-scanning is a key innovation as the resulting DTM data will be input to an innovative manufacturing process capable of milling a high-density polyurethane Physical Terrain Model (PTM) to produce a highly precise (<1mm error) replica floodplain, also at 1:1 scale. The PTM can then be set in the flume & subjected to a steady flow discharge identical to a representative over bank event observed at the HWRC study site. Precise replication of the natural floodplain morphology and flood hydrology within the flume will enable us to acquire spatially distributed 3D flow velocity data at each node on a dense 3D measurement grid, using Acoustic Doppler Velocimetery. The gridded flow velocity data will then be analysed to enable discrimination of the dominant terms in the momentum balance, as well as identification of mechanisms of turbulence production & dissipation, in all cases within different regions of the flow. This research represents the first study into the nature of turbulent flows over topographically complex floodplains. The project will deliver an original empirical model of floodplain flow mechanics that will be used to inform the design of robust computational simulations of forested floodplain flows, as well as to provide a benchmark model validation data set that will be made available to the community. As an ancillary product the project will also 'concept-proof' a novel and transferable terrain modelling technique that enables precise replication of complex topographic surfaces.
Collaborating research institutes, centres and groups
Mean flow and turbulence structure over exposed roots on a forested floodplain: Insights from a controlled laboratory experiment
& James Brasington, 2020 , PLoS ONE , 15 (2)