Floodplain woodland hydrodynamics
Xavier, Patricia Anne
MetadataShow full item record
Floodplain woodlands are valuable environments, providing a diverse habitat for many riparian and land-based species. It is now recognised that the continual loss of floodplain woodland has impoverished the national biodiversity of riparian environments, and measures have been brought in both nationally and through the European Commission to halt the decline. This has however, highlighted a deficiency that has existed for many years in the field of river hydraulics. The representation of complex riparian vegetation environments within river models remains an area not adequately addressed. This research presents experimental investigations into floodplain woodland vegetation, with a view to improving the representation of these vegetations within numerical models. Floodplain woodland hydrodynamics were explored with scaled-down (1:8) stag gered arrays of single stem and multi-stem model trees at planting densities of 8.8, 19.8 and 80.6 plants per m 2. The planting densities investigated correspond to the recommended planting densities cited by the Forestry Commission UK. Roughness factors, including Manning's n, the Darcy-Weisbach friction factor / and the bulk drag coefficient Ca were computed for the different model tree and planting density combinations. Velocity measurements within the arrays were investigated, and a study to determine the optimum sampling strategy was carried out to obtain representative velocity and turbulent kinetic energy measurements within the model tree arrays. The optimum sampling locations for streamwise velocity appeared to be clustered around 0.3 s and 0.7 s, where s is the lateral or longitudinal spacing between the model trees, while for turbulent kinetic energy the optimum location was 0.5 s. Full scale drag force versus velocity tests of floodplain woodland trees were carried out and a drag area parameter CdA.Uo derived. The trees experienced little to no bending at low velocities, with force varying linearly with the square of velocity, while considerable deflection was observed at higher velocities, with force varying linearly with the velocity. Physical parameters including height, diameter, mass and volume of the wood are compared against the drag area parameter, with mass and volume showing a stronger correlation than height or diameter. The increase in the drag area parameter due to the presence of foliage was also investigated. The numerical incorporation of floodplain woodland vegetation is presented with respect to two-dimensional depth-averaged numerical modelling. A reach of the River Laver in North Yorkshire, England was modelled to assess the hydraulic impact of the conversion of arable land to floodplain woodland.