Influence of saltmarsh vegetation canopies on hydrodynamics in the intertidal zone

Recent estimates of global sea level rise indicate mean values around 3.1 mm yr"1. As a result, many coastlines face an increasing risk of coastal erosion, and the threat of flooding is becoming a major concern. Unfortunately, coastal defences can be very costly with recent estimates as high as £5000 per metre length of seawall in the UK. There is a need to consider more economically feasible options, and by accounting for the ability of saltmarshes to absorb wave energy, reduce flow velocities and stabilise sediments, the costs of coastal defence structures may be significantly reduced. But first, an improved understanding of the implications of saltmarsh vegetation on hydrodynamics is fundamental to their inclusion in the design of coastal protection schemes. This includes the influence of saltmarsh vegetation on velocity and turbulence structures and the drag forces that arise due to the obstruction to the flow created by the vegetation. Two contrasting areas of coastal saltmarsh were selected for the location of a field survey to identify typical field conditions, such as bed gradients, submergence levels, vegetation types and densities. The two sites differ in that the first was non- grazed, while the second was heavily grazed to assess the impact of sheep farming on vegetation characteristics. The vegetation species, stem densities and submergence levels observed during the field survey were used as a guideline for designing a series of laboratory experiments to investigate the impact of saltmarsh vegetation on hydrodynamics. Uniform cylinder models are widely used to simulate vegetation canopies in hydrodynamic studies, yet the cylinder model can lead to an oversimplification of vegetation morphology. A comparison was made by conducting experiments under uniform flow conditions where uniform cylinder arrays and vegetation canopies were installed onto a flume bed at stem densities of 800, 1160 and 1850 stems m*2. There were differences in velocity and turbulence structures through the two types of canopy. For the same stem density, the proportion of the total flow passing through the canopy region was approximately 10% greater for the uniform cylinder arrays. The foliage found in the upper part of vegetation canopies resulted in a considerably higher level of obstruction and contributed towards reducing velocities and Reynolds stresses within the canopy. Reynolds stress penetration depths were up to 15 times greater for the uniform cylinder arrays compared to the vegetation canopies. Computational fluid dynamics models can be a useful tool for predicting the impact of saltmarsh vegetation on hydrodynamics for coastal management. Applying such models to vegetated flows requires knowledge of the drag coefficient to determine the drag term in the Navier-Stokes equations. However, in the absence of measured data, such models are often applied with the assumption that the drag coefficient is constant in value, and commonly used values include '1.0' and '1.2'. Such assumptions may be easily linked to the uniform cylinder model. However, drag coefficients calculated for Common Cordgrass ranged between 0.4 and 1.7. Values were dependent on numerous parameters, including the Reynolds number, the submergence level, the stem density and the maturity of the vegetation. Instead of the traditional drag-force approach for determining canopy hydrodynamics, a method for predicting velocity and turbulence structures based on the projected area of the vegetation was proposed. For the emergent canopy, the mean velocity was estimated by relating to a reference canopy of known projected area and mean velocity. For the submerged canopy, the surface flow layer velocity was determined effectively using the Manning's roughness approach and the depth- averaged canopy velocity was a function of the surface layer velocity and the canopy density. Velocity profile shapes for both canopies were obtained by linking the mean canopy velocity to the projected area profile. Reynolds stresses for the emergent canopy and lower part of the submerged canopy were negligible and a function of depth-averaged canopy velocity. For the upper part of a submerged canopy, Reynolds stresses were a function of the depth-averaged surface layer and canopy velocities. The proportion of the submerged canopy region experiencing higher Reynolds stresses is also a function of the vegetation density.