The influence of multi-scale spatial heterogeneity of aquatic vegetation on turbulence, mixing, and dispersion in coastal environments

Abstract

Aquatic vegetation is a common feature of coastal, estuarine, and riparian environments. Aquatic vegetation provides many ecosystem and habitat ser-vices such as acting as a buffer zone to capture sediment, particulate matter, and pollutants, thus preventing these materials from entering the open ocean. Vegetation can also dissipate hydrodynamic energy and mitigate against phys-ical hazards. However, to quantify and predict these effects across a range of forcing regimes and environments necessitates a detailed understanding of flow-vegetation interactions. Therefore in recent decades, many laboratory and numerical studies have focussed on elucidating physical processes such as turbulence generation, mixing, and dispersion in aquatic vegetated flows. However, such laboratory and numerical studies typically simplify vegetation geometries, often by representing canopies as regular arrangements of uniform cylinders, or by applying uniform time-invariant hydrodynamic forcing, which does not always capture real-world conditions. This thesis assesses the impact of spatial heterogeneity of aquatic vegetation across different spatial scales ranging from the vegetation stem-scale (O(0.01 m)) to the larger-scales ob-served in multi-patch arrays of vegetation (O(10 m)). We develop generalized theoretical models that incorporate multi-scale heterogeneity observed in the real-world aquatic vegetation.

Field measurements and model predictions of turbulent kinetic energy in canopies of sparse vegetation under tidal flows:

In the second chapter, we conducted field experiments measuring high resolution velocity data in a sparse canopy of pneumatophores coexisting with a very sparse stand of young mangrove saplings to study the small-scale turbulence (O(0.01 − 0.1 m)). There was no significant wave activity observed for the duration of the experiments making the dataset well suited for testing existing models developed from laboratory experiments under uniform currents. The measured turbulent kinetic energy (TKE) was compared with existing models which assume the eddy length scales to be on the order of the stem-diameter. In these cases the models significantly under-predicted the TKE in the real world setting. However, by modifying predictions to include the measured eddy length scales (around 15-times larger than stem diameter), the model was found to perform well, with percentage errors in predicted TKE of around 67%.The generation of the larger eddies was attributed to the presence of the saplings. This result therefore highlights the importance of characterizing the correct length scales in spatially heterogeneous vegetation canopies.

The influence of patchy vegetation on tidal flows: Field dye measurements of intra- and inter-patch mixing, residence times, and dispersion parameters:

In the third chapter, we explored medium- and large-scale (O(>1 m)) flow and mixing processes in multi-patch vegetation systems. We deployed a randomly arranged patch array of 20 identical patches on an intertidal flat and measured flow behaviours and tracer dispersion across 3 densities (which were kept the same in every patch), for different submergence levels. High-resolution spatial and temporal dye intensities obtained from UAV based dye imagery in combination with Rhodamine WT dye releases were used to estimate longitudinal and transverse dispersion coefficients. The dispersion coefficients were strongly dependent on submergence ratios with a clear inverse proportionality between dispersion coefficients and submergence levels. The intermediate density patches showed slightly greater dispersion compared to denser patches, which was attributed to the variability in the flow path of the dye clouds between different releases. Furthermore, local behaviours of the tracers were studied using the temporal variance and dye residence times at every location (video pixel). The values of both the local parameters increased from upstream to downstream inside the array with some particularly large values within close proximity of individual patches. The dye residences showed no dependence on submergence levels as the vegetation-induced dispersion compensated for the reduced bare-bed dispersion at low submergence levels. Finally, a general model for predicting dispersion coefficient as a function of submergence ratio is proposed and validated with field measurements.

The impact of the spatial distribution of patches on dispersion in submerged aquatic vegetation for varying vegetation densities and submergence ratios:

The third chapter only considered a single patch arrangement, a relative patch diameter of Λ = 100, and constant system dimensions of 6×6 m. Therefore, in the fourth chapter we conducted numerical simulations of 8 different patch arrangements using Horizontal Large Eddy Simulation (HLES) in Deflt3D combined with the in-built rigid vegetation model. A few additional simulations were also undertaken to study the impact of system dimensions and relative patch diameter on dispersion coefficients. The results showed significant variability in dispersion coefficients between the different patch arrangements. The results confirmed the inverse relation-ship between dispersion coefficients and submergence levels in the field and also showed the impact of vegetation density with higher densities producing larger dispersion values. However, for greater submergence levels (Sr > 2) the impact of both the vegetation densities and patch arrangements were negligible. The results from different relative patch diameters also implied that the frontal area density is a better parameter to characterize patch regimes instead of the solid volume fraction. Furthermore, the results from the longer patch arrays (12×12 m) showed that the dispersion coefficients reduced with increase in vegetation densities beyond a certain density, which is the opposite trend to that observed in the shorter patch arrays. Finally, a general dispersion model for predicting dispersion coefficients based on frontal area density and submergence levels is proposed for patchy aquatic vegetation.

The results from the thesis have helped to elucidate the importance of char-acterizing spatial heterogeneity of aquatic vegetation across different scales to understand the scale-dependent physical processes. Furthermore, significant contributions were made by developing the existing and newer models for predicting turbulence and dispersion in aquatic vegetation systems which can be used to understand and manage coastal and wetland aquatic systems.