Doctor of Philosophy
University of Wollongong. School of Earth and Environmental Science
Davies, Gareth, Downstream changes in the cross-sectional shape of tidal channels, Doctor of Philosophy thesis, University of Wollongong. School of Earth and Environmental Science, University of Wollongong, 2011. http://ro.uow.edu.au/theses/3266
Tide-dominated estuaries are typically muddy, and have a well-channelised morphology. Their channels characteristically exhibit a ‘tapering’ or ‘funnel-shaped’ planform shape, with a clear upstream reduction in channel width.
To better understand how the interactions of flow, sediment transport and morphology produce this characteristic shape, morphodynamic models are developed to simulate the long-term evolution of tidal channels. Theoretically, the problem is approached by first developing models which simulate the evolution of channel cross-sectional shape, and then linking these with a longitudinal hydrodynamic and sediment transport model, to simulate the evolution of an entire estuarine channel. The cross-sectional models are evaluated via a detailed comparison with the morphology of a muddy tide-dominated estuary in south-eastern Australia. Additional insight into the controls on tidal channel width is gained with a study of the width profiles of thirty tide-dominated estuaries in northern Australia, and their relations to proxies of fluvial and tidal influence.
The models of channel cross-sectional shape are strongly affected by their underlying parametrisations of flow and sediment transport. To gain insight into the effect of different parametrisations, the steady-state solutions of a range of models are investigated under steady, uniform flow. Models which assume that all sedimentis homogeneous and transported purely in suspension tend to predict the formation of stable cross-sectional morphologies which have relatively low aspect-ratios (i.e.width/depth). Their actual shape is sensitive to various assumptions about the near-bed suspended-sediment concentration, the intensity of turbulent momentum exchange, and the nature of resuspension processes. If downslope bedload transportis included in the model, then the predicted stable channel morphologies can attain much higher aspect-ratios. Their actual shape is affected by the intensity of downslope bedload transport, and by the rate of deposition from suspension. If the model is adjusted to account for near-bed increases in the sediment’s critical shear stress (e.g. due to burial), then the predicted stable cross-sectional shapes can also attain high aspect-ratios. However, in this case the model predictions dependon the channel’s assumed initial geometry.
A range of these models are extended to predict channel form in Yalimbah Creek, a cohesive tidal channel in south-eastern Australia. Model parameters are estimated using field measurements. Within the creek’s intertidal flats, numerous small (width of a few metres), low aspect-ratio channels are observed. The morphology of these can be reasonably simulated by assuming that they form through pure incision. However, larger channels tend to have increasingly high aspect-ratios. At least two models can reasonably reproduce the observed changes in channel shape with channel size: one which accounts for the lateral variation in suspended load over the cross-section, and one which includes the effects of bedrock on the channel morphology. Field measurements confirm the importance of the latter in at least some locations. However, neither model accounts for the apparent near-surface variations in the sedimentary properties of the bed, and an attempt to include this in the model results in unrealistic predictions.
In order to empirically investigate the relations between tidal estuary width profiles and tidal and fluvial processes, LANDSAT 5 imagery is used to measure the upstream changes in width within seventy-nine channels from thirty macro-tidal estuaries in northern Australia. It is shown that estuaries with a wider mouth typically have a more strongly ‘funnel-shaped’ appearance than do estuaries with a narrower mouth, and this is partially explained by consideration of the relationship between an estuary’s width, and its variations in stage and cross-sectionally averaged velocity. It is also shown that there are relationships between the characteristic distance over which the estuary’s width decreases upstream, and proxies of the tidal and fluvial influences at the estuary’s mouth.
A model is then developed to simulate the morphodynamic feedbacks controlling the downstream changes of tidal channel cross-sectional shape. This requires linking the cross-sectional models developed earlier in the thesis with a longitudinal flow and sediment transport model. The influence of river discharge and a non-erodible bedrock boundary on the predictions is investigated. It is found that a stablemorphology is only reached when fluvial discharge is included in the model. In these cases, the predicted channel width profiles show upstream rates of change which are within the range of field measurements. The presence of a bedrock boundary(which enforces a limit on the channel bed elevation) causes channels to develop a more funnel-shaped width profile, as compared to the cases without bedrock. River discharge is shown to have two subtle effects on the stable channel morphology: first, it leads to substantial changes in the channel mouth width, even though it accounts for less than 2% of the peak discharge at the channel mouth; and second, an increase in the river discharge can actually lead to a decrease in the degree of river influence in the channel, because of the complex morphodynamic feedbacks between river discharge, channel depth, and tidal propagation.