Derwent: Hydrodynamics and Sediments

Derwent

Download the Final Report (Hydrodynamic Modelling) 46.5 MB

Download the Final Report (Sediment Transport) 3.8 MB

This project was conducted under the Coastal Cathchment Inititive within the Derwent Estuary Program. Details of the hydrodynamic study can be found in:

Herzfeld, M., Parslow, J., Margvelashvili, N., Andrewartha, J., Sakov, P. (2005) Numerical hydrodynamic modelling of the Derwent Estuary. CSIRO Marine Research. 91p.

Key outcomes of the project are described below.

 

The Derwent Estuary Program (DEP) identified the need for predictive capacity to augment a decision support system for the Derwent Estuary (Figure 1). This management system is designed to aid effective decision making regarding environmental issues, and predictive models play an integral role in assessing management scenarios and achieving system understanding. Of particular importance is the issue of heavy metal pollution, and coupled hydrodynamic, sediment and heavy metal prognostic models are important tools that may assist the formulation of management strategies. The goal of the hydrodynamic component of the DEP is to provide a hydrodynamic model that may assist in management strategy evaluation within a management framework, and provide understanding into the physical dynamics of the estuary (water transports, mixing regimes and temperature / salinity distributions) and the relationship between process occurring on different time and space scales. The hydrodynamic model forms a base into which sediment, contaminant or biogeochemical models may be coupled.

The Derwent Estuary bisects the city of Hobart, capital to the state of Tasmania in southern Australia. This estuary is micro-tidal and extends approximately 60 km from the seaward end to the head of the estuary. SHOC has been applied estuary-wide to investigate impact of heavy metals. The grid used to represent the whole estuary is illustrated in Figure 2. The domain was discretized using a curvilinear grid with variable resolution, resulting in minimum resolution in the upper estuary of ~140 m and maximum of ~400 m in the lower estuary. The model uses 23 ‘z’ layers in the vertical with 0.5 m resolution near the surface to 8 m at depth. There exist 11024 (212 x 52) surface cells in the grid, only 25% of which are wet; i.e. the majority of this grid is associated with dry land.The river in the upper estuary was straightened and laterally averaged to reduce computational pressures imposed by the stability constrains if the cross-river dimension were resolved. The model was forced with river flow from the Derwent River, wind stress and surface elevations, temperature & salinity on the seaward limits of the estuary. These seaward boundary conditions were derived from a larger scale model of the region and direct measurement.

System characterization was determined using a range of diagnostic tools available in SHOC, some of which are described below. Generally, the model results confirm that the Derwent Estuary behaves as a salt wedge estuary with marine flow in bottom waters directed upstream in the estuary and a fresh water surface flow heading downstream. The head of the salt wedge is located near the Bridgewater (where the 2D river begins in Figure 1.1) under low flow and is pushed downstream under high flow conditions. Surface salinities may be less than 20 psu in the lower estuary under high flow. On diurnal timescales the tidal flow dominates the region, with flow directed up-river during the flood tide, and vice versa during the ebb. Under low flow conditions a surface reversal of the currents is seen in the upper and middle estuary during the tidal period but this is absent under high flow since the river discharge overwhelms the tidal flow. The Derwent Estuary is micro-tidal with tidal range ~1 m and having diurnal mixed character with a form factor of ~1.5 (i.e. the tide is a mixture of diurnal and semi-diurnal). The tide undergoes a neap-spring cycle of ~14 days. Tidal excursions in the estuary are variable in time and space, with an average of ~4 km and with excursions reaching double this value under favourable wind and river flow conditions. The tide is asymmetrical with ebb tide shorter and associated with stronger currents than the flood. On the stronger ebb tide the pycnocline undergoes vertical displacement that lead to large salinity fluctuations near the bottom. This displacement may be due to advection resulting from secondary cross-river circulations enhanced by a compound bathymetry and river curvature.

The mean flow over a neap-spring cycle exhibits downstream flow at the surface with magnitudes less than 0.2 ms^-1 and up-estuary flow at the bottom with magnitudes less than 0.05 ms^-1. The surface flow appears to favour the eastern bank of the river in the lower estuary. A cyclonic gyre is observed at the bottom in the lower estuary. Example of the surface and bottom Eulerian mean flow is provided in Figure 3.

The calculation of flushing times can be computed using an e-folding rate based on depletion of total mass in a given region. Flushing times varied from less than 1 day for many of the side bays to ~11 days for the whole domain. A flushing estimate for the whole domain based on the average time for neutrally buoyant particles to exit the domain was computed as ~12 days. An example of the temporal evolution of mass for Elwick Bay in the middle estuary is depicted in Figure 4.

The distributions of passive tracers released as point sources at various locations in the estuary may be post-processed to generate the spatial 5, 50 and 95 percentile distributions. These distributions provide an indication of connectivity throughout the region. The dynamics governing tracer distributions are linear, hence these distributions may be scaled proportionately with the source input. This provides an indication of potential concentrations and distributions of pollutants that may be input at the point source locations. These analyses generally supply more information than the residence time of sub-regions of the estuary, since temporal variability in the estuary is more effectively captured, and quantitative spatial distributions are provided. These analyses for the Derwent generally revealed a mixing zone of high concentration, with lower concentration throughout the remaining estuary. Tracers released in surface waters resulted in distributions of higher concentration predominantly downstream of the release site. Tracers released in bottom waters generally followed the salt wedge circulation and impacted much of the surface estuary waters. Examples of these distributions are provided in Figure 5.

Particle tracking may also be used to assess connectivity in the domain. Results indicated that particles were fairly evenly distributed throughout the whole estuary as a result of release from each site, indicating the estuary is reasonably well connected. Particles released in the salt wedge in deeper water were more evenly distributed throughout the domain in comparison to those released in surface waters, whose distributions were predominantly confined to downstream from the source (consistent with motion expected due to the residual flow). Particle trajectories may also be used to investigate the diurnal tidal forcing; particles exhibited up-estuary movement on the flood tide, and down-estuary on the ebb, except in the upper and middle estuary under high flow conditions when particle trajectories were oriented down-estuary on both flood and ebb tide. Mean tidal excursions of ~4 km were diagnosed from particle trajectories. Example of trajectories obtained from particle tracking analyses are displayed in Figure 6.