Exchange and Mixing in Cockburn Sound, Western Australia: a Seasonally Stratified, Micro-tidal, Semi-enclosed Coastal Embayment
Thesis DisciplineCivil Engineering
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
The hydrodynamics of a relatively deep semi-enclosed micro-tidal stratified coastal embayment and its adjacent waters is investigated through a combination of observational and analytical modelling techniques. The relative importance of hydrodynamic processes influenced by baroclinic and barotropic mechanisms, at time scales from hourly to annual, is detailed. The field site for the investigation is Cockburn Sound and surrounding waters, off southwest Australia. It is shown that for either a homogeneous or two-layered coastal embayment that is separated from the ocean by relatively shallow sill openings, a wind-stress will result in free surface and interface variations which approximate those of a geometrically equivalent closed basin; the similarities become greater as the sills become shallower. This is achieved through the development, validation and application of an analytical hydraulic model of throughflow, free surface and interface variation (ie upwelling), based on the Bernoulli and Momentum theorems. It is demonstrated that simple analytical methods originally developed for the calculation of mean dynamical motions and upper mixed layer deepening in closed basins can be applied, without undue loss of accuracy, for the same purposes in semi-enclosed embayments such as Cockburn Sound. The hydrodynamics of Cockburn Sound and adjacent waters (south to about Wambro Sound) is shown to progress through a characteristic annual cycle, within which it transforms from being a Region of Freshwater Influence (named the Swan-Canning ROFI) during mid-winter to spring, to a Region of Evaporation Influence (named the Perth ROEI) during autumn. The ROFI regime relates to the hydrodynamic influence of strong salinity gradients set up through coastal freshwater buoyancy fluxes derived from estuarine outflows of the Swan-Canning Estuary. The concept of a ROEI regime, which relates to the hydrodynamic influence of strong salinity gradients set up through evaporation and/or hypersaline estuarine discharges, is introduced in this thesis. South of Cockburn Sound, the coastal waters between Dawseville and Warnbro Sound are affected by estuarine discharges from the Peel-Harvey Estuary and local evaporation and hence also undergo a characteristic intra-annual ROFI to ROEI transition. The region is accordingly named the Peel-Harvey ROFI for the mid-winter to spring period and the Peel-Harvey ROEI for the autumn period. The work of this thesis supplements the global understanding ofROFls (Simpson, 1997a) and introduces the concept of the ROEl, through the baseline study of two newly defmed ROEl regimes. It is shown that wind-stress and density gradients are the primary influences on the mean basin-scale hydrodynamic behaviour of Cockburn Sound and its adjacent waters throughout the year. During Cockburn Sound's ROFl regime, full-depth mixing occurs on average about once per synoptic cycle (ie about weekly), when frontal systems, originating in the 'Roaring Forties', bring strong storm winds from the northwest and southwest quadrants (10-15 m s⁻¹). Following these events, the wind pattern tends to southwesterly-southeasterly and weakens <10 m s⁻¹), and deep-water renewal takes place as relatively dense ocean water plunges into the embayment via the sill openings. This results in the establishment of stable vertical stratification, characterised by a parent pycnocline between the relatively dense inflow and resident embayment water. Surface waters are advected out of the embayment. This process continues as the wind remains weak-moderate and swings through the southeast quadrant. The wind then continues to swing through the northeast and northwest quadrants, gradually strengthening to about 10 m s⁻¹, while driving buoyant plumes of brackish Swan-Canning Estuary water southwards into the embayment, assisted by Coriolis force. The cycle then returns to the severe mixing phase, as the next storm front passes the area. Throughout the cycle, notwithstanding severe storm mixing, incomplete vertical mixing of the water column (down to about the depth of the parent pycnocline) occurs via surface processes associated with diurnal winds and penetrative convection. In contrast, during the ROEl regime evaporative processes lead to strong vertical salinity stratification in the embayment, and full-depth mixing is not as regular as during the ROFl regime. Although the average occurrence of storms (which fully mix the embayment) appears to also be, on average, about 4-5 times per month (as per the ROFI regime), the essential factor that was clarified through the observations and analyses is that up to 3 weeks may pass between these events. The mean density of the embayment is greater than that of the adjacent ocean. Hence, mixing within the embayment is followed by relatively buoyant inflows of ocean water. Flushing of the embayment as a whole is therefore less regular than during the ROFl regime, and this influences the poorer water quality (due to eutrophication) that characterises the ROEl regime. During summer (ie between the ROPI and ROEI regimes), seabreezes and penetrative convection regularly mix the water column to the bottom (on almost a daily basis), due to the relatively strong nature of mixing agents and the relatively weak buoyancy fluxes at this time of the year. The analytical and observational clarifications of the hydrodynamic influence of density gradients, both within Cockburn Sound and between the embayment and adjacent ocean, have improved the qualitative and quantitative understanding of the nature and rates of exchange and mixing in these waters. In turn, this has provided a better understanding of the biological characteristics of the embayment, such as the annual occurrence of nuisance algal blooms and de-oxygenation during the ROEI regime. Numerical hydrodynamic models were applied in a major study of the ecology of the Perth coastal zone, ie the Southern Metropolitan Coastal Waters Study (Simpson et al, 1996). The results ofthis thesis guided the choice, application, calibration and validation of the models. It is shown that baroclinic models are appropriate for the relatively strongly stratified ROPI and ROEI regimes, and that barotropic models are suitable otherwise, such as during summer when full-depth mixing occurs almost daily and the density difference between the embayment and adjacent ocean is relatively small.