Journal of the Royal Society of Western Australia, 83: 251-254, 2000 Hydrodynamics of Leschenault Inlet, Western Australia A Charteris1 & D Deeley2 ‘WBM Oceanics Australia, PO Box 604 Collins St West, Melbourne Vic 3000 email: [email protected] 2Acacia Springs Environmental, PO Box 236 Palmyra WA 6157 Abstract A mathematical model of Leschenault Inlet and Koombana Bay was developed using finite element techniques to describe the circulation patterns in Leschenault Inlet under the influence of a variety of tidal, river inflow and wind conditions. Simulations included long term tidal characteristics, typical summer ebb and flood conditions, typical winter ebb and flood conditions and typical summer conditions with the influence of an afternoon sea breeze coinciding with a flooding tide. Tidal constituents for the Port of Bunbury, south of Leschenault Inlet provided the tidal boundary condition data for the model. The port is influenced by predominantly diurnal tides (one tide per day) with a mean spring tidal range 0.5 m. A slightly higher than mean spring tide was used for all short period simulations with a tidal amplitude of 0.3 m about mean sea level. Additionally, the four main constituents from Bunbury (M2, S2, Ol, Kl) were used to generate a long period of tides to assess the effect of the spring and neap tidal ranges on water levels in Leschenault Inlet. Results indicated a significant attenuation of the ocean tidal range in the inlet as well as a 4- to 7-hour phase lag for high and low water between the ocean and the inlet. These characteristics are a consequence of the hydraulic restriction presented by "The Cut". Mean water level in Leschenault Inlet was also found to be influenced by river flow and was generally higher during the winter months when flows from the Preston, Ferguson, CoUie and Brunswick Rivers are higher. Circulation patterns in the inlet were found to be similar for summer and winter conditions. The afternoon sea breeze had only a minor effect on circulation patterns in Leschenault Inlet. The model has not been verified against measured current data in the inlet. Nevertheless, it pro¬ vides a useful tool for comparing different tidal, wind and river flow scenarios. Keywords: Leschenault Inlet, south-western Australia, estuary, hydrodynamics, mathematical modeling. Introduction Leschenault Inlet is a large coastal lagoon located just north of Bunbury, Western Australia. It is connected to the Indian Ocean by a narrow artificial passage known as "The Cut". Water levels in the inlet are influenced by tidal variations in Koombana Bay and surface drainage water from the Collie and Brunswick Rivers to the east and the Preston and Ferguson Rivers to the south. Flows in these river systems are highly seasonal. This paper describes the circulation patterns in Leschenault Inlet under the influence of a variety of tidal, river inflow and wind conditions. Model Development A mathematical model of Leschenault Inlet and Koombana Bay has been developed using finite element techniques from the RMA suite of software (King 1997). The finite element network is shown in Fig 1, with detail at "The Cut" shown in Fig 2. The model has not been verified against measured data, nevertheless it provides a useful tool for comparing different tidal, wind and river flow scenarios. Simulations presented herein are: long term tidal characteristics, typical summer ebb and flood conditions, and typical winter ebb and flood conditions. There are published tidal constituents (i.e. the har¬ monic components of the tides: M2 = main lunar constituent for a semidaily tide, S2 = main solar constitu- Figure 1. The finite element network used in the mathematical © Royal Society of Western Australia, 2000 modelling of Leschenault Inlet and Koombana Bay. 251 Journal of the Royal Society of Western Australia, 83 (4), December 2000 Figure 2. Computation mesh detail at "The Cut", used in the math¬ Figure 3. Tide and wind boundary conditions over a tidal cycle of ematical modelling of Leschenault Inlet and Koombana Bay. 12 hours. ent for a semidaily tide, Ol = main lunar constituent for a mid and northern Leschenault Inlet for this long simula¬ daily tide, K1 = soli-lunar consituent for a mainly fort¬ tion. The mean water level in the inlet is higher during nightly constituent) for the Port of Bunbury, just south of spring tide than during neap periods. In effect, the spring Leschenault Inlet (Anon 1997). The port is influenced by tides "pump up" the mean level in the estuary to a maxi¬ predominantly diurnal tides (one tide per day) with a mum of about 0.06 m above mean sea level while during mean spring tidal range 0.5 m. A slightly higher than mean the neap periods the mean level falls to a low of about spring tide has been used for all short period simulations mean sea level. Mean water level in Leschenault Inlet is with a tidal amplitude of 0.3 m about mean sea level. Ad¬ around 0.03 m above mean sea level. During winter, ditionally, the four main constituents from Bunbury (M2, higher river inflows tend to increase this mean level to S2, Ol, Kl) have been used to generate a long period of around 0.10 m above sea level (dependent on river flow). tides to assess the effect of the spring and neap tidal ranges Fig 5 shows results from the summer simulation. Water on water levels in Leschenault Inlet. levels in the inlet and ocean are plotted along with cur¬ rent velocity through "The Cut". Flows in the tributary rivers were determined by re¬ view of river flow records over the period 1985-1995. Table These model results indicate a significant attenuation 1 summarises these data. The adopted flow for model¬ of the ocean tidal range in the inlet as well as a 4- to 7- ling purposes is approximately the 90 percentile value to hour phase lag for high and low water between the ocean allow investigation of typical higher flow conditions. and the inlet. These characteristics are a consequence of the hydraulic restriction presented by "The Cut". Peak The influence of the afternoon sea breeze on circula¬ flood tide velocities in "The Cut" occur about 1-2 hours tion patterns in Leschenault Inlet is included. A 20 kt prior to high water in the ocean and peak ebb velocities (10 m s'1) southwest wind is simulated and is timed to about 2-3 hours before low water in the ocean. Flood and coincide with the rising tide in Koombana Bay. Fig 3 ebb tidal flows in "The Cut" continue on for several hours shows the wind speed and tide track over a tidal cycle. after ocean high and low water respectively. Figs 6A-6D show circulation patterns in Leschenault Inlet under the Results influence of summer and winter river flows (respectively). A long term model simulation of 900 hours (37.5 days) Figs 6E-6F show the circulation patterns during summer was conducted to examine the effect of spring and neap with the additional influence of the afternoon sea breeze tides on water levels in Leschenault Inlet. Tidal bounda¬ coinciding with the rising ocean tide as described above. ries were generated from constituents as described above A comparison of the model results indicates that the cir¬ and ''summer" flow conditions were adopted for the tribu¬ culation patterns in Leschenault Inlet are very similar for tary rivers. Fig 4 presents results from the ocean, southern. summer, winter and the summer breeze condition. How¬ ever, circulation patterns in Koombana Bay appear to be quite different. The eddies observed are set up by the ebb Table 1. Summary of River Flows for the Collie, Brunswick, tide jet through "The Cut", the strength of which is de¬ Preston and Ferguson Rivers (1985 to 1995) pendent on the tidal range, wind conditions and river flow Flow conditions. The existence of these eddies is unconfirmed, Average 90 Percentile Adopted and their presence has little hydrodynamic influence on circulation patterns within Leschenault Inlet. Brunswick & January 7.8 10.2 60.7 Collie Rivers July 28.6 60.7 60.0 Numerical modelling of Leschenault Inlet and Koombana Bay indicates that Leschenault Inlet is influ¬ Preston & enced by a predominantly diurnal tide (one tide per day). Ferguson January 0.2 0.3 0.5 A delicate balance of forces governs water movement in Rivers July 13.9 29.9 30.0 the Bay, "The Cut" and lower inlet (van Senden 1987) in- 252 Charteris & Deeley : Hydrodynamics of Leschenault Inlet Figure 4. Mean water level variation in Leschenault Inlet over a spring tide - neap tide - spring tide cycle. Figure 5. Flow characteristics at "The Cut" for a summer simulation. Water levels in the inlet and ocean are plot¬ ted along with current velocity through "The Cut". Ocesn North ¥ South The Ci t 253 Journal of the Royal Society of Western Australia, 83 (4), December 2000 eluding surface wind stresses, buoyancy and mo¬ mentum fluxes from "The Cut", currents induced by sea level changes, bottom friction and inertia forces due to topographic constraints. The bal¬ ance of these forces has been found to vary seasonally. Peak flood tide velocities in "The Cut" occur about 1-2 hours prior to high water in the ocean. Peak ebb tide velocities in "The Cut" occur about 2-3 hours before low water in the ocean. The mean water level in Leschenault Inlet varies and reaches a high of about 0.06m above mean sea level fol¬ lowing a period of spring tides. Mean water level at about mean sea level occurs in the inlet follow¬ ing a period of neap tides. Mean water level in Leschenault Inlet is also influenced by river flow and is generally higher during the winter months when flows from the Preston, Ferguson, Collie and Brunswick Rivers are higher. Circulation patterns in the inlet are similar for summer and winter conditions. The afternoon sea breeze appears to have only a minor effect on circulation patterns in Leschenault Inlet. The circulation patterns would result in a strong component of transport of suspended sediment that would move to the prograded supratidal flats to the far north. This northerly movement of sediment is evident as northerly directed spits and other sedimentologic factors described elsewhere in this volume. The simple two-dimensional model described here used a coarse representation of bathymetry in the Inlet. Inclusion of a detailed bathymetry may make some difference to predictions because of the large shallow expanses along the eastern and western margins of the estuary. The magnitude of these differences is not expected to be great and is unlikely to greatly alter the circulation patterns described here. References Anon 1997 Australian National Tide Tables. Common¬ wealth of Australia. King I P 1997 RMA2 - A Two Dimensional Finite Ele¬ ment Model for Flow in Estuaries and Streams. van Senden D 1987 Dynamics of unsteady jets in shal¬ low receiving waters. PhD Thesis, Department of Civil Engineering, University of Western Australia, Perth. conditions, and summer conditions with a seabreeze. Arrows indicate direc¬ tion of current. Graded shades of grey indicate magnitude of current m s . 254