The nature and causes of the seiches of Moreton Bay were investigated using both field measurements and mathematical models. The characteristics of the seiches were determined from an analysis of the field data, while the mathematical models were used to determine if atmospheric pressure waves were the most important generation mechanism.
The seiches had a range of water level up to 20 cm. with a duration of up to two days. The range of the seiches was greater at Redcliffe and Tangalooma than elsewhere. The higher order modes of oscillation of More ton Bay seemed to he favoured, possibly by local enhancement of the seiche by the topography of the bay. The seiches were mainly observed during the winter and were usually associated with cold fronts and troughs. Not all the seiches occurred during episodes of strong winds, while most of the seiches were associated with small fluctuations of the atmospheric pressure.
An analytical models was developed to determine the water level response of a closed, one-dimensional channel of constant depth, to travelling wave-like disturbances of the atmospheric pressure. This model is an extension and correction of Wilson's (1972) model. The dynamic response was much greater than the hydrostatic response and was quite large for a broad range of velocities centred about the free wave velocity of the channel. The other important parameter of the atmospheric pressure forcing was the wavelength of the pressure waves. The water level response increased as the wavelength decreased. Tests of the response due to travelling wind stress patterns indicated that the atmospheric pressure was the more dominant term.
The simultaneous measurements of the water level, currents, atmospheric pressure and wind from several seiche episodes were examined. The observations could be explained in terms of local forcing of the bay by atmospheric pressure fluctuations propagating across the bay. However, there was no coherency between the measured water levels and the atmospheric pressure. Water level oscillations were also observed at stations outside Moreton Bay, leaving open the possibility of edge waves propagating into the bay causing the seiches.
A two-dimensional hydrodynamical numerical model was used to simulate the generation of seiches in Moreton Bay by travelling atmospheric pressure disturbances. Leendertse's (1967) model was modified to include forcing by atmospheric pressure and wind stress. A radiative boundary condition was specified to allow waves generated within the model to radiate out of the open boundaries. The accuracy of the model was tested against the analytical models while the energy content of the bay was also calculated to check the conservation of energy. Tests of the numerical model confirmed the importance of the speed and wavelength of the pressure waves, as well as showing the variation of the water level response at different locations around More ton Bay to different directions of the forcing. The numerical model failed to show any interaction between the seiche and the tide as observed at Tangalooma.
Simulations by the numerical model of seiche episodes, using actual pressure records, gave large amplitudes of the water level oscillation but underestimated the measured values. Using a finer grid to improve the accuracy of the computation and the representation of the bay, gave a larger response at Redcliffe and Tangalooma, However, the Tangalooma water level was distorted by a large oscillation caused by the coarse representation of the land boundary near that gridpoint. It appears that a very fine grid model and more detailed measurements would be needed to be able to use the numerical model to verify that atmospheric pressure was the dominant forcing mechanism.