CHANGING MEAN SEA-LEVEL AND TIDAL CONSTANTS ON THE WEST-COAST OF AUSTRALIA

Long time-series of sea-level observations spanning 21 years (1966-86) from four ports on the west coast of Australia are analysed in yearly sets to study the secular trends and periodic changes in the tidal harmonic constituents. Mean sea level and five harmonic consistuents-Sa from low frequency band, O-1 and K-1 from the diurnal band, M(2) and S-2 from the semi-diurnal band-are investigated for secular trend and other variations that are not accounted for by conventional methods. Secular trends at Darwin, Wyndham and Geraldton are such that the amplitudes of M(2) and S-2 tides are decreasing and their phases are retarding. The changes observed at Fremantle are similar in magnitude but opposite in sign. In the diurnal band, variations in the phases of O-1 and K-1 are similar to those of the M(2) and S-2 tides. The observed trends of the diurnal and semidiurnal tides of Wyndham and Fremantle are well above standard error. Secular changes in the annual tide are not significant. In addition to the secular trends, these constituents are also modulate by terms which are not present in the tide-generating potential and they cannot be resolved from observations of less than 18.61 years. For example, in the spectrum of the equilibrium tide, there is no nodal term associated with the S-2 constituent because it is of the solar origin, whereas in the observed tide two nodal terms appear as side bands. At some ports the amplitudes of the new nodal terms are large enough to modulate the principal tide by 2%. The mean sea level has been rising at the average rate of 1.73 mm year-(1) over this period. It is modulated by a tidal signal of 18.61 years' cycle and non-tidal signals which are spatially consistent but aperiodic. A strong correlation between the residual component of the annual mean sea level and Southern Oscillation Index suggests that non-tidal long-period perturbations of mean sea level (MSL) are mainly due to Fl Nino-Southern Oscillation effects.