Analysis of sea level fluctuations around the Australian coast with anomaly time series analysis approach

Abstract

Thirty two-year observations (1990–2022) of water level from a total of 14 high-quality acoustic tide stations around the coast of Australia deployed for the Australian Baseline Sea Level Monitoring Project were analyzed. The following three approaches were used: Fourier Transform (FT), Empirical Mode Decomposition (EMD), and Singular Spectrum Analysis (SSA). The water level anomaly was observed to have predominant annual variations with a period of about 12 months based on the Fourier transform. The intrinsic components of stations were extracted in the EMD analysis and the mean period of each of the components was calculated using the zero down crossing method. A regular association was observed between the order of modes and the mean period such that the periods increase by a factor of two on successive modes. The third method used for anomaly analysis was SSA. The number of the obtained components in this method was less than in the EMD. Moreover, the order observed for the components' period in the EMD was not seen in this method. Spectral analysis of Autocorrelation function (ACF) has demonstrated that peak frequencies are almost the same with anomaly spectra so the dominant modes in anomalies are also present in the ACF.