Periodic structures of Great Lakes levels using wavelet analysis

The recently advanced approach of wavelet transforms is applied to the analysis of lake levels. The aim of this study is to investigate the variability of lake levels in four lakes in the Great Lakes region where the method of continuous wavelet transform and global spectra are used. The analysis of lake-level variations in the time-scale domain incorporates the method of continuous wavelet transform and the global spectrum. Four lake levels, Lake Erie, Lake Michigan, Lake Ontario, and Lake Superior in the Great Lakes region were selected for the analysis. Monthly lake level records at selected locations were analyzed by wavelet transform for the period 1919 to 2004. The periodic structures of the Great Lakes levels revealed a spectrum between the 1-year and 43- year scale level. It is found that major lake levels periodicities are generally the annual cycle. Lake Michigan levels show different periodicities from Lake Erie and Lake Superior and Lake Ontario levels. Lake Michigan showed generally long-term (more than 10 years) periodicities. It was shown that the Michigan Lake shows much stronger influences of inter-annual atmospheric variability than the other three lakes. The other result was that some interesting correlations between global spectrums of the lake levels from the same climatic region were found.