Crosscorrelation of sea levels

Abstract

Crosscorrelation has been used to determine those factors that occur in common within related strings of geological data. Knowledge of these factors allows the isolation of events that either affect all the data sets or are confined to an individual data string. The crosscorrelation can be expressed in either the time or the frequency domain. However, in the frequency domain, amplitude and phase of the input functions can be calculated and utilized as separate spectra. This makes it possible to combine amplitude and phase spectra from different data sets and then calculate inverse transforms that create new time displays. The amplitude spectrum from the crosscorrelation of related data sets, in conjunction with those for the original data, allow the isolation of components common to all from those that are set-specific, while retaining the original phase spectra for each location. Inverse transforms can then display the revised spectra as time functions containing only the desired frequency components with their new amplitudes and the original phase at the location where the data was recorded. Sea level data from harbors on the Baltic sea illustrate an example of revised sea level measurements separated from local variations that affect the collection locations, yet with the sea components in the correct phase at the desired location. © 2008 Springer-Verlag Berlin Heidelberg.