Selected time series and their spectra in geosciences

dr Maciej J. Mendecki

Natural processes often generate time series data across various phenomena. One common type of time series, which can be effectively analyzed through spectral techniques, is ground motion evoked by both natural and anthropogenic events. The characteristics of these events may vary, encompassing waveforms produced by seismic events or ambient seismic noise. In seismology, spectral analysis can be used to filter waveforms for clear signals, identify dominant frequencies that reflect focal characteristics, and even recognize near-surface site effects influenced by local geology. Beyond seismology, time series analysis has applications across other areas of geoscience, where patterns may be less immediately apparent. Tree-ring analysis, as a form of time series, can reveal annual growth pulses in the frequency domain. Signal decomposition into higher-order components can be performed to uncover the regular dynamic behavior embedded within tree-ring data.