EARTHQUAKE ACCELEROGRAM PROCESSING USING SINGULAR SPECTRUM ANALYSIS

In this paper we propose the Singular Spectrum Analysis (SSA) method for earthquake accelerograms analysis. SSA is a model free technique of time series analysis, which combines advantages of other methods, such as Fourier and regression analyses. The aim of SSA is to make a decomposition of the original series into the sum of a small number of independent and interpretable components - slowly varying trend, oscillatory components and noise. Comparing data received about one earthquake by different stations it is possible to follow to Fourier spectrum and SSA spectrum changes with distance to epicenter growth. It is possible also to observe experimentally the fast fading of high frequency oscillations comparing with low frequency oscillations. Designing building constructions in the seismic regions, it is necessary to take into account the frequencies of ground vibration during earthquakes, because the coincidence of natural frequency of construction with the ground vibrations frequency may lead to resonance, increasing the load on the structure. SSA gives the possibility to receive the spectrum of ground vibrations free from noise contribution.