EMPIRICAL MODE DECOMPOSITION OF EARTHQUAKE ACCELEROGRAMS

Abstract

This article analyzes the strong motion records of past earthquakes by empirical mode decomposition (EMD) technique. The recorded earthquake acceleration time histories are decomposed into a finite number of empirical modes of oscillation. The instantaneous frequency and amplitude of these modes and evolutionary power spectral density (PSD) is estimated from the Hilbert–Huang transform (HHT). Strong motion parameters such as spectral and temporal centroid, spectral and temporal standard deviation, Arias intensity, correlation coefficient of frequency and time are derived from the evolutionary PSD. The variation of these parameters with magnitude, distance and shear wave velocity of the recording station is reported. Empirical equations to estimate these six ground motion parameters are derived from the strong motion data by regression analysis. These equations can be used by engineers to estimate the design ground motion.