SOME DYNAMICAL CHARACTERISTICS OF MICROSEISM TIME-SERIES

Abstract

Microseism time-series are non-stationary, non-linear and stochastic, and these characteristics can be reproduced by a forced non-linear damped oscillator. In the present study we show that such an oscillator is also able to explain other widely observed features, such as the variation, for a given seismic station, of the frequency of the secondary peak; the variation of the frequency of the primary peak for different seismic stations relative to the same source; the variations of amplitude of the power spectrum for stormy days with respect to quiet days; and the incoherent propagation of microseisms. Numerical simulations with the proposed phenomenological model suggest i) that the main spectral peak may be due to a competitive process between the resonant response of the medium and an external harmonic force (Longuet-Higgins model), ii) that the secondary peak may be generated by the process associated with the activity of the coastal waves or as a subharmonic of the resonant frequency and iii) that the large amplitude variations between quiet and stormy days refers in fact to variations in the source (storm) distance. From a general point of view we can say microseism activity can be interpreted as the resonant response of the Earth to atmospheric cyclonic storms coupled with the oceans.