ARMA models to represent bi-modal sea spectra

The paper analyses the results of the sea spectrum reconstruction in case of bi-modal sea states when ARMA models are used to describe the sea wave time series. The proposed spectral reconstruction procedure allows deriving a parametric model of the spectrum relying on the description of the phenomenon as the output of a linear shift invariant system with white noise as input. The method has been tested considering simulated sea states composed by different wind and swell components, spacing among different cases to estimate the effectiveness of the procedure when varying the frequency and the relative amplitude of the two components. The method shows promising results in identifying both the wind and swell components associated to the considered test cases and shows to be robust to the frequency spacing of the two modes.