Semi-probabilistic integrated model for real sea forecasting

This paper proposes an innovative hindcasting methodology of sea states based on the “Significant Wave Method” revisited in its application (Amatucci [1]) and integrated by probabilistic concepts coming from the “Equivalent Triangular Storm model” and the “Equivalent Sea model” (Boccotti [6]). The hindcasting model for the prediction of the wind waves has been changed using the real sequence of the wind states, wind stress factor and considering wave generation only in the direction of the wind, with fetch lengths averaged over small arcs or large arcs. Obtained wind waves, the proposed analytical solution allows to forecast wave climate off a fixed location by the definition of the “Resulted Hindcast Sea” and the associated “Equivalent Hindcast Sea”. This method has been applied at the test case Tyrrhenian Sea - location Ponza (Italy), within a medium-long observation period (07/1989-12/1998) by comparing real sea and resulted sea in probabilistic terms. The resulted sea seems suitable in order to get the significant wave height corresponding to an assigned return period.