On Long Term Statistics of Ocean Storms Starting From Partitioned Sea States

Abstract

The paper proposes an analysis of ocean storms carried out starting from significant wave height time series of HOMERE sea-states hindcast database based on WAVEWATCH III model. Considering that wave spectra often exhibit multiple peaks due to the coexistence of wind waves and swells, here sea states are described by partitioned sea states that can be interpreted physically as representing independent wave systems. The analysis presented here in the paper deals with the contribution of swells to the storm peaks and on how they influence the long term statistics. The sensitivity of return values of significant wave height to swell contribution is investigated via an application of the Equivalent Triangular Storm Model (ETS). The ETS model provides analytical solution for the calculation of the return period R(Hs>h) of a sea storm whose maximum significant wave height exceeds a given threshold h. The approach of ETS consists in substituting each actual storm with an ETS described by two parameters: the storm intensity, that is the triangle height and it is equal to the maximum significant wave height during the actual storm; the storm duration, that is achieved imposing the equality between the maximum expected wave height of actual and equivalent storms. It has been experimentally proved that the actual storm and associated ETS are statistically equivalent because they have the same maximum significant wave height and the same probability P(Hmax>H) that the maximum wave height exceeds a given threshold H. The sequence of ETSs obtained in this way represents the equivalent sea, while the sequence of actual storms is the actual sea. The equivalent and actual seas present the same wave risk because they are characterized by the same number of storm events, each of them with the same intensity and the same P(Hmax>H). For the proposed analysis a set of four points from open sea to the coast is considered in area of the Gulf of Biscay (France). The results show that the contribution of swells is more significant for the storms of small and medium intensity and decreases for increasing storm intensities. Further return values variability neglecting swell is less than 7% at any point for return periods up to 100 years.