7. Modelling of environment
A proper modelling of the environment (typically waves and, in some cases, wind) is fundamental in obtaining accurate estimations of ship motions. Therefore, the modelling of the environment plays a crucial role in the evaluation of ship safety. In this context, Pawłowski (2009) analysed analytical expressions of typical sea spectra used in Naval Architecture, showing that, with proper renormalization, such shapes can be approximated by families of functions usually used for describing probability density functions. The topic of extreme (freak) waves has instead been the subject of investigation by Waseda and Kinoshita (2010), who reported non-Gaussian behaviours in case of generation of short crested waves, and reviewed a series of accidents in view of the possible occurrence of freak waves, considering weather forecasting/hindcasting information. The experimental modelling of extreme waves was investigated by Hennig and van Walree (2014), who described different approaches for experimental modelling of extreme waves and investigated nonlinear effects on wave crests distributions, showing that, for a given sea state steepness, the directional wave spreading reduces the probability of occurrence of extreme wave crest heights. A direct specific link between environmental modelling and nonlinear ship motions assessment was instead provided by Rosén et al. (2013). Rosén et al. (2013) used both idealised spectra and spectra coming from forecasting/hindcasting together with simplified semi-analytical spectral methods for assessing risk of pure loss of stability and parametric rolling, showing that spectral representation can have a significant influence on the final assessment. Fig. 7-1, from Rosén et al. (2013), shows an example comparison of estimated GM variation encounter spectra when using the sea spectrum obtained from ECMWF forecast and when using the corresponding idealized spectrum defined by two wave systems (wind waves and swell).