Group length which is defined as number of waves in a group, is an important rameter to evaluate groupiness. In this study, the maximum entrop distribution G, pa y () = α γ e x p (− β n) n f G G G is applied to describe G, where α, β and γ can be determined by some reasonable constraints when n is given. For n =1, γ = 0, this distribution has the same form as the commonly used stribution derived by Longuet-Higgins (1984). Comparisons of both the maximum entropy distribution and the distribution of iggins (1984) with the laboratory wind-wave data show ives a better fit; it has the best agreement for n = 0.6.
The grouping of sea waves plays a principle part in many coastal and ocean engineering problems, including low-frequency motions and force analyses of moored vessels, the stability of fixed or floating structures, as well as surf beats in the nearshore zone (Donelan et al., 1972; Goda, 1983; Tayfun, 1989). Besides, wave groups may cause exceptional damage to ships and offshore structures. Recent interest in the subject (Tayfun, 1990; Tayfun, 1994; Kit et al., 2000; Cho and Kim, 2005) has been stimulated by the probable effect of nonlinearity on the formation of a wave group. Wave envelope function is usually used for the description of group properties. Group, defined as number of waves between two successive up-crossings of crossing a given reference leve is an useful parameter to evaluate wave groupiness. In this paper, a new PDF is derived for the distribution of group length. The derivation is based only on the maximum entropy principle, without limitation of Gaussian process hypothesis.
