Stationary and transient states of random seas

Abstract Based on the concept of an ergodic stationary stochastic process, a more complete definition of the so-called sea-state is proposed. Statistical techniques are presented which improve understanding of the stationary and transient components of the sea-surface elevation. Data from Frigg Field measurements are used extensively. In particular, a sequential method for change detection in the significant wave height process is implemented. The stochastic dependence between the duration of a sea-state and its level of energy is then studied. A first-order Markov approach is used to describe the sequences of sea-states. Finally, some theoretical formulae which predict the largest wave height as a function of the r.m.s level and the duration of the sea-state are validated.