Wave forecasts and small-vessel safety: A review of operational warning parameters

摘要

Traditionally most meteorological offices forecast height, direction and period of wind sea and swell based on phase-averaged wave models. In recent years, there has been special interest in whether it is possible to produce better forecasts, which include information about high-risk situations that are not resolved by the traditional wave parameters. Here we will review and discuss sea-state parameters and safety warning-indices that have been suggested and investigated in recent years. In this review we particularly focus on parameters that are important for small vessels. Some of the findings are:rn1. A current trend in marine forecasts, going beyond the usual parameters, is tailoring of the product to the end users. The extent to which wave forecasts are tailored to small vessels differs quite a lot among meteorological offices.rn2. Single wave and crest heights are adequately described by first- and second-order theory, respectively. Present understanding of mechanisms behind abnormally high single waves suggests that modulational instability is limited to almost unidirectional seas.rn3. Combining wave height and steepness or calculating the risk of synchronous waves is useful, especially in relation to safety of smaller vessels.rn4. Ship accident statistics suggest that the H_(m0) value of sea state is not as important as whether this value is unexpected, due to rapid development or compared to local wave climate.rn5. Severe waves can occur in areas where strong currents oppose the waves, and operational warnings exist for some areas.rn6. The best way to communicate the directional composition of the wave field still seems to be a division of the sea state into wind sea and swell.rn7. In spite of incomplete physics, the predicted level of wave dissipation can be used to highlight potentially dangerous seas in some areas.rn8. Local experience-based warnings are necessary if dangerous sea states can occur that are not resolved by prognostic wave models.