Surface waves often represent the most critical environmental parameter for offshore activities, and reliable forecasts are in many cases essential for safe operation.
Currently, decisions on safety during extreme wave situations are generally taken based on deterministic wave forecasts, and uncertainty related to forecasts are acknowledged and accounted for via safety margins. However, unmanning criteria reflecting a consistent probabilistic model for the forecast uncertainty are generally lacking.
In 1998 the European Centre for Medium-Range Weather Forecasts (ECMWF) introduced a new method for ocean wave forecasting based on the Ensemble Forecasting technique. In this paper we address how the ensemble prediction system (EPS) can be used to establish a probabilistic model applicable for platform unmanning criteria and procedures.
Before ensemble forecasting techniques can be introduced operationally for this purpose it is important to investigate how the decision process can be improved with this method. By improvement we expect the probability forecasts to yield a higher hit rate and a lower false alarm rate.
Extreme wave warnings are simulated based on a data series established for a 10-year period and the hit and false-alarm rate from deterministic and probabilistic forecasts are compared with measured data. The ensemble forecasts are used to:
• establish a model for uncertainty of forecasts for bad weather
• update safety margins used at present such that target safety is obtained if deterministic forecasts are used
• discuss a revised scheme where the EPS forecasts are applied on an operational level