Metocean extreme estimations: The sensitivity of offshore design measures to statistics' uncertainties

Most Coastal and Offshore Engineering applications have to deal with the extreme waves, estimated by an extreme value model, a statistical approach that uses the available wave data to estimate physical extremes. The usual lack of reliable long-term wave datasets, implies the need to give particular attention to the usage of the results obtained by extreme estimation methods. In this study, analysis of the aforementioned uncertainties is carried out by finding extremes using different statistical models. These models consist of unique combinations of the following settings: extremes estimation approaches, data sampling rates, probabilistic distributions adopted and parameters estimation methods. We use a hindcast dataset for a location in the Eastern Mediterranean Sea, consisting of 37 years (1979-2015) of one-hourly wave data readings. The differences in estimated extreme values illustrate the underlying statistical uncertainties. Thereafter we study the impacts of these uncertainties on the contour line location, the wave theory applicability of the design state, and finally the effects on wave forces calculated by Morison postulate on a selected case-study vertical cylindrical structure, such as an offshore riser.