A working group was set up in 1992 to assess the state of the art as to the methods for the determination of the extreme wave conditions to be used in coastal development projects within this framework, the respective methods of the participants were tested in a "round robin" configuration, for several case studies The paper presents the characteristics of these different methods, and the results of the comparisons that were performed Conclusions of high concern to engineers are drawn:
- –
Results may be presented as return periods, but also in terms of distributions of the maximum value over a given duration.
- –
In all cases, engineers need to favor methods that emphasize the goodness of fit on the tail part of the distributions.
- –
Distributions, instead of single values for the extremes of the parameter of interest, may be useful in the decision-making process.
In the present study, uncertainties related either to sampling procedures or to the representatively of the observation period are larger than the differences between the results of individual methods. However, for future applications, the prediction method should become a key factor m the quality assessment of design waves determination.
For coastal development projects, the choice of design conditions is an important step, which heavily influences the final cost of constructions and their ability to withstand storms. The choice of design wave heights is directed by extreme waves prediction, the quality of which depends to a large extend on the availability of good data at the location of interest. Yet, existing tools for extremes prediction from databases are many, and apply sometimes significantly different approaches. The technical services of the french industry for equipment, transport and tourism (STNMTE) thus decided to undertake a systematic comparison of these tools in order to investigate the differences in estimation resulting from their use and to quality each of them.