A Benchmarking Exercise on Estimating Extreme Environmental Conditions: Methodology and Baseline Results

Abstract

A wide range of methods have been proposed for the derivation of environmental contours for marine structures that must meet reliability targets. An environmental contour is a set of joint extremes of environmental conditions associated with a target return period. In general, environmental contour methods help with the prediction of some future critical combinations of environmental conditions (e.g., wind, waves, current) at a location of interest based on a limited dataset, thus allowing designers to ensure a prescribed structural reliability. In fact, some of these contour methods are specifically recommended by technical specifications and standards as part of a design process. This paper outlines the rules and procedures for a collaborative benchmarking exercise — focused on open comparison — in which researchers are invited to develop and present their own contour derivation approaches based on common datasets that will be available to all. Hindcast and observational datasets are considered and two exercises are planned: One focuses on applying environmental contour methods to a wide range of datasets and the other focuses on uncertainty characterization. Besides describing the benchmark’s methodology, this paper presents baseline results of computed contours following current recommendations. The overall goals of this endeavor are: (i) to work towards the development of more robust statistical models and contour construction methods, (ii) to encourage increased discussion in the international research community and among practitioners, and (iii) to support ongoing efforts to improve technical specifications and standards.