Bivariate Archimedean Copulas to Solve Complex Dependency in Marine Engineering Problems

Abstract

Climate change has made offshore installations suffer severe consequences from extreme marine hazards. These offshore installations were designed to withstand extreme natural hazards. Marine natural hazards commonly occur interdependently. Thus, a multivariate model to capture this complex dependence structure is necessary. Practically, modelling marine natural hazards is usually assumed to be mutually independent or correlated by a multivariate Gaussian distribution. However, this biased assumption is not relevant to capture the real dependence structure between marine parameters. Copula functions are used to model the complex dependence structure between marine parameters. A simplified case study is selected to illustrate the modelling between wave height and wind speed. Results are compared with the traditional joint probability approach to demonstrate the advantage of copula functions. The use of copula functions provides a better result to model any complex dependence structure between correlated variables.