A semi–parametric approach in the estimation of the structural risk in environmental applications

In environmental applications, the estimation of the structural risk is crucial. A statistical model for the behavior of the input variables is generally required, possibly accounting for different dependence structures among such variables. Copulas represent a suitable tool for dealing with natural extremes and non-linear dependencies. Two semi-parametric procedures for the approximation of, respectively, Extreme Value and Archimedean copulas, are proposed in order to provide a model for the estimation of the structural risk. The approximating techniques are evaluated by Monte Carlo tests, and illustrated via a case study concerning a preliminary rubble mound breakwater design.