A0789
Title: A geometric approach to extreme value theory with application to flood risk modelling
Authors: Ioannis Papastathopoulos - University of Edinburgh (United Kingdom)
Lambert De Monte - University of Edinburgh (United Kingdom) [presenting]
Ryan Campbell - Lancaster University (United Kingdom)
Haavard Rue - KAUST (Saudi Arabia)
Abstract: Extreme value theory (EVT) is a branch of probability and statistics concerned with the estimation of the probability (and the characterization of the behavior) of rare events of possibly unobserved magnitudes. A new methodology is discussed from the recently emerging geometric EVT framework and its application to the modelling of ocean levels exceeding a dyke in Newlyn, England. The approach exploits new theory results for the limiting distribution of the radial part of a radial-angular decomposition of a multivariate random vector of interest and for the distribution of directions in the multivariate space along which radial extremes occur. The framework allows for a natural extension of the concept of univariate return levels to multivariate settings, and the proposed statistical inference methodology offers great flexibility in modelling various (extremal) dependence structures.
