Simulating flood event sets using extremal principal components

Hazard event sets, a collection of synthetic extreme events over given period, are important for catastrophe modelling. This paper addresses the issue generating sets river flow northern England and southern Scotland, region which has been particularly affected by severe flooding past 20 years. We start analysing historical across 45 gauges, located within study region, using methods from value analysis, including concept extremal principal components. Our analysis reveals interesting connections between dependence structure region's topography/climate. then introduce framework is based on modelling distribution components in order to generate flow. The generative dimension-reducing that it distinctly handles their contribution describing nature region. also detail data-driven approach select optimal dimension. Synthetic flood subsequently generated efficiently sampling fitted distribution. hazard can be easily implemented practitioners our results indicate good agreement observed simulated dynamics. For considered application, we find outperforms existing statistical approaches sets.