Integrating a multivariate extreme value method within a system flood risk analysis model

Integrating a multivariate extreme value method within a system flood risk analysis model.
Wyncoll, D.and Gouldby, B.P.
Journal of Flood Risk Management, 8 (2). pp. 145-160. (2013)

Download:[img]
Preview
document citation PDF (399Kb)
Official URL:http://onlinelibrary.wiley.com/doi/10.1111/jfr3.12069/pdf
Abstract:Effective management of flooding requires models that are capable of quantifying flood risk. Quantification of flood risk involves both the quantification of probabilities of flooding and the associated consequences. Modern flood risk models account for the probabilities of extreme hydraulic loading events and also include a probabilistic representation of the performance of flood defence infrastructure and its associated reliability. The spatial and temporal variability of flood events makes probabilistic representation of the hydraulic loading conditions on the flood defences complex. In the system method used widely within England and Wales, simplifying assumptions relating to the spatial dependence of flood events are made. Recent research has shown the benefits of using improved multivariate extreme value methods to define the hydraulic loading conditions for flood risk analysis models. This paper describes the development of an improved modelling system that enhances the systems-based risk analysis model currently applied in practice, through the incorporation of a multivariate extreme value model. The improved system has been presented on a case study site in the North West of England.
Item Type:Article
Uncontrolled Keywords:Flood risk analysis, Spatial dependence, Multivariate extremes, Systems modelling, Flood defence failure, Reliability analysis
Subjects:Floods > Flood risk assessment and mapping
Floods > General
ID Code:658
Deposited On:25 Sep 2013 12:05
Last Modified:08 Feb 2017 07:15

Repository Staff Only: item control page