Green, F. and Nadal, N. and Yawn, M. and Benson, T. and Hazelwood, C. and Liu, Y. (2026) An end to end stochastic framework for extratropical cyclone storm surge extremes. In: EVAN 2026 (7th International Conference on Advances in Extreme Value Analysis and Application to Natural Hazards), 21-23 July 2026, Delft University of Technology. (Submitted)
Full text not available from this repository.Abstract
Extratropical cyclones (ECs) are a major source of extreme winds and coastal storm surges at high latitudes, yet their impacts in medium to low latitude regions remain poorly characterised due to their relative infrequency and the scarcity of observational extremes. Such regions represent a particularly challenging setting for extreme value analysis, where classical stationary or point based approaches often perform poorly and uncertainty dominates inference. This study presents a novel stochastic tracking framework for EC related storm surges, designed to support extreme estimation under data poor conditions. A low latitude, EC infrequent case study is used as a deliberate methodological stress test, enabling evaluation of how physically based, probabilistic modelling can supplement limited empirical records. The framework implements a comprehensive modelling chain, beginning with raw meteorological data and proceeding through objective and probabilistic EC detection, track identification, and stochastic simulation of EC space–time trajectories that reproduce observed variability, clustering behaviour, and non stationary features. Simulated EC tracks are translated into spatially coherent wind pressure fields, allowing track level uncertainty to be propagated into estimates of wind extremes. These wind fields are subsequently used to drive a surge response model, providing design relevant extreme water level estimates derived from dynamically plausible storm scenarios rather than marginal extrapolation alone. The large spatial extent and strongly asymmetrical structure of ECs, compared with the more compact and idealised structure of tropical cyclones, introduce additional challenges for statistical modelling, particularly in representing spatial dependence and event organisation; these are addressed through a hybrid approach utilising clustering analysis. While the breadth of the modelling chain and the number of interacting components precluded exhaustive fine tuning of individual model elements, no comparable end to end stochastic EC tracking and hazard response framework was identified in the literature at the outset of this project. The present work therefore introduces a novel first step into an important and challenging domain, demonstrating the viability of an integrated approach and providing a foundation for future refinement and application in extreme value oriented hazard and design studies.
| Item Type: | Conference or Workshop Item (Poster) |
|---|---|
| Subjects: | Coasts > General |
| Depositing User: | Helen Stevenson |
| Date Deposited: | 11 May 2026 10:47 |
| Last Modified: | 11 May 2026 10:47 |
| URI: | http://eprints.hrwallingford.com/id/eprint/1734 |
Actions (for site administrators only - login required)
 |
View Item |
