Toward a large-scale coastal storm surge model for tropical cyclones – ICCE 2026

This record contains the slides of the oral presentation given at the 39th International Conference on Coastal Engineering (ICCE 2026), held at the Galveston Island Convention Center, Galveston, Texas, USA, on 17–22 May 2026. The presentation starts from the uncertainty in future tropical cyclone (TC) risk, comparing synthetic TC models (STORM and CHAZ), climate scenarios, exposure and adaptive capacity, and motivates the need for fast, large-scale storm surge tools that can run large ensembles of synthetic TCs. It then presents a Graph Convolutional Network (GCN) model that predicts coastal storm surge peaks induced by TCs in the North Atlantic. The data preprocessing is described, including TC tracks well represented in ERA5, GTSM storm surge simulations (1980–2022), calibration against GESLA tide gauges, and a hazard tracker with event augmentation. This work is part of the TRANSCLIMA project (Detection and Global Risk Assessment of Coastal Transitional Climate and Neo-Climate Regions in a Warming World), funded by the European Union's Horizon Europe research and innovation programme under the Marie Skłodowska-Curie grant agreement No. 101111262 (HORIZON-MSCA-2022-PF-01).