This is the authors' abstract. We don't add key points for this paper.
This record contains the slides of the oral presentation given at the EGU General Assembly 2026 (EGU26), held in Vienna, Austria, and online on 3–8 May 2026, in session ITS4.36/NH13.11 (abstract EGU26-13339). The presentation introduces a deep learning framework based on Graph Convolutional Networks (GCNs) that predicts tropical cyclone (TC)-induced coastal storm surge in the North Atlantic basin, producing a coastal storm surge peak map for a TC in less than one second. This tool was trained with TC tracks well represented in ERA5 and storm surge simulations from GTSM for 1980–2022. Each TC is represented as a graph whose nodes are coastal stations and whose edges describe the connectivity of the coastline, with TC features (position, local wind and pressure, translation speed, direction relative to the coast and storm size) as inputs. Eighty percent of the TCs were used for training and 20% (47 events) for validation. The presentation also covers the preprocessing of the data, including calibration against GESLA tide gauges, and outlines next steps towards a global, multi-hazard, event-based tool. Abstract: Odériz, I., Losada, I. J. and Muis, S. (2026). CYCLONE: A superfast large-scale coastal storm surge model for Tropical Cyclones. EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13339. https://doi.org/10.5194/egusphere-egu26-13339 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).
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Odériz et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: