Meteo-HySEA

Meteo-HySEA is a GPU-based numerical model of the HySEA family specifically developed to simulate meteotsunamis, i.e., long-wave oceanic disturbances generated by atmospheric pressure perturbations and wind forcing. The model solves the 2D Shallow Water Equations in spherical coordinates, enabling accurate global-scale simulations. It incorporates spatially variable atmospheric pressure fields as external forcing, a non-constant Coriolis parameter—essential for long-lasting geophysical flows—and periodic boundary conditions to represent transoceanic tsunami propagation. Thanks to its GPU implementation, Meteo-HySEA efficiently simulates the full meteotsunami lifecycle, from generation and propagation to coastal amplification and inundation, thereby addressing one of the most critical aspects of hazard assessment: the onshore impact on communities and infrastructure.

The model has been validated against laboratory experiments reproducing Proudman resonance (e.g., Vilibić, 2008; Beisiegel & Behrens, 2022), as well as a real-world case study in the Gulf of Mexico in 2010 (Horrillo et al., 2022), using realistic bathymetric data and synthetic atmospheric pressure forcing. Furthermore, its performance in the Adriatic Sea has been benchmarked against the CPU-based AdriSC-ADCIRC system and evaluated against three well-documented meteotsunami events (2014, 2017, and 2020) at known hotspot locations: Vela Luka, Vrboska, and Stari Grad. These results using forecasted atmohspheric and real topobathymetric data show that Meteo-HySEA generally reproduces the timing, spatial patterns, and overall structure of observed events, and in some cases yields stronger responses than AdriSC-ADCIRC under identical forcing.