Author(s): Susanna Dazzi; Renato Vacondio; Matteo Pianforini

Linked Author(s): Susanna Dazzi

Keywords: Deep learning models; Flooding; Forecasting; GPU; Levee breach; Shallow water model

Abstract: In this work, the deep-learning FloodSformer model for inundation forecasting and the GPU-parallel hydrodynamic code PARFLOOD are applied to a real case study of breach-induced flooding, in order to compare the accuracy and computational performance of these alternative approaches for real-time forecasting.

DOI:

Year: 2026