Description and objectives

This project aims to develop an integrated approach combining numerical modeling, for determining exposure scenarios of coastal natural and anthropogenic components, with a conceptual modeling framework based on neural networks and statistical techniques. The activities are closely linked to the twin project (Thematic Area 3), which focuses on the study of wave motion in deep waters. The project activities are organized as follows.

The analysis will be applied to a pilot site suggested by the RETURN consortium. It will start from meteo-marine data on waves, wind, atmospheric pressure, and astronomical tides, provided as outputs of the twin project (Thematic Area 3). Additionally, topographic data related to both submerged and emerged beach areas will be considered.

Two hydrodynamic models will be employed. The first, the SWAN model, will simulate the propagation of wave parameters from the open sea to the coast. The second, the XBeach model, will be used to assess flooded areas under different scenarios associated with extreme meteo-marine conditions.

To determine coastal flooding without relying on computationally intensive numerical simulations—which are costly and incompatible with an effective early warning system—a conceptual modeling approach using artificial neural networks (ANNs) and statistical techniques will be developed.

Artificial Neural Networks (ANNs) are particularly useful for this problem, as they can map complex non-linear relationships between input and output variables, provided that a sufficient dataset of process realizations is available