Advancing Coastal Safety: HPC-based Early Warning System in Gulf of Naples

Early Warning Systems (EWSs) have been used to provide timely warnings about potential coastal hazards, such as tsunamis, storm surges, and other natural disasters. These systems are crucial in regions prone to extreme events, where early detection and timely warnings can save lives and reduce property damage. In the framework of climate change, the more frequent and intense storm surges make the existing coastal defense unsuitable for their purpose. In this context, the EWSs can be useful to enable rapid response toward increasing sea storm events and minimize their coastal impact by alerting local authorities and residents. In this work, a prototype EWS, called Shoreline Alert Model (SAM), was designed and developed to be a useful tool for hazard mitigation in the Gulf of Naples (Campania, Italy). SAM is a Python-based software with a parallelization scheme on heterogeneous architectures. This approach allows a drastic reduction in system processing time making it useful for real-time applications. Focusing on an 8-km vulnerable coastline, SAM processes static and dynamic meteo-marine and geomorphological input data and compares the output values with predefined thresholds to give an alert index. The model was validated by comparing the alert index with the effective damage suffered by the coastal sector under consideration