From hindcast to forecast: A statistical framework for real-time coastal circulation bulletins in the Gulf of Genoa

Abstract

The increasing exploitation of coastal areas requires a detailed knowledge of environmental conditions that could be crucial for different applications, such as navigation safety and environmental protection. In this framework, this study introduces an offline forecast system for coastal circulation, being littoral currents among the main triggers of nearshore dynamics. The system utilizes clustering techniques for selecting relevant climatic scenarios based on historical metocean data, which are next used to simulate coastal circulation patterns through a robust modeling chain. The representative scenarios are then associated with meteocean forecast windows updated on a regular basis at the study area, i.e., the Gulf of Genoa in the Tyrrhenian Sea. Such operation takes advantage of tailored distance metrics, allowing to detect the closest scenario to the expected conditions, providing in turn real-time and refined predictions for coastal circulations. Such an offline approach ensures reliability, especially in coastal regions where accurate circulation forecasts are crucial for environmental management, marine navigation, and disaster preparedness.