WIND AND WAVE TRAINED ARTIFICIAL NEURAL NETWORKS FOR THE FORECASTING OF WAVE CLIMATE IN HARBOUR AREA

Nowadays, maritime transportation has expanded rapidly, involving the need to enhance several navigation-related issues, particularly concerning the safety of navigation, which is significantly impacted by weather conditions. In this regard, creating a wave forecasting system could facilitate vessel movement at the harbour entrance or inside the sheltered area. Wave characteristics are usually estimated using numerical models, which generally require high computational costs, making them inadequate for nowcasting and forecasting wave climate. The current study describes the implementation of a forecasting methodology for the port area of Augusta (Sicily) based on an Artificial Neural Network (ANN) that attempts to deliver a trustworthy response and the numerical model but with a significant reduction in the computational time.