Evaluation of a clustering algorithm to enhance the nearshore wave prediction system

The continued growth of marine traffic is causing increased accidents close to ports. They may result from a variety of reasons, such as the weather. Accurate estimates of the wave climate can lower the risk of maritime accidents. The analysis of the weather forecast in port locations often uses numerical models. These models are not ideal for wave forecasting or nowcasting since they have a high computational cost. This constraint can be solved via artificial neural networks (ANNs). However, the ANNs that were already developed tended to focus on specific areas of the study region, such as piers and port entrances. More information on the wave climate over a vast area might help early warning systems. The main goal of this work is to investigate the clustering approaches that identify homogenous areas to develop a novel method for employing ANNs to evaluate nearshore wave characteristics in real-world scenarios. As a case study, this research focuses on the port of Augusta (SR), one of the most important Italian ports.