Developing an automated tide and surge measurement system in coastal regions using deep learning techniques

The most recent advances in machine learning and deep learning methodologies are generating great interest in a variety of research fields, including environmental studies. By facilitating automated and remote collection of data, these new tools have changed approaches to measuring maritime features. This research focuses on the development of a deep learning model to automatically measure tide and surge, with the goal of achieving precise results through the analysis of security camera data. The deep learning model was used to predict tide and storm surge from surveillance cameras strategically placed in two different coastal locations namely Santa Lucia in southeastern Sicily and Lignano Sabbiadoro in Friuli Venezia Giulia, Italy, adopting the Inception v3 architecture. The deep learning model uses categorization methods to assign a water level value to a given frame. This approach is especially useful in situations where typical tidal sensors are inaccessible or too far away from measurement stations, such as during extreme events requiring precise surge observations. The dataset employed to train and validate the deep learning model encompasses the full range of tide values observed in the study areas. The accuracy of the system was evaluated by comparing the predictions generated by the deep learning model with the corresponding tide gauge values. The experiments conducted clearly show that the model is quite effective in measuring tide and surge remotely, achieving an accuracy of over 90% and keeping the loss value below 1 for the deep learning model. These results highlight the model’s ability to address the absence of data collection in difficult coastal environments, providing valuable information for coastal management and hazard assessment. This research significantly contributes to the expanding realm of remote sensing and machine learning applications in environmental monitoring, enhancing understanding and decision-making in coastal regions.