Machine-guided discovery of a real-world rogue wave model.

Big data and large-scale machine learning have had a profound impact on science and engineering, particularly in fields focused on forecasting and prediction. Yet, it is still not clear how we can use the superior pattern-matching abilities of machine learning models for scientific discovery. This is because the goals of machine learning and science are generally not aligned.

Generation mechanism and prediction of an observed extreme rogue wave.

Rogue waves are individual ocean surface waves with crest height [Formula: see text] or trough-to-crest height H that are large compared to the significant wave height [Formula: see text] of the underlying sea state: [Formula: see text] or [Formula: see text]. The physics of rogue wave generation and the potential of predicting the rogue wave risk are open questions. Only a few rogue waves in high sea states have been observed directly, but they can pose a danger to marine operations, onshore and offshore structures, and beachgoers.

Real-world rogue wave probabilities.

Rogue waves are dangerous ocean waves at least twice as high as the surrounding waves. Despite an abundance of studies conducting simulations or wave tank experiments, there is so far no reliable forecast for them. In this study, we use data mining and interpretable machine learning to analyze large amounts of observational data instead (more than 1 billion waves).