Enormous waves may suddenly emerge, posing a threat to boats and oil rigs. Thanks to a machine learning algorithm leveraging centuries of data, the prediction of these rogue waves has become possible, shedding light on their formation process.
Rogue waves, which need only to be twice the height of neighboring waves, stand out for their unpredictability. Unlike catastrophic waves like tsunamis triggered by seismic events, rogue waves result from the complex interplay of currents and winds.
Although rogue waves can be perilous, they are not to be underestimated. Their random appearance can endanger maritime vessels and offshore structures. While previous research has explored methods to forecast these waves with limited lead time, a comprehensive new framework promises a more extensive early warning system.
Researchers at the University of Copenhagen kickstarted the process by inputting vast amounts of wave data into an AI model. This dataset encompassed over 700 years of information and tracked more than a billion waves recorded by buoys across 158 global locations, operating round the clock.
Johannes Gemmrich, a key figure in the study, emphasized the frequency of irregular waves in their analysis. Surprisingly, their database contained around 100,000 waves meeting the criteria of rogue waves, indicating a daily occurrence of such significant waves in specific oceanic regions.
Contrary to prior assumptions, the team’s investigation pinpointed linear superposition as the primary mechanism behind rogue wave formation. Dion Häfner, the lead author, explained that the convergence of two wave systems amplifies the likelihood of creating large crests and troughs, culminating in the emergence of massive waves.
By harnessing real-time buoy data, the team’s innovative algorithm can identify the convergence of risk factors for rogue waves in oceanic areas. This breakthrough enables shipping companies to proactively plan alternative routes based on this predictive data, mitigating potential risks effectively.
