An advanced tool that can anticipate the emergence of abnormally large and unpredictable waves at sea — known as rogue waves — up to five minutes ahead is unveiled in a study published today (July 18) in the journal Scientific Reports. The authors propose that the tool could be utilized to issue early alerts to ships and offshore platforms to allow those onboard to seek refuge, execute emergency shutdowns, or navigate to mitigate the impacts of impending rogue waves.
Creation and Training of the Artificial Neural Network
The tool developed by Thomas Breunung and Balakumar Balachandran comprises an artificial neural network that has been educated to differentiate ocean waves that will be succeeded by rogue waves, from those that will not. The authors educated the artificial neural network using a dataset encompassing 14 million 30-minute-long samples of sea surface elevation measurements from 172 buoys situated near the coastlines of the continental United States and the Pacific Islands. They utilized their tool to anticipate the emergence of rogue waves using a distinct dataset comprising of 40,000 sea surface elevation measurements from the same buoys.
Predictive Performance and Possible Enhancements
The researchers observed that their tool could accurately forecast the onset of 75% of rogue waves one minute into the future and 73% of rogue waves five minutes into the future. The tool also managed to predict the occurrence of rogue waves near two buoys not encompassed within the datasets utilized in training with 75% precision one minute into the future. This underscores that the tool might be capable of projecting rogue waves at new locations.
Future Paths in Rogue Wave Prediction
The scientists propose that the precision and lead time of their tool’s predictions could be further refined by integrating water depth, wind speed, and wave location data. Subsequent research could also enable the heights of upcoming rogue waves or the times at which they might arise to be anticipated, they suggest.
Image Source: Jirsak / Shutterstock
