Oldenburg. Extreme waves - mountains of water up to 30 metres high - repeatedly cause serious shipping accidents. However, it is still unclear how they are created. A team of scientists has now unravelled part of the mystery: the researchers have developed a method for the short-term prediction of extreme waves.

Their results were recently published in the "New Journal of Physics". The main author of the specialist article is Oldenburg doctoral candidate Ali Hadjihosseini, with Dr Matthias Wächter and Prof Dr Joachim Peinke also involved. The Volkswagen Foundation is providing financial support for the joint project coordinated by the University of Hamburg.

Extreme waves - not to be confused with tsunamis - are individual waves that occur very suddenly, often on the open sea. They are at least twice as large as the waves surrounding them and develop enormous forces on impact. Smaller ships can sink, while larger ones are at risk of being unable to manoeuvre. Experts estimate that up to ten shipwrecks a year can be attributed to these wave monsters. For a long time, extreme waves were dismissed as a sailor's yarn, but since satellite images proved their existence in the early 1990s, they have increasingly aroused the interest of scientists.

Attempts to predict extreme waves have often failed due to the problem that the phenomenon cannot be described using linear mathematics. The Oldenburg scientists have now tried a different approach: they have developed a statistical method that can be used to predict extreme waves - known as multipoint statistics.

Specifically, the experts recorded wave heights at various measuring points at different times, mainly using buoys. In order to be able to analyse the wealth of measurement data, they used a trick from probability theory known as the Markov process. The main effect: even if only part of the history of an event is known, it is possible to predict its development just as well as when analysing the entire course of events. This trick enabled the researchers to simplify their measurement data without having to sacrifice important information.

"That was the decisive step," recalls Wächter. He assumes that his research will one day ensure greater safety on the high seas. So far, however, the forecast for individual waves has only been in the range of a few seconds - far too short to be able to react as a sailor. "However, the general risk of extreme waves can already be determined well," continues Wächter. In addition, multipoint statistics can also be used to predict other phenomena, such as gusts of wind. Research into this is already underway at the Institute of Physics.