The Draupner Wave

Summary

On 1 January 1995 a downward-pointing laser on the Draupner E platform in the North Sea measured a wave 25.6 metres from trough to crest, in a sea whose significant wave height was about 12 metres. It was one of the first confirmed instrument records of a rogue wave, and it showed that waves far larger than the statistics of the day allowed were real.

Statistics of the time put a wave like it at roughly once in 10,000 years. Within a decade a satellite radar survey had found more than ten waves of that size in three weeks of data from the world's oceans, and in 2018 researchers using the University of Edinburgh's FloWave basin reproduced the Draupner wave at scale by crossing two groups of waves.

The wave is no longer a puzzle about whether such things exist. The remaining argument is about exactly which conditions produced this one.

What is documented

The measurement. The Draupner E platform stands in about 70 metres of water in the Norwegian sector of the North Sea. On 1 January 1995 a laser mounted on the platform and pointed down at the sea surface recorded the height of the water beneath it. One wave in the record stood 25.6 metres from trough to crest, with its crest 18.5 metres above the mean sea level. The significant wave height of the sea state, a standard measure of the typical larger waves, was about 12 metres.

Why it stood out. Its height was more than twice the significant wave height, and its crest about one and a half times it. Rogue waves are defined by that kind of disproportion: in the words of Mark McAllister and colleagues, by "their unexpectedly large size relative to the population of smaller waves in which they occur". A large wave in a large sea is expected. One more than twice the significant wave height was not, under the statistics then in use.

What the statistics predicted. In 2004 the European Space Agency summarised the statistics then used for ocean waves as indicating that such deviations "should occur only once every 10000 years".

The search that followed. In December 2000 the European Union began the MaxWave project on extreme waves. Using radar imagery from ESA's ERS-1 and ERS-2 satellites, the project identified more than ten individual waves above 25 metres around the globe in three weeks of data, as ESA reported in July 2004.

Why this record mattered. Other large waves had been reported, and some measured, before 1995; McAllister and colleagues describe Draupner as "one of the first confirmed field measurements of a freak wave". What set it apart was the quality of the evidence: an instrument on a fixed platform measuring the sea surface directly, and a well-described sea state around the wave. That made it the reference case for a generation of theory, numerical models and experiments.

The crossing-sea analysis. In 2011 Thomas Adcock, Paul Taylor and colleagues re-analysed the record in the Proceedings of the Royal Society A. From the second-order difference waves in the measurement they concluded that the wave "might have resulted from two wave-groups crossing", with mean directions about 90 degrees or more apart, and confirmed with nonlinear numerical simulations that such a crossing could produce it. A hindcast by the European Centre for Medium-Range Weather Forecasts showed swell running at about 80 degrees to the wind sea.

The laboratory recreation. McAllister, Samuel Draycott, Adcock, Taylor and Ton van den Bremer tested the idea at the FloWave Ocean Energy Research Facility at the University of Edinburgh, a circular basin 25 metres across, 2 metres deep and ringed by 168 wavemakers. Working at a scale of 1 to 35, they made two wave groups cross at angles of 0, 60 and 120 degrees. At full-scale equivalent, breaking capped the crest at 17.2 metres when the groups travelled together, and reached 18.4 metres at 60 degrees. At 120 degrees the crest reached 18.9 metres, matching Draupner, and the water broke as upward jets rather than in the usual way. They concluded that "breaking becomes less crest-amplitude limiting for sufficiently large crossing angles". The paper was published online in the Journal of Fluid Mechanics on 11 December 2018.

The field since. Rogue-wave research is active. In 2023 Dion Häfner, Johannes Gemmrich and Markus Jochum trained a neural network on wave-buoy observations and distilled it into an equation for rogue-wave probability that, in their tests, predicted unseen data better than existing theory. In 2025 the journal Chaos published "Rogue waves: Theory, methods, and applications, 30 years after the Draupner wave".

Leading explanations

A natural extreme of the sea state (established). The Draupner wave is accepted as real and as a product of ordinary ocean wave physics, not of any unknown process. It has been reproduced from ordinary wave components in numerical simulations and, at scale, in a laboratory basin. What changed after 1995 was, above all, the estimate of how often such waves occur.

Two wave systems crossing at a large angle (proposed by Adcock and colleagues in 2011; shown to be possible in the laboratory in 2018; whether it applied on the day is still discussed). The analysis of the record and the hindcast swell direction support it, and the Edinburgh experiment showed that crossing at a large angle allows a Draupner-sized crest without the breaking that limits waves travelling together. The experiment shows that the mechanism works. It cannot show that the sea on 1 January 1995 was in fact crossing, and later studies, including one by Luigi Cavaleri and colleagues in 2016, have taken a fresh look at the sea state on the day.

Current status

Later explained. The Draupner wave is real, well measured, and within the reach of modern wave physics. It has been reproduced numerically and at scale in a laboratory basin, and it has been joined by many other recorded rogue waves. The registry keeps it because in 1995 it was genuinely anomalous against the statistics then in use, and because the explanation is the interesting part. The detailed mechanism of this particular wave remains a subject of research, as does the forecasting of rogue waves generally.

Last reviewed: September 2026. Records are provisional. Where the evidence changes, the entry changes. Found an error? Tell us.