New Method Predicts Monster Waves

The seven-story freak wave that slammed into the cruise ship Norwegian Dawn last month was not so freakish after all. Rogue waves are more common than most people realize, and scientists are starting to predict when and where they will strike.

Government wave forecasts generally are no more accurate than weather predictions. Wave forecasts made by the United States deal with data grid points that are 15.5 miles apart, which misses the fine points crucial to boaters.

Image Gallery


Monster Waves

Now, Vijay Panchang at Texas A&M University at Galveston and his associates say they can accurately predict the daily height of waves anywhere off the coast of the United States for the next 48 hours across spaces as close as 500 yards apart.

People who live in coastal Maine already use his forecasts, where Panchang proved his point by comparing his model output with measurements made by buoys. His predictions frequently show waves as high as 30 feet, even in close-in coastal waters.

"Everyone talks about the tsunami in the Indian Ocean in December of last year, but no one even realizes how big (non-tsunami) waves can be," Panchang told LiveScience.

70-feet of water

The Norwegian Dawn, a 965-foot ocean liner , was sailing back to New York from the Bahamas on April 16 when it was struck by a storm that pounded the vessel with heavy seas and the rogue 70-foot wave. The wave smashed windows and sent furniture flying.

In a separate event, a buoy off the coast of Alabama recently recorded an average wave height of 16 meters before the gauge broke, Panchang said. Since that figure is just an average of measurement of a sea-state, the biggest wave at that location was probably twice that size--32 meters, or about 100 feet.

"There were oil platforms destroyed," said Panchang, who requires his wave mechanics students to read The Perfect Storm. "The sheer magnitude of these things amazes me."

Panchang also is developing a similar wave model prediction system for the Prince William Sound Oil Recovery Institute in the Alaska port of Valdez, site of the Exxon Valdez oil spill. That wave model should be online by next year. The website for the National Weather Service station for Houston and Galveston links to wave forecasts by Panchang's team for the Gulf of Mexico.

Panchang and his colleague Dongcheng Li will present the Maine modeling results this month at the annual meeting of the American Society of Coastal Engineers.

Rogue waves are real

In the past, some scientists wrote off "rogue waves" as rare or even mythology. However, new satellite data collected by the European Space Agency's ERS satellites has confirmed what too many ship captains have come to know. Ocean waves that rise as tall as ten-story apartment buildings are a leading cause of large ship sinkings.

The giant waves form when strong winds beat against an opposing ocean current, when waves from different storms join forces, or when swells interact in strange ways with a particular seafloor.

Severe weather has sunk more than 200 supertankers and container ships more than 200 yards long in the past two decades, an ESA analysis found.

But some past statistical work in the past showed that rogue waves could only occur every 10,000 years. So many ships and offshore platforms are built to withstand maximum wave heights of only about 50 feet.

"Two large ships sink every week on average, but the cause is never studied to the same detail as an air crash," said Wolfgang Rosenthal, a senior scientist with the GKSS Forschungszentrum research center in Germany. "It simply gets put down to 'bad weather.'"

Counting waves from space

The two ERS satellites equipped with radar were launched in 1991 and 1995 to carry out a global rogue wave census and arrive at the truth.

Without aerial, cloud-penetrating radar, scientists could only go on radar data on waves collected from oil platforms.

The radar instruments on the satellites detected the height of individual waves at the surface in 3-mile by 6-mile patches of the sea. Three weeks of data, including 30,000 of these patches or "imagettes" of the sea with their wave height information were analyzed and searched for extreme waves at the German Aerospace Center.

A scientific team counted more than ten individual giant waves around the globe more than 75 feet high during the three-week period.

Eventually, the ERS data will be used to create a worldwide atlas of rogue wave events for efforts to eventually generate forecasts like Panchang's, only globally.

Surf's up!

In coastal Maine, surface waves are the biggest energy threat to the state's 3,000-mile coastline. Waves affect those on land too, moving sediment around in ways that pulls the soil right out from under buildings, endangers bridges and other coastal structures, and changes shipping lanes and harbor shapes overnight.

To make his predictions, Panchang starts with 5-day-ahead national wave forecasts from the National Oceanic and Atmospheric Administration which give a very coarse picture of what waves will do and how high they will get. This information cannot capture the intricacies of coastal Maine.

So Panchang then forces a mathematical model, factoring in local detailed sea depth and wind speeds and directions, to come up with more detailed forecast that is far more useful to locals.

In the future, Panchang and his Texas A&M colleague Shreenivas Londhe hope to show that "learning" computers called neural networks, a form of artificial intelligence, can also do the job with help from buoys in the ocean that have been collecting wave height information at particular locations for decades.

These arrays of processors imitate the way the human brain works. The researchers' networks crank through all the data for previous years, hour by hour, to find the most likely wave pattern to follow conditions just like the current conditions.

In a test of wave height predictions for the past month in coastal Massachusetts, Galveston, and Dauphin Island off Alabama, the network "brain" has accurately forecast what buoys in those locations later recorded. This approach is useful in small, complex port areas where the mathematical approach works less well.

As for who uses his forecasts, Panchang recently received an email out of the blue from some surfers in Maine. "I have lived in Maine and I didn't even know there were surfers in Maine because the water is so cold," he said. "They said, 'There are so few opportunities for us to go surfing. So before we do that, we always look at your website and have been using your forecasts and it always works out well.'"

Panchang's wave predictions are online here.

Related stories