SIX YEARS AFTER the China dropped her cargo into the sea, the European Space Agency spent three weeks studying winter waves via satellite. The research confirmed that waves of more than 100 feet are altogether real. There's evidence, in fact, that as the planet warms their numbers may be rising, and ESA scientists believe that their data may help solve many maritime mysteries.

One of the most famous such mysteries is the sinking of the MS München, in 1978. The München, at the time a year-old state-of-the-art barge carrier, vanished while crossing the North Atlantic. Its disappearance has never been explained, but many nautical detectives suspect that it and its crew of 27 fell victim to a monsterwelle.

A true monsterwelle is defined not by size but by its mathematical improbability, an example of chaos theory in action. The seafarer's handbook Heavy Weather Sailing notes that "whilst one wave in twenty-three is over twice the height of the average wave, and one in 1,175 is over three times the average height, only one in over 300,000 exceeds four times the average height." If the average waves in a particular sea are only three feet tall, and a 12-foot breaker were to suddenly rear up in their midst, it would be a little monster. What really makes the fearsome monsterwellen fearsome isn't size; it's the element of surprise. They can come out of nowhere, even in calm seas, overwhelming a ship before the helmsman has time to escape it.

Scientists are still trying to explain monster waves. One leading theory is that they arise when a strong current—like the Gulf Stream—collides with a countervailing storm swell, or when a deep-ocean swell collides with a shallow shelf. Another theory, called "constructive interference" or "random superposition," holds that in chaotic seas two wave trains with identical wave periods and crest heights can ever so rarely combine into a kind of super train, producing monsters. Which might explain why such waves often seem to come in sets of three, a phenomenon known to sailors as "the three sisters." Three sisters 40 or 50 feet tall can be more dangerous than a single 100-foot giant. The first monstrous sister rocks the boat mightily before plunging it into a trough so deep, one eyewitness account compares it to a "hole in the ocean." Then the second or third wave sinks it.

Nonetheless, sailors are often mistaken when they identify monstrously large waves as freaks or rogues. The tsunamis that struck land all around the Indian Ocean in 2004 were well over 100 feet, but they were the result of an enormous undersea earthquake, not chaotic hydrodynamics. Typhoons and hurricanes regularly whip up waves over 70 feet high, but evidence suggests that you're more likely to encounter a monsterwelle in lower sea states than in tall ones.

The peril of the sea that modern merchant mariners are most likely to face, navigational wisdom holds, isn't a monsterwelle but something known as "synchronous rolling," so called because the natural roll period of the ship falls in sync with that of the waves. On a 2007 scientific research trip in the Labrador Sea, I met a California oiler named Beau Gouig, who explained it to me this way: "Basically what happens is when the wave crests get far enough apart, the ship starts rolling, and with each roll it takes longer to recover, so you're going farther to starboard, and then farther to port."

A few years ago, Gouig had been on a container ship caught in a typhoon. "We were sailing from Japan to L.A.," he said. "I was up on the bridge. We were in autopilot, and we were getting hammered. We did three or four 50-degree rolls, buried the bow three times. Waves swept all the lifeboat ladders off deck. I said to the captain, ‘These wave crests are getting far apart.' That's basically why you're up there, to watch for synchronous rolling. Suddenly everything on the bridge just goes voom."

The ship's captain did exactly what all captains are trained to do: He hove to, immediately turning into the oncoming waves and slowing down, letting the ship ride up and over. "All the books say that you've got only two minutes to break that cycle. You've got to make a hard turn and get the bow up on a wave or down into a trough. Another 30 seconds and I think we would have rolled right over."

The crew were all amazed, Gouig said, that no containers had been lost. "Spills happen all the time," he said. Then he told me the story of what he believed to be the most famous case of synchronous rolling: the case of the APL China.