Skip to main content

Understanding the science behind tsunamis

While experts know it’s a question of when, not if, they differ whether the U.S. East Coast is threatened

While experts know it’s a question of when, not if, they differ whether the U.S. East Coast is threatened

Underlying the mud and the grief, the leveled villages, smashed fishing boats and lives lost to the Indian Ocean tsunami in December is — for those of us who go to sea — one sobering scientific certainty. The next big tsunami is on its way, and in its wake will come more death, debris and sorrow.

Where will it be?


Pick an ocean. The potential exists wherever there are deep seas in which tsunamis can be born and across which they can race, even thousands of miles. No shoreline, from the poverty of the third world to prosperous Manhattan, can be said to be entirely safe.

How will it happen?

Another tsunami most likely will result from an earthquake like the one off Sumatra the day after Christmas, which registered 9.0 on the Richter scale. Or it could possibly follow the eruption of an underwater volcano, the triggering of a catastrophic submarine landslide, or the impact of a giant meteor plummeting into an ocean.

When will it come?

Even the best-informed scientists don’t know. Which isn’t to say tsunamis are a total mystery. Scientific research can draw fairly accurate pictures of the sources of tsunamis and their method of traveling across the seas. There is even the technology to broadcast warnings to far-flung lands within minutes of a tsunami’s birth. An early-warning system has been in place for decades to notify those countries bordering the tsunami-prone Pacific when a geological event has occurred that could set a wall of water moving as fast as a commercial jetliner. However …

“There isn’t an early warning system in the Atlantic, and there wasn’t one in the Indian Ocean,” notes Dr. Jeffrey K. Weissel, an expert in submarine landslides at the Lamont-Doherty Earth Observatory of Columbia University in New York. “The only reason we have it in the Pacific is the United States, Japan and Canada have a lot of money.”

Many of the more than 200,000 lives lost in December in a dozen nations from Indonesia to Madagascar might not have been saved by a multi-million-dollar alarm system. The earthquake that created the Indian Ocean tsunami was close to low-lying, highly populated islands and was gargantuan — large enough to stall the earth’s rotation, shortening the day by 2.676 microseconds. But the event that triggered what many may have thought was a tidal wave was just the thing that the Pacific Tsunami Warning Center in Hawaii was established to detect and broadcast.

“The initial objective of PTWC is to detect, locate and determine the seismic parameters of potentially tsunamigenic earthquakes occurring in the Pacific Basin or its immediate margins,” the center explains on its Web site ( “To accomplish this, it continuously receives seismographic data from more than 150 stations around the Pacific through cooperative data exchanges” with international agencies. “If the earthquake location, depth and magnitude criteria needed to generate a tsunami are met, a tsunami warning is issued to warn of an imminent tsunami hazard” in areas that the tsunami could reach within a few hours.

The December tsunami sped west from Sumatra nearly 3,000 miles over the Indian Ocean to reach Somalia in Africa only hours later. For most in its path, there was no warning other than the historical evidence that the potential source of a great wave was lurking beneath the sea. In this case, that source would be an earthquake.

Earthquake tsunamis

Earthquakes, both on dry land and beneath the ocean, are the result of the same geological facts. Earth has a solid rock crust 70 to 150 miles thick, which rides on top of a deeper layer made of partially melted rock. The crust is fractured like a jigsaw puzzle into segments geologists call plates. The potential for earthquakes exists where two plates abut.

“The earth’s interior is slowly turning over,” says Steven Ward, a tsunami expert at the University of California Santa Cruz and a scientist who isn’t opposed to mixing metaphors. “The common analogy is a conveyor belt. Where the conveyor belt comes up is the ocean ridges; where it goes down is the subduction zone [the meeting point of two plates]. The conveyor belt [the molten earth] is pulling the plates along.”

Another way to think of this interaction between the plates and the hot earth below them, Ward says, is to envision hot air coming out of a floor grate in a room. The hot air rises to the ceiling and, unable to climb farther, it moves along the ceiling. Under the earth’s crust, it is this movement of magma that attempts to drag the plate floating above it.

Since all the earth’s plates are bordered by other, immobile plates, when one is pushed against the other pressure builds until something has to give.

On Dec. 26, the India Plate was engaged with the Burma Plate like a heavyweight wrestler who has gained a significant advantage by diving low on a lightweight. The two had been relatively immobile for some time, but then the India Plate, reacting to a weakness in the Burma plate to its east, made its move. It suddenly thrust even lower, causing the edge of the Burma plate to rise, perhaps a meter or more according to Daniel Cox, director of the O.H. Hinsdale Wave Research Laboratory at Oregon State University in Corvallis.

If you were in a boat or a ship directly above the event, you felt almost nothing, certainly not enough to be even mildly concerned. What is a 3-foot wave in the midst of the ocean?

But initially, this was not even a wave, according to Cox. What actually happened 150 miles off Sumatra — and what happens whenever an earthquake creates a tsunami — is that a massive slab of the Burma Plate 600 miles long and 60 miles wide rose that 3 feet at once. Above the seabed the ocean was about four miles deep, which means that hundreds of cubic miles of ocean were, in a very short span, lifted on the back of the Burma Plate. Just as quickly as the water rose, creating a huge but gentle mound, gravity began its work, and the water tried to slide off the mound, Cox says.

“Water is incompressible,” Cox continues. “If you had a bottom-mounted pressure transducer, it could detect a tsunami [at a great distance] because the tsunami is felt throughout the water column, even in the deepest parts of the ocean, in contrast to a hurricane wave. If you went down several hundred meters, you wouldn’t know the [hurricane] wave exists.”

(Tsunamis are not “tidal waves” because they are in no way caused by tides. Tides are caused by the gravitational pull of the moon and sun. Most ocean waves are the result of wind pushing against the surface of water. Storm surge is caused by the wind shoving up a pile of water that rides ashore on top of a high tide.)

When the sea begins to slide off of the submarine mound created by an earthquake, the whole column of water, from the seabed to the surface, is moving. And it is pushing on the sea beside it.

When the leading edge of this “wave” has traveled a great distance — several hundred kilometers, Cox says — another forms. There are mathematical formulae to calculate the length of tsunami waves.

“The wave is moving about 500 miles an hour in the open ocean, similar to the speed of a 747,” Cox says. “As it travels, you would have warning times that are similar to how long it would take to fly across the ocean.”

But then the tsunami slows. “When the water depth decreases, the wave length also decreases and therefore the speed of the wave decreases,” Cox says.

But while the wave is slowing, its energy remains the same, he says. “The energy is getting concentrated into a smaller area, and that will increase [the wave] height. When it’s less than 1 meter in the open ocean, it can be tens of meters at the coast. Claims of a 40- or 50-foot wave in the Indian Ocean may not be exaggerated.”

Most of the worlds’ tsunamis have resulted from earthquakes, according to the International Tsunami Information Center in Hawaii. The most recent devastating tsunami created by an earthquake was in 1960. That earthquake was in Chile, and its waves caused death and destruction in Hawaii and Japan. But in recorded history alone, there are many more examples of cataclysms.

A September 1992 earthquake off Nicaragua caused little damage there, but within an hour a tsunami with average waves of 13 feet and some that reached 35 feet struck the coast, killing inhabitants who were caught by surprise.

An early morning earthquake that reached 7.4 on the Richter scale shook an area in Alaska’s Aleutian Islands in April 1946, raising a large section of the sea floor. The Alaskan mainland was protected by the Aleutians, but one island was struck by a 100-foot wave.

That tsunami spread south and west. Later that day in Hawaii, a group of children watched in amazement as the ocean withdrew from the shore near their school in the isolated village of Laupahoehoe on the Big Island. Some of them walked out on the newly exposed seabed, only to have a 55-foot wall of water come racing toward them and their school, located on a level peninsula at the base of a sheer palisade rising directly behind them. Sixteen children, five teachers and four others were trapped there and killed when they were dragged back out to sea. The same wave smashed into the city of Hilo about 20 miles to the south, killing more than 100. Farther west, 30-foot waves that had started in Alaska came ashore in the Marquesas Islands.

Other types of tsunamis

While earthquakes trigger about 80 percent of all tsunamis, there are other causes. In recorded history, two of the most important are submarine volcanic eruptions and landslides. The third — the impact of a meteor falling into an ocean — has happened only in unrecorded history and is least understood.

The U.S. Geological Survey points to the tsunami that followed the 1883 eruption of the Krakatau volcano as a prime example of that type of wave. That also happened near Sumatra, where the South Asia tsunami originated.

“A series of tsunamis washed away 165 coastal villages on Java and Sumatra, killing 36,000 people,” according to information on the USGS Cascades Volcano Observatory Web site. “The larger tsunamis were recorded by tide gauges as far away as the southern coast of the Arabian Peninsula, more than 7,000 kilometers from Krakatau.”

Landslide tsunamis

There is little chance of volcanic or earthquake tsunamis affecting the North Atlantic, according to geologists. The type of plate movement that is common in the Pacific and Indian oceans is uncommon beneath the Atlantic, and there are no submarine volcanoes that threaten to generate tsunamis, scientists say.

Still, conditions exist that could create significant submarine landslides, according to some scientists, who say those events could trigger devastating tsunamis. Researchers at the Woods Hole (Mass.) Oceanographic Institute, Columbia University, and the University of Texas, who had looked at updated underwater mapping of the U.S. East Coast, speculated in a scientific paper in 2000 that things might be happening 150 miles off the mouth of Chesapeake Bay that could result in a huge landslide and tsunami.

“We discovered some curious features,” says Columbia’s Weissel, a landslide expert, “a series of [interlinked] cracks along the outer [continental] shelf, the shelf edge. We thought these might be the beginnings of slump, a submarine landslide, in the area just north of a submarine landslide that occurred during the last ice age.”

Weissel says the ancient landslide was about 20 miles across and 15 miles deep, dipping down to the ocean basin. “A slide that size would pose a significant tsunami hazard to the adjacent coastline,” Weissel says. “These cracks if you measured along their length were about the same size.”

Following the publication of the paper by Weissel and his colleagues, Ward at the University of California Santa Cruz created a model for a tsunami generated by a landslide.

“[He] found that indeed you would get significant tsunami waves generated eastward across the Atlantic toward the coastline from Cape Hatteras to Cape Cod, a few meters high,” Weissel says.

Weissel and his colleagues returned to the Atlantic last summer. With a robotic submarine packed with instruments and set to cross the ocean floor just above the seabed, they prepared to test their theory. The scientists had thought the cracks they had seen in the mapping might have been formed by escaping methane gas. They suspected that the gas had been trapped under a thick layer of mud when the sea was 120 feet lower, during the last ice age.

“What the [instruments] showed us was that the whole area of that outer shelf was underlain by accumulations of natural gas, trapped by what looked like a delta that had formed … in the last ice age,” says Weissel.

There are two points of view on how this affects the likelihood of a submarine landslide, he says. The first is that shallow gas below the sea floor poses a hazard because the presence of gas essentially destabilizes the overlying sediment and makes it more susceptible to sliding, says Weissel. He says the opposing view is that this release of pressured gas through these cracks actually makes the area less susceptible to a landslide.

“What we found is that there is a lot of area on the outer shelf that hadn’t blown out, and the gas is still there,” says Weissel. He says they don’t know the answer to the “riddle” of whether there is a landslide and tsunami threat.

Both recorded and geologically interpreted ancient history shows what could happen if this landslide let go. “The biggest one by far — and I’m glad we weren’t around for it — is the one that occurred 7,000 or 8,000 years ago on the margin of western Norway,” Weissel says. “The crest of the landslide measures something like 200 miles long. A 200-mile section of the outer shelf of the Norwegian margin gave way.”

Geologists have found mixtures of shell debris and sand on top of upland sediments, like peat bogs, suggesting that the wave that started in Norway came ashore in Scotland, Iceland and Greenland, and was 20 feet high, Weissel says.

Seventy-five years ago, a tsunami was generated three weeks after the stock market crash that triggered the Great Depression. On Nov. 18, 1929, an earthquake shook the seabed about 150 miles south of Newfoundland. The quake registered 7.2 on the Richter scale, not enough by itself to generate a tsunami but enough to set off an enormous submarine landslide. About 200 cubic kilometers of seabed began sliding down the Laurentian Slope, at first drawing the surface of the sea above it down while pushing a wave ahead of it to the south.

Then a new wave heading north filled the hole that had been created above the slump and raced toward the Newfoundland coast, where villages were knit together at the foot of mountains. Residents had no place to go when the series of 6- to 21-foot-high waves approached. On that sparsely populated coast, 29 people died.

At the same time, the debris of the landslide was racing down the ocean slope into the abyss. With nothing to stop it, the landslide picked up speed. Moreover, the fact that it was under water made it faster than a mountain landslide. “They hydroplane a bit,” Weissel says.

The landslide also severed a dozen trans-Atlantic cables hundreds of miles from its source.

Canary threat

Following the Indian Ocean tsunami, the potential landslide that got the most attention is one that would begin above ground, on the island of La Palma in the Canary Islands off northwest Africa.

The Canaries are a jumping-off point for many cruisers sailing from Europe to the Caribbean. The islands rely heavily on tourism for economic survival, and suggestions that their beautiful mountains are dangerous aren’t exactly welcome.

But that is exactly what some scientists, including Ward, have said. Some time — five, 50 or 5,000 years from now, no one knows — one side of the volcano on La Palma will fall into the ocean, and a tsunami that could hit Manhattan with a huge wall of water will be created, Ward predicts.

“La Palma itself is a volcano that collapsed a half-million years ago,” he explains. “It has re-formed itself. We think it’s getting toward the end of its cycle.”

When the volcano erupted in 1971, a crack was created along its top. The crack is 2 miles long, 10 or 15 feet wide, and 20 feet deep, Ward says. “It’s a potentially ominous sign,” he says. “The thinking is that the crack descends into the mountain and then slopes off like a children’s slide. We’ve modeled the collapse of 100 cubic miles. Every journalist tries to find his own analog — ‘half the size of Rhode Island.’ We tried to model what happened last time.”

Half a million years ago, a similar block fell off La Palma, sliding 2 to 3 miles deep into the ocean and coming to rest 40 miles away. “It goes pretty fast,” Ward says. “It picks up speed to maybe 200 mph.”

This landslide and the following tsunami could be triggered by an earthquake, an eruption, or even heavy rainfall that saturates the mountain, Ward says. “When things get too steep, it doesn’t take much,” he says. The result, on far shores, would be waves of 10, 15 or 20 meters in height, according to Ward’s computer model. “If you shove 100 cubic miles of material into the water, you’ve got to move 100 cubic miles of water,” Ward says.

When Ward and other scientists reported their work in a scientific journal in 2001, there was immediate media coverage. Ward says the reporting was appropriate.

Others disagree. A group called the Tsunami Society, headquartered in Hawaii, discredits Ward’s prediction on its Web site ( “We would like to halt the scaremongering from these unfounded reports,” the society says. “While the active volcano of Cumbre Vieja on Las [sic] Palma is expected to erupt again, it will not send a large part of the island into the ocean, though small landslides may occur. … No such event — a mega tsunami — has occurred in either the Atlantic or Pacific oceans in recorded history. NONE.”

“La Palma is just one of a number of cases where people use numerical models that are inadequate,” says Dr. Charles L. Mader, a retired scientist from the Los Alamos National Laboratory and a member of the Tsunami Society. “Whether they are formed from a landslide like we had in the Aleutians or volcanic eruptions like Krakatau, they have short wavelengths. These things are serious threats nearby, but they will not propagate across ocean basins. The models that [predict] them across ocean basins are lacking essential physics.”

Ward defends his work. “I wrote a scientific paper that goes through a journal and [peer review],” he says. His critics have not, Ward says. “They are welcome to write a scientific paper that shows how it’s wrong.”

Until then, he stands by his calculations and waits for the tsunami that will hit the East Coast — someday.