No one knew, because nothing like that had happened in living memory. The last widely devastating tsunami in the Indian Ocean was that spawned by the eruption of Krakatoa in 1883, which killed 36,000 people. Disaster officials in the region understandably have focused on cyclones, floods, and other natural calamities that strike the region every year.
Now, in the wake of one of the worst natural disasters in recorded history, their focus has broadened. Within days of the catastrophe, with the adage "better late than never" sitting heavily on everyone's minds, commentators worldwide were calling for a tsunami warning system in the Indian Ocean akin to a successful one now operating in the Pacific. If such a system had been up and running in the Indian Ocean, experts agree that many of the thousands of lives lost in places relatively distant from the quake's epicenter, such as Thailand, Sri Lanka, and India, might have been saved.
But setting up a truly effective warning system, one that can alert coastlines even as close to a tsunami's birthplace as northwest Sumatra was last December, is a daunting task. In talking to a host of experts on tsunami detection and hazard mitigation, it's clear that numerous challenges exist to ensuring that disaster officials are well-prepared for the next big tsunami—which, incidentally, could arise at any time off the coast of the U.S. as close as it was off Sumatra.
Towards a warning bell
Before installing a warning system, experts need to identify and map areas of risk. Much of this preliminary work has been done in the Indian Ocean; some needs more effort. The work includes pinpointing tectonic faults of concern, mapping the seafloor topography, studying the effects of past tsunamis, and making computer models showing how future tsunamis might behave.
"Computer models are the brain of any warning system," says Tad Murty, a tsunami expert at the University of Manitoba at Winnipeg who recently spent two months in his native India working with Indian counterparts. "We will run a series of scenarios, all permutations and combinations of anything that can happen, so that when an earthquake occurs, the computer will match the closest scenario and then use it to help with warning."
Murty estimates an Indian Ocean warning system will cost between $250 and $400 million. That will represent the combined expenditure for all 37 or so countries that have coastline on the Indian Ocean. While the Indian system will have major differences from its Pacific counterpart in its computer models, he says, it will likely be similar to the Pacific setup in instrumentation.
The Pacific system was established two years after a 1946 quake off Alaska spawned a tsunami that killed over 100 people and caused massive destruction around the Pacific. Initially the system was based on seismic data plus inputs from a series of tide gauges that sent information about water levels around the ocean to warning centers in Hawaii, Japan, and elsewhere in the Pacific. Experts used the system to accurately "call" every major tsunami in the years since, but wishing to err on the side of conservatism, they also had a false-alarm rate until relatively recently of 75 percent.
Then, in 2002, new sensors went into place to supplement the tide gauges. Six monitoring devices anchored to the seafloor far out in the Pacific measure changes in water pressure above. When such changes indicate a tsunami passing overhead, a sensor sends a signal with data to a buoy directly above, which forwards the message via satellite to the Pacific Tsunami Warning Center (PTWC) in Hawaii. Specialists then analyze the data, decide if a potentially dangerous tsunami is under way, and take appropriate action.
The new system has already proved its value. In 2003, experts were able to call off a warning after analysts determined that the tsunami generated by a big quake off Alaska was not hazardous. Within weeks of the Indian Ocean tsunami, President George W. Bush ordered that 32 additional such buoys be added to the Pacific as well as to the Atlantic and Caribbean (neither of which currently has a tsunami warning system, even though both have on occasion spawned deadly tsunamis).
The Pacific warning system is not perfect, however. For one thing, the seismic data currently tell quickly about a quake's epicenter, but not about the alignment of the rupture. If experts had known immediately that the Sumatran rupture had been along the north-south portion of the fault, they would have known that any resulting tsunami would have traveled largely east-west and could have predicted what targets lay in its path. "Right now, there's no quick way to know which way that rupture goes," says Chip McCreery, director of the PTWC. "We need to figure that out."
Perhaps more importantly, experts at the PTWC had no way of determining in short order how large the quake was. It took about two hours, based on calculations made at Harvard's Centroid-Moment Tensor Project, for PTWC officials to learn that it was a 9.0. (Some seismologists now believe the quake might have been a 9.3.) Every undersea quake of that magnitude has triggered a lethal tsunami, says Tom Heaton, a seismologist at the California Institute of Technology. "If these guys [at the PTWC] had known 10 minutes after the earthquake that they'd had a nine, I'll bet they'd have been on the phone to CNN worldwide saying 'Magnitude nine earthquake! There's an extreme emergency in the Indian Ocean!'"
"If you feel an earthquake, if you see the water retreating, or if you hear a loud roar from the ocean, head inland fast."
Heaton stresses that it was not the fault of the PTWC but rather what he believes is a certain lack of communication between the seismic and tsunami communities. But fixing this specific problem is straightforward, he believes: The capability exists for centers such as the PTWC to detect the size of earthquakes anywhere in the world within minutes after the event, at a cost of perhaps $100,000 and several months of work to upgrade software, he says.
Earthquakes smaller than 9.0 can also launch tsunamis, and for these analysts' skills are truly put to the test. All tsunamis are different, and determining their level of threat can be tricky. Analysts can be caught between a rock and a hard place. On the one hand, experts can't afford to be wrong—to fail to warn about a tsunami that ends up taking lives. On the other hand, evacuations are costly. In Hawaii, the price tag to stage an evacuation is over $50 million, says George Curtis, a tsunami expert at the University of Hawaii at Hilo. "You have a tremendous loss of business. You have a great expense of overtime for police, firemen, road crews putting up roadblocks. It's a major disruption." And if you have too many false alarms, a cry-wolf reaction occurs, and citizens grow complacent.
Even the finest warning system, with the best technology and personnel, will be useless without the means to communicate warnings to relevant authorities. Sadly, the 2004 tsunami showed up this lack all too well in the Indian Ocean. In Thailand, for example, seismologists knew early on how big the quake was and that it probably spawned a big tsunami. "But for an hour they didn't know what to do about it," says Curtis. "They didn't know who to call." In Hawaii and other parts of the Pacific system, such communication channels to civil defense authorities are in place and tested regularly.
Learn and live
The final piece of the puzzle is education. In potentially vulnerable parts of the Pacific region, from Japan to Hawaii to Alaska, vigorous public education programs have been in place for some time. One such hot zone is the coastline stretching from Washington, through Oregon, and into northern California. This strip would lie in the line of fire of any tsunami triggered by an eruption along the Cascadia subduction zone, which runs for nearly 700 miles along the Pacific Northwest coast. Any tsunami could hit nearby shores within 20 to 40 minutes.
"The educational program they have in place there is that you should have a backpack filled with all the stuff you need to be self-sufficient for a couple of days," says Eddie Bernard, director of the Seattle-based Pacific Marine Environmental Laboratory, which developed the buoy-based warning system and is a leading center of tsunami research. "Once you feel the earthquake, you grab that backpack on your way out of your house." Both authorities and residents have been indoctrinated with where to go and what to do. (Knowing when it's safe to return after a local quake is something still needing research, Bernard says, but as a benchmark, in the 2004 event tsunami waves pummeled the Sumatran coast for 12 hours following just five minutes of shaking.)
"Every step you go inland, even if it's not up, is a step toward more safety."
Time can be the enemy of effective education, however. "Humans are amazingly adaptive animals, so those impacted will remember this horrific tsunami," says Bernard. "It's our responsibility to devise effective educational programs to ensure that the next generation does not forget." McCreery, for one, says it's vital to get tsunami hazards into school curricula. In devastated parts of the Indian Ocean, local authorities could also erect memorials to victims that double as warning signs. "The memorial could explain what the hazard is, that it's going to happen again, and what to do," he says.
Knowing oneself what to do, rather than waiting to hear from officials, can mean the difference between surviving and not. Even the most efficient detection-and-warning system might not be able to warn areas extremely close to subduction zones. The tsunami ravaged Sumatra's Banda Aceh only a half hour after the quake, and any tsunamis generated by quakes off the Aleutian islands could hit nearby Alaskan communities within 15 minutes. (The U.S. maintains a separate warning system in that state that can warn regional locations based on seismic data alone.)
For residents of such coasts, and really for anyone who ever visits a beach on vacation, it's essential to know nature's three tsunami warning signs, Bernard says. "If you feel an earthquake, if you see the water retreating, or if you hear a loud roar from the ocean, those are three natural indicators that a tsunami could be present. In any of those cases, head inland fast. Every step you go inland, even if it's not up, is a step toward more safety."