On shifting ground: how much is enough when mitigating earthquake damage?

Earthquakes now endanger more people than ever. The world population has more than doubled in the past 50 years and, by 2007, half of the planet's 6.6 billion people will be living in urban centers. Because more than 380 major cities lie on or near unstable seams in the Earth's crust, one seismologist has come to a grim conclusion: A catastrophic temblor sufficient to kill 1 million people could occur, on average, once per century.

The prediction is based on tallies of recorded earthquake occurrences over hundreds of years. If quakes of similar size and distribution were to recur in today's demographic landscape--which features megacities such as Mexico City, home to 18 million people--the consequences would be disastrous.

Roger Bilham of the University of Colorado in Boulder presented his disturbing forecast in May at a meeting of the Seismological Society held in San Juan, Puerto Rico. But Bilham and many of his colleagues argue that the forecast need not come true: Even if the earthquakes can't be stopped, most of the deaths could be avoided.

"We seem to be content with reducing the risk," says Cinna Lomnitz, a seismologist at the National University of Mexico in University City. "We don't really think about abolishing the risk, which is what we should do."

Lomnitz claims that risk abolishment is a reasonable goal, attainable with a combination of quake-resistant construction, better urban planning, and emergency preparedness. But scientists and public officials around the world are debating which measures are wise. The cost of girding structures against earthquake damage can be daunting, and the potential benefits depend on the local likelihood of a quake.

In areas at high risk, "there's no question that one should make buildings earthquake-resistant," says Seth Stein, a seismologist at Northwestern University in Evanston, Ill. "The question is how much [you can afford], and how you trade it off with other possible uses of resources."

LOW-TECH PROTECTION That balance can be hard to strike for the people of developing nations. Brian Tucker, director of the nonprofit organization GeoHazards International (GHI), says that earthquakes are often the last thing struggling people want to think about. Every day in Pakistan, for example, people face problems such as malnutrition, diarrhea, and child prostitution. "We're not trying to force earthquake safety onto the table and push other things off the table," Tucker says. "We try to inform public officials of the risk and of the options they have to address it, and then we let them decide what they want to do."

Often, Tucker argues, inexpensive engineering solutions can go a long way toward preventing quake fatalities in developing countries. Elementary seismic-reinforcement techniques, such as constructing with ductile steel and properly weighted mixes of cement and sand or adding concrete reinforcement walls and steel bracing, can keep a building standing after an earthquake.

GHI teams are using a grassroots approach to teach these practices to masons in areas around the world. One of the most successful projects is located in a village in the Kathmandu Valley of Nepal, a country that sits on a turbulent seismic zone among the Himalayas.

With a budget of $5,000 and materials donated by local industry, village masons spent 6 months embedding steel latticework in a school building's corners, in effect tying the floors to the walls. Engineers from GHI's partner, the National Society for Earthquake Technology, instructed the masons. The society is working to reduce the earthquake risk in Nepal, where there are no building codes and, according to GHI, construction standards have degraded rather than improved over the past 100 years.

The project made the schoolchildren and the community at large aware of the risk of earthquakes. In the months following the school's inauguration, some village dwellers paid to have the same kind of work done on their own homes.

"Of course, it was just a drop in the bucket," Tucker says. "It was one village, and even for a small country like Nepal, there are thousands of villages that have to be done."

Quakes today exact a far harsher penalty in terms of loss of life in developing countries such as Nepal than they do in wealthier countries. According to the Office of U.S. Foreign Disaster Assistance, the number of deaths per fatal earthquake--an average of some 12,000 people--was the same for both industrialized and developing nations in the first half of the 20th century. The figure remained the same for developing nations in the latter half of the century, while it dropped to below 2,000 people in the developed world.

Scientists at GHI have highlighted this asymmetry in a preliminary study of what they call the earthquake lethality potential of 21 cities around the world. The value for a given city estimates the number of its inhabitants who would have a 10 percent chance of being killed by an earthquake in a 50-year period. The analysis takes into account factors such as soil structures, standards of construction, the integrity of key buildings such as schools, the potential for fires and landslides, and the availability of medical care and search-and-rescue operations.

According to the GHI report, which was released in 2001, Kathmandu, Nepal, is at highest risk, with an earthquake lethality potential of 70,000. The cities in the study with the lowest fatality potential are Kobe,Japan, and Vancouver, British Columbia. A schoolchild in Kathmandu for example, has about 400 more times the chance of dying due to earthquakes than a child in Kobe does.

Commissioned by the U.S. Agency for International Development, Tucker and his colleagues at GHI are now doing a survey of 20 cities in India. For this and future work, Tucker plans to refine the methodology for evaluating cities' earthquake lethality potential, and to eventually use it to tell the public where the risk is highest.

SHAKY PREDICTIONS In one region of the United States, a debate is going on over just how high the local quake risk is--and what preventive measures will be cost-effective. In Memphis, Tenn., city officials are deciding whether to adopt a new building code that has more rigorous seismic standards than those currently in place in the city. The code stipulates that certain building elements, such as columns and beams, will need to be heavier, for example, and that a building's openings will require sturdier reinforcement.

Some local engineers estimate that such additions could add from 3 to 10 percent to construction costs. "These new building codes would require that we design here just as if we had a building sitting in San Francisco," says Joe Tomasello, a structural engineer at the Reaves Engineering Firm in Memphis. San Francisco, which lies along the active San Andreas fault, regularly experiences quakes, while Memphis quakes are rare.

Still, the city bears a risk higher than that of most U.S. cities. Memphis is located in the New Madrid seismic zone, an ax-shaped area that extends from just west of the city to southern Illinois. Deep below ground in the Earth's crust is a rift some 70 kilometers wide and 300 km long. The zone is usually seismically calm, experiencing only minor rumbles. But in the winter of 1811 and 1812, a series of three massive earthquakes shook the area surrounding New Madrid, Mo.

With an estimated force of between magnitude 7 and 8, the quakes were among the most powerful in U.S. history. The shocks set church bells aquiver in Boston, 1,600 km from the epicenter. In Tennessee, a swathe of land sank and filled with river water, forming what is now Reelfoot Lake. Whole forests were uprooted in northwest Tennessee and southeast Missouri.

According to Art Frankel and his colleagues at the U.S. Geological Survey in Denver, a quake of similar force occurs in the New Madrid seismic zone, on average, once every 500 years. The USGS team rates the chances of one of these massive quakes happening in the next 50 years at 7 to 10 percent. Their prediction is based on signs of earthquakes in soil strata they've examined that, they say, suggest strong quakes occurred in the area around the years 900 and 1400.

"The best guide to the future is what's happened in the past," Frankel says. "It would be irresponsible to dismiss the evidence."

But Stein claims that this record is not a reliable indicator of the likelihood of an earthquake in the years ahead.

In a paper published in the May 13 Eos, the weekly newsletter of the American Geophysical Union, Stein and his colleagues, including Tomasello, reported that depending on what models seismologists use, the chances of a repeat of the 1811-1812 earthquake series in the next 50 years runs between 0.01 percent and almost 20 percent. "We are playing a game of chance against nature and we have no idea what the rules are," Stein says.

The debate gets even more complex. It's not just a question of how often a quake that strong might occur but how that quake would affect the region. That's a function of ground motion, which is the main thing that building engineers worry about when designing for quake resistance.

Ground motion varies from region to region because it depends on the nature of the rock surrounding the fault. Frankel says that in the New Madrid seismic zone, seismic waves propagate with less attenuation than they do around the San Andreas fault, in part because the New Madrid crust is less fragmented and because it's cooler than the western crust. Though there is no fixed ratio, most seismologists agree that a lower-magnitude quake in New Madrid would produce the same amount of shaking as a higher-magnitude quake in California would.

No one is certain of the relationship between earthquake magnitude and ground shaking in the New Madrid seismic zone. In seismic-hazard maps published last year, the USGS calculated potential ground motions for the United States by taking the average of the outcomes from five ground-motion models. Each model has a different way of describing how the wave energy from a quake dissipates. The resulting maps indicate the force of ground motions that have a 2 percent or 10 percent chance of being exceeded in the New Madrid zone in the next 50 years.

Motions of more than 0.1 times the acceleration of gravity (g) can damage buildings. In Memphis, the motions that have a 2 percent chance of being exceeded are as high as 0.7g. Another way to express it: Memphis can expect this ground motion to be exceeded once in a 2,500-year period. Engineers developing the new Memphis building code opted to play it safe and prepare for ground motions almost as great.

Stein calls the USGS calculations "unconstrained by data." Frankel counters that the historical data from the New Madrid seismic zone is telling enough, and that data from the region of Bhuj, India, which is similar to the New Madrid Seismic Zone, serves as a reliable guide.

The mayor of Memphis, Willie Herenton, is campaigning against the proposed code. In a July 10 letter to Tennessee's governor, Herenton complained that "we are being told to plan for an event that may occur over a future time period longer than the time span of the Common era!" He fears that enforcing the proposed new code could stymie city development and lead to job losses.

"Until the next earthquake, the community is worse off," Stein says, "whereas, there will be benefits when the earthquake occurs." At press time, Memphis officials were days away from voting on whether or not to adopt the new code.

Whatever that decision may be, the U.S. Veterans Administration is acting with caution in the New Madrid seismic zone. Construction workers are in the process of amputating the top nine floors of the VA hospital in Memphis. Engineers figured it would be easier to lop off most of the building than to seismically reinforce every floor, and they've been doing the demolition work while doctors carry on treating patients in the lower levels.

By October, 5 stories will remain where 14 once stood. These remaining levels will be buttressed with steel bracing and concrete reinforcement walls that can withstand the lateral forces that impinge on a building during an earthquake. The whole project, which includes the construction of a new, quake-resistant wing, will cost more than $100 million.

The renovation work is part of the VA's 30-year, nationwide push to make safe all of its "essential facilities"--including hospitals and nursing homes--in areas where there is a sizable seismic risk. About 400 VA buildings are being demolished or strengthened or are waiting for designated overhauls. The initiative was prompted by the deaths of 46 people in the collapse of two VA hospital buildings during the 1971 San Fernando earthquake in California.

That same quake kicked off a flurry of activity among engineers, resulting in an extended menu of methods for making buildings stand up to earthquakes. The menu includes the most common practice of fitting steel and concrete bracing--used in the Memphis VA hospital--and more high-tech interventions, such as seismic dampers. These giant shock absorbers and bearings isolate a building from ground shaking.

Lomnitz predicts that it's just a matter of time before these features are standard in all structures. Everyone living in an earthquake-prone region, he argues, should be able to count on their building to ride quake waves. He acknowledges that realizing his vision will take time and a "knuckling down."

In the meantime, Bilham is optimistic. UN statistics suggest increasing urbanization worldwide: In the next 30 years, almost all of the world's population increase--about 2 billion people--will be in the urban areas of less-developed nations. That will mean the construction of hundreds of millions of new homes. Bilham says that this provides the perfect opportunity to plan and build cities to appropriate standards.

"I can't imagine living on a planet in the year 3000 where we will tolerate buildings falling down just because an earthquake happens," he says.