Megacities are something new on the planet. Earthquakes are something very old. The two are a lethal combination, as seen in the recent tragedy in Port-au-Prince, where more than 200,000 people perished – a catastrophe that scientists say is certain to be repeated somewhere, and probably soon, with death tolls that once again stagger the mind.

In 1800, there was just one city with more than a million people – Beijing. Now there are 381 urban areas with at least 1 million inhabitants. Urbanization crossed a threshold last year when, for the first time, more people lived in city settings than rural ones. About 403 million people live in cities that face a significant seismic hazard, according to a recent study by seismologist Roger Bilham, of the University of Colorado.

The next Big One could strike Tokyo, Istanbul, Tehran, Mexico City, New Delhi, Kathmandu or the two metropolises near California’s San Andreas Fault, Los Angeles and San Francisco. Or it could devastate Dhaka, Jakarta, Karachi, Manila, Cairo, Osaka, Lima or Bogota. The list goes on and on.

”You can name about 25 cities that are like Port-au-Prince. They’re not going to shake but every 250 years (on average). But if you can name 25 . . . you’re going to have an event like this every 10 years,” said David Wald, a seismologist with the U.S. Geological Survey.

In many vulnerable cities, people are effectively stacked on top of one another in buildings designed as if earthquakes don’t happen. It is not the tremor that kills people in an earthquake but the buildings, routinely constructed on the cheap, using faulty designs and, in some cities, overseen by corrupt inspectors. The difference between life and death is often a matter of how much sand went into the cement or how much steel into a supporting column. Earthquakes might be viewed as acts of God, but their lethality is often a function of masonry.

”In recent earthquakes, buildings have acted as weapons of mass destruction,” Bilham writes in the journal Nature.

For years, earthquake scientists have shouted their warnings about the strong likelihood that a major quake would level an impoverished city and kill hundreds of thousands of people. They have said, for example, that Nepal’s Kathmandu, where masonry structures expand so haphazardly that some eventually cantilever over narrow city streets, is every bit as vulnerable as the surrounding Himalayas are majestic. They have said that a million people could die in a major quake in Tehran, Iran.

What’s impossible, however, is knowing precisely which of these cities will be the next to crumble. Or when. For all practical purposes, scientists can’t predict earthquakes.

The theory of plate tectonics, largely developed since the 1960s, explains why earthquakes happen in general. The major plates of the earth’s crust move constantly, creeping along at about the speed of fingernail growth. They rarely move smoothly past one another but are usually locked in place. On a strike-slip fault of the type that ruptured in Haiti, strain builds on the fault line for decades or centuries. The fault in Haiti had not ruptured in 240 years.

Scientists can map faults and estimate how much strain has accumulated since the last quake. What they can’t do is say that a given fault will ”break” tomorrow or next year or 10 years from now. Any calculation of earthquake probabilities has a lot of slop in the numbers.

”The problem is, the slop is huge on a human time scale,” said Susan Hough, a seismologist with the U.S. Geological Survey. ”We’re wired to deal with the immediate. We’re not geared to plan and stress about things likely to happen in 30 years.”

Some large earthquakes have small precursors, called foreshocks, but others happen without warning. There is one famous case of earthquake prediction, in Haicheng, China, in 1975. A local official sounded the alarm after many foreshocks and reports of snakes emerging from hibernation. But that prediction was more akin to a hunch than a scientific argument. There have been countless, less publicized instances when predicted earthquakes did not materialize.

As Hough notes in her book ”Predicting the Unpredictable,” the successful prediction of earthquakes was an official government mandate in Mao Zedong’s China, but no one foresaw the killer quake that took at least 240,000 lives in Tangshan in 1976.

Port-au-Prince had not been hit with a major quake since the days of French rule in the 18th century. Only in recent years have scientists mapped the fault that runs near the city.

”Just the beginning of work had been done. But enough was known that it could produce a big earthquake,” said Carol Prentice, a geologist with the U.S. Geological Survey. ”We knew it would be bad, but I didn’t imagine that it would be this bad.”

Prentice had been to the island of Hispaniola and had studied a different, roughly parallel fault called the Septentrional, which runs along the island’s northern edge. It had been difficult getting into Haiti, she said, so she and her fellow scientists focused their work on the Dominican Republic. That fault is another time bomb, threatening the Dominican city of Santiago, with a population of more than 1 million, Prentice said.

The entire Caribbean is seismically active. So is much of Central America. The next Big One could be on the isthmus of Panama, where Panama City sits just six miles from a major fault that hasn’t ruptured in four centuries.

Or the next catastrophe could be in Caracas, Venezuela, where millions of people live in poverty near a boundary of two tectonic plates, including the one that created the fault that broke in Haiti. The last catastrophic quake was 198 years ago.

Another seismic bull’s-eye is Mexico City, which sits on the worst possible soil, a drained lake bed that will intensify seismic waves. The devastating earthquake of 1985, which killed about 10,000 people, was centered hundreds of miles away but managed to ring Mexico City like a bell.

Earthquakes can turn up closer to home than many Americans realize. Several major tremors have been recorded off the East Coast, including near Newfoundland in 1929 and Boston in 1755. Charleston, S.C., had a quake in 1886 that killed 60 people.

Hough, of the USGS, said it might be that all three earthquakes were associated with the edge of the continental shelf and that any coastal city, including Washington, could get rattled by a quake someday.