Deep under Pakistan-administered Kashmir, rocks
broke, faults slipped, and the earth shook with such
violence on 8 October that more than 70 000 people died
and more than 3 million were left homeless [see photo,
"Devastated"].
But what happened in the weeks and days and hours
leading up to that horrible event? Were there any signs
that such devastation was coming? We think there were,
but owing to a satellite malfunction we can't say for
sure.
How many lives could have been saved in that one
event alone if we'd known of the earthquake 10 minutes
in advance? An hour? A day?
Currently, predictions are vague at best. By studying
historical earthquake records, monitoring the motion of
the earth's crust by satellite, and measuring with
strain monitors below the earth's surface, researchers
can project a high probability of an earthquake in a
certain area within about 30 years. But short-term
earthquake forecasting just hasn't worked.
Accurate short-term forecasts would save lives and
enable businesses to recover sooner. With just a
10-minute warning, trains could move out of tunnels, and
people could move to safer parts of buildings or flee
unsafe buildings. With an hour's warning, people could
shut off the water and gas lines coming into their homes
and move to safety. In industry, workers could shut down
dangerous processes and back up critical data; those in
potentially dangerous positions, such as refinery
employees and high-rise construction workers, could
evacuate. Local government officials could alert
emergency-response personnel and move critical equipment
and vehicles outdoors. With a day's warning, people
could collect their families and congregate in a safe
location, bringing food, water, and fuel with them.
Local and state governments could place emergency teams
and equipment strategically and evacuate bridges and
tunnels.
It seems that earthquakes should be predictable.
After all, we can predict hurricanes and floods using
detailed satellite imagery and sophisticated computer
models. Using advanced Doppler radar, we can even tell
minutes ahead of time that a tornado will form.
Accurate earthquake warnings are, at last, within
reach. They will come not from the mechanical
phenomena—measurements of the movement of the earth's
crust—that have been the focus of decades of study,
but, rather, from electromagnetic phenomena. And,
remarkably, these predictions will come from signals
gathered not only at the earth's surface but also far
above it, in the ionosphere.
For decades,
researchers have detected strange phenomena
in the form of odd radio noise and eerie lights in the
sky in the weeks, hours, and days preceding earthquakes.
But only recently have experts started systematically
monitoring those phenomena and correlating them to
earthquakes.
A light or glow in the sky sometimes heralds a big
earthquake. On 17 January 1995, for example, there were
23 reported sightings in Kobe, Japan, of a white, blue,
or orange light extending some 200 meters in the air and
spreading 1 to 8 kilometers across the ground. Hours
later a 6.9-magnitude earthquake killed more than 5500
people. Sky watchers and geologists have documented
similar lights before earthquakes elsewhere in Japan
since the 1960s and in Canada in 1988.
Another sign of an impending quake is a disturbance
in the ultralow frequency (ULF) radio band—1 hertz and
below—noticed in the weeks and more dramatically in the
hours before an earthquake. Researchers at Stanford
University, in California, documented such signals
before the 1989 Loma Prieta quake, which devastated the
San Francisco Bay Area, demolishing houses, fracturing
freeways, and killing 63 people.
Both the lights and the radio waves appear to be
electromagnetic disturbances that happen when
crystalline rocks are deformed—or even broken—by the
slow grinding of the earth that occurs just before the
dramatic slip that is an earthquake. Although a rock in
its normal state is, of course, an insulator, this
cracking creates tremendous electric currents in the
ground, which travel to the surface and into the air.