JEFFREY BROWN: Next, preparing for and predicting earthquakes.
California and other Western states conducted major earthquake drills today, just hours before a 3.9 tremor hit the San Francisco area. The quake was centered on the University of California, Berkeley, campus. There were no initial signs of major damage or injuries.
It was the fourth year California held the drill exercise and the largest yet, with more than eight million people participating.
To dramatize what could happen, authorities brought earthquake simulators to cities including Hayward, which sits on a fault expected to rupture again within the next few decades. The simulator was designed to mimic some of the effects of a 7.5 quake. Emergency officials demonstrated recommended responses, such as the drop-and-cover technique.
Preparing is one thing. Having the time and ability to warn people in advance is another.
NewsHour science correspondent Miles O'Brien reports on efforts to make those predictions after the Tohuku earthquake hit Japan earlier this year.
MILES O'BRIEN: It is a trial that has scientists all over the world in an uproar: six seismologists and one government official charged with manslaughter because they didn't predict this, the 2009 earthquake in L'Aquila, Italy, which killed more than 300.
Many here are comparing it to the trial of Galileo in 1633. He was found guilty of heresy for suggesting the Earth wasn't the center of the universe.
"It's not a Galileo trial," says a victim's lawyer. "But it is definitely a trial to find out if there were some responsibilities, some omissions, behaviors, or wrongdoings on the part of the scientists."
If found guilty, the accused could face 15 years in prison.
Should scientists be held criminally accountable for these kinds of things?
JIM MORI, Kyoto University: I don't think they should be if they are - if they are doing their job.
MILES O'BRIEN: Jim Mori is a professor of seismology at Kyoto University.
JIM MORI: We don't know enough to make a good prediction, and I don't think you can be held responsible for something that you really are not aware of.
MILES O'BRIEN: If any nation could predict an earthquake, it would likely be Japan. Scientists here have been hard at work on this elusive goal since the Great Kanto Earthquake of 1923, which killed 145,000 and prompted the founding of the Earthquake Research Institute at Tokyo University.
Seismologist Naoshi Hirata is the director of the Earthquake Prediction Research Center here.
How difficult a challenge is it to predict an earthquake with some certainty?
NAOSHI HIRATA, Earthquake Prediction Research Center: Earthquake is a kind of -- a kind of catastrophe in the rock. And there is a breakage. And it's very difficult to determine, very precisely determine the place and what is going on.
MILES O'BRIEN: But they are trying mightily, as professor Hirata showed me.
NAOSHI HIRATA: This is the distribution of small earthquake below Japanese island.
MILES O'BRIEN: Using Japan's uniquely dense network of seismometers, researchers have created a detailed map of previous earthquakes. Connect the dots, and you have a 3-D map of the faults that threaten Japan.
It is a very clear view in the rear-view mirror, but not a crystal ball.
Could you say, March 11, 2:46 p.m., magnitude 9.0, take cover? Are we going to ever get to that point?
NAOSHI HIRATA: No, I don't think so, because the magnitude 9 earthquake is so huge, probably 500-kilometer-long and 200-kilometer-wide, it's impossible to put all the sensors on such huge areas.
MILES O'BRIEN: There was a time when seismologists were much more sanguine about earthquake predictions.
JIM MORI: About 40 years ago, most seismologists really thought that within maybe a few years or 10 years, earthquake prediction would be a reality.
MILES O'BRIEN: In 1975, a Chinese seismologist predicted the 7.3 magnitude Haicheng earthquake the day before it happened, prompting a mass evacuation, saving countless lives.
It is the only documented prediction of a major earthquake ever, but, today, scientists believe it was little more than good luck.
Forty years of looking hard later, where are we?
JIM MORI: It's just turned out to be a lot more complicated, a lot harder. The Earth seems to be a lot more heterogeneous. And so the signs or the clear changes are not so clear.
And so we just have a very difficult time now determining what's happening before an earthquake, what's just sort of part of the natural complexity of changes that are going on in the earth or around faults before an earthquake. And we haven't been able to find that sort of magic precursor.
MILES O'BRIEN: Ah, the magic precursor. Over the year, there have been no shortage of odd notions, unusual cloud patterns, surges in electromagnetic radiation or radon gas, and any number of odd animal behaviors, frogs or snakes on the march, even catfish doing a jig.
Bob Geller is a professor of geophysics at the University of Tokyo.
BOB GELLER, University of Tokyo: The animals aren't nearly as sensitive as scientific instruments for making measurements.
MILES O'BRIEN: The day I met Professor Geller, he was on the front page of a big national newspaper, after he called on the Japanese Parliament to repeal a 1978 law that mandates government funding and a concerted effort to predict the Tokai earthquake, Tokyo's big one.
How much money is being spent on an annual basis in Japan in pursuit of this quixotic goal?
BOB GELLER: Well, I will give you a round number of $100 million or so.
MILES O'BRIEN: A year?
BOB GELLER: Yes.
MILES O'BRIEN: Geller believes the money would be better spent retrofitting old structures to be more quake-resistant.
Hard as predictions may be, Japan does operate the most sophisticated earthquake and tsunami early warning system in the world. I visited the command center at the Japanese Meteorological Agency in Tokyo.
While I was there, an aftershock of the devastating March 11 earthquake got the crew's attention, but it wasn't big enough to sound an alarm.
Akihiko Wakayama is the man in charge.
He told me: "When we detect the big earthquake, within a few minutes, we will send out the signals of the earthquake happening. And if a tsunami is also involved, we also send out a tsunami warning."
The alerts go to the media, millions of cell phones and automatically stop elevators, hazardous industrial mechanisms and bullet trains. Here is how it works. Every earthquake creates a primary or P-wave and a secondary or S-wave. The latter does the damage. The former, only detectable by seismometers, has a much smaller amplitude and thus moves much faster.
So, the early warning system takes advantage of a gap, not unlike the time between lightning and thunder. An early warning system is ideally suited for an earthquake where the epicenter is a fairly good distance away from the people you would like to warn.
I am in Kobe, Japan right now. And in 1995, the great earthquake here caused extensive damage and killed about 6,000 people. This is some of the damage that has been preserved here at a memorial park on the port. In the case of the Kobe earthquake, the epicenter was about 20 kilometers from where I am standing right now, putting this city effectively in the early warning system blind spot.
As it turns out, Japan's early warning system didn't provide very accurate warnings of the magnitude 9.0 Tohoku event. It predicted a magnitude 7.9, a huge difference in the logarithmic scale of earthquake intensity.
MASUMI YAMADA, Kyoto University: This is at the beginning. You see the magnitude? So, far, it's like...
MILES O'BRIEN: Engineering professor Masumi Yamada of Kyoto University showed me the problem. The March 11 magnitude 9.0 earthquake unzipped a fault 500 kilometers long off the Pacific coast of Japan, but not all at once. It was a long chain reaction.
MASUMI YAMADA: Our current system assumes this earthquake just occurs only a point. So that caused a problem.
MILES O'BRIEN: Meanwhile, the ruptures kept coming, rolling like a southbound freight train toward Tokyo.
So are you worried that people might lose faith in the system?
MASUMI YAMADA: We definitely need some improvement of the system, so that it's useful -- it's still useful information for people.
MILES O'BRIEN: The inaccurate warning had disastrous consequences, as it predicted a tsunami that wouldn't have overtopped most seawalls, lulling people into a false sense of security. So even an accurate early warning system, much less a way to predict earthquakes with some specificity further in advance, remains elusive.
Is there something that science is missing right now that will be discovered eventually? Or is this one of these chaotic things that might be impossible to predict?
JIM MORI: It might be very difficult to predict, because any earthquake, even the magnitude 9.0 earthquake, starts from a very small source, and then grows to a huge dimension. And so, if you want to predict a magnitude 9.0 ahead of time, you have to actually produce -- predict that very small beginning.
And, right now, it's very difficult to distinguish that small beginning of magnitude 9.0 from a small beginning of a magnitude 5.0 or 6.0.
MILES O'BRIEN: So in a nation that is perhaps the most earthquake-savvy in the world, they have learned a lesson in the limits of seismology and science in general. Fortunately for the scientists here, it won't be a lesson paid in prison.
JEFFREY BROWN: That's the first of three stories Miles will be filing from Japan for us.
In his next reports, he will look more closely at tsunamis and detecting radiation after the nuclear disaster.