Earthquake predicted for this area:

A UCLA team had predicted a 50-50 chance of a magnitude 6.4 earthquake to hit the San Gabriel Mountains and northeast by 5-September, 2004.

No Sign Yet of Predicted Big California Earthquake
Saturday,  Sept. 4, 2004, 1:56 PM EDT
By Jill Serjeant
[http://www.reuters.com/newsArticle.jhtml?type=scienceNews&storyID=6151477&section=news]
 
LOS ANGELES (Reuters) - The clock is running out on a highly publicized prediction that a major earthquake will rip through Southern California by Sunday.
 
But even if the earth does not move by the weekend, seismologists largely agree that the forecast had done more good than harm by reviving interest in the controversial science of quake prediction.
 
A UCLA team startled Californians and the wider scientific world in January by predicting there was 50-50 chance of a 6.4 magnitude or larger quake hitting a 12,000 square mile mostly desert area east of Los Angeles by Sept. 5.
 
They used an algorithm, or mathematical pattern recognition formula, developed by team member Professor Vladimir Keilis-Borok which had already successfully forecast a 6.5 temblor in central California in December 2003 and the 8.1 magnitude quake that struck the Japanese island of Hokkaido in September last year.
 
The public forecast of a quake in southern California shook up the world of seismology where reliable earthquake prediction was largely discredited after the 1980s when another big prediction embarrassingly fell flat.
 
But a panel of earthquake experts deemed the prediction valid although too untested to warrant emergency measures in a state which measures up to 60 earthquakes a day -- few of them even rattling a window.
 
As one seismologist said, "Even if he was right, what can you do in a six-month window? You are far more likely to be murdered in L.A. than die in an earthquake in California."
 
With just days to go before the deadline, seismologists agreed there was only a 2 percent chance of the quake now striking in the predicted time in an area comprising the Mojave desert, Palm Springs and San Bernardino.
 
"It doesn't look like the earthquake will fill this window," said John Vidale, director of the UCLA Institute of Geophysics and Planetary Physics where Keilis-Borok and the seismology team work.
 
"I'm sure he (Keilis-Borok) is not happy if his prediction is not fulfilled because he sees this as crusade against the skeptics who have been digging in their heels for decades."
 
But Vidale said seismologists were now coming around to the idea that earthquake prediction was a valid science.
 
ADVANCEMENT TO SCIENCE
"Scientifically, a proper test cannot be a mistake and this one was a proper test. As long as it is a proper scientific hypothesis and it is disprovable, it is an advancement to the science," said Lucy Jones, scientist in charge of the U.S. Geological Survey for Southern California.
 
Seismologists noted however that one of their biggest challenges was predicting which of the 35,000 earthquakes that occur every year in California would cause vast damage.
 
"We don't know how they start, or how they stop, and how they stop determines how big they are. That is focus of my research," said Dr. Egill Hauksson, a seismologist at Caltech in Pasadena.
 
"It is impossible to successfully predict earthquakes when we don't understand the physics of earthquakes," he said.
 
Jones, who also chairs the California Seismic Safety Commission, said possible emergency precautions such as mass evacuations, lowering reservoir waters, canceling vacations for emergency workers and warning drivers to stay off freeways were inappropriate in this case because of the vast area and the large time-window.
"Even if Keilis-Borok was perfectly correct, there is not much you can do with this information," Jones said.
 
Vidale said that even if there were a quake on Sept. 10 in the predicted area, or a big quake in Los Angeles tomorrow "technically that would be a miss."
 
"We would all be happier if there was some harmless earthquake that filled this prediction, but the earth tells us how it behaves. We don't tell it."

Then there's Parkfield:

The area around Parkfield, in central, coastal California is the most seismically active along the San Andreas Fault. Researchers had predicted an earthquake to occur there before 1998, and had "instrumented" the fault for this event. The earthquake finally, hit, about 6 years after predicted.

Here is a link about the earthquake.