When a magnitude 5.6 quake rattled through the Bay area on Oct., 30, 2007, Tom Bleier and his team rushed to their headquarters in Palo Alto.
The quake was later found to have been centered near "Alum Rock Park" near San Mateo.
They looked at the peaks and valleys sent to them from specially designed monitoring stations around the state, but did those stations actually "predict" the event?
"We looked at it a little bit closer and we saw these very strange pulsations," said Bleier, CEO and Chief Technology Officer of QuakeFinder.com.
The pulses happened five days in a row, two whole weeks before the quake hit Alum Rock.
The pulses are generated, says Bleier, from electro-magnetic energy coming off of stones and boulders beneath the earth, and through the air above his monitors.
Bleier says it's like "underground lightning" set off by movement precluding a major event.
"We're not claiming to be able to do it yet," said Bleier, "but man, the pattern is there."
Before the earth moves, millions of things happen below the soil's surface. Part of the challenge of earthquake prediction has been cataloguing those events, in a timely fashion.
But Bleier's critics, most notably one geophysicist from the U.S. Geological survey say the pulse findings are inconclusive at best.
Too many things generate measurable impulses, they said, from the sun, to non-moving rocks, even to man-made events, like trains or traffic.
But Bleier says his data is newer, and more precise, that the data from the USGS, and he says he just needs a couple more events of magnitude five or larger to work with.
"The earthquake has a very distinctive pulsation," Bleier said. "It's usually about three to four seconds long. The shape of it is very different."
This week his team is in LA to add three more monitors near the dreaded San Andreas fault.
He said if he sees the same "pulses" there, he'll call California's Office of Emergency Services and tell them to prepare for a major quake within days.