golden gate bridge
A major Hayward quake could destroy 150,000 homes and wreck Oakland and Berkeley, besides disturbing the office towers in San Francisco's financial district Robert Galbraith/Reuters

California's 'big one' could happen any day but scientists just cannot say when, a geophysicist said, clarifying a comment that set off a scare following a minor quake in the region on Tuesday.

The 4.0 earthquake that struck near Fremont had resulted in media 'aftershocks' warning of an inevitable major quake along the Hayward Fault.

A major quake on the fault is overdue by seven years going by historical data.

Speaking to Discovery News, U.S. Geological Survey (USGS) geophysicist Tom Brocher, former director of the agency's Earthquake Science Center, clarified his comments that had set off internet shock waves.

There is no scientifically accepted basis for predicting earthquakes besides broad forecasts based on history and the structure of the fault, he pointed out.

The most recent analysis estimates that over the next 30 years, the Hayward Fault has a 31 percent probability of producing an earthquake of 6.7 or higher in magnitude.

"What we know for sure is that the Hayward Fault is capable of producing a major quake at any time," said the scientist, who was the lead author of a 2008 USGS report warning northern Californians of the need to prepare for such an event.

Not as well known as the San Andreas fault, the Hayward Fault is actually one of the world's most dangerous earthquake faults as it runs through a densely populated area.

It has produced five major quakes over the past 700 years, occurring about 140 years apart.
This would indicate the region should have experienced a 'big one' in 2008, since its last major quake occurred in 1868.

A major Hayward quake could destroy thousand homes and wreck Oakland and Berkeley, besides affecting the high-rise towers in San Francisco's financial district.

While big data analytics that use satellites and study atmosphere abnormalities to predict quakes in time are improving, most predictions are still based on historical information and precursors.