A Creepy Situation

The slow creeping motion of seismic faults could hold the key to predicting earthquake patterns, according to a new study by Scripps Institution of Oceanography at UC San Diego postdoctoral scholar Yoshihiro Kaneko and colleagues at California Institute of Technology.

A new earthquake simulation model developed by the researchers produced scenarios of earthquake rupture patterns and showed how a creeping section impedes rupturing and is often a permanent barrier to large earthquakes.

Many faults around the world, including California’s most famous seismic landmark, the San Andreas Fault, have locked regions and sections that slowly and continuously creep.  The plates at the central portion of the San Andrea Fault, from San Juan Bautista to Parkfield, Calif., are gradually moving past each other three centimeters per year and only produce minor quakes. The north and south sections are firmly locked in place and release stress in the form of large earthquakes.

In one of Kaneko’s simulations, an earthquake ruptures through both the two locked segments and a central creeping patch.  A subsequent earthquake at the same site 29 years later starts at the same location but is stopped by the central creeping patch.

As a result, postseismic movements trigger smaller earthquakes in the adjacent locked segment less than a week later, which leads to a complex pattern of smaller earthquakes and slow slip events in the right locked segment 30 to 60 years after the initial earthquake.

“This simulation illustrates that the central creeping patch creates complexity of large earthquakes in the model, acting as an occasional barrier to earthquakes and causing clustering of large events,” said Kaneko.

Scientists have long known that faults that creep are less likely to have large earthquakes. The study’s authors believe this new information generated from the simulation model could help scientists move closer to the possibility of forecasting patterns of large earthquakes on fault systems.

"This study convincingly demonstrates the importance of variations in friction properties on faults, and how they may control the occurrence of both earthquakes and slow deformation between earthquakes,” said Kaneko’s faculty advisor, Scripps geophysicist Yuri Fialko. “It opens up a possibility of evaluating future behavior of seismic ruptures based on a combination of long-term geodetic and seismologic observations and computer simulations."

The results were published in the April 25 online edition of the journal Nature Geoscience.

-- Annie Reisewitz