At the Seismological Society of America’s Annual Meeting, researchers posed a seemingly simple question: how wide are faults?
Using data compiled from single earthquakes across the world, Christie Rowe of the Nevada Seismological Laboratory at the University of Nevada, Reno and Alex Hatem of the U.S. Geological Survey sought a more comprehensive answer, one that considers both surface and deep traces of seismic rupture and creep.
By compiling observations of recent earthquakes, Rowe and Hatem conclude that from Turkey to California, it’s not just a single strand of a fault but quite often a branching network of fault strands involved in an earthquake, making the fault zone hundreds of meters wide.
“So that suggests that significant parts of the broad array of fractures that develops over many earthquakes can be activated in a single earthquake,” said Rowe, who noted that this width sometimes roughly corresponds to the width of Alquist-Priolo zones established for safe building in California.
“We want to know how this might change things like the shaking patterns that you would expect, or how much radiated energy you get from an earthquake,” Rowe explained. “Because it’s not the same if you have slip distributed on many strands as when it is all on one strand of the fault.”
At the same time, the researchers found that the width of creep zones at these earthquakes are much narrower, both near the surface and 10-25 kilometers deep in the earth. The creep zones, between 2 and 10 meters wide, “may be the most localized behavior a fault does,” Rowe said.
The study emphasizes the importance of thinking of faults in a more three-dimensional manner, said Rowe.
“As a geologist, it's always kind of been a cognitive disconnect for me when I talk to earthquake modelers who have these two-dimensional features that they model earthquakes on,” she said. “Because the sheer resistance, the strength or the friction, comes from a volume of rock that's deforming during an earthquake or in between earthquakes. So the size of that volume controls the strength of the fault in some really tangible ways.”
The researchers used a variety of data in their study, including rupture maps, creeping zone width from surveys of slowly shifting monuments along faults and satellite observations, the locations of earthquake aftershocks, low velocity damage zone widths, and the zones delineated by certain types of rock such as pseudotachylyte, ultramylonite and mylonite that are a signature of creep and deformation.
The findings also have implications for how scientists study past earthquakes to calculate earthquake recurrence intervals on faults, Rowe noted.
Slip rates and recurrence intervals can be constrained using localized measurements, but it can be difficult to disentangle the slip that occurred during an earthquake and aseismic slip that occurred after the event. The 2014 Napa, California earthquake is a good example of this phenomenon, said Rowe, noting that almost half of the slip measured after that event occurred slowly after the earthquake.
But if the Napa earthquake occurred thousands of years ago and researchers came across its traces in the rock record, “you would just see a bigger earthquake. You might lump all of that slip as a single event,” Rowe said.
Creep isn’t always accounted for in calculating recurrence intervals, “so finding out that creep zones are quite narrow means that we should be aware that we could be convolving creep with seismic slip when we look at those paleoseismic records,” she added.