Here I thought I was just lucky. For the first 40 years of my life I bounced between northern and southern California and almost always managed to be at the opposite end of the state when a big quake struck along the San Andreas Fault or one of its many close relatives. I missed the quakes because of the way the San Andreas and its related faults let loose only at the far ends of the state and never at the same time – or so I thought. Now, it turns out, the San Andreas may be capable of making people all along its 800-mile length miserable at the same time, according to new research published in the latest issue of Nature.
For decades now the San Andreas has been thought to have northern and southern areas where it binds up (“couples” in seismological parlance) and builds up stress that lets loose in large, infrequent, earthquakes. In contrast, the central section of the fault has been more prone to steady, regular creeping movements (decoupled), which keeps it from building up stresses and avoids big ruptures. The same creeping stress relief was thought to be at work along faults in other parts of the world, including Japan and Taiwan.
But the 1999 7.6-magnitude Chi-Chi earthquake in Taiwan flaunted that theory. Then, in 2011, the Japanese 9.0 magnitude Tohoku-Oki earthquake (and its disastrous tsunamis) broke the model utterly, and furrowed the brows of seismologists everywhere.
Now two seismologists are proposing a new model to explain how parts of faults that seem to be creeping along benignly and not building up dangerous stress, can also slip and trigger large ruptures in adjacent parts of the fault. The underlying mechanism, as I understand it, is the frictional heating of fluids in the pores of the rocks in the fault zone to the point that the rocks weaken and the friction that keeps the fault from rupturing is overcome – causing a large slip, or rupture, rather than gradual creep. They have based the model on extensive laboratory work using actual fault zone rocks from Taiwan.
The two seismologists – Hiroyuki Noda of the Japan Agency for Marine-Earth Science and Technology, and Nadia Lapusta of the California Institute of Technology – write:
“Our results demonstrate how fluids can cause stable fault behaviour to become destructive through thermal fluid pressurization during rapid slips. In fact, experimental studies have shown that wet, clay-rich sediments, which are likely to be present in shallow portions of subduction zones, are indeed susceptible to thermal pressurization….
Findings such as these have important implications for seismic hazard, because they suggest that currently creeping (decoupled) fault regions, which are thought to be stable and aseismic, may participate in destructive events and host large seismic slip.”
An obvious question you might have at this point is: How can this be true for the San Andreas when we’ve never seen it behave this way? The researchers anticipated that question:
“After large events, such regions would stay locked for a while, but would eventually accumulate enough stress to start creeping again, obscuring the evidence of their destructive past.
Hence, using only the relatively short recorded seismic and geodetic history to estimate seismic hazard is inadequate, as demonstrated by the unexpected 2011 Tohoku-Oki earthquake.
For example, on the basis of fewer than 100 years of recorded history, the creeping section of the San Andreas fault in California is thought to provide a barrier between the southern and northern locked parts of the fault, essentially preventing a massive earthquake that would involve all of California from San Diego and Los Angeles to San Francisco. However, our study provides a plausible physical mechanism by which that segment could join such a massive event.”
What else can I say, except YIKES!
IMAGE: A collapsed building in San Francisco’s Marina District caused by the 1989 Loma Prieta earthquake. Credit: NOAA/NGDC, D. Perkins, U.S. Geological Survey