May 5, 2012
by John Vidale
A paper this month (Jeff McGuire et al., Nature Geoscience, 5, 336-341, 2012) shows a fascinating progression of fault failure, providing a new clue as to how indications of activity precursory to large earthquakes might appear. The observation comes from the Gofar transform fault, which is on the East Pacific Rise offshore central America.
The first figure (click on it to see a bigger version) is topography on the left, accompanied by a simple cartoon of the tectonics and a more accurate cartoon. The fault is a transform, which means it has strike-slip earthquakes, and rather than break all at once, tends to have magnitude 6 earthquakes on the orange patches and intense swarms of earthquakes on the red patch.
The timing of the M6 earthquakes in the two locations is shown by the red and yellow dots on the timeline of earthquakes in the region below. The timeline shows that the M6 events on the two segments occur close in time, and each break roughly every four years. Noticing the pattern, McGuire put a set of seismometers on the ocean floor at the right time to neatly capture the 2007-2008 sequence of earthquakes. The interesting part comes between the 2007 yellow dot and the 2008 red dot, specifically in the week before the 2008 M6 earthquake.
The third figure shows a cross-section of the seismicity in the region of the two M6 earthquakes. The red and dark blue dots (below the blue triangles with the numbers 6 and 10) are in the locations of the two orange semi-circles above - those are the patches of fault that break in M6 earthquakes, with red here corresponding to red in the timeline and dark blue corresponding to yellow in the timeline (sorry about the inconsistent color scheme).
Swarms have been seen prior to earthquakes before, that is not the new part of this observation, but the details of this swarm are cool. Apparently, after the blue patch failed in an M6 in 2007, fault movement began on the swarm area (triangle 8, around the yellow dots), and intensified in the week before the M6 around the red patch. The properties of the rock in the yellow patch changed noticeably - the seismic wavespeed decreased by 3% - during the week of activity. The interpretation is that the swarm area has the unforeseen properties in that it (1) takes a week to yield, (2) it yields until the neighboring fault patch breaks, and (3) we can tell that significant yielding is occurring by monitoring the temporal change in seismic wavespeed. Further, we can try to figure out from the observed changes in the physical properties of the fault what allows this stretch of fault to display such well-mannered behavior.
The swarm extends much deeper than the M6 events are inferred to have broken, and the fluctuation in seismic wavespeed is inferred to show that myriad tiny cracks are opening during the week-long burst of strain, perhaps enhancing fluid flow to facilitate redistribution of fluid pressures.
This is still a scientific work in progress. First, we need to figure out how universal is the presence of faults patches that reveal their yielding in such observable ways. Second, I hope Jeff is ready to look more closely at the next iteration of this sequence in 2012 (or maybe 2016?) to get a sharper picture of the action so we better know how to look for it elsewhere.
- Is Mount St. Helens seismicity increasing?
- Warm weather triggers snow avalanches at St. Helens
- Seismic recordings of a gas explosion in North Bend, WA
- Seismic signals generated by the March 22nd Oso Landslide
- Legacy web site content returns
- Ice avalanches on Cascade volcanoes
- A New View On What's Shaking on the Cascade Volcanoes
- Seismic Spectrograms - A new way to look at wiggles
- The final football game analysis
- The Football Game Experiment Continues
- July (1)
- May (1)
- April (1)
- March (2)
- February (3)
- January (4)
- October (1)
- September (1)
- August (1)
- June (1)
- April (3)
- March (4)
- February (1)
- January (2)
- December (1)
- November (2)
- October (3)
- September (1)
- August (3)
- July (2)
- June (4)
- May (4)
- April (2)
- The wech-o-meter takes over all of Cascadia
- Keystone Cops: Italy prosecutes seismologists for failure to predict deadly quake
- UFOs in eastern Washington? No, rather UTEs (Unidentified Terrestrial Events)
- New Sodo Seattle Liquefaction Array Installed
- Why we should constantly watch the deformation of the seafloor
- Mystery chirp near Newberry Volcano
- Planting seismographs causes earthquakes? or maybe ice-quakes?
- Tunneling rumbles south under Capitol Hill
- 15 years of mostly silent magma inflation near Three Sisters, Oregon
- Mount Hood earthquake swarm of Feb 23, 2012
- Web glitches: duplicate (and even triplicate!) earthquakes
- How earthquake magnitude scales work
- Mine blast masquerades as volcanic tremor
- The Spokane Swarm about 10 years ago
- Another hum around Mount St. Helens
- Slow slip: A new kind of earthquake under our feet
- PNSN and social media
- 3am M3.4 earthquake in St. Helens Seismic Zone
- The wrong kind of volcano noise
- Fast chatter on Rainier an hour ago
- Can slush-mageddon trigger earthquakes?
- Rainier Repeating Earthquakes Update and Comparison with Weather Patterns
- 22-minutes drumbeat icequakes(?)
- Mount Rainier popping away
- Repeating Earthquakes on Mount Rainier - are glaciers the culprit?
- Debunking another SEC football myth by the PAC-12
- One year ago, Seattle Seahawks 12th Man Earthquake
- The odds this year of a megaquake on the Pacific Northwest coast
- Is the plague of great earthquakes this decade a sign of increased danger?
- Nile Valley landslide talks to PNSN seismologists
- Good vs evil in central US earthquake hazard analysis
- Why does a volcano scream?
- Predicting big quakes from patterns of little ones
- 1-hour warning for Japanese M9 earthquake?
- Sound Transit train under Interlaken keeps a rollin'
- Invisible changes under the hood at the PNSN
- Sound Transit Tunneling Noise
- "Visionary" toads
- Earthquake early warning in the PNW
- November (1)
- December (13)