By Michael R. Forrest
SCEC InstaNet News Feature Writer
The words "Slow Earthquake" conjure
up an image of a fault moving silently, stealthily, slipping
over the course of several minutes to a few weeks, generally
avoiding the scrutiny of geologists and geophysicists by generating
few - if any - waves.
Relatively little research has been conducted
on slow events, mostly because of their elusive nature. This
article presents two mini-interviews: one with Malcolm Johnston,
a US Geological Survey scientist who was part of the team that
detected the 1992 slow earthquake on the San Andreas fault, and
the other with Yehuda Bock, a geodecist with Scripps Institute
of Oceanography, who questions the terminology.
Kanamori and Hauksson (1992) applied
the "slow earthquake" moniker to a ML 3.5 earthquake
in the Santa Maria Basin, which lasted an unusual two minutes.
Ihmle and Jordan (1994) examined 107 shallow focus earthquakes
and found that 20 were preceded by "episodes of slow, smooth
deformation immediately before their high-frequency origin times,"
which they also referred to as slow earthquakes.
In Japan, Kawasaki, et al. (1995) detected a large, ultra-slow
earthquake subsequent to the July 18, 1992, Sanriku-Oki, Japan
earthquake (MS 6.9) which corresponded to a MW 7.3-7.7 event
(quartz-tube extensometers recorded the crustal strain). Back
in the United States, Linde et. al. (1996) revealed that in December
of 1992 two borehole strain meters located near the San Andreas
fault (in the transitional zone between the locked and creeping
segments of the fault in central California) recorded a series
of "strange signals." These were interpreted as the
result of a slow earthquake.
Other research examined a slow earthquake occurring in Nicaragua,
1992, (Ihmle, 1996 a,b), with one paper dismissing the slow earthquake
concept (Aki, 1992).
Interview with Malcolm Johnston, US
MF In 1992 your group observed a slow earthquake
on the San Andreas fault in the transitional area between the
creeping and locked segment that lasted about a week and displaced
a few centimeters. Is there any new evidence that such earthquakes
might be commonplace on other segments of the fault, or any of
its daughters, such as the Calaveras, San Jacinto, Palos Verdes,
or Newport Inglewood?
MJ We have observed four clear examples
of slow earthquakes. We have observed many other possible slow
events on just one instrument. This probably results from the
fact that we have just a few instruments and these are focused
largely on the San Andreas fault. A number of these have occurred
on the Calaveras (San Francisco Bay region). We have not seen
any events on the San Jacinto, Palos Verdes, or Newport Inglewood
faults. As you suggest, a reason might be because we have no
instruments installed anywhere near these faults.
MF The slow earthquakes on the San Andreas
were recorded to depths of as much as 8 km. One would imagine
that truly slow earthquakes might be extremely common in the
ductile middle crust, but that they'd be impossible to detect?
MJ I agree that these events might be
common in the ductile middle crust. We have the sensitivity to
detect them but have not "seen" them as yet. This may
be because they are not there, or because they occur much slower
than those in the upper 10 km.
MF If slow earthquakes are more commonplace
than previously thought, this would suggest that earthquake hazards
on many faults might be overestimated?
MJ From what we have seen so far, the
total moment release from slow earthquakes in the few places
where we have adequate instrumentation is still small compared
with the plate accumulation rate, though it is comparable to
the seismic moment release. However, I think that it is really
too soon to make a judgment about this. Thus the hazard may or
may not be overestimated.
MF Are there any studies underway to see
how often slow earthquakes might occur on strike-slip faults?
MJ Yes, we are trying to determine just
how often slow earthquakes occur on all types of faults not just
strike slip types.
MF What is the ideal setting for a slow
MJ I don't think we know yet what the
ideal setting is for a slow earthquake.
Interview with Yehuda Bock, Scripps
Institute of Oceanography
MF Is there possibly no difference between
slow earthquakes and simple tectonic motion? Is it all semantics?
YB In my opinion, yes, it is only a
matter of semantics and a way for seismologists to expand the
frequency domain in which they have always operated.
MF If there is a difference, should the
"creep" at Parkfield [California] be considered
a semi-continuous slow earthquake?
YB No. Creep is not an earthquake. It
releases strain but in an aseismic manner. It can only be measured
MF Do you have any thoughts on the large
"slow earthquakes" which seismologists haveseen in
subduction zones - and, should one assign moment magnitudes
to such "earthquakes?" (Should one assign a "magnitude"
to a large episode of tectonic motion?)
YB This is an interesting question.
Some strain is released aseismically. We've
seen this after the Landers earthquake, too, from continuous
GPS measurements. Therefore, the magnitude of the earthquake
(including all its aftershocks) is not an adequate measure, without
taking into account aseismic deformation, such as postseismic
MF What might be SCIGN's role in possible
proving/disproving the existence of slow earthquakes?
YB SCIGN will be able to measure aseismic
deformation lasting from as short as seconds to years. As I indicated
before we will also be able to measure
earthquake-induced ground motions directly. A seismometer measures
ground accelerations and then doubly-integrates to compute ground
motion. Should we then call earthquakes "ultra-fast"
deformation? No, of course not. We need to distinguish between
seismic and aseismic deformation.
MF Do you have other comments/observations?
I've always been somewhat amused by
this strange nomenclature ["slow earthquakes"] which
has been coined by seismologists. Earthquakes are one consequence
of plate motions and interactions (other consequences include
mountains and basins, for example). They are not a cause but
only an effect (a secondary issue is whether earthquakes trigger
other earthquakes - except for aftershocks after a main event,
this is still an open question). Geodesy measures plate motions
and crustal deformation directly by measuring precise 3-D positions
of geodetic markers from once every few seconds to once every
few years). Geodesy measures aseismic deformation (although we
are getting to the point where we will be able to measure earthquake-induced
ground motions directly, not accelerations that are measured
by seismometers). Seismology measures seismic motions (it requires
earthquakes or artificial earthquakes such as explosions).
References, Slow Earthquakes
Compiled by Michael R. Forrest
SCEC InstaNet News Feature Writer
Aki K., 1992, Higher-Order Interrelations
Between Seismogenic Structures and Earthquake Processes, Tectonophysics
211: (1-4) 1-12, Sep 30.
Ihmle P.F. and T.H. Jordan, 1994, Teleseismic
Search for Slow Precursors to Large Earthquakes, Science 266:
(5190) 1547-1551, Dec 2.
Ihmle P.F., 1996, Frequency-Dependent Relocation
of the 1992 Nicaragua Slow Earthquake: An Empirical Green's Function
Approach, Geophysical Journal International, 127: (1) 75-85 Oct.
Ihmle P.F., 1996, Monte Carlo Slip Inversion
in the Frequency Domain: Application to the 1992 Nicaragua Slow
Earthquake, Geophysical Research Letters, 23: (9) 913-916, May.
Kanamori H. and E. Hauksson, 1992, A Slow
Earthquake in the Santa-Maria Basin, California, Bull. Seis.
Soc. 82: (5) 2087-2096, Oct.
Kawasaki I., Asai Y., Tamura Y., Sagiya
T., Mikami N., Okada Y., Sakata M., and Kasahara M., 1995, The
1992 Sanriku-Oki, Japan, Ultra-Slow Earthquake, Journal of Physics
of the Earth, 43: (2) 105-116.
Linde A.T., Gladwin M.T., Johnston M.J.S.,
Gwyther R.L., and R.B. Bilham., 1996, A Slow Earthquake Sequence
on the San Andreas Fault, Nature, 383: (6595) 65-68 Sep 5.