Geology Reference
In-Depth Information
(see figure A) when the direction of slip is
reversed! These slip transients span one to
two weeks and occur about every 14 months.
Although the surface displacement is
typically a few millimeters, modeling of
these data suggests that each slow slip event
represents 2-4 cm of slip along the subduc-
tion interface between the up-dip locked
zone and the down-dip freely slipping zone
(see figures B and C). Initially, these events
were termed “silent earthquakes” because
no seismogenic signature was recognized
from them. Subsequent studies suggested
that fluid flow and seismic tremors (low-
frequency shaking) are caused by (or asso-
ciated with) slow subduction zone
earthquakes (Brown
et al
., 2005). The dis-
covery of slow earthquakes on the Cascadia
subduction zone prompted a search for sim-
ilar behavior elsewhere using continuous
GPS and seismic arrays, and, in less than a
decade, either slow earthquakes or episodic
tremor attributed to slow earthquakes have
been discovered in subduction zones in
Alaska, New Zealand, Chile, Japan, Mexico,
and Costa Rica, as well as along the San
Andreas Fault.
A big question now concerns the cause of
the apparent periodic behavior of slow earth-
quakes. Is this an intrinsic property of slip on
large fault interfaces, or are there external
drivers? One might imagine that a process
that periodically increases or decreases the
load across the fault could modulate friction
and fault slip. A potential source for such
periodic loading is the
Chandler wobble
: the
irregular wobbling of the Earth around its
rotational axis due to the fluid nature of its
core and oceans. In Cascadia, the periodicity
of the Chandler wobble matches that of the
slow earthquakes (see figure D). Stress
anomalies that could be caused by the
Chandler wobble were recently calculated
(Shen
et al
., 2005) for the transition zone
on the subduction interface where slow
earthquakes occur (see figure B). Although
little is known about the sensitivity of these
faults to small stress changes, the magnitude
of these anomalies may be sufficient to trig-
ger episodes of slow slip.
A different study on the Cocos Plate
subduction zone along the western coast of
Mexico argued that climate may modulate
slow earthquakes, which display a nearly
annual periodicity there (Lowry, 2006). In
this site, hydrologic loading (due to seasonal
rainfall variations) is predicted to produce
shear stress changes (see figure E) that are
twice as great as those predicted for
Chandler wobble. Neither of these models
produces a perfect match to the slow
earthquake record, but each tends to show
stresses peaking before the slow-slip events,
a pattern consistent with loading and slip
on better-studied faults. A challenge in
coming years as more examples of slow slip
are uncovered will be to discover whether
one or multiple causal mechanisms are
responsible for this unexpected, but striking
behavior.
campaign-mode GPS are much less than those
using permanent GPS instrumentation for the
same array. For rapid acquisition of geodetic
data across a broad region, campaigns with
mobile receivers are highly effective. Survey
sites are usually marked by a monument fixed in
an immobile substrate, such as bedrock, so that
the GPS receiver can be positioned over precisely
the same point on subsequent surveys. The
precision of GPS measurements is in part a
function of the duration of data collection. For
any given occupation of a single GPS site, a
longer interval of data collection will yield more
precise positioning information. Good-quality
data collected for 24 hours can have an
uncertainty of greater than 4-5 mm. If data are
