Geology Reference
In-Depth Information
interseismic interval. In this model, all of the
elastic strain is recovered in each seismic event,
so that no permanent strain remains within the
blocks that slip past each other on opposite
sides of the fault.
Data on pre- and coseismic deformation prior
to and following the 1946
M
=
8 earthquake
in the Nankai Trough of southwestern Japan
provide support for aspects of this model
(Fig. 4.3). Measured interseismic strain replicates
the shape, but is opposite in vertical direction to
the coseismic strain caused by that large
subduction-zone earthquake. This behavior is
clearly consistent with Reid's (1910) model.
Considerable topography, however, exists across
the deformed zone and suggests that coseismic
subsidence does not fully compensate all of the
interseismic uplift: some permanent strain is
represented by this topography.
Coseismic
Strain Release
U
x
Interseismic
Strain Accumulation
horizontal
displacement
U
x
fault plane
d
u
d
u
y
y
d
u
x
/
d
d
U
x
/
d
y
y
2
D-shear
strain
y
y
2D
dislocation model:
cross section
y
y
locking
depth
Slip
d
u
D
D
Slip rate
d
u /
d
t
Fig. 4.2
Model of the earthquake cycle for a
strike-slip fault.
The far-field strain (large arrows, top panel) of one
block with respect to the other remains constant
through time, as does aseismic slip (d
u
/d
t
, bottom
panel) in the ductile zone below the locking depth (
D
).
Interseismic displacement is greatest farthest from the
fault, but most of the shear strain (d
u
x
/d
y
, middle panel)
occurs within two locking depths (2
D
) of the fault.
During coseismic displacement, the greatest
displacement (d
u
) occurs along and near the fault and
compensates for the “slip deficit” developed during
the interseismic interval. The coseismic shear
strain is equal and opposite to the interseismic shear
strain. Modified after Thatcher (1986b).
Alternative earthquake models
As a result of observations related to numer-
ous earthquakes, Reid's (1910) simple model
of the earthquake cycle has been significantly
modified. First, the deformation observed
along faults, such as folds and fault offsets,
indicates that not all of the pre-faulting strain
is recovered. In other words, as seen in the
example from Japan (Fig. 4.3), some perma-
nent, unrecoverable strain is commonly pre-
sent. Second, the interseismic interval may
sometimes be further subdivided into a post-
seismic interval and an interseismic one. The
post-seismic interval immediately follows an
earthquake and is one during which strain
accumulates more rapidly than in the subse-
quent interseismic interval (Stein and Ekstrom,
1992). In some recent earthquakes, the amount
of post-seismic deformation and energy release
has been shown to be equal to that of the
coseismic event (Heki
et al.
, 1997). Newly
recognized “slow” earthquakes (Beroza and
Jordan, 1990; Kanamori and Kikuchi, 1993)
in which deformation occurs over periods of
hours to years can release vast amounts of
energy and cause large-scale deformation
without generating the catastrophic energy
releases of typical earthquakes. Third, new
straight marker that was oriented perpendicular
to the fault would be bent into a sigmoidal
shape. The two-dimensional shear strain, which
could be envisioned as the amount of bending
of the formerly straight marker, is greatest near
the fault (Fig. 4.2). This bending can be consid-
ered as “elastic strain” in that it is recoverable,
rather than permanent. When the frictional
strength of the fault is exceeded by the imposed
stress, the fault ruptures during an earthquake.
Coseismic displacement is greatest along and
adjacent to the fault. In the context of the
earthquake-cycle model, slip is just enough to
balance the slip deficit and restore the marker to
a linear trend, perpendicular to the fault, but
now offset by the amount of relative motion
of the two crustal blocks during the entire
