Stress, faults, and folds - Tectonic Geomorphology - page 78

times larger than the rup ture patch itself can
experience significant changes. Thus, slip on
one fault can increase (or decrease) the stresses
across other faults. Such loading can push other
faults closer to failure, thereby causing triggered
slip (Pollitz and Sacks, 2002). Alternatively, by
lowering stress on some faults, the time until
these faults rupture is expected to increase
(Miller, 1996).

motions, several different scenarios can be envi-
sioned to accommodate the regional strain, only
some of which would involve characteristic
earthquakes (Schwartz and Coppersmith, 1984).
In one scenario (Fig. 4.7A), displacements and
rupture lengths are randomly distributed in
such a fashion that, over time, all regional strain
is accommodated, and an essentially uniform
slip rate occurs along the length of the fault. In
this variable-slip scenario, the displacement
experienced at any given point, the size of the
earthquake, and the position of the rupture
segment each vary unpredictably between suc-
cessive events. Consequently, no faults exhibit
characteristic earthquakes in this scenario. A
second scenario (Fig. 4.7B) suggests that, at any
given point, the displacement is consistent from
one earthquake to the next, although displace-
ment can vary along strike. Over time in this
uniform-slip model, a consistent slip rate also
occurs all along the fault. In addition, the
large earthquakes have a repetitive pattern in
terms of rupture length and displacement varia-
tions along the rupture, and moderate-sized
earthquakes occur more frequently. In the char-
acteristic earthquake model (Fig. 4.7C), consist-
ent displacement at a point also occurs from
one event to the next. Over time, however, the
slip rate varies along the fault because the
cumulative displacement varies along the length
of the fault. Each large earthquake represents a
repetition of the rupture location, length, and
displacement pattern of the previous large
earthquake, and moderate-sized earthquakes
are sufficiently infrequent that they only accom-
modate a fraction of the residual slip variation
along the fault.
In viewing these models, it is easy to envision
how characteristic earthquakes along normal or
reverse faults could progressively build an irreg-
ular topography. If the site of greatest structural
displacement along a fault were to be essentially
fixed in the landscape through time, mountain
peaks and basin depocenters would occur in
predictable positions above the zone of maxi-
mum uplift or subsidence. Both structure con-
tours and the landscape topography might be
expected to be closely related to these repetitive
cycles of displacement.

Characteristic earthquakes
and fault models

It has been proposed that some faults are
typified by characteristic earthquakes , in which
a fault or a segment of a fault ruptures repeat-
edly and displays approximately the same
amount and distribution of slip during each
successive event (Schwartz and Coppersmith,
1984). If a fault did indeed display characteris-
tic earthquakes, knowledge of a single faulting
event would provide a remarkable under-
standing of both previous and future rupture
patterns, because the strain build-up and
release, stress drops during faulting, variations
in displacements along the fault, and the length
of the rupture would be approximately dupli-
cated in successive earthquakes. Thus, tremen-
dous predictive power may reside in faults that
rupture via characteristic earthquakes. If rup-
tures along a fault were controlled by stable
asperities and barriers (i.e., consistent failure
stress and terminating stress) that persisted
from one earthquake to the next, this stability
could provide a mechanism for generating sim-
ilar slip distributions along the fault in multiple
events.
In order to test whether or not a fault is
typified by characteristic earthquakes, either a
suite of well-documented historical earthquakes
or information from paleoseismic studies has to
be used to reconstruct the temporal distribution
of events and the spatial pattern of slip in any
particular event in the past. The patterns of dis-
placements of past ruptures need to be com-
pared with respect to the length and position of
the rupture and the distribution of displace-
ment along the fault. Given a regional strain
field, such as that controlled by relative plate