Cyclic stressing and seismicity at coupled subduction zones
M. A. J. Taylor1, G. Zheng1, J. R. Rice1,
R. Dmowska1, and W. D. Stuart2
1 Department of Earth and Planetary Sciences and Division of
Engineering and Applied Sciences, Harvard University, Cambridge, MA 02138,
U.S.A.
2 USGS, 345 Middlefield Road, Menlo Park, CA 94025, U.S.A.
EOS Trans. Amer. Geophys. Union, 76, No. 46, F533, 1995.
The finite element method is used to analyze the variation of stress
and seismicity at a coupled subduction zone during an earthquake cycle.
A generic 2D plane-strain model is used with fixed dip and plate thickness
and periodic earthquake slip is imposed along the thrust interface consistent
with the long term rate of plate convergence and degree of coupling.
The stress field is simulated by periodic fluctuations in time superimposed
on a time-averaged field. The oceanic plate, descending slab and
continental lithosphere are assumed to respond elastically to these fluctuations,
and the remaining mantle under and between plates to respond as Maxwell
viscoelastic. The computed fluctuating stresses are generally consistent
with observed earthquake mechanism variations with time since a great thrust
event. In particular, trench-normal extensional earthquakes tend
to occur early in the earthquake cycle in the outer rise, but occur more
abundantly late in the cycle in the subducting slab below the main thrust
zone [e.g. Lay et al., PEPI 1989]. Compressional earthquakes,
when they occur at all, have the opposite pattern. In the outer rise
the time averaged stresses can be approximated by a bending field with
stresses near to conditions of normal faulting at shallow depths and compressional
failure at greater depths. The periodic fluctuating stresses, although
small (on the order of a few bars for representative values of model parameters)
are thus sufficient to modulate the time averaged stresses and so induce
failure in these regions, effectively controlling the timing of seismicity.
Increasing the relaxation time of the mantle or the slab thickness, or
decreasing the dip angle, reduces the magnitude of early post-seismic strain
rates and causes the fluctuating part of the stress to change sense, from
tensional to compressional or vice versa, later in the cycle.
For purely elastic response outside the thrust zone, the fluctuating
part of horizontal stress in the outer rise increases to tension at the
time of the main thrust event and then decreases at a constant compressive
rate throughout the cycle. When mantle viscoelasticity of a short
enough relaxation time is included, we find that at a few slab thicknesses
from the trench, the tension continues to increase slightly for a period
after the main event before a significantly compressive stressing rate
resumes. Near the trench, the stressing rate is still notably compressive
throughout the cycle, which is inconsistent with the extended time scale
of seismicity there. However, when a thin aseismically slipping zone
with viscous rheology is introduced between the trench and deeper portions
of the thrust interface [as in Dmowska et al., EOS 1994], the extensional
stress in the outer rise continues to increase in the early part of the
cycle, consistent with the extended period of tensional earthquakes in
that region. The duration of the increase in tension is controlled
by the relaxation properties of the aseismic zone. For example, for
a thrust contact of length 100 km (the shallowest 37 km being aseismic),
we find that in order to produce stress histories in the outer rise about
20 km from the trench, which continue to rise for times around 20 years
after the great thrust event before stressing rates turn compressional
there, 
/
h
must be of the order of 0.1 yr/km for a recurrence time of 100 years (h
is width and
is viscosity of the aseismic zone,
is rigidity of the surroundings). For stresses to rise for around
10 years, /h
is of the order of 0.05 yr/km. The /h
values are about 50% larger for recurrence intervals of 50 years.
return
to Publications List