Earthquakes and geological structures of the St. Lawrence Rift System - Intraplate Earthquakes - page 89

geological formations could lead to enhanced stresses at mid-crustal levels. Another hypoth-
esis is that areas that are currently active represent aftershock zones of large historic or
prehistoric earthquakes (Ebel et al ., 2000 ) .
If the SLRS connection only partially explains the St. Lawrence seismicity, what factors
can render portions of the SLRS faults inherently weak and/or enhance the local stress
levels?
Assuming that the level of tectonic stresses is everywhere similar in Eastern Canada,
there are a few factors that can weaken the faults at depth and favour reactivation. In
the brittle regime, the reactivation of a fault is controlled by the stress orientation, the
friction coefficient on the fault, and the pore-fluid pressure acting on the fault interface. The
reverse-faulting regime that prevails down to the mid-crust ( 30 km) in Eastern Canada

implies stress differences that are difficult to conceive. At 25 km, for example, a hydrostatic
pore-fluid pressure ( λ

0.4) and a coefficient of friction of 0.75 imply a critical stress
difference for sliding of about 1,200 MPa in an ideally oriented reactivated thrust fault.
These high stress differences are about one order of magnitude larger than the upper limit
of 100-200 MPa usually assumed for crustal stresses. This suggests that high pore-fluid
pressure and/or low coefficient of friction must exist along reactivated faults to give rise
to the SLRS mid-crustal seismicity (Lamontagne and Ranalli, 1996 ) . Another factor is the
orientation of the faults with respect to the maximum compressive stress (which is assumed
to be sub-horizontal in Eastern Canada). Hence, only faults that strike nearly perpendicular
to the maximum stress axis have a higher probability of being reactivated as reverse
faults.
If, on the other hand, it is assumed that all faults are equally weak to the first order,
then factors must be sought that can locally increase the stresses. Stress can concentrate in
areas with lateral mass anomalies within the lithosphere ( Goodacre and Hasegawa 1980 ;
Assameur and Mareschal, 1995 ) . In the latter study, the fact that two out of three regions
with the highest induced stress differences remain aseismic suggests that mass anomalies
can favour but not control earthquake occurrences. Other models that have been suggested
to explain the occurrence of seismicity within continental interiors include localized stress
concentration around weak intrusions (Campbell, 1978 ) , intersecting faults (Talwani, 1988 ;
Gangopadhyay and Talwani, 2007 ) and ductile shear zones in the lower crust (Zoback,
1983 ) . Some models relate seismicity to elevated temperatures at depth (Liu and Zoback,
1997 ) . In these models, plate-driving forces are largely supported by the (seismogenic)
upper crust; the lower crust is weakened as a result of higher temperatures and the total
strength of the lithosphere is reduced. Other models refer to regional stress fields perturbed
by forces associated with lithospheric flexure after deglaciation (Stein et al ., 1979 ; Quinlan,
1984 ; Grollimund and Zoback 2001 ) . Based on strain determination for Eastern Canada,
deglaciation is also used by James and Bent ( 1994 ) and Mazzotti and Townend ( 2010 )
to explain perturbed stress environments. A concentration of post-glacial rebound stresses
can occur in local zones of weakness, perhaps containing low-friction faults (Mazzotti and
Townend, 2010 ) . This model suggests local sources of weakness to reconcile the apparent
reorientation of maximum compressive stresses in the CSZ and in the LSLSZ.

=