Earthquakes and geological structures of the St. Lawrence Rift System - Intraplate Earthquakes - page 88

except near the 1935 Temiscaming epicentre ( Figure 4.2 ) , where only a small portion of
the regional graben faults are active (Bent et al ., 2002 ) .
Our evidence in the SLRS suggests that the relation between earthquakes and rift faults
breaks down for the lower magnitude earthquakes (probably smaller than M 6 and certainly
less than M 4.5), where local stresses and/or fault weaknesses contribute to earthquake
occurrences. This approach was also proposed on a global scale by Schulte and Mooney
( 2005 ) and Mooney et al . ( 2012 ) . In the SLSR, and in the CSZ in particular, a study of
focal mechanisms for magnitude 2 to 5 has shown that numerous fault orientations are
reactivated in the smaller earthquakes, including mixed thrust-strike-slip events. When
studied in detail, earthquakes in sub-areas of the CSZ do not occur on rift faults, or on
some impact structure faults (Lamontagne and Ranalli, 1997 ) . Baird et al . (2009, 2010)
have emphasized the importance of local variations in stress level within the CSZ and of
strength differences between the crater and the rift faults outside the crater. They showed
that much of the background seismicity pattern can be explained by the intersection of weak
faults of the St. Lawrence rift with the damage zone created by the Charlevoix impact. In
terms of strain measurements, the first interpretations of crustal strain along the SLRS were
supportive of a homogeneous stress system, but more recent results have shown that local
stresses are anomalous in active areas such as the CSZ (Mazzotti and Townend, 2010 ) . For
smaller Eastern Canadian earthquakes, the direction of the maximum compressive stress
varies, which suggests local perturbations in the stress field in contrast to studies that favour
a regionally homogeneous eastern North American stress field (Zoback and Zoback, 1991 ) .
Another possibility for the creation or reactivation of local faults is the role of sub-
crustal processes, such as what was proposed for the NW-SE band of earthquakes of the
WQSZ. Since no surface geological feature is seen along this trend, it was proposed that
the seismicity could be due to an extension of the New England Seamount Chain track or
the passage of this region over the Great Meteor hotspot between 140 and 120 million years
ago ( Figure 4.2 ; Sykes, 1978 ; Crough, 1981 ; Adams and Basham, 1989 ) . Ma and Eaton
( 2007 ) have suggested that the passage over a hotspot, possibly imaged by lithospheric
velocity anomalies at 200 km depth, may explain the enhanced level of seismicity either by
thermal rejuvenation of pre-existing faults or by stress concentration caused by mid-crustal
strength contrast between mafic and felsic rock. If this hypothesis is valid for the WQSZ,
the hotspot would have generated crustal fractures (Adams and Basham, 1989 ) through
the generation of thermo-elastic stress (e.g., Marret and Emerman, 1992 ) . Two problems
remain to be resolved before the hotspot-seismicity link is accepted in this area: the hotspot
track has no geological expression (faulting, metamorphic grade changes, intrusives) in the
WQSZ (Sleep, 1990 ) and there is very little seismic activity along most of the length of the
hotspot track, including along the Monteregian Hills region.
Numerous hypotheses have been advanced to account for the local concentration of
WQSZ earthquakes. For example, it was proposed that the NW-SE band of earthquakes in
WQSZ represented crustal zones of weakness delineated by the drainage pattern (Goodacre
et al ., 1993 ) . A possible correlation was also suggested with positive aeromagnetic anoma-
lies and the Helikian (Mid-Proterozoic) paragneisses. The weakness of upper-crustal