Geoscience Reference
In-Depth Information
dipping away from the epicentre (Du Berger et al ., 1991 ) , this earthquake is now referred
to a SLRS earthquake. To exemplify the difficulty in correlating faults and earthquakes,
the 1982 M 5.8 Miramichi and the 1989 M 6.3 Ungava earthquakes, both shallow focus
earthquakes, do not correlate with any fault of regional extent (Wetmiller et al ., 1984 ;
Lamontagne and Graham, 1993 ) . These events are reminders that faults with dimensions
necessary to generate an M 6 earthquake (about 10 km) exist almost everywhere in the
Canadian Shield, even where rift faults are not present. Consequently, we conclude that
smaller magnitude earthquakes ( M < 4.5), which represent the vast majority of earthquakes
in Eastern Canada, do not necessarily represent reactivated SLRS faults.
We believe that some global studies have partaken in the confusion because they did
not look at the SLRS at the local level. For example, Schulte and Mooney ( 2005 ) used
a continent-wide view to suggest a spatial correlation of the SLRS seismicity with the
Appalachian front to the east (which is known to be inactive). They also correlate SLRS
with the epicentre of the 1732 Montreal M 5.8 (which they rate as M 6.3) earthquake,
whichGouin ( 2001 ) locates in Northern New York State from the same intensity data. We
argue that the current earthquake database does not have a sufficiently large number of
earthquakes with reliable epicentres to strongly support a correlation between earthquakes
and the entire SLRS. Another example of a continental-scale view of seismicity is the
suggested correlation between earthquakes and low-angle thrust faults (“sutures”) of the
Grenville Province (Van Lanen andMooney, 2006 ) .We note that numerous aseismic areas of
the Grenville Province have these low-angle thrust faults. Recently, Mooney et al .( 2012 )
suggested a correlation between crustal seismicity and younger lithosphere surrounding
the ancient cratons, with aseismic zones corresponding to areas with thick (and cold)
lithosphere.
One must note that the strongest support of a SLRS-seismicity connection is based on
the locations of earthquake epicentres, and hypocentres in the CSZ, where SLRS faults are
present. We believe that the regional-scale model, suggesting a seismic hazard based on the
sole presence of SLRS faults, may not hold if one considers the local scale of the various
seismic zones. In the CSZ, for example, the hypocentre-rift-fault connection suggested
by Anglin ( 1984 ) is based on about 6 years of earthquake recording by a local network.
Today, after some 35 years of recording, a more complex picture emerges, with numerous
sub-zones, each one with its special focal depth distribution, level of activity, and focal
mechanism complexity. There is some control by the SLRS faults, but they appear to bound
seismically active blocks rather than being active themselves (Lamontagne, 1999 ) .
Another problem with the rift fault hypothesis is that active regions are separated by
weakly active areas ( Figure 4.1 ) . Tuttle and Atkinson ( 2010 ) did not find any evidence of
Holocene earthquake activity in the Quebec City to Trois-Rivieres segment, despite the
presence of SLRS faults. These authors suggest that the seismic activity at Charlevoix may
be localized. This suggests that local factors lead to seismicity in recognized seismic zones,
not uniquely the presence of SLRS faults. In the WQSZ as well, the seismicity-rift-fault
correlation does not explain that most earthquakes locate well to the north of the graben
faults in a NW-SE alignment. In general, the Ottawa-Bonnechere graben is weakly seismic
Search WWH ::




Custom Search