The Charlevoix Seismic Zone:
Overview of the earthquake activity and the geological environment
1.1 Introduction
Located some 100 km downstream from Quebec City, the Charlevoix Seismic Zone (CSZ)(1) is the most seismically active region of eastern Canada (Figure 1.1). Historically, the zone has been subject to five earthquakes of magnitude 6 or larger: in 1663 (magnitude 7.0); 1791 (magnitude 6.0); 1860 (magnitude 6.0); 1870 (magnitude 6.5); and 1925 (M(2) 6.2) (magnitudes as listed in Anglin et al., 1990; Bent, 1992). The earthquake potential of the CSZ led the Earth Physics Branch (EPB; now part of the Geological Survey of Canada: GSC) to conduct two field surveys, in 1970 and 1974 respectively (Leblanc et al., 1973; Leblanc and Buchbinder, 1977). The 63 hypocentres located during these two surveys clearly defined the boundaries of the CSZ, i.e. an active zone about 30 by 85 km, elongated along the St. Lawrence River, enclosing the towns of Baie-St-Paul, La Malbaie and La Pocatière (Figure 1.2). In 1977, the EPB deployed a seven-station local seismograph network centred on the active zone defined by the field surveys. Between 1978 and 1997 inclusively, the network detected nearly 2200 local earthquakes, of which 54 exceeded Nuttli Magnitude (mN) 3.0 (Figure 1.3). The current CSZ network detects more than 200 earthquakes per year. Based on historical and current earthquake rates, the CSZ is the highest seismic hazard zone in continental eastern Canada.
The hypocentres located (3) over the years have provided an insight into the CSZ seismotectonics. Most earthquakes cluster along or between the mapped Iapetan faults (also called St. Lawrence paleo-rift faults; Anglin, 1984; Adams et al., 1995). Under the north shore of the St. Lawrence River, for example, hypocentres define a SE steeply dipping plane. When projected to the surface along the trend and dip of that steeply dipping plane, the hypocentres are bounded by two Iapetan faults (the St-Laurent and the Gouffre NW faults; Figure 1.4). Based on this evidence, Anglin (1984) concluded that "almost all the activity in this section of the St. Lawrence River valley is related to preexisting steeply dipping faults that strike along the direction of the valley" (p. 602). Outside the CSZ, the Iapetan fault system extends along the St. Lawrence, Saguenay and Ottawa rivers. Therefore, in these areas, the earthquake hazard may be higher than suggested by the low level of historical seismicity. This geological control of earthquake occurrences has been included in the recent national seismic hazard maps of Canada (Adams et al., 1996).
1.2 Objectives and outline of the dissertation
This dissertation uses newly available information to examine the factors that control the occurrence and spatial distribution of CSZ earthquakes. The current knowledge of CSZ seismotectonics is mainly based on the findings of Anglin (1984). Since then, the scientific database has increased considerably in quantity as well as quality. While Anglin (1984) relied on some 250 hypocentres recorded between October 1977 and May 1983, 1300 additional earthquakes were located between June 1983 and December 1997, using mostly three-component digital data. These high quality records can be used to study focal mechanisms, earthquake clusters, and crustal velocities. In addition, since the work by Anglin (1984), advances have occurred in several related fields: crustal processes leading to fault instabilities; earthquake clustering; and temperature distribution controlling rock rheology. Finally, faults, especially those under the St. Lawrence River, can now be inferred by improved techniques to analyze the gravity and magnetic fields, together with radar images and seismic profiles.
The dissertation is divided into six chapters. In Chapter 1, introductory information on the CSZ geology and earthquakes is presented. The section on seismicity covers the earthquake monitoring, locations, magnitude, frequency and distribution of events. In Chapter 2, factors that control earthquake occurrences (temperature distribution, rheology, stress differences) are presented and applied to the CSZ. In Chapter 3, an update to structural geology of the CSZ is presented and compared with the seismicity. The section includes the description of new faults, mainly under the St. Lawrence River, as revealed with new data sets and techniques. In Chapter 4, focal mechanisms, earthquake groups and multiplets are integrated into a seismotectonic analysis of the CSZ. In Chapter 5, local earthquake data are used to invert local and regional velocity models. Finally, in Chapter 6, the factors leading to seismicity in the CSZ are discussed based on the results of the previous chapters.
1.3 Geology of the CSZ
Geomorphologically, the CSZ is an area of contrasting topography, with elevations varying between sea level and 1171 meters (Figure 1.5A). While the south shore is a gently rolling landscape, the north shore is a mixture of rugged highlands, plateaus and valleys, separated by dramatic changes in elevation (Figure 1.6/Figure 1.6BCD/Figure 1.6EF). To the west and northwest of the CSZ, the Laurentian Plateau is an area of high elevation (up to 1171 m above sea level), cut by numerous steep valleys (Figure 1.5B). To the north and east, tectonic events and glacial erosion created a gentler relief. To the north of the CSZ, the plateau is cut by a series of EW normal faults related to the creation of the Saguenay Graben in late Precambrian to Ordovician (700-500Ma). Along the St. Lawrence River, low relief generally corresponds to the hanging wall of normal faults of similar age. In addition, a Devonian meteorite impact (350Ma) has shattered the plateau, creating a semi-circular depression 56 km in diameter. The centre of the crater is a 768 m high central peak, Mont des Eboulements, which is surrounded by an interior plateau of up to 15 km radius, and by a peripheral depression of up to 27.5 km radius (Rondot, 1989). In the Quaternary, the multiple passages of the glaciers have preferentially eroded the shattered impact structure and the regional fault zones to the north and west of the CSZ. The St. Lawrence River covers about 25% of the surface of the CSZ. The deepest areas can be 150 m deep near the north shore. Southeast of Ile-aux-Coudres and Ile-aux-Lièvres, the river is generally shallow and the underwater relief very gentle.
Four main geological assemblages make up the geology of the CSZ: the Precambrian Shield of Grenvillian age, the Ordovician St. Lawrence platform, the Appalachian nappes and the Quaternary deposits (Rondot, 1979; Figure 1.7). The main lithologies of the Precambrian Shield are the charnockite-mangerite assemblages, meta-sediments, anorthosite intrusives and local gabbro dykes within fault zones. The charnockite-mangerite assemblages of the Laurentides Park Charnockite Complex cover most of the western part of the region. Locally, the complex is intruded by magmatic bodies, such as the St-Urbain Anorthosite. The Ordovician sedimentary rocks, mainly limestones, rest unconformably on the Precambrian basement, as thin erosion remnants on the north shore, and as thick sedimentary sequences under the St. Lawrence River. The Appalachian sequences are made up of Cambrian sedimentary rocks (sandstones, mudstones) thrust over the Precambrian and Ordovician rocks. The surface expression of this thrust plane is often referred to as Logan's Line. Finally, Quaternary deposits cover the previous assemblages. On land, these Quaternary deposits are generally thin but pervasive sequences of glacial till and marine clays. Offshore, deposits hundreds of meters thick are found in elongated valleys filled with multi-episodic interglacial sedimentary sequences, covering more than 250,000 years of the St. Lawrence River's history (Occhietti et al., 1997).
All CSZ earthquakes occur within the Precambrian basement (Leblanc and Buchbinder, 1977). The seismogenic basement is cut by faults created during four major tectonic events: the Grenvillian collision (1100 to 900 Ma); the rifting episode related to the opening of the Iapetus Ocean ( 700 Ma); the Taconian reactivation of these faults at the closing of that ocean ( 450 Ma); and finally, a Devonian meteor impact ( 350 Ma; Rondot, 1979). In addition to being mapped in the field, most major faults correspond to strong lineaments in remote sensing imagery. Under the St. Lawrence River, geological structures in the seismogenic Precambrian basement are hidden by several kilometres of Appalachian nappes (Lyons et al., 1980) and hundreds of meters of Quaternary sediments. Chapter 3 describes how some fault positions can be determined with gravity, magnetics, remote sensing and seismic profiles.
1.4 Earthquakes of the CSZ
1.4.1 Earthquake monitoring
Between 1977 and 1998, the Charlevoix region has been the only eastern Canadian seismic zone with a seismograph network sufficiently dense for routine hypocentre determinations. Since October 1977, the area has been monitored by a microseismic network of between 6 to 8 stations located on both shores of the St. Lawrence River (Figures 1.8 and Figure 1.9). Between 1977 and 1988, the Charlevoix seismograph network consisted of a six to seven component vertical short-period analogue network. The signals from the stations were radio-transmitted to a central node where they were archived on a magnetic tape together with a time signal from a local clock. Weekly, the tapes were sent to the head office in Ottawa. Based on events seen on the LMQ analogue records, an analyst (generally F. Anglin of the EPB) would digitize the analogue data at 60 samples/s and print a selected time window of the data with a time scale of 1.5 cm/s (Anglin and Buchbinder, 1981; Figure 1.10. The Y-scale of the traces (gain) was adjusted to ease the phase picking on paper play-outs. Two stations completed the Charlevoix network, one analogue station on the north shore (LMQ) and one station on the south shore (at first, a low gain analogue station, POC, then, an Eastern Canadian Telemetered Network (ECTN) station, LPQ).
In November 1988, the Charlevoix local network became the Charlevoix Local Telemetered Network (CLTN), a digital three-component short-period array. The data, sampled at 80 Hz, were radio transmitted to a central node where timing was added. There, a detection algorithm analyzed each vertical component for seismic events, and, for each trigger, stored a time slice of the whole CLTN data. Daily, the data were transferred to Ottawa and analyzed with an interactive analysis package (Figure 1.10C). Later, in January 1994, the station LMQ became a digital broadband Canadian National Seismograph Network (CNSN) station, with GPS timing and continuous archival of the data in Ottawa. Finally, in August 1994, each CLTN station was upgraded to a high gain short-period digital instrument with 100 Hz sampling rate and GPS timing. As with any CNSN station, an automatic trigger algorithm continuously scans the CLTN data for events of interest. To detect very small events, LMQ analogue seismograms are also examined. To complete the CSZ network, regional stations provide first motion data and peak amplitudes for magnitude calculations (Figure 1.11).
Between June and November 1996, additional stations were deployed to help define focal mechanisms of micro-earthquakes and to provide additional constraints on the simultaneous velocity inversion for this Ph.D. dissertation. Up to 8 additional sites were occupied by portable seismographs (Lamontagne et al., 1997; Figure 1.12). Six stations were digital three-component short-period instruments sampling at 200 Hz and two were analogue MEQ-800 instruments, used for picking first motions. The network detected some 120 events during the survey.
1.4.2 Hypocentre Locations
CSZ hypocentres are routinely located with direct P (Pg) and S (Sg) phases recorded by the local Charlevoix network. Phases from stations outside the CSZ are not used due to the poorly constrained regional velocity model. The GSC program "GRL", based on a grid search algorithm, locates hypocentres with a homogeneous half-space velocity model (6.2 km/s for Pg; 3.57 km/s for Sg). This velocity model does not consider the lower velocity Appalachian rocks, on which three stations are located. With a wedge of lower velocity material, most hypocentres shift by 0.3 to 1.5 km towards the southeast (Lamontagne, 1987). In this thesis, most hypocentral maps show earthquake locations computed with the homogeneous velocity model. The seismic velocity aspects are examined in more details in Chapter 5.
Most CSZ events locate within the seismograph network, with epicentral distances varying between 0 and about 80 km. The average Root Mean Square (RMS) for the solutions is about 0.08 s, with 90% of the RMS less than 0.12 s. These errors are due to the imprecision of phase picking and to the imperfect velocity model. The solutions have average formal errors of ± 0.6 km in epicentral location and ± 1 km in depth, for events within the network (Anglin, 1984). These errors can be larger for events outside the network.
The picking precision of Pg and Sg phases has changed over the years. In the location process, weights are assigned to the phases according to their quality (most Charlevoix stations are quality "A" (± 0.25 second; weight = 4), while some can be B (± 1 second; weight = 1) or C (± 4 seconds; weight = 0.25). Since the original analogue Charlevoix array had only vertical seismometers, it is probable that Sg phases were not as precise as they have been since November 1988, when the CLTN stations became three-component. Since then, most Pg arrivals are picked on the vertical component, while most Sg phases are measured off the horizontal components.
1.4.3 Magnitudes for CSZ events
Since the beginning of operation of the Charlevoix local network, in October 1977, and December 1997, 2164 earthquakes were located in the CSZ. As for other eastern Canadian earthquakes, the magnitude scales ML (Richter or local magnitude) and mN (Nuttli) are used for CSZ earthquakes. Located CSZ earthquakes vary in size between magnitude ML -0.5 and mN 5.0. At the Geological Survey of Canada (GSC), the mN and the ML scales are used to measure eastern Canadian earthquakes. The following descriptions of the two scales are based on Drysdale et al. (1990).
Nuttli (1973) defines the mbLg scale for two distance ranges: 50 to 400 km and 400 to 800 km. In eastern Canada, the first formula did not give acceptable results in its defined appropriate distance range. Thus, the mN scale used by the GSC is an extrapolation of the second formula to epicentral distances between 50 and 400 km (R.J. Wetmiller, pers. comm.). The formula applied by the GSC is as follows
where D is the epicentral distance in km; A is the half peak-to-peak trace amplitude of the S phase in millimetres; T is the period of the trace at the maximum; K is the trace magnification in thousands for that period.
The mN scale is used for events recorded by stations beyond 50 km. On a given seismic trace, the peak-to-peak trace amplitude of the Lg wave is measured if not over attenuated. If recorded only within 50 km epicentral distance, the event is rated on the ML scale. The ML scale is defined as (Gutenberg and Richter, 1956)
where K is the trace magnification in thousands at period T, KW is the magnification of the Wood-Anderson seismograph for that period, and log10 A0(D) is the calibration function specified by Richter (1958). This scale was defined for California earthquakes, where the attenuation is higher than in the Canadian Shield. According to Leblanc et al. (1973), the ML scale can be deficient at distances of less than 50 km, especially in respect to the attenuation of high frequency waves in the near-field, and the local variations of site amplification, radiation pattern and rock properties. Consequently, the ML scale is used only as a relative measure of the size of CSZ earthquakes.
For the CSZ, the two magnitude scales are approximately related as (Figure 1.13)
This relation was obtained by applying with robust statistics (Huber, 1981) to the data set described in the figure caption of Figure 1.13. This empirical relation remains approximate due to the large data scatter. Ideally, a uniform magnitude scale for the CSZ earthquakes could be defined with a spectral level analysis (moment magnitude), which we did not attempt in this study. Until then, predictions based on variations of b-values for micro-earthquakes cannot be properly evaluated.
For the time period 1970-1997, the number of located earthquakes is closely related to the number and sensitivity of the local stations (Figure 1.14). Prior to the deployment of the Charlevoix local network in October 1977, only a few tens of earthquakes could be located with the regional stations. The 1970 and 1974 field surveys revealed a high level of earthquake activity, undetected by the regional stations. With the local network, the annual number of located earthquakes increased to nearly 100 in 1978. At the same time, the magnitude threshold decreased from mN 3.0-3.5 to mN 2.0 (Figure 1.14).
As described above, two magnitude scales are used for CSZ events. Between 1977 and 1997, a gradual increase in the number of mN magnitude events has occurred (Figure 1.15). This increase is closely related to changes in the number and the type of stations at regional distances. In the early 1980's, the mN scale, extrapolated down to 50 km epicentral distance, started to be used routinely in eastern Canada. At the same time, additional stations of the ECTN were installed within 200 km of the CSZ. These factors explain the increase in mN 2.0 events and the corresponding decrease in ML 1.0 events at around year 1980 (Figure 1.15). Starting in December 1988, the magnitude of smaller mN events started to be calculated with the installation of the ECTN station Lac Daran (DAQ), located some 90 km west of the CSZ. In August 1994, the CLTN and station DAQ became part of the CNSN network. The continuous archiving of the data allowed the location of small events, previously undetectable, and the calculation of mN magnitudes with DAQ data (Figure 1.15). These two factors led to an increase in the number of located events (ALL) and mN 1.0 (M1+) events. Finally, the 1996 June to November field survey caused a peak in the number of located events, largely due to location of many ML < 0 events.
The initial year of complete reporting of different magnitude levels can be estimated. According to Basham et al. (1982), all mN 2.8 earthquakes have been located since 1968 (Table 1.1). In the period 1970-1976, however, a rather small number of earthquakes with mN 3.0 was recorded (three years out of six with no earthquakes; Figure 1.14). Two possibilities exist: a quiescent period, or a period with a detection threshold > mN 3.0. The apparent increase in the number of mN 3.0 events after LMQ and the local network became operational (in November 1976) is more in line with the second interpretation. The location threshold for events in the 1968-1977 period is therefore conservatively taken to be mN 3.5.
Table 1.1 Estimated first year of complete reporting of magnitude levels in the CSZ
(Basham et al., 1982). The magnitudes are defined to one-tenth unit; each category includes
earthquakes in a half-magnitude range; ex: magnitude 5.5 includes mag. 5.3- 5.7.
3.0 |
3.5 |
4.0 |
4.5 |
5.0 |
5.5 |
6.0 |
6.5 |
7.0 |
1968 | 1963 |
1937 |
1928 |
1920 |
1900 |
1800 |
1660 |
1660 |
For the period from October 1977 to December 1997, the magnitude of complete reporting is closely related to the early stages of the Charlevoix local network, between October 1977 and November 1988. During that period, the analogue records of station LMQ served as a visual detector of CSZ earthquakes, all located within 70 km epicentral distance. The detection of an event on the LMQ records occurred for an amplitude of about 2 mm (zero-to-peak). For comparison, on 19970217 at 19:47, an mN 1.7 event, located 55 km from LMQ, gave a peak-to-peak amplitude of 6.7 mm on LMQ. From this, we can estimate that a 2-mm amplitude translates into a mN 1.2 (ML -0.3). For the period 1977-1997, the location completeness is therefore conservatively taken to be mN 1.5 (or ML 0.2 based on Equation 1.3). After the conversion to the CNSN in August 1994, the location threshold is probably lower: during the 1996 field survey, most events missed by routine analysis, but detected by the temporary stations, were smaller than ML < 0.0 (Lamontagne et al., 1997).
1.4.4 Frequency of occurrence of CSZ earthquakes
For the period 1977-1997, nearly 2000 earthquakes have been located in the CSZ with magnitudes varying between ML -0.5 and mN 5.0 (Table 1.2). The distribution of earthquakes in time and space was examined interactively without finding any conclusive pattern. To illustrate this, the epicentre map was divided into sub-zones within which hypocentres showed similar characteristics (closely grouped hypocentres with similar depth distribution; Figure 1.16). While some zones show almost continuous activity (zone 10 for example), others have sporadic activity followed by quiescence (zones 3 and 15 for example; Figure 1.17/Figure 1.17B). Magnitude mN 4.0 events do not seem to follow a definite pattern with other events in their surroundings; they can start, continue or end a period of increased activity.
An earthquake occurrence rate larger than 30 events per 30 days was thought to provide a warning for an imminent magnitude > 5.5 earthquake (Buchbinder et al., 1988). The activity plot by these authors was based on events detected on LMQ (and not necessarily located by the network). Since their data are lost and the exercise ended in the late 80's, I decided to produce a similar plot with the located events, keeping in mind that the local network capability controls the total numbers (Figure 1.18). During the time period 1977 to 1997, the number of located earthquakes varies between 0 and about 30 per 30-day period, similar to what was reported for the period 1977-1985. No rate increase is noticed prior to an event of magnitude mN 4.0.
Using the magnitude completeness and the ML-mN relation, the magnitude-recurrence curve was determined (Figure 1.19). For the period October 1977 to December 1997, the log(cumulative rate) is a linear function of the magnitude, up to about magnitude 4. The best estimate of the slope is 2.14. At about magnitude > 4, the slope changes, implying a higher rate for larger events than suggested by the low magnitude earthquakes. This may suggest that the magnitude > 4 events have overestimated magnitudes, or that there is a real change in occurrence rates at higher magnitudes.
Table 1.2 Number and yearly rate of located CSZ events by magnitude range for the period October 1977 to December 1997. Due to the ML-mN magnitude overlap, the numbers are the total number of earthquakes of both magnitude types. Consequently, some annual rate statistics were not computed where they are considered not representative (N.R.).
All located events |
M 0.0 (ML and mN) |
M 1.0 (ML and mN) |
M 2.0 (mostly mN) |
M 3.0 (mN) |
M 4.0 (mN) |
M 5.0 (mN) | |
Number |
|
|
|
|
|
|
|
Average Yearly Rate |
N.R. |
|
|
|
|
|
|
1.4.5 Spatial distribution and focal mechanisms of CSZ earthquakes
To define the characteristics of the earthquake activity of the CSZ, a hypocentre data set must be defined. From the 2164 earthquakes located between October 1977 and December 1997, some 1450 hypocentres were selected based on at least 5 local stations with a minimum of 8 Pg and Sg phases (see Figure 1.3). This selection insures an acceptable precision to most hypocentre locations (same as in Anglin, 1984). Considering their distribution in respect to depth, two-thirds of Charlevoix earthquakes occur between 7 and 17 km (Figure 1.20), with more than 80% at less than 15 km. Interestingly, this is also the depth range where two large events occurred (1925 mb 6.5: 10 ± 2 km; Bent, 1992; and 1979 mbLg 5.0: 10 ± 2 km; Hasegawa and Wetmiller, 1980). The concentration of large earthquakes at 10 km depth may correspond to the strongest part of the crust (Wetmiller and Adams, 1990).
Across the CSZ, the earthquake depth distribution varies; shallow (<5 km) and deep (> 20 km) events, for example, concentrate in certain areas Figures 1.21 A / B / C / D / E. On a section perpendicular to the St. Lawrence River (Figure 1.21C), the hypocentres define two major groups, one that dips steeply to the southeast, and one located under the River without obvious alignment. The section parallel to the River shows the focal depth variations along that trend (Figure 1.21C). Events with magnitude mN 4.0 concentrate near the two zones of magnitude 4.0 for the period 1924-1978 (Stevens, 1980; Figure 1.23A). Earthquakes of smaller magnitude (mN 3.0 events; Figure 1.23B; mN 2.0 events; Figure 1.23C) do not cluster in one particular subarea.
CSZ focal mechanisms are quite variable in orientation and, to a lesser extent, faulting style (Figure 1.24). It is generally assumed that, on the average, most Charlevoix earthquakes occur as thrust events on preexisting SE steeply (60o) dipping faults. These focal mechanisms, along with additional ones, are examined and discussed in Chapter 4.
Footnotes
1. The name Charlevoix associated with the seismic zone is somewhat misleading: most of the earthquakes occur under the St. Lawrence River, between Charlevoix County on the north shore and Kamouraska County on the south shore. To consider this aspect, the CSZ is sometimes referred to as the Charlevoix-Kamouraska Seismic Zone. In this dissertation, we will use the traditional name CSZ of Basham et al. (1982).
2. M is the moment magnitude, formerly listed as Mw in some publications.
3. In this dissertation, the term located refers to an event for which an epicentre (and often an hypocentre) can be determined. It differs from the terms recorded and detected earthquake, which represents an event recorded on one station or more, but not necessarily located. Since only located earthquakes are kept in the GSC database, the discussion will only refer to located earthquakes.
Figure 1.1 Earthquakes in eastern Canada. Earthquakes were selected on a completeness basis to ensure a fair representation of the seismicity (historical and instrumentally recorded; Anglin et al., 1990). Damaging earthquakes, with magnitudes in brackets, include: 1663 and 1925 in Charlevoix; 1732 near Montréal; 1929 Grand Banks; 1935 Timiskaming; 1944 Cornwall-Massena; and 1988 Saguenay. Most eastern Canadian earthquakes occur in these seismic zones: Grand Banks; Western Quebec; Lower St. Lawrence and Charlevoix.
Figure 1.2 Boundaries of the Charlevoix Seismic Zone (CSZ) after Basham et al., (1982). Some place names referred to in the text are shown.
Figure 1.3 Elevation in meters and earthquake hypocentres of the CSZ for the period October 1977-December 1997. Only events with 8 phases from 5 stations are shown. Triangles are the stations of the Charlevoix network. Major faults (black lines) from Rondot (1979).
Figure 1.4 CSZ hypocentres of the period October 1977 to May 1983 projected to the surface along a plane oriented N52oE and dipping 70oSE (from Anglin, 1984). Major faults (black lines) from Rondot (1979). Annotations A and B refer to faults that appeared to bound the seismic activity. The 6-fathom line appeared to bound the earthquakes to the SE.
Figure 1.5 Topography of the CSZ with elevation in meters. (A) With major faults of Rondot (1979) and seismograph stations (triangles). (B) Chromo-stereoscopic image of the north shore of the CSZ integrating RADARSAT-SAR ortho-image with the digital elevation model. The colour range varies from -50 m in blue to 1082 m in red. The texture of the land surface comes from the RADAR data (compliment of Thierry Toutin, Canada Centre for Remote Sensing; Toutin and Rivard, 1997).
Figure 1.6 Scenes of the Charlevoix Seismic Zone. (A) Location map; (B) Gouffre river valley which represents a part of the peripheral graben of the impact structure; (C) Linear coastline of the north shore, created by normal faults; (D) The Malbaie river U-shape valley, created by a fault and deeply eroded by the glaciers; (E) View of the impact structure from La Malbaie and of its central peak, the Mont des Éboulements; (F) View of the impact structure from Baie-St-Paul. The Ile-aux-Coudres to the left belongs to the Appalachian nappes.
Figure 1.7 Main geological features of the CSZ overlaying the elevation model. Tectonostratigraphic units are grouped as Precambrian units (dotted), Ordovician sedimentary rocks (bricks) and Appalachian rocks (slanted bricks). Logan's Line (LL) represents a structural discontinuity between the complexly deformed rocks to the southeast (the Appalachian nappes) and the undisturbed rocks to the northwest (the Canadian Shield and its Ordovician sedimentary cover). Major faults positions (black lines) of the north shore are from Rondot (1979).
Figure 1.8 Time-history of the Charlevoix network between 1970 and 1998. Three field experiments took place: in 1970: Leblanc et al., 1973; in 1974: Leblanc and Buchbinder, 1977; and in 1996; Lamontagne et al., 1997. Acronyms used: (1) LPQ (La Pocatière), SHQ (Saint-Hilarion), LMQ (La Malbaie); and A56, co-located with LMQ, are seismograph stations; (2) CLTN: Charlevoix Local Telemetered Network; (3) CNSN: Canadian National Seismograph Network; (4) sp: short period; (5) bb: broad band; (6) CSZ network: Charlevoix Short-Period Local Network (6 stations).
Figure 1.9 Location of the permanent seismograph stations in the CSZ. Some stations were relocated over the first few years, without changing the general distribution of stations. Since 1985, the stations are: A11; A16; A21; A54; A61; A64. Other stations are: broadband (LMQ); ECTN digital short-period (LPQ); and analogue (POC; LMQ until 1993; SHQ).
Figure 1.10 Examples of how arrival times have been picked over the years. Traces of the 19840214 09:01 mN 2.5 event from which Pg and Sg phases were picked during the period October 1977 to November 1988. (A) High gain display. The top and bottom X axes are second marks. The six traces correspond to stations A10; A16; A20; A64; A54 and A61. Arrival times were measured by interpolating between the two vertical 10 second lines. Since only the vertical traces existed, Pg phases were generally more evident than the Sg phases. (B) Low gain display. Some Sg phases were picked on this plot. C) Example of a computer display used to pick Pg and Sg phases after 1993 (software "dan" by Nanometrix Inc). The icons (top) represent the various options to get, display or analyse data. The lower three traces are the three components of station A54 (Z: vertical; N: north-south; E: east-west). The time scale for these three traces can be seen at the bottom (40 seconds total). The top trace represents an enlarged view of 2 seconds of the trace A54..EHZ. Phases are picked in the zoom window, with adjustable X and Y scaling.
Figure 1.11 Regional stations surrounding the CSZ (triangles), digital (DPQ; DAQ) and analogue (CIQ; QCQ; SLQ).
Figure 1.12 Seismograph stations in the CSZ (triangles) during the 1996 summer field survey; red, permanent; blue, digital portable; black, analogue; Lamontagne et al., 1997).
Figure 1.13 Correlation between mN and ML magnitudes, computed with stations beyond 50 km epicentral distance for mN, and calculated from local Charlevoix stations at 25 to 50 km epicentral distance for ML. A total of 888 data points were used, which implies that many data points are superimposed especially between ML -0.5 and mN 1.0. The optimum curve is obtained with a linear regression routine.
Figure 1.14 Distribution of located earthquakes for the period October 1977 - December 1997 with respect to different magnitude (M) thresholds: (top): for all events (ALL); M 0; and M 1; (bottom): for M 2, M 3 and M 4 (circles). No distinction is made between the ML and mN magnitude scales. Major changes in the local network are shown in the upper diagram.
Figure 1.15 Number of events for various magnitude thresholds: (1- top): mN magnitude; (2- bottom) ML magnitude.
Figure 1.16 Sub-zones around earthquake clusters with similar depth distribution. These sub-zones are discussed in Chapter 4.
Figure 1.17 Time-history of the activity of the various sub-zones of Figure 1.16 for: (A) the period October 1977 - December 1987; (B) the period January 1988 to December 1997.
Figure 1.18 A 30 day histogram of located earthquakes in the CSZ, between 1970 and 1997. Events of magnitude mN 4.0 are shown with stars. In March 1989, an earthquake doublet occurred (shown by a star at Y = 2).
Figure 1.19 Magnitude-recurrence curves for the CSZ. The data points in red represent some 1374 earthquakes that meet the magnitude completeness criteria of Table 1.1 (for events until October 1977), plus our magnitude completeness criteria defined in Section 1.4.4 (for events between October 1977 and December 1997). The red curves represent the magnitude-recurrence curves (minimum-average-maximum) based on the data set. The blue dots and the blue curves represent the obtained curves without the data points that meet our completeness criteria for 1977-1997 (less than 300 points).
Figure 1.20 Focal depth distribution of the Charlevoix earthquakes for the period October 1977 - December 1997. Events are grouped in one kilometer intervals. Two curves are drawn: one for the number of earthquakes at a given depth, and one for the cumulative percentage.
Figure 1.21 CSZ hypocentres and stations (dots and triangles respectively) plotted on an elevation map of the CSZ: (A) with 0 to 5 km depth; (B) 5 to 10 km depth; (C) 10 to 15 km depth; (D) 15 to 20 km depth; (E) deeper than 20 km. Dot colours represent the focal depth of the events.
Figure 1.22 CSZ hypocentres: A) hypocentres (dots) and stations (triangles) with the end points of the two cross-sections A-A' and B-B' plotted on an elevation map of the CSZ ; (B) Vertical cross-section perpendicular to the St. Lawrence River. The width of the active zones may represent the actual width of the zone where fractures are reactivated as well as variations of fault trends. The red circles are the magnitude 4.0 earthquakes; (C) Vertical cross-section parallel to the St. Lawrence River. The red circles are the magnitude 4.0 earthquakes.
Figure 1.23 Hypocentres by magnitude ranges. On an elevation map of the CSZ, stations (triangles) and hypocentres (circles) plotted for various magnitude ranges: (A) mN 4.0 (the two circles are the two areas of large earthquakes of Stevens, 1980); (B) mN 3.0; (C) magnitudes (mN and ML) 2.0. Circle colours represent the focal depth of the events.
Figure 1.24 Lower hemisphere focal mechanisms in the CSZ. The zones of compressional readings are shaded. The maximum and minimum pressure axes are shown as P and T, respectively. The dates of the events and the magnitudes (M) of the two largest events are shown: (a) from Bent, 1992; (b) from Hasegawa and Wetmiller, 1980; (c) from Lamontagne, 1987; (d, e) from Wetmiller and Adams, 1990; (f, h, i, j , l, m) from Leblanc and Buchbinder, 1977; (k, n) from Adams et al, 1988; (g, o) from Adams et al., 1989.