Montreal jolted by M3.8 earthquake - small by historical standards
Holy complex intraplate geology, Batman!
Citation: Hubbard, J. and Bradley, K., 2024. Montreal jolted by M3.8 earthquake - small by historical standards. Earthquake Insights, https://doi.org/10.62481/c6952f3d
Earthquake Insights is an ad-free newsletter written by two earthquake scientists. Our posts are written for a general audience, with some advanced science thrown in! To get these posts delivered by email, become a free subscriber. If you would like to support our work here, please also consider a paid subscription.
A magnitude 3.8 earthquake struck ~80 km northeast of Montreal, Canada, at 5:43 AM local time on September 1, 2024 (09:43 UTC). The earthquake was felt across the St. Lawrence Valley area, as well as in northern New York State, with more than 750 reports to the USGS. Maximum shaking intensities near the epicenter reached ~IV-V (light to moderate). The USGS PAGER estimates that more than 8 million people experienced at least weak shaking, including residents of Montreal. We are not aware of any reports of damage.
What was the magnitude?
While the USGS reports a magnitude of 3.8, Earthquakes Canada is instead reporting the earthquake as a magnitude 4.6. In this case, the difference arises from different ways of measuring the magnitude using the recorded seismic waves. The USGS is calculating the Moment Magnitude, using the whole waveform recorded at regional seismometers to estimate the total moment of the earthquake. In contrast, Earthquakes Canada is using MN, or Nuttli Magnitude, a different kind of magnitude that is based on a specific component of the seismograms, which also takes into account the rock types in eastern North America. (Eastern North America needs its own local magnitude scale, because the cold, old, hard bedrock in the east can more effectively transmit seismic waves than the younger, softer bedrock in western North America, where the “local magnitude” or ML is sometimes used.) In general, Nuttli Magnitudes are higher than Moment Magnitudes. Confusingly, Nuttli Magnitudes are also sometimes written as Mblg.
Moment magnitude is generally considered a more accurate measure of the size of an earthquake. However, it can be difficult to calculate for small events, so you will still see these alternate magnitudes commonly used by various reporting agencies.
We don’t usually run into the Nuttli magnitude scale, for the obvious reason that felt earthquakes aren’t very common in eastern North America. This M3.8 (yes, we will use the moment magnitude, since it is available) is still pretty borderline — certainly not a damaging event.
How does magnitude 3.8 compare to past earthquakes? First, let’s look at earthquakes recorded since 1980, which is about when the networks improved enough to allow detection of smallish earthquakes here.
For the map area below, we found twelve earthquakes above M4.5, with one even reaching M5.9 - suggesting that the recent M3.8 is not particularly impressive. But wait! It turns out that while the M5.9 is a moment magnitude, some of those events are also measured on the Nuttli magnitude scale — like this M4.5 on March 14, 1996. And that’s not all; this M4.6 on October 19, 1990 reflects the body-wave magnitude, or mb — apparently not adjusted for those hard rocks of eastern North America. So, perhaps we can say that events similar to or larger than the recent earthquake occur about every four years, for a box ~600 km square around the epicenter — or at least, that’s what we’ve seen over the last ~45 years.
This quibbling over small magnitude events starts to seem a little silly when we extend our time range backward, to include historical events.
That map area hasn’t only produced M4 and M5 earthquakes - but also five events above magnitude 6: M7.3-7.9 (in 1663), M7.0 (in 1791), M6.0 (in 1860), M6.5 (in 1870), and M6.2 (in 1925). Those are BIG earthquakes for eastern North America — even taking into account magnitude uncertainties for old events. And they all seem to have occurred in one specific location, northeast of the recent M3.8.
What could possibly be going on?
A geological story
When you’re looking at intraplate earthquakes, which don’t match to faults at the surface, it’s wise to look at the bedrock geology.
Eastern North America is not on a plate boundary; this is an intraplate setting. However, the crust here is riddled with ancient fractures, dating from times long past when this region was a plate boundary. The Saint Lawrence River parallels one of these ancient fault zones — a huge rift that formed about half a billion years ago, as part of the repeating cycle of continents colliding and splitting apart. This rift is now named the Saint Lawrence Rift System. Ancient faults that formed as the crust stretched out are still lying deep beneath the Earth’s surface, covered by the sediments that filled the rift itself.
The Saint Lawrence Rift didn’t get the last geological word in, however. After the rifting ended, the eastern edge of North America was shoved backward over itself by a continent-continent collision farther east. This collision caused folding and thrusting of the upper crust and development of a continent-scale mountain chain. Huge rafts of crust were piled up onto the older Saint Lawrence Rift, covering up the rift sediments and slicing-and-dicing older structures. But by the time these thrusts had advanced westward to the rift, the collision was about out of steam; thus, we are left with a geological coincidence where an ancient rift and an ancient thrust system are found in the same area — just along the Saint Lawrence River.
As a side note, the modern Saint Lawrence River is all that remains of the great Champlain Sea. This interior seaway formed as the last glacial ice sheets retreated, and encompassed all of the Great Lakes and some major Canadian cities, reaching to the Atlantic Ocean. Continuing the chain of geological coincidence, the seaway was located right on top of the Saint Lawrence Rift System. Want to find a fossil whale skeleton? The clay-soil woodlands of southeastern Canada aren’t as bad a place to look as you might have thought!
A great rifting followed by a great collision seems like it should be enough for any continental interior — but this area is full of surprises. Between those two events was another striking (ha) event: a huge asteroid impact, right on top of the rift. A stony asteroid, ~2 km-wide and moving at hyper-velocity, landed right on top the rift, leaving behind the Charlevoix Impact Structure. That’s about one fifth the diameter of the asteroid that killed the dinosaurs (but it struck about 100 million years before they evolved) — plenty big enough to cause some damage.
The impact structure is still visible today: a semicircular valley on the northwestern bank of the river, extending beneath the water. The area of the impact structure is unusually smooth compared to the rugged terrain around it, probably due to the erosive effect of ice sheets on the busted-up or otherwise altered rocks. At the center of the ring, there is a peak that is interpreted to have formed due to elastic rebound after the crust above was ablated away during the impact (although you should remember that the peak itself - as a topographic feature - must be quite young, the geology that helps it stand up is very old).
We cribbed a cross section from an exhaustive discussion of the petroleum resources of this area (Lavoie et al., 2009), and labeled the three main geological features that might be related to ongoing earthquakes in the area. First, the deeply buried normal faults of the Saint Lawrence Rift. Second, the thrust faults of the Appalachian collision. Third, the giant asteroid impact. Not shown are the Pleistocene ice sheets.
Seismicity in an ancient terrain
We often see smallish earthquakes within old fault zones, even far from plate boundaries. There are several possible causes. First, the ancient faults are still zones of weakness. Various stresses can cause these faults to fail; here these stresses are probably either transmitted from the plate edges as the plate is jostled by its neighbors, or are associated with slow, upward rebound of the crust as it adjusts to the melting of the ice sheets, which weighed it down about 20,000 years ago. Second, fault zones can be great conduits for fluids to move through the crust, which can cause seismicity. Third, great fault zones near the surface can mark major disturbances at depth, within the hot mantle, which may still be slowly evolving even though the colder surface geology has died out long ago. So, it isn’t uncommon for little earthquakes to happen in these areas.
More concerningly, we sometimes see large earthquakes within old fault zones, too. With five earthquakes above M6 along the Saint Lawrence Rift, one reaching M7.3-7.9, that seems like a real thing to worry about. Those are BIG earthquakes for eastern North America — even taking into the uncertain magnitudes of the old events.
If you are an earthquake aficionado, some alarm bells might already be going off. If you walked up to an earthquake geologist and said “Hey there friend, remind me — what was that really big earthquake that happened hundreds of years ago in North America, along an ancient rift zone, right beneath a major river?”, most of them would respond with something like “Oh, you mean the New Madrid earthquakes!” That’s because the 1811-1812 New Madrid sequence was a tremendous seismic event that would be highly destructive if repeated today, and is pretty famous. The earthquakes occurred along the failed Reelfoot Rift, below the Mississippi River, in a region where tectonic crustal deformation is almost too slow to measure. So, we have some similar pattern of earthquakes and geology to think about.
Intriguingly, the largest earthquakes along the Saint Lawrence Rift all seem to have happened in the same place — within the Charlevoix Seismic Zone, located within and around the Charlevoix impact area. This area doesn’t just produce large earthquakes; it also experiences an unusually high rate of small and moderate earthquakes as well.
What could be causing this odd concentration of earthquakes? Several studies have suggested that the crustal damage zone in the impact area somehow concentrates seismicity. In 2009, Baird et al. used numerical models to argue that the the rift faults might be able to “channel” high stresses into the area of the impact structure, resulting in small earthquakes within the impact zone, and larger earthquakes along the rift faults near the edges of the crater. A more detailed study of recorded earthquakes noted that seismicity within the meteorite structure was not localized on specific structures; they interpreted this as evidence of the crust being highly fractured, with weak faults in all directions (Yu et al., 2016).
For most earthquake puzzles, we can find clues by comparing similar structures from different regions. Unfortunately, that is not possible here: it is pretty unusual for impacts to occur directly on top of ancient rifts. Furthermore, Baird et al. (2009) noted that this is one of only two impact structures that are seismically active; the other — the Vredefort crater in South Africa — only produces seismicity because of deep rock bursts in gold mines!
The 1663 earthquake produced a dramatic eyewitness account. Jesuit Father Jérôme Lalemant wrote:
Roofs seemed to bend down in one direction, and then back again in the other; Bells range of their own accord; beams, joists, and boards creaked; and the earth leaped up, and made the palisade stakes dance in a way that would have seemed incredible, had we not witnessed it in different places.
The disturbance was much greater in the forests, where there seemed to be a battle between the trees, which crashed against one another, — not merely their branches, but even, one would have said, their trunks being torn from their places to leap one upon another.
War seemed to be waged even by the Mountains, some of them being uprooted, to be hurled against others, and leaving yawning chasms in the places whence they had sprung. At times, too, they buried the trees, with which they were covered, deep in the ground up to their topmost branches; and at other times they would plant them, branches downward, which would then take the place of the roots, leaving only a forest of upturned trunks.
An earthquake, or the last march of the ents? Hard to know for sure.
Today’s M3.8 earthquake is a tiny baby compared to bigger ones that can apparently arise in this region. There is a lot of geological complexity in the area, which makes it hard to point a finger at a specific cause. However, it does seem likely that a much larger earthquake will eventually occur in this area, which should provide a lot of information on the locations and behaviors of the active faults. In the meantime, we recommend that people check the security of their palisade stakes.
References:
Baird, A.F., McKinnon, S.D. and Godin, L., 2009. Stress channelling and partitioning of seismicity in the Charlevoix seismic zone, Québec, Canada. Geophysical Journal International, 179(1), pp.559-568. https://doi.org/10.1111/j.1365-246X.2009.04275.x
Lalement, J. (1663), The Great Earthquake of 1663. https://louisianalineage.com/Earthquake1663.htm
Lavoie, D., Pinet, N., Dietrich, J., Hannigan, P., Castonguay, S., Hamblin, A.P. and Giles, P., 2009. Petroleum resource assessment, Paleozoic successions of the St. Lawrence Platform and Appalachians of eastern Canada. https://publications.gc.ca/site/eng/9.821438/publication.html
Sonley, E. and Atkinson, G.M., 2005. Empirical relationship between moment magnitude and Nuttli magnitude for small-magnitude earthquakes in southeastern Canada. Seismological Research Letters, 76(6), pp.752-755. https://doi.org/10.1785/gssrl.76.6.752
Tuttle, M.P. and Atkinson, G.M., 2010. Localization of large earthquakes in the Charlevoix seismic zone, Quebec, Canada, during the past 10,000 years. Seismological Research Letters, 81(1), pp.140-147. https://doi.org/10.1785/gssrl.81.1.140
Yu, H., Liu, Y., Harrington, R.M. and Lamontagne, M., 2016. Seismicity along St. Lawrence Paleorift faults overprinted by a meteorite impact structure in Charlevoix, Québec, Eastern Canada. Bulletin of the Seismological Society of America, 106(6), pp.2663-2673. https://doi.org/10.1785/0120160036
Thank you for such a detailed summary of conditions in the area!!!!!!!!!!! I really enjoyed this. Interesting on potential effect of meteor impact.
Hats off to the Canadians. There doesn't appear to be any overeaction by people to the quakes and all reports seem similar and reasonable. Pretty good for a populace that gets such few earthquakes.
Would a meteor impact like that shatter the underlying rock and generate a large series of radial cracks or faults?