Earthquake swarm beneath the Aegean Sea

Greek government is responding to elevated risk of large earthquake

Για να διαβάσετε αυτήν την ανάρτηση στα ελληνικά (αυτόματη μετάφραση από την Google), κάντε κλικ εδώ.

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.

Edit, Feb. 5, 2025: See our latest post, with updates and visualizations of seismic patterns.

A notable swarm of small to moderate earthquakes has been rattling the seafloor between the Greek islands of Santorini and Amorgos over the last week. Seismicity started on January 27th, with events below M3; since the 29th, the maximum magnitudes have grown to M5.1. A number of the earthquakes have been large enough to feel on nearby islands. Shaking from the M5.1 earthquake, which occurred in the afternoon of February 3rd, was reported in both Athens to the northwest and Crete to the south.

Figure 1: Earthquakes since January 1, 2025 in the region between Santorini and Amorgos, Greece. Earthquakes are projected to the right onto a timeline since January 25, and shown on a compiled timeline.

While each individual event might not be considered serious, the increasing rate of seismicity has raised some concerns. The faults in this region are capable of very large events: they sourced the deadly July 9, 1956 M7.8 Amorgos earthquake, which was followed 13 minutes later by a M7.2 large aftershock. The intense shaking from these earthquakes, and the rapid arrival of a huge tsunami wave, caused major damage throughout the central Aegean islands.

Although the source faults of these large ruptures have long been uncertain, a group of scientists from France, Greece, and Spain recently conducted remote submarine explorations around the epicentral location (Leclerc et al., 2024). The group found one fault with a freshly exposed surface at the base, with an average of 12.7 meters of implied slip — a plausible match for the 1956 earthquake.

Figure 2: 3D view of a 500-meter-long section of the Amorgos Fault, showing the path of the submarine that explored for evidence of recent slip. Figure 2 of Leclerc et al. 2024.

The question on everyone’s mind is: how will this earthquake swarm evolve, and could it potentially trigger a larger earthquake — perhaps one like the M7.8 that occurred in 1956?

Large earthquakes are rare. Leclerc et al. estimated that the 1956 earthquake accommodated 9 meters of horizontal extension, on average. GPS measurements on the islands show that only about 4 mm of extension are occurring each year — so it should, on average, take about 2,250 years to balance an earthquake like the one in 1956, if it accounts for most of the slip within this faulting zone. Unfortunately, large earthquakes are also irregular, and we cannot rely on these kinds of estimates to inform hazard: while the average rate of earthquakes might balance tectonic forcing, the gaps between individual events might be much longer or shorter. Furthermore, the 1956 earthquake rupture probably did not reach across the whole fault zone, so faults along strike — past the tips of the rupture — might have more accumulated stress.

With this in mind, a number of precautionary measures are being implemented in the affected area, targeted are mitigating impacts in the event of a larger earthquake. A common refrain in earthquake science is: “Earthquakes don’t kill people, buildings kill people.” So, local schools are temporarily closed, and residents are advised to avoid large indoor gatherings and stay away from abandoned buildings and steep terrain. People are also being advised to avoid coastlines in the event of a tsunamigenic earthquake. Emergency crews have also been preemptively deployed.

Understanding swarms

We have written about earthquake swarms a number of times before, so we can draw some insights from those past events.

First, what is an earthquake swarm? Most earthquake sequences that we discuss feature a large earthquake — a “mainshock” — that is followed by aftershocks, and in some cases preceded by smaller foreshocks. In a sequence like this, the relationships between the earthquakes are largely understood: each earthquake is associated with movement of the crust that either increases or decreases the stress on nearby faults. These stress changes can also drive motions of fluid in the crust, which can themselves change the effective stress on faults. Ultimately, these stress changes can trigger more earthquakes, with the effects slowly fading over time. While the aftershock rate eventually decays away, each aftershock still has a small chance of triggering another earthquake larger than itself.

That is not what we are currently seeing in Greece. In contrast, the earthquakes seem to be getting bigger over time, and there are a lot of earthquakes of similar magnitudes. The seismicity is not decaying. This indicates that there is some other physical process triggering seismicity, not just the stress changes associated with each individual event. Swarms are typically either magmatic — related to the movement of magma in a volcanic setting — or associated with the movement of fluids in the crust. Although these earthquakes are not so far from Santorini (an active and historically prominent volcano), they are actually associated with a different tectonic system to the northeast. That fault system has a prior history of earthquake swarms, interpreted as fluid-related.

The challenge is that, unlike mainshock-aftershock sequences, we don’t have any real “scientific laws” for swarms, and we usually cannot monitor the movement of fluids deep in the crust (except by their seismic effects). While we can observe that the largest earthquake today was bigger than the largest one yesterday, and that the maximum magnitudes seem to be growing over time, that does not come with an expectation that the pattern will continue.

What can we learn from past events? While there are examples of seismic swarms that appear to have triggered (in some complex way) very large earthquakes — like the M7.5 earthquake that struck Noto, Japan on January 1, 2024 — there are many more that have not. Among those is a swarm that rattled eastern Taiwan several weeks after the April 3, 2024 M7.4 earthquake. That swarm included multiple earthquakes above magnitude 6, and was extremely disquieting: it occurred near the tips of several large faults with histories of large earthquakes, which had already been stressed by the M7.4 mainshock. And yet, the swarm dissipated after a few days, and the aftershock sequence resumed the normal behavior.

And in yet other cases, swarms were followed by delayed triggering of a larger earthquake — so it is important to remember that even if the swarm stops, there may still be some period of elevated risk.

Figure 3: Three examples of seismic swarms that either did or did not trigger a larger earthquake.

So, it isn’t easy to say what will happen with this swarm in the Greek islands. One possibility is that the pulse of fluids responsible will migrate upwards and out of the fault zone, and the seismicity will stop. It is also possible that either the fluids or the earthquakes within the swarm will trigger a larger rupture on one of the main faults. That second possibility is more unlikely, simply because large earthquakes are rare. However, the risk of a large earthquake on this fault system is certainly elevated above normal, and the governmental response seems completely appropriate.

Typically, tectonic earthquake swarms might last for a few days to weeks, but do not usually persist much longer than that, with some notable exceptions (like the swarm in Noto, Japan).

Tectonic setting

Why are these earthquakes happening? To understand the forces at work, we have to zoom out and consider the larger picture.

The earthquakes are occurring within the northeast-southwest-oriented Santorini-Amorgos fault zone, named after the nearby islands of Santorini and Amorgos. However, these faults are just a minor part of a huge and complex system.

Greece is encircled to the west and south by a long, curved subduction zone called the Hellenic Trench. The very old, very cold oceanic crust that lies hidden beneath the eastern Mediterranean Sea is sinking below Greece. However, this sinking slab isn’t just sliding steadily downward into the mantle: it is also pulling downward and away, in a process known as “slab rollback.” In response, the overriding crust — i.e. Greece — is being torn apart, stretching and thinning in order to fill the space left by the retreating slab. This stretching causes Crete to move toward Africa, compared with northern Greece.

Figure 4: Cartoon model showing the process of slab roll-back beneath Greece adn western Turkey, and the resulting extension of the overriding crust. Figure 5 of Meng et al. (2021).

The geological details of how the Aegean crust has evolved over time are fascinating, and are a major focus of tectonics research in the area. Suffice it to say, the stretching of the crust has evolved over space and time, and the scars of the deformation are superimposed in many different geological structures.

The Santorini-Amorgos fault zone is one area where active crustal extension is concentrated today. Along this fault zone, a network of oblique normal faults dissect the “landscape” — which is really mostly underwater (a “seascape”?). The falling blocks are marked by deep submarine valleys, in some places more than 700 meters below sea level. The uplifting blocks are marked by ridge tops that rise above sea level to form beautiful (and populated) islands.

Figure 5: Block model illustrating some of the faults making up the Santorini-Amorgos fault zone. Figure 15 of Hooft et al. (2017).

A number of studies have been conducted to map out these faults. Since they mostly lie underwater, one of the best approaches is to use seismic imaging, which involves sending sound waves through the ocean and into the ocean floor, and then detecting the reflections and refractions at a set of detectors. The source of sound waves (an airgun) and the detectors (hydrophones) are dragged behind a ship.

Here is one of the seismic images that was collected of the fault system (note that this image is vertically stretched, so the fault dips are not accurate).

Figure 6: Seismic reflection profile across the eastern part of Anydros Basin. Top: interpreted. Bottom: uninterpreted. Downdropping of the hanging walls has created deep valleys that have progressively filled with sediment. Figure 5 of Tsampouraki-Kraounaki et al. (2021).

You can immediately see how one area in the center of the profile has tilted downward, and the resulting valley has been progressively filled up with sediments. This tilting and subsidence is due to slip on a large normal fault. Here is a map of faults across the Santorini-Amorgos fault system, which was created using this kind of data:

Figure 7: Detailed fault map of the Santorini-Amorgos fault zone. Figure 9a of Tsampouraki-Kraounaki et al. (2021).

It is clear that the Amorgos fault, thought to be responsible for the 1956 earthquake, is not the only player in the area. This is pretty common in extensional systems, since the entire crust has to thin over a wide region, rather than just along one fault.

What about volcanoes?

As you can see from the above map, the volcano of Santorini itself lies within the southwestern end of this crustal rift system. This is the kind of setting where tectonicists and volcanologists can all get together to worry about how their pet systems might interact. However, a lot of questions remain about those kinds of interactions.

The crescent-shaped island of Santorini is in fact the rim of a volcanic caldera, left behind after a huge eruption around 1600 BCE. Minor eruptions since then have left their mark on the small, uninhabited island of Nea Kameni, located in the center of the caldera. Like many geologists, we took an educational vacation to Santorini in our early 20’s, including a visit to Nea Kameni.

Figure 8: A relatively youthful and wind-blown Kyle exploring the also relatively young lava flows on Nea Kameni.

While some very small earthquakes have been detected around Santorini over the last months, these do not seem directly related to the larger swarm to the northeast. However, news reports on the subject might lead to some confusion. Many reports focus on the northeastern swarm, but pair it with a discussion of the volcanic history of Santorini, without a clear distinction; some also mention the less notable seismicity within the caldera itself. So, to be clear:

• There has been some minor seismicity within and around Santorini’s caldera, starting around the middle of 2024. Maximum magnitudes have reached ~M3. That seismicity is within the range of normal, and there is no associated concern about an eruption. The present advisories are not in response to this seismicity.

• There has been a larger swarm of earthquakes, centered ~25 km northeast of Santorini (halfway to Amorgos). This swarm started around January 27th 2025, with maximum magnitudes reaching M5.1. This seismicity is thought to be tectonic in origin, not volcanic. Because the largest recent earthquakes in Greece occurred along this fault system about 70 years ago, there is a known precedent for a highly dangerous event in this area. Thus, the advisories are in response to this (still low-level) seismicity.

References

Andinisari, R., Konstantinou, K.I. and Ranjan, P., 2021. Seismicity along the Santorini-Amorgos zone and its relationship with active tectonics and fluid distribution. Physics of the Earth and Planetary Interiors, 312, p.106660, https://doi.org/10.1016/j.pepi.2021.106660

Hooft, E.E., Nomikou, P., Toomey, D.R., Lampridou, D., Getz, C., Christopoulou, M.E., O'Hara, D., Arnoux, G.M., Bodmer, M., Gray, M. and Heath, B.A., 2017. Backarc tectonism, volcanism, and mass wasting shape seafloor morphology in the Santorini-Christiana-Amorgos region of the Hellenic Volcanic Arc. Tectonophysics, 712, pp.396-414, https://doi.org/10.1016/j.tecto.2017.06.005

Hubbard, J. and Bradley, K., 2024. Deadly M7.4 earthquake strikes Taiwan. Earthquake Insights, https://doi.org/10.62481/c4a3297f

Hubbard, J. and Bradley, K., 2024. M7.5 earthquake strikes western Japan, triggers tsunami. Earthquake Insights, https://doi.org/10.62481/e8bf9b2e

Hubbard, J. and Bradley, K., 2024. Seismic swarm near southern tip of April 3 M7.4 in Taiwan. Earthquake Insights, https://doi.org/10.62481/b28eb629

Hubbard, J. and Bradley, K., 2024. Swarm offshore Vancouver Island punctuated by M6.4 earthquake. Earthquake Insights, https://doi.org/10.62481/a23a9f0a

Leclerc, F., Palagonia, S., Feuillet, N., Nomikou, P., Lampridou, D., Barrière, P., Dano, A., Ochoa, E., Gracias, N. and Escartin, J., 2024. Large seafloor rupture caused by the 1956 Amorgos tsunamigenic earthquake, Greece. Communications Earth & Environment, 5(1), p.663, https://doi.org/10.1038/s43247-024-01839-0

Meng, J., Sinoplu, O., Zhou, Z., Tokay, B., Kusky, T., Bozkurt, E. and Wang, L., 2021. Greece and Turkey Shaken by African tectonic retreat. Scientific Reports, 11(1), p.6486, https://doi.org/10.1038/s41598-021-86063-y

Tsampouraki-Kraounaki, K., Sakellariou, D., Rousakis, G., Morfis, I., Panagiotopoulos, I., Livanos, I., Manta, K., Paraschos, F. and Papatheodorou, G., 2021. The Santorini-Amorgos Shear Zone: Evidence for Dextral Transtension in the South Aegean Back-Arc Region, Greece. Geosciences, 11(5), p.216, https://doi.org/10.3390/geosciences11050216