The seismic swarm in Greece gets more complicated

New data reveal unexpected volcano-tectonic interactions

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 independent 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.

This is an ongoing and evolving situation.

Updates from the Greek Civil Protection are available here.

The latest emergency announcement (2025-02-14) from the Interdisciplinary Committee for Risk and Crisis Management of the University of Athens is available here.

---

NOTE: This post was originally published on February 14, 2025, and included visualizations of a new, preliminary seismic dataset. The owner of the dataset had initially granted us access, with permission to create and publish figures, but not to publish the catalog itself. However, later there was concern that our post could interfere with scientific publication. In this updated post, that catalog has been replaced with public catalog data from the EMSC and from the University of Athens. Although the story remains essentially the same, we have adjusted the text and figures to ensure that all of our interpretations are clearly visible in the publicly available data. We apologize for the confusion. However, we think those data, and the method that produced them, will have profound implications for the study of complex seismic events now and in the future, and we look forward to seeing the data fully interpreted and published.

---

The seismic crisis in Greece continues unabated. With states of emergency declared at both Santorini and Amorgos, many residents and tourists have evacuated, without certainty about what is driving the unusual seismic behavior, if and how it relates to the local volcanoes, and when it might stop. This is a challenging situation for all involved.

Fortunately, scientists have started to tease apart the story about why these earthquakes are occurring.

Here, we will take a look at some of the new data and muse about the big-picture questions.

First, it is important to note that the data we discuss are being rapidly disseminated by seismologists, geodesists, and volcanologists who are highly experienced in the volcanic and tectonic systems that have become active. The spirit of rapid data sharing and discussion is clearly alive and well.

Along with the data, these scientists have presented some early interpretations. We will provide links to the available information that we have found, and we strongly recommend that interested readers seek out those original information sources. Our goal for this post is not to confirm or argue with any of these ideas, but rather to use the newly presented information to put together a plausible, but not necessarily correct, timeline of events. Over time, peer-reviewed publications will certainly emerge that should provide a higher degree of confidence.

Ok, so what is new?

1. Ground surface deformation has been seen at Santorini

New data has demonstrated that there has been motion of the ground surface at Santorini over the last days, weeks, and months. The best data come from three continuously running GPS stations located on Santorini.

While our first posts were a bit cagey about the possible role of the Santorini volcano in this crisis, these data clearly point to motion of magma at relatively shallow depth beneath the caldera. The data are presented in a rapid report to EMSC by Pierre Briole (CNRS, France), in coordination with scientists at three Greek institutions: Athanassios Ganas (NOA), Panagiotis Elias (UPAT) and Vassilis Sakkas (NKUA).

The report contains very clear logic, and is well worth reading. The single figure that captures the story best is from a GPS station called SANT, located almost 400 meters above sea level along the eastern rim of the caldera.

Side note: because the station locations must be reported with high accuracy, we can usually figure out where the stations actually are. Some Google Earth sleuthing suggests that SANT is located in the back yard of a school; our guess is that the antenna can be seen on top of a small building in the rear courtyard. This is quite typical, as GPS stations need to be located in a stable place with reliable access, and with a clear view of the sky. A critical skill for a geodesist is the development of good long-term relationships with the people who steward the instruments.

Figure 1: Location and photo of the building where GPS station SANT is located (longitude 25.42261°, latitude 36.43360°).

So, what do we see in the data? We can see the up, north, and east components on the following plot from the report. This plot covers the period from January 1, 2024 up to February 11, 2025.

Figure 2: Time series of the up, north, and east location measurements of GPS station SANT from 2024/01/01 through 2025/02/11. Figure 4 (top) of Briole (2025).

From January 2024 through July 2024, the station recorded a very slow drift that doesn’t require any explanation. GPS coordinates are allowed to drift a bit, because it is extremely hard to keep the reference frame perfectly stable. Around July-August 2024, the flat trends suddenly changed. (We note that around the same time, the scatter of the data points decreased; we don’t know why.) These changes would have been hard to detect at the time, as they are just changes in slope rather than a sudden jump. We see that SANT started to move east (the black dots), up (the gray dots), with smaller motion to the north (the white dots).

This trend continued until late January 2025, when there was a sudden change. The eastward motion accelerated, a new northward motion strongly accelerated, and the station started moving down. The report includes an image that shows this latest change in more detail. The change seems to occur around January 27-29, 2025.

Figure 3: Time series of the up, north, and east daily measurements of GPS station SANT from 2025/01/01 through 2025/02/11.Figure 4 (bottom) of Briole (2025).

Two other GPS stations are discussed in the report, located further to the southeast. These show changes in motion at about the same times, although the changes are somewhat less impressive. The agreement means that it’s not just one site acting up - the ground really is moving.

There is a good context for interpreting these GPS data: a volcanic crisis that occurred beneath Santorini in 2011-2012. (As far as we can tell, a volcanic crisis is simply a period of volcanic activity that impacts society and requires active management by officials.) During that time, many temporary GPS stations were deployed that observed the effects of a magmatic intrusion beneath the caldera. When magma is injected up into the shallow crust, it forces the ground above upward and outward. This inflation of the volcano is a classic signature of magma movement (as we have seen in the repeated cycles of inflation and eruption in southwest Iceland over the last year).

The Briole report identifies the most likely location of this magmatic intrusion, using the three GPS stations on Santorini to triangulate a location:

Figure 4: An initial attempt to match the direction of motion at the three GPS stations on Santorini to a point source. Figure 6 of Briole (2025).

As expected, the source is within the caldera, and is not too far away from the 2011 source.

How do these changes relate to seismicity?

As we discussed in our earlier two posts, there was a minor outburst of earthquakes beneath the Santorini caldera that took off during the third week of September, 2024. Because the GPS data seem to deviate from baseline around August (as measured via that delicate instrument, the eyeball), it looks like the earthquakes started after the ground began to deform.

The intense seismic swarm northeast of Santorini started later: around January 27, 2025. While the swarm is not co-located with Santorini, the observation of ground deformation at Santorini at about that same time does suggest that the swarm is indeed somehow related to magma motion beneath the volcano. More on that in the next section!

So, it appears that there was a small inflation event that started in July-August 2024, which was overprinted or replaced in late January 2025, when the seismic swarm in the Aegean Sea began. While SANT began to move much faster horizontally to the north and east, it also seems to have started to sink back down. This is hard to explain, and suggests that there may be additional movement of magma or tectonic displacement outside of Santorini caldera.

2. Updated seismicity

The hypocenters in the catalog fall within a large volume, suggesting that many faults (or even magmatic dykes) are active; telling the difference is a significant problem. We will talk more about what the geology might be like down there, later.

We find that timelines and cross sections make it easier to tell the story.

Here is an overall view. In this figure, seismicity from the EMSC catalog since May 1, 2024 is colored by time. You can see:

1. Two persistent clusters prior to the development of the swarm. One cluster is below Santorini. The second, fainter cluster lies to the northeast, along the southern flank of the submerged Kolumbo volcano.

2. The giant earthquake swarm began around this northeastern cluster, where a transitional swarm lasted for a few days.

3. The major swarm then took off to the northeast.

Figure 5: Seismicity since 2024-05-01 between Santorini and Amorgos. Earthquakes are colored by time, and projected too the right onto a timeline. Seismic catalog is from the EMSC.

Let’s take a more careful look at the pre-swarm seismic clusters. The cluster of events below Santorini is pretty shallow, at <8 kilometers depth. The cluster near Kolumbo is much more diffuse, and some of the events are deeper. (Kolumbo is visible in the bathymetry, with a classical conical shape and a crater.)

Figure 6: Seismicity from May 1 2024 through January 15 2025, from the EMSC, colored by depth. Earthquakes are projected to the right onto a timeline and shown on a cross-section below the map.

A few small earthquakes were detected below Santorini starting in late June; the cluster really started in the third week of September, 2024. (This is consistent with the seismic catalog from the University of Athens.) The cluster below Kolumbo also started in mid-September. Volcanologists have long discussed the relationship between Santorini and Kolumbo, and this sequence will surely add to that discussion.

The transitional swarm originated near the Kolumbo cluster on January 27. A few days later, seismicity expanded northeastward into the mapped fault zone that extends towards the island of Amorgos. The figure below shows a timeline of the swarm up to almost the present day (it’s hard to write as fast as data can come in!). We will annotate a similar figure toward the end of the post. We also drew three cross sections through the data. The points are colored by time.

Figure 7: Seismicity since 2024-05-01, from the EMSC catalog. Earthquakes prior to January 24 are shown in black; earthquakes since January 24 are colored by time. Events are projected onto a timeline since Jan 24. Three cross-sections are shown below the map.

3. Comparing the seismicity to the GPS

When we have multiple independent datasets, we usually want to match them up. In the old days, we might print out some images and tape them together. In the new days, we just screenshot them and throw them into Keynote for a quick rescaling, thus accomplishing the same thing.

We made a new timeline like that in the figure below, focused on the period January 1 - February 11, 2025. This covers the time when the swarm really got going. This time, we used the earthquake data from the catalog provided by the University of Athens. When we aligned the GPS and seismic timeline figures and “taped” them together, we found an impressive correlation:

Figure 8: Comparison of the data from GPS station SANT on Santorini with a seismic timeline similar to the one from Figure 8.

The early part of the seismic swarm looks different from the later part: smaller events in a more constrained area in the deeper crust. Eventually, the swarm propagated northward in some discrete jumps, and started to activate faults in the shallow crust. This produced numerous magnitude 4+ earthquakes, within a much larger volume of the shallower crust.

Story time: A possible interpretation

So, what does all of this data tell us?

We think that the Santorini-Amorgos seismic crisis may have undergone three stages, so far. This is very preliminary, and we mainly want to show how our thought process works. We are sure that many other researchers are coming up with the same kinds of stories right now, better constrained by more thorough analyses. For instance, the seismic waveforms of the earthquakes likely carry a lot of information that we haven’t ourselves considered, and there is a huge literature on the volcanic systems that we just haven’t read. So take this all with a grain of salt.

Anyway, here’s our proposed story. After we tell the tale, we will have a graphic that tries to tell it again.

First, magmatic inflation began at shallow depth beneath the Santorini caldera around the third week of September of 2024. This caused GPS sites on the island to quietly move upward and away from the caldera — a classic inflation signal. In a typical event, this would have been identified and routinely monitored, and would probably have been only of academic or local interest unless an eruption actually happened. The earthquakes associated with this inflation were certainly noted by scientists, but rightly did not cause concern in the public sphere.

At the same time, a persistent cluster of earthquakes began below the southern flank of the underwater Kolumbo volcano, northeast of Santorini. These earthquakes were also quite small.

The earliest signs of the oncoming earthquake swarm appeared to the northeast of Santorini on January 26. An accelerating earthquake swarm started in the middle to lower crust. The Greek government was appraised of these events by watchful seismologists, and convened an emergency meeting on January 29. We don’t know who first discovered the events, or whether sensible shaking alerted people to the start of the swarm. The first news reports of the oncoming crisis that we are aware of appeared on January 30.

We suspect that this swarm was caused by the intrusion of a magmatic dyke into the lower crust, spreading northeastward from the general magmatic center beneath Santorini and Kolumbo. The earthquakes in the resulting swarm arrived in pulses, but did not exceed magnitude 4.

The crustal deformation caused by this deep event added to the existing inflation signal at Santorini, causing the GPS sites there to accelerate rapidly. This could be why the GPS data at Santorini can’t be fully explained by a 2011/2012 type event — events occurring outside the caldera are more important to this acceleration phase.

The magmatic intrusion in turn drove hot fluids (e.g. water and CO2) northeastward and to shallower crustal depths, where the active faults between Santorini and Amorgos were ready and waiting. As the pressurized fluids migrated into these fault zones, they decreased the clamping stress on the faults, allowing them to begin to slip seismically. The high pre-existing tectonic stresses on these shallow crustal faults allowed them to produce larger earthquakes, up to M5.2 (so far).

Here’s a graphic that pulls together these ideas, using the period January 25-February 6.

If we haven’t said it often enough, this is all just our initial take on the story and we humbly request information from those more in-the-know!

Figure 9: General interpretation of the seismic/volcanic crisis between Santorini and Amorgos.

Take-aways

This magmatic/seismic sequence is truly extraordinary, because it seems to have operated across all levels of the crust. There is a remarkable evolution in seismic behavior over both space and time. Relatively small signals of magmatic action at Santorini, which might have been of interest but not necessarily great concern, have somehow evolved into truly impressive activity beneath the Aegean Sea.

Our cross sections generally show the inside of the earth as a clean expanse of white or light brown. In reality, there is an extremely complex stack of rocks that have been built up over hundreds of millions of years, if not billions of years, of continental tectonics and magmatism. The structural geology, fault networks, dyke networks, sills, folds, etc. that make up the crust can probably never be seen. But events like this, in which little earthquakes each reveal something about the crust they happen in, can possibly tell us a lot about what’s down there, and how it behaves.

One of the reasons that this swarm is so complicated is that the Aegean crust has been stretched out and thinned by extensional forces, which are still acting today. The hot lower and middle crust has responded to this extension by flowing like molasses, while the cold upper crust on top has fractured along many brittle faults. There is also evidence that some of the faults extend into the mantle: the 1956 Amorgos earthquake did rupture to the surface, but seismological studies indicate that its focal depth was quite deep, near or maybe even below the base of the crust.

Untangling this tale will probably take years of research, although we expect to see rapid results being published over the coming months. Luckily, there are many long-standing research collaborations between geodesists, volcanologists, seismologists, and geologists, from Greece and from abroad, who are very interested in understanding how the whole system works. It will be really interesting to see what tales unfold from these studies, over the next months, years, and even decades.

Looking forward

What do we expect for the future, based on our make-believe story?

We are not in the business of predicting the future. However, our conceptual tale does give us a few things to think about.

First, it is unclear from the University of Athens earthquake data whether the seismic activity at Santorini continues or has abated. There is a clear need for improved monitoring in this area, and we are confident that plans for that are already well underway. It seems that lessons learned from the 2011/2012 magmatic event might still be needed in the near future.

Recent studies of the Santorini-Kolumbo system have recommended establishing an observatory to monitor the magma chambers, as these are active volcanoes with the potential to generate explosive eruptions (e.g. Chrapkiewicz et al., 2022). This proposal seems particularly relevant now, especially considering the implications of the renewed activity of the shallow faults.

Second, the seismic swarm to the northeast is likely occurring because deep magmatic intrusion is pumping hydrothermal, metamorphic, or magmatic fluids into the fault system between Santorini and Amorgos. Although the shaking caused by these earthquakes is distressing, it is probably more a symptom than a cause. Their distance from the populated islands and their moderate size has minimized any direct damage so far. Because the deep part of the swarm near Kolumbo and Santorini looks like it has already ended, it seems likely that the shallow part will simply diffuse away over time. If renewed intrusion occurred, the shallow swarm might also renew (or it might not, if the fluids have already been expelled from the source area (boy is this all complicated)). Of course, this is all speculative until the real experts have their say.

Third, we still do not have any insight into whether the seismic swarm will escalate into a large earthquake. While we have read some statements that suggest this swarm has “foreshock-like” features, we do not believe that it is possible to make that assertion. We are simply without much precedent for this event, and we think it is important to hold to standard seismological principles. Yes, more earthquakes does usually mean a higher chance of a larger earthquake, although the actual probability increase may be quite small. No, seismic swarms do not usually culminate in a large earthquake. The chance of a large triggered event will decrease as the swarm abates, but will never drop to zero, as this region hosts large faults with a history of large earthquakes.

Finally, we want to again thank the scientists who are making their data and rapid analyses available in almost real time. The coordinated response to this event is revealing the great depth and expertise of the many institutions and people who care about geology, natural hazards, and human impacts in Greece. We take all the blame for any misinterpretations we may have made, and give all the credit to the people who make these kinds of posts possible! Σας ευχαριστώ!

Please let us know in the comments if you have any comments or corrections.

References:

Briole, P., in coordination with Ganas, A., Elias, P., Sakkas, V., 2025. Santorini seismo-volcanic event: GNSS time series and preliminary models, version 20250213b. https://www.emsc-csem.org/Files/event/1765158/20250213b-santorini-seismo-volcanic-crisis-gnss.pdf

Brüstle, A., Friederich, W., Meier, T. and Gross, C., 2014. Focal mechanism and depth of the 1956 Amorgos twin earthquakes from waveform matching of analogue seismograms. Solid Earth, 5(2), pp.1027-1044. https://doi.org/10.5194/se-5-1027-2014

Chrapkiewicz, K., Paulatto, M., Heath, B.A., Hooft, E.E.E., Nomikou, P., Papazachos, C.B., Schmid, F., Toomey, D.R., Warner, M.R. and Morgan, J.V., 2022. Magma chamber detected beneath an arc volcano with full‐waveform inversion of active‐source seismic data. Geochemistry, Geophysics, Geosystems, 23(11), p.e2022GC010475. https://doi.org/10.1029/2022GC010475

Hubbard, J. and Bradley, K., 2025. Earthquake swarm beneath the Aegean Sea, Earthquake Insights, https://earthquakeinsights.substack.com/p/earthquake-swarm-beneath-the-aegean

Hubbard, J. and Bradley, K., 2025. Update on the seismic swarm in Greece, Earthquake Insights, https://earthquakeinsights.substack.com/p/update-on-the-seismic-swarm-in-greece

Leclerc, F., Brenguier, F., Briole, P., 2025. L’ObservatTerra. Un peu de science pour comprendre la crise sismique de janvier-février 2025 à Santorin. Video available here: