Tourists have left, schools are closed. Santorini, source of 2% of the Greece GDP, has been evacuated. There has been continuous shaking for two weeks. The epicentre is 20 km from Santorini and the earthquakes are mainly below M5, but there is a risk that this could lead to a larger event. In any case, constant shaking puts anyone’s nerves up. But what is going on? Is there a reason for this swarm and is it in any way associated with the volcanic history of Santorini and nearby Kolumbo?

Santorini lies in a complicated region. To the south lies a subduction trench where old oceanic crust subducts below the Aegean Sea. That is the main cause of the volcanic arc of which Santorini is the jewel. To the north is the approaching behemoth of the Anatolian plate, the only continental plate entirely located in one country. The Anatolian plate is moving southwest into the Aegean Sea. The northern boundary, the North Anatolian fault, is extending into the Aegean Sea but hasn’t reached Greece yet. The area around Santorini is pressured in different directions. The subducting plate below is pulling north. The subduction zone is itself moving south (rolling back), pulling the Aegean in the opposite direction. And Anatolia is pushing southwest. No wonder this is a broken, fault-ridden region. The mess is centred on Santorini. The plate of the Aegean sea is being split apart under the stress. The split runs from Santorini to the northeast into the Aegean Sea.

The Aegean Sea has a history of large earthquakes. The major eruption of 1600BC which enlarged the caldera of Santorini was preceded by a large earthquake, perhaps 50 years before, which caused widespread destruction. That eruption itself also had tectonic precursor activity, and we know that the island had been fully evacuated shortly before the explosion.

The most recent major earthquake in the region was a double one in 1956. On July 9 there was an M7.2-M7.5 earthquake near Amorgos. (The precise size is not well known: the oft-quoted M7.7 is the top end of the possible range.) The depth was 25 km, which is the depth of the Moho here. It was followed by a deeper earthquake near Santorini which was probably mid to high M6 – again the size is not well known and the quoted size of M7.2 is an upper limit. (The second event can be seen as an aftershock if it was a magnitude less than the main event, or a separate event if it was stronger than that.) This was the strongest earthquake to hit Europe during the 20th century; most of the damage was at Santorini. There was a tsunami across the Aegean Sea which reached a run-up height of 25 meters on Amorgis. The tsunami was assumed to have been caused by subsea landslides, however recently an offset of 10-15 meters has been discovered which is the remnant from that earthquake: the offset may be sufficient to explain the tsunami.

Faults

The island of Amorgos, northeast of Santorini has a steep cliff on the southern side. The cliff descends into a subsea trough which is visible on google maps. The trough is 35 km wide and extends from Santorini to the northeast over a length of close to 100 km. The trough is 750 meters deep. The cliff of Amorgos is part of the northern edge of the trough, called the Amorgos fault. At the far side is the Anafi-Astypalaea fault (I hope I have the spelling right) which runs past the island of Anafi. There are several faults within the trough. One of these is the Santorini-Amorgos fault, which was assumed to be the source of the 1956 earthquake took place.

Mapping of the sea floor has shown a sharp scarp along the Amorgos faultwith a step height of 10 to 15 meters. This would fit an M7 earthquake, and therefore this fault now appears the likely source of the 1956 earthquake. It also fits the peak run-up height of the tsunami at Amorgos. During the earthquake, the fault accommodated extension of around 9 meters. The (disjointed) reports of the tsunami indicate that it started with the water -and perhaps the seafloor along the scarp- going down, consistent with extension. The earthquake was part of the widening of the trough, the break-up of the Aegean plate.

This fault has been quiet since. The extension of the trough has been measured at 4mm/yr. This would indicate that an event of this size may recur every 2000 years or so. There have been earthquakes at Amorgos in 1733 and 1891 which may may have been on this fault or more probably on other faults in the region. The most recent large event was in 42 AD. (Santorini erupted four years later but whether that is related is not known.) (see footnote) And before that, there was the large earthquake around 1650BC. This pattern would fit the 2000 years recurrence time. But this is speculative since we do not know which fault(s) failed in these events.

There are several other faults in this region. The Santorini-Amorgos fault shows frequent small earthquakes and earthquake swarms and is probably able to release its stress adequately. The maximum event from this fault has been calculated as M6.9 but that will be reduced by the frequent microseismicity. But the other faults are too quiet and are likely to release their stress in fewer but larger earthquakes. This has not happened since at least 1919. There is a notable risk of earthquakes of M6 from these faults. This is the reason for the caution by the Greek authorities and is why Santorini was, as a precaution, evacuated.

Current events

The greek word ‘seismos’ (σεισμός) means shaking. It is not specific to earthquakes: it can also be used for storms or even for commotion. It is the perfect word for an earthquake swarm.

Such swarms are quite common within the trough along the Santorini-Amorgis fault. They tend to last for weeks or months, with events of M2 to M3 along a small part of the fault. The cause probably varies. In some cases, especially near Santorini, pressure from fluids (magmatic or hydrothermal) may be involved. In other cases it just resolves some build-up stress.

The plot below shows earthquakes of M2 or stronger in this region. It was created using the database of the University of Athens, available from http://www.geophysics.geol.uoa.gr/stations/maps/recent_eq_2d.htm. The colour refers to the depth: the shallower ones (less than 15 km, i.e., in the crust) are in red, orange is 15 km or more, yellow is more than 30 km, green more than 60 km and blue more than 100 km. Do be aware that both the location and depth can be inaccurate: some of the scatter in the locations is instrumental. (Also, if using the USGS earthquake maps be aware that their locations can be off by 10 km or more in this area. It can be better to use local resources.)

The annual maps show distinct swarms. In 2018/2019 there was a swarm southwest of Amorgos, with some deeper earthquakes which lasted from November to December 2018. There was also a swarm just south of Santorini in January 2019. Santorini itself was neatly avoided, suggesting this was purely tectonic. 2020 was calm, but in 2021 there was a swarm northeast of Santorini. It started in February and continued episodically until May 2022. 2023 was again quiet and 2024 saw a small uptick in the region between Santorini and Amorgos. But notably, there was a swarm focussed on the Santorini caldera, in the same region where there had been activity and uplift in 2011/2012.

The weak swarm in 2021/2022 occurred at the Kolumbo volcanic centre. This region has shown such activity before. The faults extend to this area and it is difficult to distinguish seismic activity on the faults from activity associated with the volcano. The earthquakes were mainly below the magma chamber (which is 6-7 km deep) and it has been suggested that they were caused by magma migration into the chamber. The volcanic activity occurs on a line running SW to NE which is parallel to the faults and the trough, indicating a relation. The faults allow easier magma migration. This is the reason why tectonic swarms carry a potential (but unproven) connection to eruptions. The formation of the Nea Kameni island in the Santorini caldera was associated with the earthquake of 1707. But the much larger earthquake of 1956 did not cause any volcanic activity. This is perhaps because the 1956 event was on the Amorgos fault, while the Kolumbo-Kameni line of voclanoes is associated with the southern extension of the Santorini-Amorgos fault. This is speculation, but the microseismicity may indicate that this fault is better lubricated with fluid!

Back to 2025. Why are we so excited about a swarm in an area where they are frequent? The following plot will show. It is the same as the above, but now for 2025 up to Feb 8 only.

The 2025 swarm far exceeds the earlier ones – and has done so within only two weeks. Over the two years of 2021 and 2022 there were 170 registered earthquakes of M2 or stronger. Since the sudden start of the 2025 swarm on 27 January until mid-day on 9 February, there have been 1006 such earthquakes. Of these, two were M5 and 120 were M4. I see no evidence yet for a decrease in activity. There have been 6 M4’s already today (Feb 9).

The earthquakes of M4 and M5 are presumably the ones with the most accurate locations. They are located at depths of 6 to 8 km, across a region at the centre of the trough extending over some 10 km. The location agrees with the Santorini-Amorgos fault. Note that the region is also well separated from the Kolumbo volcanic centre. There has been no known volcanic activity in this region, to the best of my knowledge. And, as our own expert Mike Ross commented, none of the events show any indication of fluids: they are solidly tectonic. This is a fault doing what it does best: faulting.

There can be little doubt that if this activity had come in one big event, it would have been a damaging earthquake of (I guess) high M5 to M6. The Santorini-Amorgos fault is releasing pent-up stress in a long-lasting tantrum. A tantrum is better than an explosion!

Future

Does that mean everything is fine and people can return home, invite the tourists back in and re-open schools? That is not for VC to decide! But in my opinion, it is better to wait a little longer. There has been no earthquake activity near Santorini. Is this movement increasing the stress there, with a larger likelihood of a more major event? It is too early to tell: give it another week. Could this affect the other faults in the region, which have been locked for over a century? That is possible but a much harder problem to handle. It could just mean an elevated risk of a large earthquake over the next year: it is impossible to continue an evacuation for that long with such a vague reason. it will instead require earthquake-proving as many buildings as possible.

Does it increase the risk of a volcanic eruption in the region? It may take years before we know. There has been a little increase mid 2024 in the Santorini caldera which perhaps was related to the build up towards the current seismic storm – it is hard to know. If the current swarm is releasing stress and is enjoying a bit of rifting, it may make magma rise a bit easier even as far away as Nea Kameni. But that is not a significant change to the situation before. The Santorini caldera has erupted 5 times in the 18^th to 20^th century. It is reasonable to assume there will be one or two eruptions in the 21^st century as well. That expectation does not change.

But there is one thing to note. This swarm happened in the winter when few tourists were present. The 15,000 inhabitants could be evacuated. But how about a busy summer day when people can’t move from the crowds? How quickly could an evacuation take place at that time? That is something which I expect Greek authorities are thinking about – and perhaps losing a bit of sleep over. They have done well, in my opinion. But they may have to rise further to the occasion.

The bottom line is that this seismo-storm is a good thing. It releases stress, it attracts attention to future dangers, and it is non-volcanic. What more could we wish for?

Albert, February 2025

Further reading

Santorini: Beauty and the Beast

Footnote: This earthquake in 42 AD is mentioned in Leclerq et al, 2024, based on a catalogue of historical events. Having looked up what is known about it, I have severe doubts about the veracity of this claim. It seems this event, if it existed, may have happened in Turkey rather than in the region of Santorini.