Minoan eruption


The Minoan eruption (also Thera or Santorini eruption) is the name given to the late Bronze Age eruption of the Aegean volcanic island of Thera (now Santorini), which in the 17th or 16th century BC destroyed the settlement of Akrotiri, which was closely associated with Minoan culture (the view, often held until the 1960s, that it brought about the demise of Minoan culture on Crete,

The pyroclastics ejected during the eruption can be found in archaeological sites throughout the eastern Mediterranean and thus provide a fixed point in the stratigraphy. The dating of the eruption is controversial; there were about 100 years between the historiographically and scientifically determined dates. However, since a refinement of the scientific methodology, the radiocarbon dates can be reconciled with the historiographic findings.

The volcano of Santorini is a result of plate tectonic processes. It is part of a volcanic island arc in the southern Aegean Sea, which lies above a subduction zone formed by the collision of the African and Eurasian plates.

The core of the island consists of metamorphic rocks aged about 200-40 million years. Today, they are visible on the surface only at the highest elevation, Profitis Ilias (567 m), but lie beneath younger strata in four places on the southern island. The rest of the island consists of volcanic rocks formed in at least twelve medium and larger and other smaller eruptions since the Pleistocene, i.e. in the last 1.8 million years. These are mainly pyroclastic deposits, but five lava flows can be traced throughout the area. Age determinations of the rocks suggest an interval of 20,000 years between major eruptions and 5,000 years between minor eruptions.

Volcanism in the area of Santorini began about 2 million years ago, when the first eruptions from the seabed occurred in the area of the Akrotiri Peninsula and probably also at the site of the Christiania Islands, 20 km southwest of Santorini. The island of Santorini is the result of a complex history of volcanic eruptions during this period, during which the island repeatedly changed its shape and size. About 400,000 years ago, the center of volcanic activity shifted to the center of the present-day caldera. The most characteristic type of activity in the last 400,000 years was the cyclic construction of shield volcanoes, which were formed about 3,600 years ago by large explosive and destructive events such as the eruption that had a strong impact on the cultures of the Mediterranean Sea, especially in the east. In detail, the volcanic evolution of Santorini can be divided into six main stages:

Modern investigations show that the archipelago already had approximately its present shape in Minoan times (including an island in the middle of the caldera), which it received as a result of the Cape Riva eruption about 21,000 years ago.

In 1939, the Greek archaeologist Spyridon Marinatos published a theory according to which the eruption of the Thera volcano had led to the demise of the Minoan culture on Crete. For Marinatos, the Thera eruption must have resembled that of the Indonesian volcano Krakatau, which claimed the lives of some 36,000 people in 1883. In addition to a rain of ash that had darkened the sky for a radius of several hundred kilometers, the tidal wave resulting from the eruption was a particularly important parallel for him. With heights of up to 15 meters, the wave triggered by Krakatoa in 1883 had washed over the coast of the neighboring islands and destroyed numerous towns. Marinatos assumed a similarly devastating flooding of the coasts of Crete by the Thera eruption and suspected in it the cause for the decline of the Minoan culture.

In the meantime, traces of tidal waves have been identified in some places on the northeast coast of Crete. So in Pseira, Palaikastro and Papadiokambos. Even on the coast of Israel tsunami traces were found and dated. The excavations of Palaikastro show that the whole place was flooded and destroyed, but later it was at least partially rebuilt, so the Minoan culture still existed.

The extent of the eruption assumed by Marinatos – he assumed four times the amount of tephra (80-120 km³) compared to the Krakatau eruption (20-30 km³), which would correspond to a magnitude 7 eruption on the Volcanic Explosivity Index (VEI) – was revised downward over the years. Since the thickness of ash layers on neighboring islands also did not confirm Marinatos” assumption, a smaller eruption (30 km³) of magnitude VEI 6 was assumed. Pollen analysis of sediment layers before and after the Thera eruption also indicated minimal changes in regional vegetation and thus a relatively small eruption.

However, in 2002, ash layers were found which, due to their thickness, are understood to indicate an eruption more than twice as strong (up to 100 km³ of tephra). Further investigations of the seafloor around Santorini in 2006 identified deposits of pyroclastic flows of considerable thickness. The new estimate based on this now gave a total volume of 60 km³ of magma, which safely raised the strength again to 7 according to VEI.

Today, the eruption is divided into four major phases. It was preceded by several earthquakes. The inhabitants then left the island. They had enough time to take their valuables with them. During the excavations of the city of Akrotiri, no corpses, jewelry or elaborate tools were found. Shortly after the earthquakes, Akrotiri was apparently visited again. Attempts were made to salvage undestroyed pithoi (storage containers) and pieces of furniture, to tear down walls that were in danger of collapsing, and to sort building materials for reuse.

However, the salvage operation was canceled, the rescuers fled again, leaving behind the already provided storage containers and furniture. The cause is considered to be the first case of pyroclastics. It was only small amounts of volcanic ash and lapilli from a vent almost exactly in the center of the island. After that, a pause occurred. Since tufts of grass were detected on some wall stumps in Akrotiri, a dormant period of several months is speculated.

The first output of pumice

The first phase of the eruption proper consisted of a plinian eruption with the ejection of light pumice and ash. Deposition occurred at about 3 cmmin, and the maximum thickness of the layer was 7 m. Where the ashes collected under steep slopes, 11 m could be reached. The output started with white material and changed to pink, into which rock fragments in bright yellow, orange and red tones were increasingly intercalated. The colors originate from the increasing temperatures of the rock when hitting the ground or previous layers, respectively.

The energy of this phase is considered to be rather low. The material was ejected by volcanic gases; initially, water had not yet entered the vent. This phase is said to have lasted between one and eight hours. Only in the uppermost layers of the first phase did pyroclastic flows mix into the loose deposits – the lava had come into contact with seawater.

Pyroclastic flows

When cracks breaking open in the rock allowed seawater to enter the volcanic vent and evaporate, a phreatomagmatic explosion occurred with the energy of the eruption multiplied. The volcano was now able to eject much heavier material, but its deposits were also much more unevenly distributed.

The second phase began with the eruption of round lapilli about 10 mm in diameter, mixed with ash and a few larger chunks. Deposits from this eruption reach a thickness of 5.90 m on Thirasia in the west and only about 10 cm at the very east of the island. This is followed by a layer of only 1-18 cm of white ash and another thick layer between 6 m in the west and 15 cm in the east and southeast. This second layer is composed of lapilli with intercalated volcanic bombs, ranging in size from a few centimeters to blocks 5 m in diameter. The blocks consist mainly of black, smooth lava, which was also typical for earlier volcanic eruptions on Santorini, for example at the Skaros rock.

The second phase lasted about one hour. The volcanic vent ruptured in a southerly direction, as can be concluded from the orientation of some deposits.

Phreatomagmatic deposits

In the third phase of the eruption, the largest output of volcanic material took place. The pyroclastics flowed as a continuous stream and swept away boulders of enormous size. The blocks reached diameters of 20 m in this phase, 0.5-2 m are typical. They consist of porphyritic dacite and to a small extent of material comparable with obsidian.

The blocks are embedded in ash flows, flows of lapilli and, towards the end, flows of mud from pumice with high water content. In some places in the southeast of the island, the deposits of the third phase reach a thickness of 55 meters.

The vent shifted northward again during this phase. The seawater entering the vent mixed with the volcanic material and, according to one interpretation, formed an enormous mass of hot mud called lahar. It is said to have overflowed the up to 400 m high walls of the caldera. So much material was ejected that the resulting cavity collapsed and the island above it collapsed. This formed the northern half of today”s caldera. On the outside of the island, the volcanic flows flowed into the sea, widening it to form shallow coastal plains.

Ignimbrite, lahar and rubble streams

With the fourth phase ended the outbreak. It is multiform. The deposition of ignimbrite layers alternated with lahar flows, ash flows and huge amounts of debris. It is also possible that ash clouds were ejected in between. Most of the material flowed off to the edges of the island: While at the caldera only about 1 m thick layers are attributed to the fourth phase, they form alluvial fans of up to 40 m thickness outside, depending on the terrain profile.

The rocks of the fourth phase are smaller than before, the maximum size does not exceed 2 m anymore. Also it can be proved that at two places in the south lahar currents flowed back into the caldera. The energy of the eruption must have decreased significantly. McCoyHeiken assume that only now, at the very end of the eruption, the ring of the island collapsed, the northwestern channel between the main island and Thirasia was formed and the rock in the south of Thirasia collapsed. Only the rocky islet of Aspronisi, remnant of an earlier eruption, remained standing.

The deposition of theraeic tephra in virtually the entire eastern Mediterranean – from Nichoria in Messenia and the Black Sea – provides a unique fixed point for the synchronization of various relative chronologies from these regions. At the same time, this makes virtually the entire absolute chronology of the Late Bronze Age in the eastern Mediterranean, as well as synchronous chronologies in much of the rest of Europe and the Near East, dependent on the dating of this eruption, which is why, understandably, the question of the dating of the Minoan eruption is one of the most hotly contested in archaeological research today.

Especially since the 1980s, numerous investigations with the most diverse methods essentially led to a division of opinions into two camps: on the one side the representatives of the “late dating” (1530-1520 B.C.) and accordingly the “short chronology”, on the other side those of the “early dating” (1628-1620 B.C.) and the “long chronology”. It is also noteworthy that the “fronts” are not between natural sciences and humanities, but across all camps. However, the debate, which is largely conducted in high-profile scientific journals such as Nature and Science, has not yet found a definitive answer.

Archaeological-historiographic method

Marinatos originally roughly dated the Minoan eruption to 1500 B.C. ± 50 years, since he also assumed this period for the demise of the Minoan palace centers on Crete. Although excavations in the following decades showed that the Minoan civilization did not perish suddenly, but only from about 1450 B.C. over a period of probably several decades, the dating of the Minoan Eruption in the late 16th century B.C. proved to be the most probable from an archaeological point of view. This is because, in the meantime, finds came to light on Crete (e.g., evolved vase painting styles) that, on the one hand, no longer occur on Santorini, but, on the other hand, clearly predate the collapse of the Minoan civilization and came to light on Crete above deposits of ash that probably originated from the eruption.

The relative chronology of the Minoan culture, which was already worked out by Arthur Evans and since then was refined more and more, was last connected among other things in 1989 by Peter Warren and Vronwy Hankey with the quite secured, absolute chronology of Egypt. Accordingly, the phase “Middle Minoan III” (MM III) is connected with the Hyksos period, the phase “Late Minoan IA” (SM IA) with the end of the Second Intermediate Period and “Late Minoan IB” (SM IB) with the time of Hatshepsut and Thutmosis III. If one sets with this argumentation the Minoan Eruption about 30 years before end of the phase SM IA, this results in a period of 1530 to 1500 B.C.

Other archaeologists bring arguments for an early dating of the Minoan eruption, such as Wolf-Dietrich Niemeier, the excavator of the palace of Tel Kabri in Palestine, who points out that a doorstep in the building destroyed in 1600 BC corresponds completely to the one uncovered at Akrotiri. Likewise, the wall paintings showed clear stylistic links to the frescoes at Thera. Niemeier therefore supports the “long chronology” and a shift of the end of SM IA from 1500 to 1600. Results of the excavation at Tell el-cAjjul in the Gaza Strip point in the same direction. Since an early dating would have the consequence that beside the Minoan also the Egyptian chronology, which is considered as very safe, would have to be revised – and with it all chronologies dependent on it in the Near East and completely Europe -, leading Egyptologists and in particular Manfred Bietak spoke out decidedly against it. Bietak found the same offset at Tell el-Daba between the 14C dating and the placement in the relative chronology of Egypt. He dates the Minoan eruption on the basis of a very controversial assignment of excavation layers (stratum C2 in Tell el-Daba) into the reign of Thutmosis III around 1450 BCE (short chronology).

The ceramic style known as White Slip plays a special role: it was found in relatively chronologically datable layers equally on Santorini before the eruption, in Cyprus and the Hyksos capital Auaris in present-day Egypt. If the pieces can be placed in a chronological order of development, they would not only allow the synchronization of cultural areas, but also clarify the question of the early or later dating of the Minoan Eruption.

Since the political conditions in Egypt and Mesopotamia were in a state of upheaval around the middle of the 2nd millennium B.C., there is no clear written evidence of the catastrophe with which a historiographical determination of the date would be possible. Thus an Egyptian inscription, the so-called “tempest stele” of Ahmose I remains controversial. This – also formally – highly unusual description of a natural catastrophe reports of tremendous roaring and darkness for days in whole Egypt, which reminds very much of typical accompanying phenomena of a heavy volcanic eruption, e. g. The time of the catastrophe lies between the 11th and the 22nd year of the reign of Ahmose, thus 1539-1528 B.C. (after Beckerath) or 1519-1508 B.C. (after Schneider) or 1528-1517 B.C. (after Hornung, Krauss & Warburton). Should the described “storm” have been caused by the Minoan eruption, this offers a dating from a historiographical point of view. However, since so far no tephra layers of the Minoan eruption during the reign of Ahmose in Auaris or other places of Lower Egypt were proven, that “tempest” can also be interpreted symbolically as a state of desolation in Egypt after the end of the Hyksos period.

Another piece in this puzzle is the Papyrus Ipuwer, which contains a very similar description of a natural disaster and is dated to about 1670 (± 40) B.C.E. Because of the very similar descriptions in the Papyrus Ipuwer and the tempest stele, the dating of the reign of Ahmose I after the heliacal rising of Sirius is not undisputed, as well as the above mentioned dating of the Minoan eruption to the time of Thutmosis III.

Scientific methods

The “classical” dating of the Minoan eruption to ca. 153000 BC, determined on the basis of historical methods, was first questioned in 1987, when the evaluation of ice cores from Greenland at that time dated the only major volcanic eruption of the mid-2nd millennium BC to ca. 1645 BC (± 20 years).

The elevated concentration of sulfuric acid found in strata from this period could not be clearly linked to Thera, but was taken as the “most likely candidate for the Minoan eruption” based on the assumption that there had been no other major eruption in the 2nd millennium BC. The assumption that the Minoan eruption was large enough to leave acidic residue even on Greenland was based on Marinatos” original theory of an eruption comparable to Tambora. An eruption of this size, however, had to entail equally short-term changes in climate, a so-called volcanic winter, as had occurred during the largest known eruption in historical times – Tambora in 1815 (see Year Without a Summer).

As early as 1984, dendrochronological examination of longleaf pines in California”s White Mountains (see Bristlecone Pines Chronology) revealed an unusually narrow tree ring dating to 1627 B.C., indicating an extremely cold summer. The inference that this might have been the result of the Minoan eruption was not yet drawn in 1984. This was not done until 1988 – against the background of the Greenland ice core analysis, when an examination of Irish oaks also revealed a sequence of unusually narrow annual rings beginning in 1628 BC. A further investigation in 1996 with wood samples from Anatolia confirmed the climatic anomaly, two above average wide annual rings could point to unusually mild and humid summers. Most recently, in 2000, a study of several pine logs from a peat bog in Sweden found further evidence of climate change.

A direct assignment of the climate change of the 1620s B.C. to the Minoan eruption was not possible with the findings. Thus, astronomical changes or the eruption of another volcano are much more likely as the cause of the tree-ring anomalies and the acid peak in the Greenland ice sheet. For example, in 1990 Canadian researchers proposed the Avellino eruption of Mount Vesuvius, which they dated to 1660 BC (± 43 years) using radiocarbon dating (14C). An eruption of Mount St. Helens was also dated to the 17th century BC.

In 1998, investigations showed that the particles of volcanic glass found in ice cores in 1987 did not chemically match the eruption on Santorini. In 2004, these particles were assigned to the Mount Aniakchak eruption in Alaska using newer analytical methods. This was contradicted since then, the distribution of elements and isotopes of the acid peaks would fit well to the data from Santorini, the high calcium values in clay shards of Santorini would not have to be found necessarily also in the ashes in the Greenland ice, so that it could be with the particles nevertheless around traces of the Minoan eruption.

Some recent 14C dating, in turn, argues for the years 1620 to 1600 B.C.: The 2006 radiocarbon dating of the branch of an olive tree buried by the volcanic eruption on Thera, found in November 2002 in the pumice layer of the island, gave an age of 1613 B.C. ± 13 years. The evidence of leaves shows that the branch was buried alive by the outburst. This was the first time that the individual annual rings of the branch were individually 14C-dated, and their known time intervals significantly reduced the confidence intervals. In 2007, another piece of the same branch and a second, longer and superficially charred branch with several side branches were discovered only nine meters from the first site, which had not previously been dated. Objections were raised against the results because olive trees do not form distinct annual rings, whereupon the authors of the dating pointed out that their result was still unambiguous even without the confidence intervals, only as an assured sequence of samples.

The temporal discrepancy between the findings in the Greenland ice of 1645 B.C. and the 14C data from the 1620s could be put into perspective if one places and analyzes a corresponding curve of the beryllium isotope 10Be next to the classical 14C data. There was a time shift of exactly 20 years, which would make the acid peaks in the ice in the analysis fit much more closely to the presumed data from Santorini.

In 2006, archaeological finds from tsunami deposits at Palaikastro, Crete, using again refined methods, yielded an age of about 1650 ± 30 B.C. The tsunami deposits contain bones of farm animals and pottery together with volcanic ashes from the eruption, thus allowing the application and comparison of three different dating methods.

It is unclear how the Minoan eruption directly or indirectly affected the civilization of the Minoans, since they left neither written nor pictorial representations of the catastrophe. The archaeological evidence already mentioned “only” speaks against a sudden destruction of the Minoan civilization by the eruption, they cannot say more. As the southernmost Cycladic island, Santorini was the only one that could be reached within a day”s journey from Crete and was the central stepping stone for Minoan trade to the north. A network model of Bronze Age maritime trade in the Aegean suggests that the destruction of the Akrotiri base triggered increased short-term trade efforts via alternative routes. In the long term, however, the increased effort would have significantly limited long-distance trade, so that the decline of Minoan culture may have been indirectly promoted by the volcanic eruption.

Except for the aforementioned disputed stele of Pharaoh Ahmose, there is no contemporary evidence of the Minoan Eruption that allows us to draw conclusions about its impact.

It is also unclear whether the Minoan Eruption was reflected in later myths. Thus, numerous local myths reporting floods, as well as the myth of the Deucalion Flood, have been associated with the Minoan Eruption. In all of them, a god”s battle with Poseidon is reported as flooding the land. However, none of these myths explicitly speaks of a volcanic eruption. Therefore, only through partly tortuous interpretation as well as with the assumption of a catastrophic flood after the eruption can be associated with Thera. Interestingly, the Parian Chronicle dates the Deucalionic Flood to 15291528 B.C. and thus lies within the time span of the archaeological-historiographical method.

Talos, who appears in the Argonaut saga, was also interpreted as a reflection of the Minoan eruption: a bronze giant who guards Crete and throws boulders at enemy ships. Richard Hennig assumes that this myth originated in the decades immediately preceding the eruption, when the island”s volcano showed more or less strong activity.

Also the biblical Ten Plagues of the 2nd Book of Moses are associated by different authors with the consequences (Historical Exodus Research) of the Minoan Eruption.

As early as the 1960s, the Greek seismologist Angelos Galanopoulos suspected that the eruption was a model for the sinking of the island state of Atlantis, which Plato described in his works Timaeus and Critias.

36.349444444425.3993083333Coordinates: 36° 20′ 58″ N, 25° 23′ 58″ E


  1. Minoische Eruption
  2. Minoan eruption