When Did the Thera Volcano Erupt? — The Century Between Carbon and Pottery

The Year an Island Was Buried

About 3,600 years ago, a massive volcanic eruption occurred on Thera, an island in the southern Aegean Sea. Thera is the old name for the island we call Santorini today. Akrotiri, a Bronze Age town on the island's southern coast, was buried under thick pumice and volcanic ash. Across the sea to the south on the island of Crete, the Minoan civilization—a Bronze Age culture that traded by sea centered around its palaces—was flourishing, which is why this eruption is commonly called the "Minoan eruption." Around the same time, Egypt was leaving behind a chronology linking together the regnal years of its kings, but nowhere is it recorded which king's reign coincided with this eruption. The question this article follows is singular: in what year did the eruption happen?

There is a reason scholars spent decades clinging to a single year. The layer of volcanic ash left by the eruption serves as a baseline that aligns the chronologies of the Aegean, Egypt, and the Near East, acting as a tick mark that pins natural records—such as polar ice, cave stalagmites, and lake sediments—to historical dates (DOI: 10.1126/sciadv.aar8241). If this single tick mark moves, the timetables of various civilizations stacked on top of it move together.

However, two methods have read this tick mark differently. One is archaeological dating relying on pottery styles and Egyptian chronology, and the other is radiocarbon dating. Radiocarbon dating is a method that measures the radioactive carbon remaining in something that was once alive, counting backward to see how much time has passed since it stopped taking in new carbon. Archaeological dating pointed to the mid-16th to early 15th century BCE, while radiocarbon dating pointed to the late 17th to early 16th century BCE (DOI: 10.1126/sciadv.aar8241). The distance between the two answers was estimated at 100 to 150 years, and the reason was unknown (DOI: 10.1017/s003382220005579x). This article traces where that gap came from and how it narrowed or widened again as new evidence emerged.

Pottery and the Dates Set by Pharaohs

The pottery used by the people of Akrotiri right before the eruption belongs to the "Late Minoan IA" period, among the phases ordered by the sequence of Cretan pottery styles. The eruption occurred near the end of this period (DOI: 10.32028/jga.v9i.2744). This is a sequence. It tells us which pottery style the eruption came after, but not what year that was.

There is an old method for turning a sequence into years. Cross-dating is a method where, when an object from a place with a known date and an object from a place with an unknown date are found together in the same spot, the date of the known side is borrowed for the unknown side. The place in the eastern Mediterranean with a known date was Egypt. Egypt's dynastic chronology was made by stringing together the regnal years of kings read from records and artifacts, and is anchored to an actual calendar by a small number of ancient astronomical observations (DOI: 10.1126/science.1189395).

The number produced by this method is 1500 BC. In a Greek tomb, pottery from the Late Minoan IA period was found together with an Egyptian scarab-shaped seal, leading to the estimate that this pottery style was no later than around 1500 BC. In 1939, Greek archaeologist Spyridon Marinatos proposed this year 1500 as the year of the eruption (DOI: 10.1017/cbo9780511607837.009). Items from the next period, Late Minoan IB, came out in contexts linked to the time of Pharaoh Thutmose III of the Eighteenth Dynasty of the New Kingdom of Egypt, and because his reign is placed in the first half of the 15th century, the end of the IA period had to come before that (DOI: 10.32028/jga.v9i.2744). The author of that paper views this year 1500 as a number that rounds off this minimum date, and evaluates that it did not come from any specific evidence (DOI: 10.32028/jga.v9i.2744).

The Egyptian chronology itself was tested with carbon afterward, and did not collapse significantly. In dates calculated by combining radiocarbon measurements of 211 short-lived plant samples with the kings' reign information, the New Kingdom began between 1570 and 1544 BC. This is a slightly earlier value than some existing historical estimates (DOI: 10.1126/science.1189395). Accepting these results, one study reports that the possibility of major flaws in the Egyptian historical chronology has been ruled out, yet writes that archaeology's "low" chronology and carbon's supported "high" chronology still stand opposed in the Aegean (DOI: 10.1017/s0033822200047196). Therefore, the ground for the dispute narrows down to the link connecting the Egyptian chronology and the Aegean strata. The results of testing that link with carbon did not converge on one side. At the Tell el-Dab'a site east of the Nile Delta, the carbon age of the eruption matched the carbon values of the Egyptian Eighteenth Dynasty strata, yielding the result that the traditional correspondence connecting the Aegean and Egypt is mostly correct (DOI: 10.1017/s003382220005579x). Another research team wrote that the core of the problem appears to lie in the correspondence matching archaeological strata and political history (DOI: 10.1017/rdc.2017.88).

The year 1500 BC is a date borrowed from the chronology of an Egyptian king through the link of pottery. A borrowed date is only as solid as the link is solid.

Carbon Pushes Dates Up by a Century

In the 1970s, charred organic matter from the Akrotiri excavation site was sent to the radiocarbon laboratory at the University of Pennsylvania (DOI: 10.1017/cbo9780511607837.009). This involved measuring the radioactive carbon remaining in the sample, and that radioactive carbon is called carbon-14. Because the amount of carbon-14 in the air has varied across different eras, simply calculating the measured value does not yield the actual year. A calibration curve is a reference table that measures the carbon-14 of samples of known ages to trace what the carbon-14 levels in the air were year by year. Trees whose exact growing years are known by counting their tree rings provide the material for this reference table.

The freshly measured values of the Akrotiri sample seemed to match traditional dates, but when transferred to the tree-ring reference table, they yielded dates earlier than 1600 BC. Initially, an explanation emerged that old carbon from volcanic gases leaking nearby had contaminated the sample (DOI: 10.1017/cbo9780511607837.009). A 1976 review summarized that the Aegean carbon dates from the second millennium BC were consistent with one another and matched traditional dates well in the later part of the Late Bronze Age, but were distinctly early around its threshold in Late Minoan IA. The author noted that artifact evidence did not support these early dates and that current evidence could not explain them (DOI: 10.2307/503574).

Starting in the 1980s, arguments emerged that read this discrepancy from the opposite direction. A 1987 study re-examined pottery links connecting the Aegean and Egypt, concluding that the Aegean Late Bronze Age began during the Hyksos period prior to the Egyptian New Kingdom, namely the Second Intermediate Period (DOI: 10.1111/j.1475-4754.1987.tb00396.x). A 1988 comprehensive analysis deemed the traditional date of around 1500 too late and placed the eruption in the late 17th century, noting that 1628–1626 BC, the date proposed in 1984 for frost-damage traces within tree rings, also seemed viable as the year of the eruption (DOI: 10.1558/jmea.v1i1.17). In 2002, a report using carefully selected organic samples concluded that the eruption most likely occurred around 1650–1620 BC (DOI: 10.1017/s0003598x00091183). In 2006, a chronology was published that calculated carbon dates and pottery chronological sequences of the surrounding region together, concluding that the eruption must have been in the late 17th century and that the Late Minoan IA, IB, and II periods combined should be pushed back by about 100 years (DOI: 10.1126/science.1125682).

In that same year, 2006, a sample appeared that seemed like the culmination of this trend. An olive tree branch was discovered in the island's caldera wall, which researchers believed had been buried alive while standing upright in pumice during the eruption (DOI: 10.1126/science.1125087). Wiggle matching is a method of measuring the carbon-14 values of multiple segments with known sequential order from a single tree, and fitting the shape of their rises and falls directly onto the undulations of the calibration curve. The researchers applied this method to this branch. Because the calibration curve does not descend smoothly in a single direction but rather undulates, a single sample has limitations in precision, but multiple values with a sequence draw the same undulations as the curve and can thus be matched against it (DOI: 10.1017/s0033822200035967). The researchers found 72 tree rings using X-ray computed tomography and measured and matched four consecutive segments (DOI: 10.2307/jj.35529235.28). The result was a range of 1627–1600 BC, estimated to encompass the date with a 95.4% probability. This is a century earlier than the dates derived from the traditional Egyptian chronology (DOI: 10.1126/science.1125087). The researchers later noted this as 1613 BC ±13 years, adding that because the branch grew at an elevation of 150m and over 2.5km away from an active volcanic zone, the possibility of it being affected by old carbon from volcanic gases was low (DOI: 10.2307/jj.35529235.28).

This branch quickly became a target. In a 2014 print debate in the archaeology journal Antiquity, one research team argued, based on the results of investigating modern olive tree rings on Santorini, that this date was difficult to believe for several reasons (DOI: 10.1017/s0003598x00050365). According to the rebuttal by the researchers who measured the branch, this critique combined botanical investigations with other arguments in an attempt to defend the traditional "low" chronology around 1500 BC (DOI: 10.1017/s0003598x00050390). A commentary supporting the critique went as follows. It stated that the branch yielding a late 17th-century date itself was not surprising, because most other short-lived samples from Akrotiri fell in that late century. However, because olive trees do not produce tree rings every year, performing wiggle matching by connecting non-existent tree rings and matching them to the calibration curve was considered an illusion of precision. It also stated that there was no evidence the branch was alive at the time of the eruption, nor did the outermost bark side remain. It concluded that using carbon to date the Thera eruption inherently clashed sharply with historical and archaeological chronologies, making it far more problematic than scientists acknowledge (DOI: 10.1017/s0003598x00050389). The researchers who measured the branch countered in their rebuttal by refuting the main arguments of the critique point by point and responding that the critique had ignored other carbon dating studies regarding the eruption (DOI: 10.1017/s0003598x00050390).

Other carbon-side studies deliberately pushed the scope for later dates to their utmost limits. Regional differences in carbon values were only on the scale of a few years, and volcanic gas contamination remained a hypothetical stage that had never been observed in the field. Even if all the assumptions pointing to later dates were accepted, the late 17th century remained the most probable. However, looking at the Santorini data alone, the possibility of the mid-16th century (approximately 1593–1530 BC) grew stronger than in previous analyses (DOI: 10.1017/s0033822200047202). The strength of the carbon-side answer lay in the fact that multiple samples pointed in the same direction. Regarding the single sample that appeared most precise, disputes arose simultaneously over whether that sample itself could be trusted and whether carbon dating could contend with archaeological dating.

The Volcanic Ash in the Ice Belonged to a Different Volcano

While carbon and pottery were clashing, a third clock stepped in. Ice cores are pillars of ice drilled straight down from the top to the bottom of the ice sheets in Greenland or Antarctica, where snow falling year after year forms layers that allow researchers to count the years. When a large volcano erupts, it leaves a layer high in acidity in that year's layer. A 1980 paper introducing a method to read these acidic layers lined up the acidic layers of Thera alongside historical eruptions at around 1400 BCE (DOI: 10.1017/s0022143000015227). Attaching the name of a specific volcano to the acidic layers in the ice became another thread in this debate over the next twenty years or so.

In 2003, an argument emerged that tiny volcanic glass shards from the 1645 BCE (±4 years) layer of the Greenland GRIP ice core came from Thera. If that argument were correct, the year of the eruption could be pinned down to a single year (DOI: 10.1029/2003gc000608). A rebuttal that same year showed that the matching method was flawed and that, based on chemical data, the shards did not belong to Thera (DOI: 10.1029/2003gc000608). A study the following year reported that the glass in the ice and the volcanic ash of Thera differed in the concentrations of several elements and thus could not be the same, while every element was strikingly similar to the volcanic ash of the Aniakchak volcano in Alaska. The authors nevertheless noted that the date of the Thera eruption remained unresolved (DOI: 10.1029/2003gc000672). A 2008 commentary refuted the Greenland chronology dating this layer to the Thera eruption, reported that analyzing individual glass shards also pointed to Aniakchak, and noted that this interpretation supported radiocarbon dates placing the eruption around 1630–1600 BCE (DOI: 10.1029/2007jd008970). A 2012 study reported that North American eruptions likely contributed to the acid signal around this time, concluding that researchers should exercise greater caution when assigning acid signals to specific eruptions (DOI: 10.1029/2012jd017698).

In 2022, the counting of years in the Greenland ice was revised once again. A new chronology, recounted by matching multiple ice cores, came out about 13 years younger than the previous chronology around 3,800 years ago, bringing it into alignment with IntCal20, the 2020 edition of the radiocarbon calibration curve. However, the authors found no decisive evidence to match the Thera eruption to the acid signal in the ice, and newly observed a single cooling event that was not pronounced in the acid signal around 1600 BCE. In the authors' words, clear traces of Thera in the Greenland ice remain elusive (DOI: 10.5194/cp-18-1125-2022).

That same year, a different research team aligned ice records from Antarctica and Greenland to date each volcanic event during this period. Finely sliced sulfur components and volcanic ash showed that the volcano that lofted sulfur high into the sky in the 1628 BCE layer was the Aniakchak II eruption in Alaska, and the authors noted that this result put an end to decades of speculation that the 1628 event was Thera. The candidates for Thera left by the authors are sulfur emission events around 1611 BCE, 1561, 1558, and 1555 BCE, and 1538 BCE, and because all of these events involve relatively small amounts of sulfur, the authors believe that Thera's capacity to shake the climate was far smaller than the 1815 eruption of Mount Tambora in Indonesia (DOI: 10.1093/pnasnexus/pgac048).

A candidate signal also turned up in trees. A study that anchored tree-ring records from the Eastern Mediterranean Bronze and Iron Ages to precise calendar years found traces of environmental changes in the chemistry of those tree rings during growth around 1560 BCE. The authors noted that while this requires further verification, it could be evidence of the Thera eruption (DOI: 10.1073/pnas.1917445117). Ice and trees have yielded a list of candidate dates, but which of them is Thera has not yet been named.

Let's Redraw the Reference Table, the Intervals Moved

In 2018, the answer from the carbon side moved from within. Researchers remeasured carbon-14 year by year for 1700–1500 BCE using tree rings with known calendar years. Those values disagreed with the international calibration curve in use at the time, and that discrepancy shifted the calibrated date of Thera toward the 16th century (DOI: 10.1126/sciadv.aar8241). The values obtained when the Irish oak used in the 2018 measurement was remeasured in a different Swiss laboratory matched the initial measurement, but left new questions about differences in carbon values from region to region (DOI: 10.1017/rdc.2020.14).

Regional differences became a new variable in this debate. A 2012 review found no evidence for regional differences except that they were very small across the Aegean and Anatolia regions (DOI: 10.1017/s0033822200047202). In a 2020 comparison with improved measurement precision, Mediterranean carbon values showed small but decisive differences, particularly at positions where the reference table flips or flattens. These differences can shift dates, including Egyptian history and the Thera eruption, by up to several decades (DOI: 10.1126/sciadv.aaz1096). Another study that same year reported discrepancies of 0 to 22 years between IntCal20 and carbon-14 ages in Mediterranean and Anatolian trees, citing generational differences in measurement technology and the duration and timing of the growing season of the trees as causes (DOI: 10.1038/s41598-020-69287-2).

In 2026, researchers in Groningen, Netherlands, note that differences of opinion have grown since IntCal20 was released. This is because as year-by-year measurements increased, the possibility that atmospheric carbon-14 rose and fell during this period came to light. The researchers added about 90 new year-by-year measurements for 1660–1507 BCE, and found no persuasive evidence that carbon-14 was produced in large amounts suddenly during this period. Instead, they confirmed small differences among laboratory data that may have stemmed from tree species, region, or sample handling methods (DOI: 10.1017/rdc.2025.10186).

When the reference table changes, the year for the same measured value also changes. The four values of the olive branch from 2006 remain the same, but every time the calibration curve improves, its calibrated date has changed. One study reported that average values obtained by pooling multiple tree rings should not be matched directly to a year-by-year reference table, and re-matched the reference table by averaging it to match the length of that segment. The result was a range running from the late 17th to the mid-16th century, slightly younger than previous estimates (DOI: 10.1017/rdc.2023.35). The answer on the carbon side moves when the reference table changes, even if no new samples are measured.

Olive Shrubs on a Flat Section

In 2023, on Therasia, an island right next to Thera, a charred olive shrub buried in the sediment of the same eruption was reported. This shrub came out a little younger than the twigs from 2006. The research team combined the calibration results and the counted growth increments of the twigs, reporting that the probability of the eruption being in the mid-16th century has grown and that timing overlaps with several volcanic sulfur signals in ice cores (DOI: 10.1038/s41598-023-33696-w).

A flat section is a spot where the calibration curve lies almost flat for decades, so carbon-14 values come out similar for any sample that died in between. The measured value of this shrub fell across a flat section spanning from around 1620 to 1540 BC (DOI: 10.1017/rdc.2024.44).

Let's follow step by step what happens in this spot. We measure a single carbon-14 value from a piece of the shrub. We match that value to the calibration curve and find the years when that value could occur. If the curve is on a sloping section, the matching years gather to just a few years. On a flat section, nearly eighty years from around 1620 to around 1540 all yield similar values. So if calibrated without other information, the probability of the year the sample died spreads wide across the entire flat section, yielding a wide range of little use (DOI: 10.1017/rdc.2024.44). Among the candidates put forth by the ice, the periods around 1611 and 1561 both fall within this section (DOI: 10.1093/pnasnexus/pgac048). This is why it is difficult to tell the two candidates apart using only carbon values without sequence information.

To narrow down the spread-out probability, information outside the measured values is needed. The order of strata, the context from which the sample came, and the nature of the sample itself can be such information (DOI: 10.1017/rdc.2024.44). A Bayesian age model is a calculation that narrows down the range of each date by putting sequence information—such as "this layer piled up earlier than that layer"—together into multiple measured values. For a shrub, the sequence that the inner piece of a branch grew before the outer piece, and the assumption that multiple branches died all at once in the same eruption, can be such information. A 2024 reanalysis reported that the shrub likely died from the eruption and put this assumption into the calculation (DOI: 10.32028/jga.v9i.2744).

The 2024 reanalysis diverged right at this juncture. Manning, the researcher who published the reanalysis, is from the side that has argued for an early date for the Aegean since the 1980s (DOI: 10.32028/jga.v9i.2744). He views the range first obtained by the 2023 report as an inconclusive range "spanning the late 17th century and the entire 16th century," and summarizes that arguments supporting the mid-16th century followed afterward. He put in more plausible sequence information to calculate again, and reports the range with a high probability of containing the eruption year as between the late 17th century and the early 16th century (DOI: 10.1017/rdc.2024.44). The same measured values of the same shrub, depending on what sequence information is entered, produced two readings: "toward the mid-16th century" and "from the late 17th to the early 16th century." Which sequence information is "appropriate" is the substance of the dispute now.

Did the Eruption Collapse the Minoan Civilization?

The story that the eruption ended the Minoan civilization of Crete is more widely known outside of this debate. A 1976 interpretation suggested that Akrotiri was not abandoned until around 1470 BC, and since the Late Minoan IB destruction layers at various Cretan sites date to around the same time, both destructions could be explained by a single eruption. The author suggested that the pattern of destruction in Crete fits natural forces like shockwaves, tsunamis, and volcanic ash better than invaders (DOI: 10.2307/502934).

The current relative sequence differs from that picture. The eruption occurred near the end of the Late Minoan IA period (DOI: 10.32028/jga.v9i.2744), and even in the tsunami deposits at Palaikastro on the eastern coast of Crete, the latest artifacts were IA pottery (DOI: 10.1017/s003382220005579x). The IB period, which has been counted among the destruction layers of various Cretan sites, is the period that follows. Evidence that the damage from the eruption was severe is found in many places. At Malia on the northern coast of Crete, a tsunami reached the outskirts of the palace center, pushing water up to 400 meters inland, and the researchers wrote that this result opens a way to newly study the consequences for the Minoan civilization (DOI: 10.1038/s41598-021-94859-1). Today, at Çeşme on the western coast of Turkey, human and dog bones were found within the tsunami debris, and that deposit turned out to be no earlier than 1612 BC (DOI: 10.1073/pnas.2114213118). Tsunami deposits from this eruption have also been identified on the continental shelf off the coast of Israel, and their dating range encompasses both the "high" and "low" chronologies (DOI: 10.1130/g25704a.1).

There is time between the damage and the collapse. Accepting evidence that Crete continued to prosper even after the eruption, one research team used a maritime trade network model to show that the effect of losing a major port of call could appear with a time lag. This is a path where transportation costs rise, shipping routes gradually dwindle, and the economy eventually collapses (DOI: 10.1017/s0003598x00068459). This is a possible path shown by the model, not an observation that it actually happened that way. How many years passed between the eruption and the collapse depends on the question of this article, namely the date of the eruption and the lengths of the IA and IB periods. If these periods are stretched by about 100 years, like the 2006 carbon chronology, their lengths are recalibrated (DOI: 10.1126/science.1125682). While we can state the sequence that the eruption preceded the IB-period destruction of Crete, the causal link that the eruption collapsed the Minoan civilization remains a matter of interpretation.

What Has Been Narrowed Down and What Remains

An evaluation in 2024 suggested that the margin of dispute has narrowed from over a century to about 50 years. The evaluation was issued by Manning, the very same researcher who produced the shrub reanalysis mentioned earlier. According to his summary, the remaining options after 2020 are around 1611 BCE, where a major eruption is captured in the ice records (or broadly around 1600 BCE, if Thera is not captured in the ice records examined so far), and around 1561 BCE, where another major eruption is captured. Either way, the Late Minoan IA period coincides with the Hyksos period in Egypt, namely the Second Intermediate Period. Manning himself considers the 1611 side more likely (DOI: 10.32028/jga.v9i.2744). The research team that reported on the shrubs favors the mid-16th-century side (DOI: 10.1038/s41598-023-33696-w). Another research team, which compared the strata of Egyptian and Levant ruins using only carbon dating, suggested that the eruption could fall within the Second Intermediate Period (DOI: 10.1017/rdc.2017.88). A review as early as 2012 argued that the discrepancy between archaeological dates and carbon dates could be reduced to 30 to 50 years (DOI: 10.1017/s0033822200047196).

What remains is also clear. The candidate dates around 1611, 1561, and 1538 BCE lie within or at the edge of the plateau section of the calibration curve (roughly 1620–1540 BCE), and cannot be pinpointed to a single year by carbon alone without sequence information (DOI: 10.1017/rdc.2024.44) (DOI: 10.1093/pnasnexus/pgac048). Evidence that Mediterranean carbon values differ slightly from the reference table can shift dates by decades in this flat spot (DOI: 10.1126/sciadv.aaz1096). The traces of Thera in the ice have not yet been named (DOI: 10.5194/cp-18-1125-2022). In 2009, a research team proposed not mixing the two systems, but using them side by side. They suggested recording archaeological dates and carbon dates separately following the Egyptian chronology, and establishing a sort of "calibration curve" between the Egyptian chronology and carbon dates of the second millennium BCE to also find the causes of the discrepancies (DOI: 10.1017/s003382220005579x).

What this debate leaves behind is not a single year, but how to read the range attached to a single year. 1627–1600 BCE, 1613 ± 13 years, "the late 17th to early 16th century," and "the mid-16th century side" all make sense only when you hear together which sample was matched against which reference table and with what sequence information. The principle by which radiocarbon measures the age of a sample and a tree-ring reference table translates that value into a calendar year is covered from the beginning in "Dating with Carbon — Radiocarbon Dates, Read with an Hourglass and Tree Rings". How to read dates out of records, like the Egyptian dynastic chronology, and its limitations are covered in "Reading Historical Records — Grandmother's Diary and That Day's Newspaper".

Source Papers

• "Annual radiocarbon record indicates 16th century BCE date for the Thera eruption" (2018) — DOI: 10.1126/sciadv.aar8241 — Single-year carbon records from 1700 to 1500 BCE disagree with the international calibration curve, shifting the Thera calibrated date toward the 16th century. Resolves the conflict between carbon (late 17th to early 16th century) and archaeology (mid-16th to early 15th century).
• "The Minoan Santorini Eruption and Tsunami Deposits in Palaikastro (Crete): Dating by Geology, Archaeology, 14C, and Egyptian Chronology" (2009) — DOI: 10.1017/s003382220005579x — The latest artifacts in the Palaikastro tsunami deposit are LM IA pottery; the cause of the 100-to-150-year gap between carbon and Egyptian chronology is unknown. Proposes running the two dating systems in parallel and an Egyptian-carbon "calibration curve."
• "Thera, the Aegean, Egypt, the Hyksos and Anatolia: rethinking the orthodox synchronisations and histories" (2024) — DOI: 10.32028/jga.v9i.2744 — Origins of the traditional 1500 date (LM IB artifacts and Thutmose III); post-2020 options around 1611/1600 or around 1561; dispute width narrowed to about 50 years, with the author considering 1611 more likely.
• "Radiocarbon-Based Chronology for Dynastic Egypt" (2010) — DOI: 10.1126/science.1189395 — Carbon measurements of 211 short-lived plants plus a Bayesian model place the start of the New Kingdom at 1570–1544 BCE; a 19-year difference in the Nile valley is estimated to be a growing-season effect.
• "The Date of the Minoan Santorini Eruption: Quantifying the "Offset"" (2012) — DOI: 10.1017/s0033822200047196 — Rules out major flaws in the Egyptian historical chronology, but leaves the high-versus-low chronology debate open, raising the possibility that the offset could shrink to 30–50 years.
• "The Minoan Santorini Eruption and its14C Position in Archaeological Strata: Preliminary Comparison Between Ashkelon and Tell El-Dabca" (2017) — DOI: 10.1017/rdc.2017.88 — The core of the problem is how archaeological strata map to political history; the eruption could fall within the Second Intermediate Period.
• "Case Study: The elusiveness of time" (2000) — DOI: 10.1017/cbo9780511607837.009 — The date of 1500 derived from LM IA pottery and Egyptian seals, proposed by Marinatos (1939), with 1970s Akrotiri carbon measurements coming out earlier than 1600 after calibration.
• "Carbon-14 and the Beginning of the Late Bronze Age in the Aegean" (1976) — DOI: 10.2307/503574 — Aegean carbon dates are consistent, with only the LM IA period clearly named; unsupported by artifacts and unexplainable by current evidence.
• "RESEARCH NOTES AND APPLICATION REPORTS DATING THE AEGEAN LATE BRONZE AGE WITH RADIOCARBON" (1987) — DOI: 10.1111/j.1475-4754.1987.tb00396.x — Re-examining pottery loops leads to the conclusion that the Aegean Late Bronze Age began during the Hyksos period.
• "The Bronze Age Eruption of Thera" (1988) — DOI: 10.1558/jmea.v1i1.17 — The traditional 1500 date is too late, suggesting a late 17th-century eruption; frost rings in 1628–1626 seem plausible.
• "New evidence for an early date for the Aegean Late Bronze Age and Thera eruption" (2002) — DOI: 10.1017/s0003598x00091183 — Based on LM IA and IB samples, the most likely date for the eruption is around 1650–1620 BCE.
• "Chronology for the Aegean Late Bronze Age 1700-1400 B.C." (2006) — DOI: 10.1126/science.1125682 — Bayesian analysis places the eruption in the late 17th century, extending the LM IA, IB, and II periods by about 100 years.
• "Santorini Eruption Radiocarbon Dated to 1627-1600 B.C." (2006) — DOI: 10.1126/science.1125087 — Wiggle-matching of an olive branch buried in pumice yields 1627–1600 BCE (95.4%), a century earlier than dates in the Egyptian chronology.
• "The Minoan eruption of Santorini radiocarbon dated by an olive tree buried by the eruption" (2009) — DOI: 10.2307/jj.35529235.28 — 1613 ±13 CE, 72 tree rings across four joined segments; located 150 m above sea level and over 2.5 km from the volcanic zone, making the old carbon effect unlikely.
• "Wiggle-Match Dating of Tree-Ring Sequences" (2004) — DOI: 10.1017/s0033822200035967 — The limits of single dates on a wiggly calibration curve and the principle of time-series matching (wiggle-matching).
• "The olive-branch dating of the Santorini eruption" (2014) — DOI: 10.1017/s0003598x00050365 — Introduces a debate in the pages of Antiquity, arguing that the olive branch dates are hard to trust for various reasons.
• "Radiocarbon and the date of the Thera eruption" (2014) — DOI: 10.1017/s0003598x00050389 — A commentary supporting the criticisms. While assignment to the late 17th century is not surprising in itself, the precision of wiggle-matching performed on an olive tree lacking annual rings is an illusion; lacking evidence of survival and bark-side wood, dating Thera with carbon sharply conflicts with historical and archaeological dates, making it problematic.
• "The Thera olive branch, Akrotiri (Thera) and Palaikastro (Crete): comparing radiocarbon results of the Santorini eruption" (2014) — DOI: 10.1017/s0003598x00050390 — Rebuttal by the branch researchers: critics are trying to defend the low date around 1500, refuting main arguments and responding that other carbon studies were ignored.
• "Considerations of the Scale of Radiocarbon Offsets in the East Mediterranean, and Considering a Case for the Latest (Most Recent) Likely Date for the Santorini Eruption" (2012) — DOI: 10.1017/s0033822200047202 — Regional differences are on the order of a few years; the volcanic gas effect remains hypothetical; the late 17th century is still most likely, but Santorini data alone raise the possibility of the mid-16th century (approx. 1593–1530 BCE).
• "Acidity of Polar Ice Cores in Relation to Absolute Dating, Past Volcanism, and Radio–Echoes" (1980) — DOI: 10.1017/s0022143000015227 — Method for detecting annual ice core layers and volcanic acid layers, placing Thera around 1400 BCE.
• "Volcanic ash retrieved from the GRIP ice core is not from Thera" (2003) — DOI: 10.1029/2003gc000608 — The method matching Greenland volcanic ash to Thera was flawed; chemical data suggest it is not Thera.
• "Identification of Aniakchak (Alaska) tephra in Greenland ice core challenges the 1645 BC date for Minoan eruption of Santorini" (2004) — DOI: 10.1029/2003gc000672 — Glass in the 1645 (±4) layer differs from Thera and resembles Aniakchak, leaving the Thera date unresolved.
• "Comment on "A synchronized dating of three Greenland ice cores throughout the Holocene" by B. M. Vinther et al.: No Minoan tephra in the 1642 B.C. layer of the GRIP ice core" (2008) — DOI: 10.1029/2007jd008970 — Single-grain analysis also points to Aniakchak for the 1642 volcanic ash layer, an interpretation that supports carbon dates around 1630–1600 BCE.
• "Holocene tephras highlight complexity of volcanic signals in Greenland ice cores" (2012) — DOI: 10.1029/2012jd017698 — North American eruptions may have contributed to 17th-century BCE acid signals, urging caution in assigning acid signals.
• "A multi-ice-core, annual-layer-counted Greenland ice-core chronology for the last 3800 years: GICC21" (2022) — DOI: 10.5194/cp-18-1125-2022 — The new chronology is about 13 years younger at 3,800 years ago, aligns with IntCal20, has no decisive evidence for Thera matching, and shows a cooling event around 1600 BCE.
• "Geochemical ice-core constraints on the timing and climatic impact of Aniakchak II (1628 BCE) and Thera (Minoan) volcanic eruptions" (2022) — DOI: 10.1093/pnasnexus/pgac048 — Sulfur in the 1628 BCE layer is Aniakchak II; Thera candidates are around 1611, 1561/1558/1555, and 1538, all with climate-forcing potential far smaller than Tambora.
• "Securing timelines in the ancient Mediterranean using multiproxy annual tree-ring data" (2020) — DOI: 10.1073/pnas.1917445117 — Anchoring Eastern Mediterranean tree rings to calendar years; chemical changes around 1560 BCE need further proof but could be evidence of Thera.
• "Annual Variation in Atmospheric 14C Between 1700 BC and 1480 BC" (2020) — DOI: 10.1017/rdc.2020.14 — Re-measurement of Irish oak matches existing measurements, raising questions about regional differences.
• "Mediterranean radiocarbon offsets and calendar dates for prehistory" (2020) — DOI: 10.1126/sciadv.aaz1096 — Small but decisive differences in Mediterranean carbon values where standard tables are inverted or flat, which can shift dates by up to several decades.
• "Radiocarbon offsets and old world chronology as relevant to Mesopotamia, Egypt, Anatolia and Thera (Santorini)" (2020) — DOI: 10.1038/s41598-020-69287-2 — 0 to 22 radiocarbon-14-year difference with IntCal20, caused by measurement technology generations and the growing season.
• "Investigating potential radiocarbon anomalies around the time of the Minoan eruption of Thera: A new high-resolution dataset from Groningen" (2026) — DOI: 10.1017/rdc.2025.10186 — Widening differences of opinion following IntCal20, about 90 single-year samples from 1660 to 1507 BCE, no evidence of sudden production increases, and small differences among datasets.
• "CALIBRATION OF MULTIPLE TREE-RING BLOCKS AND ITS IMPLICATION ON THE DEBATE OF MINOAN ERUPTION OF SANTORINI AROUND 17TH–16TH CENTURY BCE" (2023) — DOI: 10.1017/rdc.2023.35 — Olive branch calibrated dates change with every curve improvement, pointing to the late 17th to mid-16th century via moving-average curves.
• "Olive shrub buried on Therasia supports a mid-16th century BCE date for the Thera eruption" (2023) — DOI: 10.1038/s41598-023-33696-w — The Therasia olive shrub is slightly younger than the branch, increasing the probability of a mid-16th-century date and overlapping with ice core sulfur signals.
• "Problems of Dating Spread on Radiocarbon Calibration Curve Plateaus: The 1620–1540 BC Example and the Dating of the Therasia Olive Shrub Samples and Thera Volcanic Eruption" (2024) — DOI: 10.1017/rdc.2024.44 — Probabilities spread out on the 1620–1540 BC plateau; re-analysis incorporating sequential information (Manning) places the likely range in the late 17th to early 16th century.
• "Thera and the Devastation of Minoan Crete: A New Interpretation of the Evidence" (1976) — DOI: 10.2307/502934 — An interpretation attributing Akrotiri and the LM IB destruction of Crete to a single eruption around 1470.
• "Discovery of a tsunami deposit from the Bronze Age Santorini eruption at Malia (Crete): impact, chronology, extension" (2021) — DOI: 10.1038/s41598-021-94859-1 — Tsunami reached the outskirts of the Malia palace center, up to 400 m inland.
• "Volcanic ash, victims, and tsunami debris from the Late Bronze Age Thera eruption discovered at Çeşme-Bağlararası (Turkey)" (2021) — DOI: 10.1073/pnas.2114213118 — Human and dog bones found in tsunami debris, which cannot be earlier than 1612 BCE.
• "Tsunami waves generated by the Santorini eruption reached Eastern Mediterranean shores" (2009) — DOI: 10.1130/g25704a.1 — Tsunami sedimentary layers on the Israeli continental shelf, with dates encompassing both high and low chronologies.
• "The Theran eruption and Minoan palatial collapse: new interpretations gained from modelling the maritime network" (2011) — DOI: 10.1017/s0003598x00068459 — Continued Cretan prosperity after the eruption, presenting a pathway of delayed economic collapse via the loss of port-of-call nodes through maritime network modeling.