Before the gods fall, the cold comes. That is how Snorri Sturluson tells it around 1220: three winters in a row, snow driving from every direction, a sun that is of no use, and before that three years of war. The word for it, fimbulvetr, already appears in an Eddic poem older than Snorri. For some decades archaeologists and climate scientists have asked whether a real catastrophe lies behind the image: the volcanic eruptions of 536 and around 540, which pushed the North into what may have been the coldest decade of the last two thousand years. This article sets out the passages, checks what the natural sciences have actually measured, and keeps evidence and interpretation apart. The short answer first: the climate crisis is documented, and the Fimbulwinter as a text is documented. That the one tells of the other cannot be proven.
Fimbulvetr, anglicised as Fimbulwinter, is an Old Norse compound of fimbul- and vetr, "winter". The first element never stands on its own. It intensifies, roughly in the sense of "mighty", "monstrous" or "great", and it turns up in other words of Eddic poetry: the Völuspá calls Odin Fimbultýr, the "mighty god", in one of its final stanzas, and the Hávamál speak of fimbulljóð, mighty songs Odin claims to have learned. Where fimbul- comes from historically is unknown; the dictionaries give its meaning, not its origin.
The word fimbulvetr is rare in medieval literature. Exactly two texts name the great winter as a sign of the world's end: a poem of the Poetic Edda and Snorri's prose narrative, which knows and quotes that poem. An overview of both Eddas is in the article on the Edda.
The oldest mention is in Vafþrúðnismál, the lay of Vafþrúðnir. Odin, calling himself Gagnráðr, visits a wise giant and challenges him to a contest of knowledge in which the loser forfeits his head. The questions move from the making of the world to its end. In stanza 44 Odin asks which humans will be alive when the famous fimbulvetr has passed among men. In Old Norse the two decisive lines read hvat lifir manna, þá er inn mæra líðr / fimbulvetr með firum? The adjective mærr means "famous" or "glorious". In the poem, then, the winter is already a known quantity, something people talk about. The stanza does not explain it; it assumes it.
Vafþrúðnir answers in stanza 45 with two names: Líf and Lífþrasir hide in Hoddmímis holt, the wood of Hoddmímir, live on morning dew and become the ancestors of a new humanity.
Fjölð ek fór, fjölð ek freistaðak, / fjölð ek of reynda regin: / hvat lifir manna, þá er inn mæra líðr / fimbulvetr með firum? … Líf ok Lifþrasir, en þau leynask munu / í holti Hoddmímis; / morgindöggvar þau sér at mat hafa, / en þaðan af aldir alask.
"Much have I travelled, much have I tried, much have I tested the powers: what of mankind will live when the famous mighty winter passes among men? … Líf and Lífþrasir, and they will hide in Hoddmímir's wood; morning dews they will have for food, and from them generations will grow."Vafþrúðnismál, stanzas 44 and 45; normalised Old Norse text read via a retrieval tool on voluspa.org; English version my own translation
Two things stand out. Older translators often did not keep the word at all; Karl Simrock's German version (first published 1851) renders it as "Schreckenswinter", "winter of terror". And the poem says nothing about duration, nothing about three winters, nothing about missing summers. It names one great winter and two survivors. That is all. The claim, common even in good accounts, that the poem describes three winters without summer really rests on Snorri.
When the poem was composed is open. A date in the tenth century is common but not certain. It survives in the Codex Regius, an Icelandic manuscript from around 1270.
The full version is in Gylfaginning, the first narrative part of the Prose Edda, which Snorri Sturluson wrote around 1220 to 1225. After Baldr's death and the binding of Loki the talk turns to the end. Chapter 51 (in the usual division; some editions number differently) says, in my own rendering of the normalised Old Norse: A winter comes that is called fimbulvetr. Then snow drives from all directions. The frosts are great and the winds sharp. The sun is of no use. These winters go three together, with no summer between. But before that come three other winters in which there are great battles over the whole world. Then brothers kill each other out of greed, and no one spares father or son in killing and in breaking kinship.
The sentence about the sun is only three words in the original: Ekki nýtr sólar. It can be read as "the sun no longer helps" or "there is no benefit from the sun". Snorri does not say the sun is darkened; he says it no longer has any effect.
The text names the Fimbulwinter first, but then, with the word áðr, "before", it puts the three war winters ahead of it. Count it out and you get six winters: three years of war and kin-slaying, then three years of cold without summer. Only after that does the wolf swallow the sun, a second wolf seize the moon, the stars fall, the earth shake and Fenrir break loose. The wolves are unnamed at this point; Snorri introduced them in chapter 12 as Skǫll and Hati (more in the article on Skoll and Hati).
To back this up, Snorri immediately quotes a stanza of Völuspá that begins "Brothers will fight". The three war winters connect the brother-war stanza of Völuspá with the Fimbulwinter of Vafþrúðnismál. Whether Snorri had a further, now lost source for the number three and the absent summers, or whether he systematised the material himself, cannot be decided. Both are possible, and scholars disagree.
The North's most famous apocalyptic poem, Völuspá, "the prophecy of the seeress", does not know the word fimbulvetr. Two stanzas come close to the image, though.
The first is the stanza on the wolf brood in the Iron Wood (stanza 41 in the widespread numbering of Neckel and Kuhn, 40 in other editions). One of that brood fills itself with the life-blood of doomed men and reddens the seat of the gods with gore. Then comes: svǫrt verða sólskin of sumur eptir, veðr ǫll válynd, "black becomes the sunshine in the summers after, all weathers treacherous". Here there really is darkened sunshine, and in summers at that. For the 536 thesis it is the most tempting line in the whole Edda. One caveat belongs with it: according to Parsons' edition, the Codex Regius reads svǫrt var ða here, a past tense, while the younger Hauksbók has verða. Whether the seeress predicts or looks back depends on the manuscript.
The second is the brother-war stanza (45 after Neckel and Kuhn). Brothers will fight and kill each other, kin will break the bonds of family, it is an axe-age, sword-age, wind-age, wolf-age before the world falls, and no one will spare another. This stanza says nothing of cold, snow or missing summers. Snorri made his three war winters out of it.
Anyone who writes that Völuspá depicts the Fimbulwinter is merging three texts into one. For the question of a historical core the separation is decisive, because every detail one wants to compare with 536 stands in a different text of a different age. The overall course of the end is in the article on Ragnarök.
There are no written records from the North in the sixth century. From the Mediterranean there are, and they agree strikingly. The Byzantine historian Procopius, who accompanied the general Belisarius as secretary in the wars against the Vandals and Goths, notes for 536 that the sun gave its light without brightness all year, like the moon, and looked as if it were in eclipse.
Cassiodorus, a senior official in Ostrogothic Italy, writes in a letter to his deputy Ambrosius (Variae 12.25) that the sun appeared bluish, that bodies cast no shadow at noon, that there had been a winter without storms, a spring without mildness and a summer without warmth, and that frost threatened the harvest.
These texts describe a veil in the upper atmosphere of the kind large volcanic eruptions produce: fine droplets of sulphuric acid, known as aerosols, scatter sunlight back into space. For a long time its origin was unclear; popular catastrophe theories of the 1990s met with scepticism among specialists.
The decisive step came in 2015 with a study in Nature led by the glaciologist Michael Sigl (Sigl et al., "Timing and climate forcing of volcanic eruptions for the past 2,500 years", Nature 523, 2015, pp. 543–549). Ice cores from Greenland and Antarctica contain annual layers, and sulphate from large eruptions settles in them. The team showed that older ice-core chronologies were off by several years in the first millennium and recalibrated them. For the sixth century the result was a double event: a major eruption in the high northern latitudes in 536 and an even stronger, probably tropical eruption around 539/540.
A year later a team led by the dendroclimatologist Ulf Büntgen presented temperature reconstructions from tree rings in the Russian Altai and the Alps in Nature Geoscience (Büntgen et al., "Cooling and societal change during the Late Antique Little Ice Age from 536 to around 660 AD", Nature Geoscience 9, 2016, pp. 231–236). Both series show a long, simultaneous cooling after the eruptions of 536, 540 and 547. The 540s are the coldest decade of the reconstruction in the Altai and the second coldest in the Alps. The authors coined the term "Late Antique Little Ice Age" and dated it from 536 to around 660.
Also in 2016 a group led by Matthew Toohey ran the double event through climate models (Toohey et al., Climatic Change, 2016). According to the GEOMAR press release, the Northern Hemisphere cooled on average by up to two degrees, and northern Europe, Scandinavia in particular, was among the regions where agriculture was most likely to suffer.
Which volcanoes were responsible is still open. For 536, glass particles in an ice core from the Swiss Alps have pointed to Iceland; for 539/540 Ilopango in El Salvador was proposed, but a later dating puts that eruption in 431. I know this debate only from summaries. The signal in the ice is secure, the mountain is not. Whether the plague that raged in the Mediterranean from 541 is connected with the cold is also disputed; Büntgen and his co-authors themselves urge caution about causal claims.
Around the middle of the sixth century the archaeological record in the North changes markedly. In Sweden, by the usual division, the Vendel Period begins, named after rich boat graves in Uppland.
The Swedish archaeologists Bo Gräslund and Neil Price published a much-cited paper in Antiquity in 2012 (Gräslund and Price, "Twilight of the gods? The 'dust veil event' of AD 536 in critical perspective", Antiquity 86, 2012, pp. 428–443). They set aside the question of what caused the dust veil and put the climate data next to the Scandinavian record: abandoned farms, shrinking settled areas, deserted land in districts that had been densely inhabited before. They consider a severe population collapse likely and suggest that the experience echoed in poetry and imagery. I have read only the abstract and summaries.
Regional studies sit alongside this. Daniel Löwenborg examined the cemeteries of the Mälaren valley in 2012 and reads the evidence as a crisis after which rights to abandoned land were renegotiated; a small elite profited, and the Vendel-period burial mounds may have made this new ownership visible. For southern Norway, Kjetil Loftsgarden and Frode Iversen of the Museum of Cultural History in Oslo analysed almost 7,000 dated burials; according to their 2024 summary, the number of graves falls by more than three quarters after the sixth century. Pollen profiles from Norwegian bogs show pastures growing over and forest returning, for instance at Storesætra in Stryn.
The figures sound clear-cut. They are not. Graves are not a census: if fewer of the dead are buried with grave goods, that may mean fewer people or different customs. Abandoned farms may have been merged rather than left empty; Iversen himself suggests that small, unproductive farms were given up and land concentrated. The effects varied by region, and some areas seem to have come through well. Estimates that up to half the population vanished in parts of Scandinavia circulate in popular books and interviews; they should not be read as a measured quantity. Whether the plague reached the North is not documented.
At the end of the Migration Period, in the early to mid sixth century, a striking amount of gold goes into the ground in southern Scandinavia: neck rings, bars, gold bracteates, which are thin pendants stamped on one side after the model of Roman medallions. The Danish archaeologist Morten Axboe has proposed reading some of these hoards as offerings meant to bring back the sunlight.
It remains an interpretation. Hoards can rarely be dated to the year; usually only their youngest pieces can be dated, which gives a point before which the hoard cannot have been buried, not an exact date. Gold can also be hidden out of fear of war. In the Vindelev hoard, found in Jutland in 2020, the bracteates are placed in the second half of the fifth century; the moment of deposition is unknown. A link with 536 is a conceivable background there, not a result. For the techniques behind such work, see the article on goldsmithing; the tiny gold-foil figures found at cult sites are a separate group of finds with their own interpretation (see Guldgubber).
The idea of seeing a real climate event behind the Fimbulwinter is older than the ice cores. Around 1910 the Swedish botanist Rutger Sernander linked the myth to a climatic deterioration at the end of the Nordic Bronze Age, around 650 BC; Since 2012 the debate has shifted to the sixth century, above all through Gräslund and Price.
What speaks in favour? First, the match at the core: a winter that does not end, summers that are no summers, a sun without strength. Second, the situation of the North. Farming at the edge of the possible suffers far more from two or three lost summers than farming in the Mediterranean. Third, the size of the break in the archaeological record, which fits the climate data well in time. Fourth, the darkening line in Völuspá, which speaks of black sunshine in summers. And fifth, Snorri's combination of cold, war and the breakdown of kinship, which one might expect after a famine, even though in his account the war comes before the winter.
What speaks against it, or calls for caution? The gap in time is huge. Between 536 and the Codex Regius lie a good 700 years, between 536 and Snorri almost 700. Even if Vafþrúðnismál is placed in the tenth century, around four centuries of oral tradition remain for which there is no intermediate link at all. A myth changes in every generation and can absorb other experiences. Hard winters came to the North again and again. Then there is comparative religion: a devastating winter before a new beginning also appears in the Iranian Avesta, where Yima is warned of deadly winters and shelters people and animals in an enclosure. The Dane Axel Olrik drew this comparison at the start of the twentieth century. The motif does not, then, necessarily require one particular historical event.
There is also a rival. In 2018 a team led by Clive Oppenheimer connected the Icelandic Eldgjá eruption of 939/940 with the imagery of Völuspá, with darkened sun and fire. The same stanza can thus be linked with two events four hundred years apart. The Rök runestone from Östergötland was also interpreted in 2020 as a text about fear of a new climate catastrophe, a well-argued but disputed reading.
On balance a hypothesis remains which many specialists consider plausible and which cannot be proven with the available means. Nor can it be refuted. The climate crisis itself is not in question. It is as well documented as anything from the sixth century can be.
The word lives on in the Scandinavian languages. Fimbulvinter in Danish, Norwegian and Swedish means an unusually long, harsh winter, with no end of the world attached.
The best-known modern treatment is Santa Monica Studio's God of War series. In the 2018 game, released on 20 April, the Fimbulwinter sets in at the end; the sequel God of War Ragnarök of 9 November 2022 takes place in this cold, which according to the first game lasts three years. The characters connect the winter with Baldur's death without the game supplying a precise cause. The three years correspond to Snorri's three winters; the war winters before them and a causal link to Baldr's death are not in the sources in this form. In Snorri the winter comes after Baldr's death in the order of the narrative, but it is one sign among many, not a consequence. What else the game does with the mythology is examined in God of War fact check.
In neither source is the Fimbulwinter an end point. In Vafþrúðnismál it is the time Líf and Lífþrasir survive; the poem goes straight on to ask about the new sun, the daughter of the old (see Sunna). In Snorri the winter is followed by the wolves, the breaking loose of Fenrir, the battle and the world fire kindled by Surtr. Snorri moves the hiding of Líf and Lífþrasir into the time of the fire rather than the cold, a small shift from the poem. Then a green earth rises from the sea; the article After Ragnarök tells that part.
What is the Fimbulwinter? The great winter before Ragnarök. The word is in Vafþrúðnismál; the elaboration with three winters without summer is Snorri's, in Gylfaginning (ch. 51).
How long does the Fimbulwinter last? According to Snorri, three winters in a row without summer, preceded by three winters of great wars. The older poem gives no duration.
Is the Fimbulwinter in Völuspá? No, the word does not occur there. Völuspá does know darkened sunshine in summers and the brother-war; Snorri connects the brother-war with the winter.
What happened in 536? A major volcanic eruption in the Northern Hemisphere, followed by another around 539/540, veiled the sun. Tree rings show cooling over decades. Which volcanoes they were is unresolved.
So is the Fimbulwinter historical? The climate crisis is historical. That the myth remembers it is a plausible but unprovable hypothesis.
Ragnarök · After Ragnarök · Líf and Lífþrasir · Vafþrúðnismál · Völuspá · Skoll and Hati · Vendel Period · Vindelev hoard · Rök runestone · God of War fact check
Ragnarök · What comes after Ragnarök · Lif and Lifthrasir – the human pair who survive Ragnarök · Vafþrúðnismál – Odin's Head-Wager with the Wisest Giant · The Vendel Period.
What does the strongest god do when his hammer has been taken from him? He sets out anyway. The journey to Geirröd …
Read more →Ginnungagap is the yawning void that existed before the world in Norse mythology: a chasm between icy Niflheim in the …
Read more →Gleipnir is the third and final fetter the gods used to bind the wolf Fenrir: a ribbon as smooth and soft as silk, made …
Read more →A Danish king buys two giantesses, sets them to a magic mill and has them grind gold and peace without ever letting …
Read more →