Okmok is a known hazard. The volcano occupies its own half of Umnak, an isolated part of the Aleutian islands. Okmok is possibly the most active of the 40-odd Aleutian volcanoes. Over the past 8600 years it has produced over 50 ash layers from separate explosions, and minor eruptions happen every other decade. AVO has done an extensive hazard analysis. Among the listed dangers are debris avalanches, lahars, pyroclastic flows which could reach up to 30 km away and impact neighbouring Unalaska, and ballistic projectiles thrown 5 km from the caldera. The potential of this volcano is shown by the thick and deeply eroded pyroclastic layer which covers the entire region. But now it seems one hazard was overlooked by the investigators. As all risk assessments, it only looked at what was expected. The unexpected is never included. The big one they missed was Okmok’s potential to topple Europe’s republics. Admittedly, this is not a standard item on risk assessments: it was an easy one to overlook.
A map of Umnak shows Okmok as the northeastern half of the island, connected by an isthmus. The volcano covers some 120 km2. Half of this is taken up by its central caldera, 9 km across. The caldera rim peaks at 1150 m, and the caldera itself is at around 500 meters above sea, albeit bumpy. Okmok is a shield volcano, with gentle slopes. Even though its part of the island is all just Okmok, it is not the highest volcano there: that is Tulik, 1250 meters tall, which is a cone that grew on a flank vent. There are some 30 smaller parasitic cones scattered around the island, including a number inside the caldera. Recent eruptions have come from these caldera cones.
A large eruption in 2008 became quite disruptive. This was not so much to the locals as there was only such person living around Okmok and he was safely evacuated, but the impact was on the wider area. It turned out that Okmok is located directly underneath the main North Pacific flight path. The ash closed the area for flights whilst the pumice did the same for shipping.
The caldera seems badly oversized for the volcano. Mapping of deposits on Umnak has shown that it formed during two main eruptions, the first carbon-dated to 11,500 BP, and the second one to 2050+-50 BP. Okmok-I, as the first eruption is called, consisted of a series of large eruptions which together produced a pyroclastic layer 100 meters thick. One flow reached Unalaska island, 20 km away. The caldera which formed at this time quickly acquired a lake. The volume of the eruption is around 30 km3 DRE, possibly more if a substantial amount ended up in the sea. It may have been a VEI-7.
Several significant eruptions occurred over the centuries before the second main eruption, often explosive because of the presence of water. On at least one of these occasions pyroclastic flows again reached Unalaska.
But the next eruption was much more significant. The Okmok-II eruption left 80 meters of pyroclastic flow deposit near the rim of the caldera, and 30 meters or more on the coast; it buried the Okmok-I ejecta. Unalaska was reached again, and this time it was also hit by a tsunami. The eruption occured inside the caldera at a time it was filled by a lake. The eruption was perhaps half the size of Okmok-I. After this, the volcano remained active with numerous smaller eruptions, similar to its present state. The lake refilled to a depth of 150 meters, but some 300 to 400 years ago the caldera rim gave way and the lake emptied northward in a catastrophic flood. At some point in time another 10-meter thick pyroclastic deposit formed, originating from Tulik, but this is not dated. The 2008 eruption reached VEI-4. Okmok-I may have been VEI-7, and the Okmok-II eruption was VEI-6.
The two Okmok eruptions were large, but not unprecedented. Across the globe, an eruption like Okmok-II may happen every one or two centuries. Okmok-I was more significant, something that may happen perhaps 2 or 3 times per millennium. It will have had worldwide impacts. For the smaller Okmok-II eruption, you would expect some impact on the climate, but the main effects would be near the volcano. And Okmok is very isolated. It should have been a safe eruption.
But something strange happened. Half a world away, an empire had reached a turning point. And Okmok exploded into a crisis.
Rome, 44 BC
The Roman empire is known for its emperors. Augustus and Nero are perhaps the most famous of these, one a builder, one a destroyer, one evil, one a complex character. But this is not how the empire started. It had begun as a republic with a leading clan of aristocrats but no single authoritarian leader. Decisions were made collectively by the senate, and leaders (consuls) were elected for limited periods. The wars with Carthage, with Hannibal as the best known adversary, were fought under this system. But it could be tricky. Collective decision making was slow and could falter in a time of crisis. The senate therefore had the option of electing a dictator, a person to rule unopposed for a finite period. This relied not just on the willingness of the person involved to step down at the end, but also on the fact that the loyalty of the troops was to Rome, not to the dictator. If a dictator would develop political ambitions beyond the crisis, a quiet word by the aristocrats to the army leaders would solve the problem. But as the empire grew, the military became more powerful than the senate and the system cracked. Julius Ceasar was the most brilliant of the military leaders, and he became fed up with the slow and inadequate decision making by the senate. When he crossed the Rubicon he took legions loyal to him into the capital, marching into direct conflict with the senate.
War followed, and the leader of the senate had to flee. Ceasar was elected dictator for the period of the crisis he had initiated himself. But his ambition went beyond that and in early 44 BC Ceasar declared himself Dictator Perpetuus or dictator for life. The senate was not impressed, and on the ides of March (March 15) Ceasar was lured into the senate, attacked, and killed.
The senate had expected the population to support them in this. But in fact, Ceasar was popular with the people: the population rioted and burned down the senate, leaving the perpetrators with little choice but to flee. They went to the east of the empire, Macedonia and Syria, where they assembled their armies. Over the next year, the politics and occasional battles left a triumvirate in charge in Rome, led by general Mark Antony with young Octavian as the second most important. Octavian had been adopted by Julius Ceasar as his son, albeit only in his will – i.e. posthumously; he saw himself as the chosen successor. Fifteen years of civil war followed during which the fled senators were decisively defeated, and many other senators were killed or made destitute.
During the long war Mark Antony became involved with the Egyptian ruler Cleopatra, like Ceasar before him. (Cleopatra was the last of the Ptolemaic rulers of Egypt, the final vestige of Alexander’s empire. Although she is often depicted as traditional Egyptian, the family was from Macedonia). Mark Antony moved his attention to the eastern half of the empire, the ancient heart of civilization, perhaps more cultured than Rome, while Octavian governed the western half of the empire, perhaps less important but including Rome itself. At times the two worked together but both were too ambitious for that to last. Octavian acquired himself a top general, an old school friend called Agrippa. Octavian himself was not gifted in battle strategy and he was physically not strong enough to take part in fighting. Agrippa’s military brilliance and Octavian’s gift at populism turned the tide in their favour. Around 31 BC Antony’s empire collapsed, he and Cleopatra committed suicide, and Octavian’s rule became undisputed.
In 27 BC Octavian became the first emperor, adopting the name of Augustus. The senate continued, but Rome’s time as a republic had ended.
The senate changed with aristocracy giving way to meritocracy for acquiring its members; it now even included wealthy members from the conquered provinces. A considerable minimum income was required to be considered for membership. Augustus eventually became a competent leader who managed the empire well. He build a system of government that lasted. Step by step, he replaced the existing structures by a working -and by and large honest – bureaucracy. Many aspects of our modern state were designed by Augustus. He may have lacked the brilliancy of Ceasar, but by going slow he build something that lasted far longer. But from now on, the empire would always be dependent on the armies. After his death the emperorship was passed on in the family, until all semblance of competence was lost. But even afterwards, the emperor was selected based on popularity with the legions and not on competence in government.
The Roman empire with its newly designed bureaucracy lasted for centuries, but it had become an autocracy. Augustus rarely used his autocratic powers: he governed by consulting often and looking for consent. But his successors made more use of the unlimited power the position of emperor gave, and this was dangerous when an incompetent or unscrupulous leader emerged. Eventually, the empire fell from within.
But this conflict in the Roman world was not only about personal ambitions and perceived destiny with a touch of romance. The natural world also played a role. Reports describing the beginning of the civil strife comment how the sun seemed dim, even dark. When Octavian entered Rome in early May of 44 BC, stars could be seen in the daytime around a dim sun, with a halo looking like “wreaths made from ears of wheat” and rings of changing colour. This was presented as a favourable omen. And in July of that year, during Ceasar’s funeral celebrations, a comet was seen in daylight for seven successive days. Years later, depictions of Julius Ceasar would often include a comet behind him.
The poet Virgil later recalled this time as one of “wars that grow in the dark like cancer.”
Contemporary writers attributed the dim sun to an eruption from Etna which was said to have happened shortly before Ceasar’s assassination, where lava streamed into the sea 18 km away, and hot ash fell 70 km away.
But was it Etna? We have no other record of this eruption, and it is hard to see how even a significant Etna eruption (VEI 3) could lead to a year of darkness. And now it seems that the real culprit was that volcano a world away from the Roman empire: Okmok.
Volcanic eruptions can seem obvious to recognize. The next time Grimsvotn explodes, the news will go around the world in minutes. Pinatubo is a familiar name to us, even though few of us will ever go there. But this volcanic familiarity is recent. The first major volcanic eruption that became world news was Krakatau. The one before, Tambora in 1815, was largely unnoticed, even though it killed more people than any other natural disaster ever did in the British empire. The disastrous year without summer was only linked to its eruption a century later. And the major eruption in 1808 was never recognized, even though it played a role in ending Napoleon’s adventures in Russia. 200 years later, we still have no idea which volcano was responsible for the (perhaps) second largest eruption of the last 300 years. Even for Okmok, the significant 2008 eruption was missed by AVO and was reported by passing planes.
But where our recording is lacking, the natural world takes over. A record of old eruptions is kept on ice. The deep ice cores of Greenland and Antarctica show thin layers of sulphate deposits, each from a different eruption. The second best record is from tree rings, which show when the climate worsens. In a now-famous paper in 2015, Sigl et al combined these two to get accurate dates for large eruptions over the past 2500 years. The plot is shown below. The well-known eruptions are clearly visible: Ilopongo in 540, Eldgja in 939, and Rinjani in 1257. Others are unknown: for instance, the VLE (very large event) around 430 BC is possibly the second largest eruption over the past 2500 years, but we do not know which volcano this was. A hint can be given by comparing Greenland and Antarctica: if only one shows it it was a high-latitude eruption towards that pole, if both show it it was near the tropics. The 430 BC eruption left a larger signal in Antarctica and so may have been at subtropical southern latitudes. That is all we know.
The 44BC eruption clearly shows up, and it is clearly a northern eruption. How large? That depends on how far from Greenland it was. For instance, the Tianchi 945 eruption was large, but shows up poorly probably because the wind blew the ejecta away from Greenland, while Hekla 1108 was smaller but left much more of a signal because of proximity.
The actual volcano can be identified if any tephra is found in the ice core. This is rare: dust grains rarely travel far; sulphate (much smaller particles) are distributed via the stratosphere but tephra can’t float up there and falls back. Rinjani was identified through ice core tephra but that was an exceptional eruption. But by a lucky break, tephra was found in the ice core layer of the 44BC eruption. And it unequivocally, without a doubt, pointed at Okmok. The alkali weight percentage of 4%-6% and the SiO2 percentage of 52%-60% is a close match for the pyroclastic deposits from Okmok-II, and not with any other candidate for this eruption. As a volcanic fingerprint, this is about as good as it gets.
This eruption is also known from tree rings. A large eruption especially affects spring and summer temperatures (it reduces sunlight and this has the most effect when there is most sunlight). This is exactly when trees grow, and thus a bad eruption shows up in the thickness (or wellness) of those rings. An eruption causes a sudden narrowing of the rings, followed by a gradual recovery. And this is indeed seen, and strongly so, in North American trees. This also points at a North American volcano as the cause.
The tree rings in American pine of 43 BC and 42 BC are narrow, and show evidence of frost damage. This indicates an exceptionally cold summer. The frosted tree rings in 42 BC were noted already in 1984. There was a partial recovery in 41 BC and 40 BC. Trees in Scandinavia and Austria confirm that the cold summer weather extended to those regions, with temperatures of 2 to 3 degrees below normal during these two years. Even a bristlecone pine in California’s White Mountains show a frost ring, indicating frost in early September 43 BC.
The description of the events in Rome in 44 BC seem to fit well with the Okmok-II eruption, with reports of the dimming of the sun which is reminiscent of stratospheric sulphate. It is a convincing story. But there is one slight problem. The evidence of the tree rings indicates that the Okmok eruption happened in early 43 BC. The growth ring of 44BC is normal (even wider than normal in the image shown above) and only in 43 BC did conditions deteriorate. But in that case, how could the sun dim already in 44 BC? It has been suggested there were two Okmok eruptions, with an initial one in 45 BC. But this seems exceedingly unlikely. It would have to be another VEI 6 and these don’t come that often. Was it Etna after all? Or is the situation more complex?
One of the Greenland ice cores does show a sulphate peak 2.5 years before Okmok, fairly strong but short-lived. It was only seen in Northern Greenland. This suggests it originated close to Greenland, and a link to Iceland appears plausible while a link to Etna is unlikely. But if the sulphate was limited to just one part of Greenland, it is perhaps unlikely that it reached Rome, and very unlikely that it had widespread repercussions across the Roman empire.
There is something funny about the descriptions of the events in 44 BC. The stars were reported to have been visible in daylight. That cannot be true, and this nice story must be fake news. Anything that dims the sun will also dim the stars. Worse, some of the intercepted light from the sun is scattered in all directions and makes the sky milky white. Seeing stars in daytime under such conditions is impossible. Other documents refer to a bright-coloured halo around the sun. That sounds more like a corona: an edge-coloured disk caused by water droplets in the air. This is a weather-related phenomenon, and does not require a volcano.
There is also something strange about the comet, which in July was seen in daytime for seven days. To be visible in the daytime it must have been among the brightest comets of the past 2500 years. But no one else saw it. Chinese reports indicate a comet had been visible in May. In July that comet would already have been on the way out, and much fainter. It has been suggested that the comet suffered an outburst. But this is rare, and it would have had to be an extreme outburst to make the comet so bright. Strangely, the images showing Ceasar with a symbolic comet were from many years later. Earlier images instead show a star. It seems likely that the comet was a story that came later. It never happened. One of the most famous comets of antiquity was an invention.
Could the same be true for the dimness of the sun? Was this something that did not happen in 44 BC, but one or two years later? Octavian was well aware of the power of symbolism. Did he change the story to strengthen the symbolism, and thereby his claim to power? Perhaps it was a bit of both, with some aftereffects of the Etna eruption of February(?) 44 BC playing a role, with added embellishment or even invention. It is an interesting complexity that we have three eruptions around this time, from Etna, a possible volcano in Iceland, and the big one from Okmok, and we have reports of veiled skies and halos in Italy, but that it is very hard to link these. The story does not quite add up.
Could our timeline for the eruption be off? The ice core dates are indeed uncertain by 1 or 2 years. A Chinese record of cold temperatures, which has been argued to be related to Okmok, is uncertain by 5 years and should probably be ignored. The precise dates for the eruption are actually obtained from the tree rings, and the ice cores are shifted to bring them into agreement with the tree rings. The tree rings are counted back from the 1257 Rinjani eruption. An entire ring may be missing after an exceptionally year: for instance the oak chronology in Europe misses a year in 1816, the ‘year without summer’. Several independent tree chronologies are therefore used to cross-match, and the result of this is believed to be accurate to within 1 year, going back as far as 12,000 BC. There seems little room to move Okmok-II to the preceding year.
But strangely, although we have descriptions of a dimmed sun from 44 BC, there are none for the following year when Okmok actually erupted. Plutarch writes about the year of Ceasar’s death that “the brightness of the sun was darkened, the which all that year through rose very pale, and shined not out, whereby it gave but small heat: therefore the air being very cloudy and dark, by the weakness of the heat that could not come forth, did cause the earth to bring forth but raw and unripe fruit, which rotted before it could ripe“. But the tree rings clearly indicate that the significant impact on the climate of this eruption occured a year later. It is not unlikely that in the written documents, events that happened in 43 BC where transferred to the previous year, embellishing the story about the fall of Ceasar and the rise of Octavian.
There are several allusions to poor weather and harvests around this time. The clearest of these is Plutarch, mentioned above. One of the murders of Ceasar, Marcus Brutus (made famous by Shakespeare with the speech of the dying Ceasar ‘et tu, Brutus’), fled to the east where he is said to have marched through snow storms. During the subsequent war, Mark Antony and Octavian suffered from cold weather in Philippi in 42 BC. Other descriptions agree that the winter of November 44 to February 43 BC was a cold one. However, this area of the Balkans has much much colder and snowier winters than Rome, and the complainants may just have been unfamiliar with the local conditions.
Shortage of food became a growing problem. The opposing armies had trouble getting food after the battle of April 43 BC in northern Italy, and again around the battle of Philippi in October 42 BC when both armies tried to buy food from Egypt but were rebuffed, because of shortages there.
Rome itself also suffered from a lack of food after 43 BC. A naval blockade by Sextus Pompei was hindering food imports into Italy between 43 BC and 37 BC. This became a major problem. Rome was not self-sufficient in food, and depended on food imports by sea. It required around 150,000 tons of wheat per year. Many citizens received free or subsidized supplies of grain. Octavian found the population in near revolt. The blockade was resolved only when Sextus Pompei was invited to join the ruling triumvirate.
The real reason for the shortage in Rome may have been the fall in production in Egypt where a famine developed in 43/42BC, rather than the blockade. Egypt’s food production is closely linked to the Nile summer floods, and these we know can fail after major volcanic eruptions. One should not overinterpret this: there was also famine in Egypt in 48 BC – that was actually a worse year for lack of flooding than after Okmok. The Nile is not the most reliable of rivers. Egypt used major grain storage facilities to guard against the years when the floods would not come. These storages were used in both years to feed the population.
Seneca wrote : It is well established that in the reign of Cleopatra the Nile did not flood for two successive years, the tenth and eleventh of her reign. They say that this was a sign of the loss of power for the two rulers of the world, for the empire of Antony and Cleopatra did fall. (The tenth and the eleventh years of Cleopatra’s reign correspond to 43-41 BCE. Her reign began sometime between March 52 BC and March 51 BC.)
Later, the priest Kallimachus wrote: Kallimachos, the [..] revenue officer for the district of Thebes, and gymnasiarch and cavalry-commander, previously having taken over the city, which had been ruined [as a result of] manifold [disastrous] circumstances, tended it carefully [and maintained it] unburdened [in] complete peace [..] the severe famine caused by a crop failure like none hitherto recorded, and when the city had been almost crushed by [need] …he voluntarily contributed to the salvation of each of the local inhabitants… …The famine, however, continued in the present year and became even worse [..] a failure of the flood and misery far worse than ever before reigning throughout the whole [land] and the condition of the city being wholly critical….
(Taken from https://www.nature.com/articles/s41467-017-00957-y
Thebes was the old capital, located where nowadays Luxor is found. By this time it was in decline.
Let’s put some more detail on this. The Nile floods are fed by the monsoon rains in the Ethiopian highlands. The floods reach Egypt in June, but the major event starts mid-July (or used to – the Nile floods no longer). This irrigates the crops, and the harvest happens at the end of the floods around October. The Nile floods are known to be negatively affected by northern eruptions: they failed after Eldgjá (939), Laki (1783–1784) and Katmai (1912). The Okmok eruption in early 43 BC would have impacted food production after the summer of that year. But the first reports of shortages are from April of that year. It seems there was a problem already – or perhaps Cleopatra did not at that time wish to take sides in the Roman civil war.
The comment on the Nile flood failing in the tenth and eleventh year of Cleopatra’s reign is a clear sign of when the impacts happened. It puts the first failure in 43 BC, and this favours an Okmok eruption in early 43 BC. Even though uncertainty remains, events in 44 BC or even early 43 BC cannot easily be explained by the Aleutian eruption.
What can we conclude? In the battle between the Roman republic and the distant volcano, did Okmok side with the military usurpers? Did Okmok turn Rome into a classical version of Ragnarok? (Rokmorok?) Here, we have to declare a draw. The beginning of the war, from the assassination of Ceasar to the battles in Northern Italy of early 43 BC, were free of volcanic influence. Reports of the dim sun, and the optical effects in the sky were exaggerated or conveniently re-dated, for reasons of propaganda.
The real wars did not begin until 42 BC when the main battle was fought in October 42 BC at Philippi. Was this fought under the cloud of Okmok? If so, the impact was not so much from the weather but from the lack of food. After a VEI-6 in early 43 BC, by October 42 BC the skies should have largely cleared, unless the eruption was particularly sulphur-rich.
The next crisis came in 41 BC when Rome itself was near revolt from lack of food. This was probably related to Okmok, as an aftereffect. Octavian managed to use this crisis to strengthen his position.
And perhaps there was another fall-out from the food crisis. For why did Cleopatra attach herself to Mark Antony, long before the outcome of the struggles became clear? One can guess that Egypt was in a bad way, as shown by the report from Kallimachus, with two major famines in quick succession. She urgently needed an ally, and Mark Antony was available.
So did the Republic fall because of Okmok? Clearly not. The Republic lost its main battles well before the Okmok eruption had its worst impact. The eruption brought chaos to the next few years. Octavian used this chaos best, and he came out the winner. This was how the Roman Empire was born. Okmok played a role. But it was neither the cause, nor determined the outcome.
Republics can survive one disaster. The VEI-7 eruptions of Tambora and Rinjani had a massive impact but they left no lasting world-wide damage. The double eruption of 536/540 AD did major damage to the later Roman Empire, because of the epidemic that came in its wake. Okmok was just a single eruption – there was no double blow and Rome may have had a hard time but it managed to complete a transition that was already inevitable. But not everyone was so lucky. The major casualty of Okmok was not Rome, but Egypt, already weakened and in decline. The volcano’s final victim was perhaps Egypt’s last queen: Cleopatra.
Albert, July 2020
Preliminary volcano-hazard assessment for Okmok Volcano, Umnak Island, Alaska
Extreme climate after massive eruption of Alaska’s Okmok volcano in 43 BCE and effects on the late Roman Republic and Ptolemaic Kingdom. McConnell et al, 2020, https://doi.org/10.1073/pnas.2002722117 (access blocked)
Nile summer flooding triggers revolt and constrains interstate conflict in ancient Egypt. Manning et al, Nature Communications 8, 900 (2017). https://doi.org/10.1038/s41467-017-00957-y (open access)
Augustus: The Life of Rome’s First Emperor. Anthony Everitt https://erenow.net/biographies/augustusthelifeofromefirstemperor/