An Alaskan volcano, climate change, and the history of ancient Rome

Several factors influence the variability of Earth’s climate. The Sun’s activity rises and falls with its 11-year sunspot cycle. Earth’s tilt, spin, and orbit likewise oscillate through so-called Milankovitch cycles, which influence the amount and distribution of solar energy that reaches Earth’s surface (see the article by Mark Maslin, Physics Today, May 2020, page 48). The Milankovitch cycles induce ice ages, which take thousands of years to develop—hardly a rapid response. By contrast, volcanic eruptions are sudden, dramatic events that can shock a society for years.

During a large eruption, billions of tons of rock, gas, and ash are thrust into the atmosphere. Small, insoluble particles of tephra rain out within days to weeks. But the gas, much of which is sulfur dioxide, can reach the stratosphere, where it may linger for years. In the thin dry air, SO₂ slowly becomes oxidized to sulfate aerosols. Depending on the eruption’s location, those sulfates can get caught up in stratospheric winds, which disperse them throughout one or both of Earth’s hemispheres. Extremely reflective, the sulfates scatter incoming solar radiation and cool Earth’s surface. But eventually they, too, fall out, some settling on snow near one of the poles.

As the snow becomes compressed into ice over the millennia, gases and molecular compounds become trapped, layer by layer, and form a rich, time-stamped record of the atmosphere’s chemical composition. Industrial pollution, dust, pollen, ash from biomass burning, and volcanic residue can all be found there. Since the 1960s, researchers have drilled cylindrical ice cores in glaciers and ice sheets to mine that archive.

Early on, the ice cores were cut into short slices, and individual layers were melted and analyzed one by one as a function of ice depth. But the technology has matured; today, researchers can continuously melt and analyze ice cores simultaneously for much more accurate dating. “Each year is much more clearly resolved,” McConnell says. “And because nearly everything we measure in the ice shows seasonal variation, we can use multiple parameters to identify those annual layers.” His system, shown in figure 2, holds an ice core over a heated ceramic plate. Grooves in the plate channel meltwater from the innermost part of the core to a mass spectrometer to measure the concentration of dozens of elements.

McConnell and his collaborators did not set out to look for the source of sudden climate change in the ancient Mediterranean region. Two years ago, they were analyzing lead pollution in ice cores for a project on the history of silver production. Silver mining and smelting operations waxed and waned in the ancient world. (About 10 000 g of argentiferous lead ore was used to produce 1 g of silver.) McConnell’s team used the rise and fall of lead’s concentration in the cores as a marker of ancient economic activity, war, and plague.³ 

While reviewing data for that project, McConnell and his former postdoc Michael Sigl (now at the University of Bern) came across an unusually well-preserved layer of tephra in one of the ice cores used in the pollution study. Excited about finding possibly unknown volcanic eruptions, they decided to investigate it further. Their subsequent mass spectrometry measurements of ice cores (figure 3) revealed volcanic fallout from two distinct eruptions—the first starting in early 45 BCE and the second in early 43 BCE, almost a year after Julius Caesar’s assassination. The first eruption probably took place in Iceland, the collaboration surmised. But it was short-lived and likely had little climate impact. The second, attributable to the Okmok volcano, was a whopper: It had a greater fallout of sulfuric acid, which persisted for 2.5 years.

The sharp spike in 3–10 µm insoluble tephra shards that McConnell and Sigl had spotted early on coincided with the early stages of the 43 BCE sulfur peak. The temporal narrowness of the spike pinned the eruption to January or February of that year, before the growing season in the Northern Hemisphere. Coauthor Gill Plunkett of Queen’s University Belfast conducted electron microprobe analyses on 35 tephra shards filtered from the melted ice cores to confirm the volcano’s identity.

Earth has about 1500 potentially active volcanos, not counting the belt of spreading centers on the ocean floor. But their magmas have geochemical fingerprints that reflect each one’s distinct location in Earth’s crust. The geochemistry of the ice-laden shards and Okmok’s contemporary basalt matched well enough for Plunkett to rule out several other possible volcanos.

The 2.5-year spread of the volcanic fallout coincides with air-temperature anomalies recorded in tree rings and cave stalagmites. Those climate proxies confirmed that the eruption’s effects were pervasive in the Northern Hemisphere. The ring widths of temperature-sensitive trees in Scandinavia and Austria each revealed summertime cooling of more than 2 °C in 43 and 42 BCE. Radiocarbon dating of stalagmite layers in China’s Shihua Cave showed a similarly pronounced temperature drop. And a rare frost ring found in bristlecone pine trees from the mountains of North America confirmed below-freezing temperatures there in early September of 43 BCE.

The researchers’ simulations suggest that in some regions those cooling trends persisted into the early 30s BCE. Indeed, the simulated temperature dropped by as much as 7 °C in parts of southern Europe and northern Africa. Rainfall patterns shifted as well, with summer precipitation levels some 50–120% above normal throughout southern Europe and autumn precipitation in some regions reaching 400% above normal.

The strong temperature gradient between the Northern and Southern Hemispheres could have shifted the intertropical convergence zone, the region where tropical trade winds meet. That shift, in turn, would have moved the East African monsoon rainfalls southward. The headwaters of the Blue Nile in the Ethiopian Highlands are the source of more than 85% of the summer floods that year after year delivered irrigation and silt to the lower reaches of the Nile in Egypt. With those highlands drier than normal, according to the simulations, little flooding likely occurred in 43 or 42 BCE.

Greek historian Appian recorded the lack of flooding; he cited Cleopatra’s reluctance to provide help to Rome due to Egypt’s famine and pestilence. Although establishing direct causal linkages between crop failures and political decisions is difficult, McConnell and collaborators argue that Okmok’s severe environmental impact likely contributed to social unrest. Rome’s interest in Egypt as a breadbasket was magnified, no doubt, by the stress on a largely agrarian society, and Egypt’s ability to withstand Rome was diminished by famine.² Egypt was absorbed into the Roman empire after Cleopatra’s suicide in 30 BCE.

No one knows whether the birth of the Roman Empire would have happened without Okmok’s influence. But the extreme climate the volcano produced could have hastened it.