A river of molten lava flows across a steaming black basalt landscape. The river flows from a volcanic rift near the top right of the image toward the bottom left. The fissure is filled with brighter and hotter lava and steam and gas billows up from it. The sky on the horizon is a hazy blue-gray.
The Columbia River flood basalt eruptions produced extensive lava flows that lasted for 2 million years, but they likely didn’t impact global climate for nearly that long. Flood basalt eruptions are extended periods of near-continuous but low-intensity volcanism punctuated by large explosive events, somewhat resembling the lava flows from Kīlauea Volcano’s Lower East Rift Zone, above. Credit: USGS Hawaiian Volcano Observatory, Public Domain

Tambora, Krakatoa, Pinatubo, Hunga Tonga–Hunga Ha’apai: Some of the largest volcanic eruptions of the past 2 centuries featured extensive eruption histories that altered ecosystems, released tons of volcanic gas into the atmosphere, and, in some cases, changed global climate.

Now imagine the volume of each of those eruptions multiplied 10,000- fold and drawn out for hundreds or thousands of years. That approaches the scale of the smallest and most recent continental flood basalt eruptions, which created the Columbia River Basalt Group in the U.S. Pacific Northwest.

This map of the U.S. Pacific Northwest shows a swath of basalt formations, the Columbia River Basalt Group, that stretch across Washington, Oregon, Idaho, and northern Nevada. The extend of basalt is highlighted in light yellow on a gray background that has topographic features. Red labels and blue lines mark the Northern Nevada Rift dikes, the Steens dikes, the Monument dike swarm, and the Chief Joseph dike swarm. A cyan dashed line and label marks the cratonic margin that runs vertically near the Idaho-Washington-Oregon border. Magenta lines mark aeromagnetic anomalies that run across the Nevada-Oregon border.
This map shows the extent of the Columbia River basalt group as well as some of its prominent geologic features. Credit: Victor Camp and Martin Ross/Wikimedia Commons , CC BY-SA 3.0

“It’s really more of a volcanic period rather than a single event,” said planetary atmospheric scientist Scott Guzewich of NASA Goddard Space Flight Center in Greenbelt, Md. He described flood basalt eruptions as a cross between Kīlauea and Hunga Tonga–Hunga Ha’apai: an extended period of near-continuous but low-intensity volcanism punctuated by large explosive events. “There were many tens, hundreds, thousands of events, each of which was extremely large by modern standards, that occurred within a larger volcanic complex that persisted for tens of thousands of years.”

The eruptions that created the Columbia River Basalt Group lasted from 17 to 15 million years ago, covered 210,000 square kilometers of land with basalt, and released 300 gigatons of sulfur dioxide (SO2) into the atmosphere—that’s about 20,000 times the amount released by Mount Pinatubo, Philippines, in 1991. Flood basalt eruptions, all of which were larger than the Columbia River eruption, historically have been followed by periods of volcanic winter and then significant warming, which has also led to some of Earth’s most significant extinction events. (The creation of the Columbia River Basalt Group was closely followed by the Miocene Climate Optimum and the middle Miocene extinction.)

The warming period has often been attributed to carbon dioxide (CO2) released during eruptive periods, but the CO2 levels inferred from Columbia River rock records wouldn’t produce the level of warming those same rocks suggest. Guzewich and his colleagues explored that posteruption warming discrepancy by simulating how another common volcanic gas, SO2, would alter global climate if emitted in Columbia River quantities. To their surprise, they found that that amount of SO2 altered the atmospheric structure in a way that led to significant warming, at least for a short while.

Cooling Below, Warming Above

The researchers used a global climate model that let them simulate the coupled response of the stratosphere, troposphere, global atmosphere circulation, ocean, and sea ice to a large-scale volcanic eruption. Because such a model requires so much computing power, they simulated a scaled-down version of the Columbia River eruptions, injecting 30 gigatons of SO2 into the upper troposphere and lower stratosphere over a period of 4 years and then tracking the response for 16 years. That’s about a tenth of the estimated SO2 quantity from the Columbia River eruptions over about a tenth to a hundredth of the time.

“The stratosphere gets about 10,000% more humid than it otherwise would be, and that’s really bad.”

As seen with modern volcanic eruptions, the simulations showed that SO2 quickly evolved into a very thick layer of sulfate aerosols that acted like an umbrella, cooling Earth’s surface by preventing sunlight from reaching the ground. Global temperatures cooled by 2°C–3°C on average but changed seasonally: Northern summer temperatures were about 30°C colder, and northern winter temperatures were up to 15°C warmer.

Guzewich said that the team expected that an aerosol layer that thick would tip Earth’s climate into a global ice age, but even as the sulfate aerosols cooled the surface, they also caused more dramatic changes to the upper atmosphere. Aerosols kept sunlight from reaching the ground by reflecting it back toward space, significantly warming the upper troposphere and lower stratosphere. That region of the atmosphere, known as the tropopause, is typically much colder than the air below and keeps humid surface air from rising into the stratosphere. But that much warming from the sulfate aerosols temporarily eliminated the tropopause and allowed significant quantities of water vapor (H2O) into the stratosphere.

“The stratosphere gets about 10,000% more humid than it otherwise would be, and that’s really bad,” Guzewich said. Warming from that water vapor (a very powerful greenhouse gas) quickly overtook the cooling effect of the aerosols. “It warmed the surface by about 5°C–6°C which, on a global sense, is pretty dramatic. That’s perhaps twice of what we’re on pace to do with anthropogenic climate change, and it’s doing that over a period of a couple years instead of decades or centuries.”

Stratified basalt mountain ridges partially covered with green vegetation are visible from the summit of Steens Mountain. The lower-right diagonal half of the image shows the mountainside ridges and the upper-left diagonal half shows further basalt formations in the distance.
The Steens Mountain summit in eastern Oregon looks down on the oldest basalts associated with the Columbia River flood basalt eruptions. These basalts are an estimated 17 million years old. Credit: Victor Camp/San Diego State University/USGS , Public Domain

“The question of the timing, duration, and magnitude of cooling induced by sulfur aerosols has been an important question when considering the climatic effects of flood basalt eruptions,” explained geochronologist Jennifer Kasbohm of Yale University in New Haven, Conn. The researchers “show that cooling is quite short-lived, as anticipated, but crucially that SO2 emissions lead to changes in stratospheric water vapor that actually counterbalances this cooling and leads to warming in following years.” Kasbohm was not involved in this research.

“This new model also shows the geographic areas that would be most affected by warming and cooling, which may be fruitful for future paleoclimate studies, and the importance of seasonality,” Kasbohm added. Four years after the simulated eruption, North America and Siberia were 15°C–30°C warmer than normal. Eight years after the event, Antarctica was 20°C–25°C warmer. Monthly temperatures reached 49°C (120°F) in the Amazon and east central Australia and exceeded 55°C (131°F) in southwest Asia. The protective stratospheric ozone layer was decimated.

The researchers published these results in Geophysical Research Letters in February and presented this research at the 2022 Lunar and Planetary Science Conference on 9 March.

A Resilient Climate

But even volcanism as extensive and long-lived as the Columbia River eruptions couldn’t alter Earth’s climate for very long. The oppressive aerosol shroud completely dissipated within 4 years of the eruption, with heavy sulfate aerosols particles simply falling out of the sky. By year 16 posteruption, global surface temperatures returned to normal, the stratosphere dried out again, and the ozone layer recovered.

“Even with as massive of a perturbation to the climate as this produces, the climate gets back to normal surprisingly quickly,” Guzewich said. “If you’re thinking about a disaster on Earth in modern times, it shows that the climate is somewhat resilient, that even this large of an effect doesn’t push Earth past a tipping point.” Future simulations will test the effects of adding CO2 to the mix, varying the quantity of SO2 emitted, and injecting the volcanic gases lower into the atmosphere.

“The largest member of the Columbia Ridge Basalt Group emplaced 40,000 cubic kilometers of lava—that’s 40,000 Mount St. Helens eruptions!”

“While direct SO2-led warming has been briefly considered before as a greenhouse gas–type response to explain, for example, warming after the 1783 Laki eruption in some regions, generation of stratospheric H2O from SO2 abundance has not,” said Stephen Self, a volcanologist at the University of California, Berkeley. Self was not involved with this research. “This is a well-considered and somewhat unexpected result.”

Kasbohm, who has conducted field research studying the Columbia River basalt eruptions, said that she was skeptical about how realistic the 4-year duration of the simulated eruption was. “The Wapshilla Ridge Member, as the largest member of the Columbia Ridge Basalt Group, emplaced 40,000 cubic kilometers of lava—that’s 40,000 Mount St. Helens eruptions! While a 4-year duration is technically permitted for the Wapshilla Ridge…that is far shorter than I would anticipate.” She said that a scenario with intermittent eruptions over 14–100 years is more realistic. “While it may be computationally expensive, I would appreciate seeing these eruptive parameters modeled over 10, 100, and 1,000 years, which I think would provide more realistic estimates and better quantify the more likely climatic response.”

—Kimberly M. S. Cartier (@AstroKimCartier), Staff Writer

Citation: Cartier, K. M. S. (2022), The surprising greenhouse gas that caused volcanic summer, Eos, 103, https://doi.org/10.1029/2022EO220156. Published on 21 March 2022.
Text © 2022. AGU. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.