Climate Assumption Busted: Mega-Volcano, Falling CO₂

A recent piece in Eos highlights a result that will likely make climate modelers a bit uncomfortable: during one of Earth’s more dramatic volcanic episodes, atmospheric CO₂ didn’t spike—it fell. And not by a trivial amount, either.

We’re talking about roughly a 50% drop in atmospheric CO₂ during the emplacement of the Emeishan large igneous province (LIP) about 260 million years ago.

That’s not a rounding error. That’s a regime shift.

The standard storyline goes something like this: large-scale volcanism releases vast quantities of CO₂, which warms the planet, disrupts ecosystems, and potentially contributes to mass extinction events.

The article itself acknowledges this prevailing view:

“Volcanism is commonly associated with an uptick in atmospheric carbon dioxide (CO₂) levels…”

And for large igneous provinces—events capable of covering continental-scale areas in lava—this assumption has been treated as almost axiomatic.

As one cited researcher put it:

“We’re talking about a scale of magmatic activity that dwarfs things like even a Yellowstone supereruption.”

So naturally, the expectation is more volcanism → more CO₂ → more warming.

Except that’s not what the data show here.

Using carbon isotope proxies from marine sediments (specifically fossilized algae and cyanobacteria), the researchers reconstructed atmospheric CO₂ levels across the relevant interval.

What they found:

• CO₂ hovered around ~700 ppm for ~7 million years
• Then dropped to ~350 ppm over ~3 million years
• Bottomed out for roughly 500,000 years
• This decline coincided with the early and main phases of volcanism

In other words, CO₂ didn’t rise ahead of the eruption—it collapsed.

The authors themselves express surprise:

“It was a big surprise to find that decrease in CO₂… We’re trying to make sense of that.”

That’s a refreshingly honest admission.

The explanation offered centers on crustal uplift preceding and accompanying volcanism.

As magma pushes upward, it deforms the crust into a dome-like structure. In this case:

• Uplift estimated at ~1,000 meters
• Affected area roughly 800 km across
• Result: enhanced exposure of rock surfaces

And here’s the key part—more exposed rock means more chemical weathering, which acts as a CO₂ sink.

“Any landform protruding above its neighbors will tend to be preferentially eroded and weathered… processes that can pull CO₂ out of Earth’s atmosphere.”

So before the volcanic system really gets going in terms of gas emissions, it may first draw down CO₂ via intensified weathering.

That flips the usual cause-and-effect assumption on its head.

The article includes a notable line:

“The conventional picture… is therefore clearly simplistic.”

Because if large igneous provinces—arguably among the most extreme natural CO₂ sources—can coincide with substantial CO₂ declines, then the climate system is clearly governed by competing processes, not a single dominant driver.

Weathering, uplift, ocean chemistry, biological uptake—all interacting across different timescales.

To their credit, other researchers quoted in the piece raise important caveats.

Morgan Schaller notes:

“Maybe you just don’t see spikes because you don’t have high enough fidelity to see them.”

That’s a legitimate concern. Proxy records averaged over long intervals can smooth out short-lived CO₂ spikes.

And Paul Wignall adds:

“If you start getting this result elsewhere, you’d have to start paying attention.”

In other words, replication matters. One case study doesn’t rewrite Earth system science—but it does raise questions.

What stands out here isn’t just the CO₂ drop—it’s the timing and mechanism.

Before volcanism acts as a source, tectonic processes may act as a sink amplifier.

That introduces a dynamic sequence:

• Uplift → increased weathering → CO₂ drawdown
• Volcanism → CO₂ release (possibly later or episodic)
• Net effect depends on timing, magnitude, and feedbacks

That’s a system with lags, thresholds, and competing fluxes—not a simple input-output machine.

It’s tempting to file this under “interesting deep-time geology,” but the implications are broader.

Modern climate discussions often rely on simplified analogies:

• Volcanoes = CO₂ source
• CO₂ = temperature driver

But this study underscores that geological processes can reverse expected signals, CO₂ levels can decline during periods of intense tectonic activity, and feedbacks like weathering can be large enough to dominate.

The Eos article presents the findings cautiously, but the takeaway is clear enough. When even large-scale volcanism does not produce the expected CO₂ response, it highlights just how much complexity is built into the climate system.

As always, more data—and especially replication across other large igneous provinces—will be essential. But for now, this result serves as a reminder that Earth’s climate system does not always follow the simplified relationships often assumed in broader discussions.