The Volcano That Dimmed the Sun

OlaWelcome to Fault Lines, I'm Ola.

AmaraAnd I'm Amara. Okay, so get this: today we are going thirty five years back.

OlaJune fifteenth nineteen ninety one, Central Luzon, Philippines, a mountain that nobody thought was a threat just detonated.

AmaraAn ash column punching forty kilometers into the stratosphere, seventeen megatons of SO2 into the sky! hi!

OlaAnd the Philippines just declared June fifteenth a special non working holiday in Pampanga this year. Manila Bulletin is reporting that PHIVOLCS unveiled a commemorative marker at Clark Freeport earlier this week.

AmaraThirty five years, and the science from the single eruption is still actively contested.

OlaThat's the thing nobody talks about; everyone repeats the canonical zero point five degrees Celsius. ... of global cooling, but a twenty twenty four reassessment published in ScienceDirect raises real questions about whether that number is right.

AmaraSo the baseline the entire geoengineering research community leans on might be off?

OlaWe'll dig into what that actually means, but we're not stopping at nineteen ninety one.

AmaraOh no; we're going back seventy four thousand years, Toba, the super volcano that some researchers Argued nearly

OlaWow

Amaraended us.

Ola!" Argued the archaeology is a lot messier than the headline.

AmaraAnd further still the Siberian Traps, the Deccan Traps, mass extinctions where the error bars are so wide. You could drive a geological era through them.

OlaFollow the evidence, not the rhetoric; that's what we're here for.

AmaraRight, Okay, so June fifteenth nineteen ninety one, Thirty-five years ago, Let's set the scene.

OlaLet's go. Picture June fifteenth nineteen ninety one, Central Luzon, Philippines. It's mid afternoon and the sky doesn't go dark the way you'd expect a storm to hurl in. It goes white-ash, not clouds. Typhoon Yunya is hammering the coast at the same time, so you've got this surreal mix of ash laden rain, pumice falling like gravel, lightning and the ground shaking every ten minutes.

AmaraThat combination is almost hard to process; a typhoon and the second largest volcanic eruption of the twentieth century, same afternoon. Same afternoon, and the USGS fact sheet on this puts it plainly: "More than five cubic kilometers of material ejected, ash column up to thirty five kilometers into the air; the summit collapsed into a caldera two point five kilometers across.

OlaThus so Pinatubo basically ate itself.

AmaraSo what was the warning sequence, because that's the part that doesn't get enough attention. April: earthquake swarms start. June seventh: the first magma reaches the surface. June twelfth: an ash column hits twenty kilometers. Three days later: the climactic eruption.

OlaSix weeks of escalation, and PHIVOLCS and USGS Yes; were watching every step. The USGS fact sheet credits those warnings with saving at least five thousand lives.

AmaraAt least five thousand, though over eight hundred forty seven people still died and more than two hundred thousand families were displaced. Indigenous Aeta communities lost everything; towns in Pampanga, Zambales and Tarlac were buried, not just by the eruption itself,

Speaker 3but by the eruption of the entire mountain.

AmaraBut by lahars-the volcanic mud flows that kept coming for years after.

OlaThe lahars almost did more damage than the eruption.

AmaraThe USGS noted pyroclastic deposits filling valleys two hundred meters deep; rain reactivates those deposits every monsoon season; towns stayed buried for years.

OlaSo thirty-five years on, this week, what's happening in the Philippines?

AmaraPHIVOLCS unveiled a commemorative marker on June ninth at the former Clark Air Base parade grounds right there in Pampanga. According to the Manila Bulletin, scientists from the USGS Volcano Disaster Assistance Program were there alongside Filipino officials. The Philippine government declared June fifteenth twenty twenty-six a special non-working holiday in Pampanga.

OlaA marker at Clark makes sense; that's where the monitoring teams were when the eruption happened.

AmaraExactly; and the Manila Bulletin piece notes the marker honors not just the scientists but community members and media who got the warnings out.

OlaFollow the incentives on that, right? The reason five thousand people lived is because scientists had credibility with local officials, and those officials acted on it.

AmaraWhich brings you to the part of the story that reaches well beyond

Speaker 3your own experience.

AmaraWell beyond Luzon, the eruption column punched forty kilometers into the stratosphere, Ola; the ash fell as far as the Indian Ocean.

Speaker 4Forty kilometers—that's not just a Philippine disaster, at that point.

AmaraNow, that's a planet scale injection event and that raises the question I keep coming back to: If we can trace what one eruption did to global temperatures, exactly how well do we actually understand the mechanism well enough to replicate it.

Speaker 4So the question we ended on: Does that ash column actually do the climate work? and the answer is: no, almost none of it.

AmaraRight. And this is where the chemistry gets wild. The ash, enormous as it looks, falls out of the atmosphere within days. It's heavy. Gravity wins.

Speaker 4So what does the real damage?

AmaraSo two, sulfur dioxide. Pinatubo injected roughly seventeen million tons

Speaker 3of sulfur dioxide.

AmaraTons of it above the troposphere into the stratosphere-that's where things get interesting.

Speaker 4Because, once it's up there, it doesn't just sit.

AmaraNo, it oxidizes. SO2 reacts with hydroxyl radicals and water vapor, and you get tiny sulfuric acid droplets, sulfate aerosols. And those are phenomenally good at scattering incoming sunlight back into space.

Speaker 4How fast does that spread?

AmaraTwenty-two days-the aerosol cloud circled the entire globe in twenty-two days; within a year it had nearly

Speaker 3doubled.

OlaNear total global coverage!

AmaraTwenty-two days, one volcano in the Philippines, and within three weeks you've got a planetary sunshade.

OlaThat's the thing nobody really grasps until you sit with that number, and the effect was measurable-the aerosol layer cut surface sunlight by roughly ten percent. Peak cooling in the Northern Hemisphere hit around zero point five to zero point six degrees Celsius. About eighteen months after the eruption-so December, nineteen ninety two.

AmaraHmm; and the altitude of the ejection is the key variable here, right? Equatorial eruption high enough to punch above the troposphere?

OlaExactly! Below the troposphere the atmosphere scrubs everything out through rain, weather, normal circulation. Above it you're in a different regime-no weather, much slower mixing. The aerosols just persist.

AmaraWhich is why Pinatubo's latitude mattered: near the equator the stratospheric winds carry material across both hemispheres.

OlaA high latitude eruption stays mostly in one hemisphere; an equatorial one punches globally. Location and altitude together-that's the combination that makes a climate event.

AmaraOkay, so now I want to flag something because I'd be doing listeners a disservice

Speaker 3if I didn't give you a little bit of background.

AmaraThe service, if I didn't!--A twenty twenty four paper in the Journal of Atmospheric and Solar-Terrestrial Physics took another look at that zero point five degree figure.

OlaWhat did they find?

Speaker 4They argue that natural climate variability has been improperly folded into the Pinatubo signal in some earlier studies. The true forced cooling might be smaller than the canonical number.

OlaWhich is uncomfortable because Pinatubo is our best monitored eruption ever. ever: satellites, ground stations, we had everything in nineteen ninety one; if we can't cleanly isolate the signal there...

Speaker 4Follow the incentives, not the rhetoric, right? If the best case study is still contested, what does that mean for reading eruptions that happened before instrumentation existed?

OlaOr before humans existed at all, because here's where scale becomes a real problem: Pinatubo put seventeen million tons of SO2 up there and cooled the planet by half a degree. Now imagine an eruption that dwarfs that by a factor of hundreds.

Speaker 4Five hundred times the volume of ejecta!

OlaAnd that eruption happened seventy four thousand years ago (Sumatra) and the question of what it actually did to the planet is genuinely unsettled.

Speaker 4That uncertainty is the whole next thread, because the story scientists told about Toba for decades turns out to be a lot shakier than anyone expected.

OlaSo flip that to the extreme: Toba, Sumatra, roughly 74,000 years ago, about 2,800 cubic kilometers of material ejected. According to Wikipedia and multiple research sources, that's more than 500 times the volume of Pinatubo.

Speaker 5Five hundred times.

Ola500 times. And the original hypothesis, the Toba catastrophe theory, said that triggered a volcanic winter lasting up to six years. Global temperatures dropping three to three point five degrees Celsius. Human population possibly squeezed down to as few as ten thousand individuals.

Speaker 4Yeah, I know the theory; and here's the thing nobody talks about: it's crumbling.

OlaWait. Crumbling?

Speaker 4Seriously, dig into the archaeology from India, sites in the Jurreru Valley, the Middle Son Valley. According to ScienceDirect research published in 2011, you find Middle Paleolithic stone tool assemblages below the Toba ash layer and above it. Same tool kit, same techniques. The volcanic winter was supposed to have wiped the slate clean. The tools say otherwise.

OlaOkay, but a few sites surviving isn't the same as saying nothing happened. Pinatubo devastated local populations even while the global signal was modest.

AmaraFair, but this isn't just one site, and the genetics don't cooperate either. The bottleneck signal in human DNA doesn't line up with seventy-four thousand years ago. According to Wikipedia's summary of the Toba catastrophe theory, non-African populations hit their lowest effective population size around 40,000 to 60,000 years ago. That's 10,000 to 30,000 years after Toba.

OlaSo the timing's off.

AmaraThe timing is off, and then there's the SO2 problem, which is the real kicker. The early Toba models just multiplied Pinatubo's SO2 injection and scaled it up with volume. Turns out SO2 doesn't scale linearly with erupted rock. Research published by Robock and colleagues and cited across multiple climate modelling studies ran simulations injecting up to nine hundred times the Pinatubo SO2 value. None of those runs initiated glaciation.

OlaNone of them.

AmaraNone.

Speaker 5None." And petrological analysis of the Toba tuff now suggests the actual SO2 injection was far lower than those models assumed, perhaps closer to a Tambora scale event. Early simulations overestimated it by one to two orders of magnitude.

Speaker 6Okay, so I love a good catastrophe story as much as anyone, but I hear you: the physics doesn't cooperate. Bigger eruption doesn't automatically mean proportionally more SO2 in the stratosphere. Aerosols get less efficient at bigger loadings. They settle faster; the radiative cooling actually plateaus.

Speaker 4Exactly, and that's the fundamental problem with using Pinatubo as a yardstick. Pinatubo we measured in real time; satellites, weather stations, ice cores confirming within years. Toba is reconstruction all the way down, seventy-four thousand years of geological noise between us and the signal.

Speaker 6So what do we actually know?

Speaker 4We know it was enormous; we know there's a sulfate spike in Greenland ice cores; we know there was probably meaningful cooling. But the six year winter, the human near extinction, that's a story that ran way ahead of the evidence.

Speaker 6Which sets up an uncomfortable question: if even Toba, a single explosive eruption just vastly larger than Pinatubo, is this hard to read, what happens when the mechanism is completely different? Not a pulse, but a drip lasting two million years.

Speaker 4The Siberian Traps

Speaker 6The Siberian Traps. And that's a whole other level of uncertainty. We're not talking aerosols and sunlight anymore.

Speaker 4So the Siberian Traps, two million years of eruptions, five million square kilometres of flood basalt, two hundred and fifty one point nine million years ago. The end Permian extinction that followed killed somewhere between ninety and ninety six percent of all marine species and seventy percent of land vertebrates. And here's the thing: it almost

OlaMost certainly didn't work like a scaled up Pinatubo.

AmaraRight; and that's the part I find genuinely wild, because the instinct is to say, "OK, more SO2, bigger cooling effect." But the Siberian Traps weren't one eruption: they were a drip sustained outgassing over geologic time.

OlaExactly; the sulfate aerosols from each pulse would have rained out in a year or two. Short volcanic winters, yes. But then CO2 keeps accumulating because the atmosphere doesn't flush CO2 on that time scale. You get this brutal oscillation of volcanic winter during each peak, then severe warming in the intervals between them.

AmaraSo ecosystems couldn't stabilize either way.

OlaThat's the horror of it. And here's my methodological problem: with Pinatubo we had satellites, weather stations. Direct aerosol measurements, with the Siberian Traps, were reconstructing all of this from isotope ratios and rock chemistry, with error bars in the tens of thousands of years.

AmaraBut, O Allah! hold on. The correlation between large igneous provinces and the five big mass extinctions is too consistent to explain away with measurement error. Siberian Traps and the end Permian, Deccan Traps at sixty six million years ago and the

Speaker 3K-Pg.

AmaraAnd the Cretaceous-Paleogene boundary? That pattern keeps showing up.

OlaI'm not dismissing the correlation, I'm asking how much causal weight we can actually assign. Take the Deccan debate.

AmaraOh, this is where it gets good.

OlaYeah. So the Deccan Traps erupted in pulse phases straddling the K-Pg boundary, with the highest sulphur lavas concentrated in about zero point four million years before the boundary. Research published in Science Advances found those pulses could have caused repeated temperature drops of up to around ten degrees Celsius, recurring volcanic winters stressing ecosystems. Before Chicxulub even arrived.

AmaraA Stressed Pulse model: the volcano wore down ecosystems, the asteroid finished them.

OlaMaybe, but a twenty twenty PNAS study concluded Deccan alone couldn't generate the extinction signal; in some scenarios the volcanism actually buffered the impact by warming things back up between pulses.

AmaraI'm on both sides a little bit on that one.

OlaHonestly, me too, and that ambiguity is the It's the point. The further back you go the harder it is to separate cause from noise; with the Siberian Traps our error bars are tens of thousands of years wide. That's not a rounding error, that's entire ecological collapses fitting inside the margin.

AmaraSo, the Pinatubo data which we spent all this time on-the two year aerosol signal we measured in real time-that's almost uniquely legible compared to anything in the

Speaker 3deep past.

AmaraIn the deep record-

OlaWhich is what makes what comes next so uncomfortable, because some researchers want to use Pinatubo as the calibration point for deliberately replicating the effect-intentionally.

AmaraStratospheric Aerosol Injection

OlaAnd if our Pinatubo baseline is off, that whole calculation shifts. So here's the actual question geo engineers are wrestling with: Pinatubo cooled the planet roughly half a degree for about three years. SAI proposals want to replicate that continuously, indefinitely, with planes or balloons pumping SO2 into the stratosphere.

AmaraAnd that word 'indefinitely' is doing a lot of heavy lifting.

OlaRight, because Pinatubo just stopped. The aerosol layer faded naturally; SAI can't stop. If you terminate it abruptly, you get what researchers call termination shock.

AmaraWhich is basically all the warming you suppressed

Speaker 4Yeah.

Amararushing back over maybe a decade instead of a century.

OlaScienceDirect modelling shows a temperature reduction of one to one and a half degrees Celsius would rebound over just five to ten years after After termination the rate of warming faster than climate change alone.

AmaraSo you'd be committing every future generation to keep running the programme for ever; no exits.

OlaNo clean exits; and here's where the twenty twenty four ScienceDirect reassessment becomes genuinely uncomfortable. If the real Pinatubo cooling was smaller than the canonical figure, the SAI models calibrated against it. it are over confident.

AmaraEngineering to a number that's wrong!

OlaExactly; follow the incentives here: the same natural experiment that makes Pinatubo so valuable is also the only real-world data we've got; there's no second Pinatubo to check our

Speaker 5work.

Olawork against.

AmaraSo the planet ran the experiment once thirty five years ago and scientists in the Philippines are still marking the anniversary. three), and we're trying to use that single data point to engineer a planetary thermostat.

OlaThat's the problem nobody fully solves-one eruption, one measurement window, and the stakes couldn't be higher. Okay, so Mount Pinatubo, thirty five years later, and it still has things to teach us.

AmaraRight; and I keep coming back to Amara's image of that white sky-ash, not clouds-a typhoon showing up at the same time, like the planet was just piling on.

OlaAnd then it goes planetary. Twenty two days for that aerosol cloud to circle the entire globe!

AmaraThe real gut punch for me was the geoengineering piece. Pinatubo's the only real world calibration we have for stratospheric aerosol injection, and the twenty twenty four reassessment means the baseline itself might be off!

OlaWhich is the question worth sitting with. If the best case study is still contested, what are we actually building on?

AmaraExactly.

OlaIf this episode cracked something open for you, subscribe and leave a review. Got a favorite epoch or a theory you think we got wrong?

AmaraAnd we have gotten things wrong before.

OlaEmail us at hello@faultlines dot com. Thanks for listening to Fault Lines.