Guest Post By Willis Eschenbach
Only a journalist truly committed to the ancient art of panic-clickbait could squeeze all the world’s existential dread into a headline like, “A Giant, Destructive Volcanic Eruption Is Set to Shake the World in the Coming Months, Bringing About the End of Mankind, Scientists Warn.” They’ve accompanied it with the following graphic, in case you weren’t adequately terrified.
The dead giveaway? “Scientists Warn.” Whenever you see those two words sandwiched together above the fold, you know you’re about to step into a wonderland of wild extrapolation, qualified maybes, and models run so far into the future they boomerang back with “robots take over” as the y-axis.
They start out as follows:
A detailed geophysical study published in Nature in by the U.S. Geological Survey (USGS) has refined our understanding of the Yellowstone supervolcano, uncovering new insights into its subsurface magma dynamics. Concurrently, climatological assessments by researchers such as Markus Stoffel (University of Geneva) have renewed discourse around the global systemic risks posed by a potential super-eruption — not only at Yellowstone, but at several other active volcanic complexes worldwide.
There’s an oddity here to start with. They’ve pushed together into one paragraph an actual scientific study of the Yellowstone caldera, and a paywalled puff piece by some random guy trying to frighten people about future eruptions. Unless you’re watching very closely to see which walnut the pea is under, it’s likely to be successful in making you think “Wow, a predicted super-eruption at Yellowstone, and the odds are high in other locations as well”.
Which does sound scary. So keep that thought in mind while we look at the first of the two parts they’ve pushed into one paragraph—the actual Yellowstone scientific study.
It’s the latest USGS study published in Nature under the very boring title “The progression of basaltic–rhyolitic melt storage at Yellowstone Caldera”. It gives us an upgraded, high-res CAT scan of Yellowstone’s magma plumbing. Instead of a giant pool of liquid doom sloshing under Wyoming, the new imaging shows a club sandwich: scattered blobs of partially molten rock, unevenly distributed, with most of the melt sitting in the northeast sector. The scale is impressive—400–500 cubic kilometers of rhyolitic magma waiting for its cosmic moment. The heat just keeps bubbling up from below, slow and relentless, and with enough time, these melt zones might even hook up into a larger reservoir. But spoiler: no scientist anywhere is claiming that’s on tomorrow’s chore list.
Which brings us to the great, headline-grabbing “16% chance (one in six) of apocalypse by 2100” further down in the popular reports—a number that, if ever printed on a lottery ticket, would bankrupt Las Vegas. From the article:
Still, climatologist Markus Stoffel and affiliated risk researchers estimate a ~16% probability of a VEI 7 or higher eruption occurring globally before the year 2100.
Except that particular prediction is not referred to by the scientists of the actual Yellowstone study, and has nothing to do with the Yellowstone study.
It comes from a some gentleman yclept Markus Stoffel. And he’s not even talking about Yellowstone. He’s talking about the entire planet. Nothing to do with Yellowstone.
And who is Markus when he’s at home? Is he a member of the team of authors of the Yellowstone study?
Nope.
Well, is he a vulcanologist?
Nope again.
He’s a climate professor at the University of Geneva. He’s published a lot, almost entirely regarding the effects of “climate change” on glaciers, mountain landslides, and mountain lakes. To quote from his bio page,
In a nutshell, my research is related to climate change impacts, time-series and dynamics of hydrogeomorphic and earth-surface processes at altitude and/or high latitudes, as well as on dendroecology and wood anatomy of trees and shrubs.
Translated, that means he mostly studies the nature and dynamics of landslides, and their effect on tree rings and tree populations.
Stoffel’s global “super-eruption” probability is based on … well … it’s hard to find that out. It’s from a paywalled opinion piece (not a peer reviewed study), and I’m not paying the monkey. It’s headlined:
The next massive volcano eruption will cause climate chaos — and we are unprepared
Volcanic activity will be experienced differently in a warmer world. Researchers need to understand these risks and how they could spiral.
Now, what he’s calling a “massive” volcanic eruption is scientifically known as a VEI-7 eruption or larger. The Volcanic Explosivity Index (VEI) is a logarithmic scale from 0 to 8 used to measure the relative explosiveness of volcanic eruptions. “Logarithmic” means that each step is ten times the previous step. So a VEI-7 eruption is ten times as explosive as a VEI-6 eruption. And it’s a hundred times as explosive as the VEI-5 Vesuvius eruption that buried Pompeii … so yes. Massive.
Near as I can tell from published reports and descriptions of the piece, the estimate uses a recurrence-interval logic. However, I can’t make that fit with real-world data. There have been 7 VEI-7 or stronger eruptions throughout relatively well-documented history since the eruption of the Akahoya Volcano in Japan in 7,300 BC until the eruption of Mt. Tambora 110 years ago. This makes the recurrence interval on the order of 1,070 years. So we’ll call it a thousand years. And this “thousand-year event” figure is widely quoted in the scientific literature about eruptions of VEI-7 or greater.
OK, math warning. Jump over the marked section if you don’t like math.
WARNING: MATH DANGER ZONE BELOW
Now, the Poisson formula for the probability P of a VEI-7 eruption within a certain number of years is calculated as
P = 1 − e−λt
where λ is the annual rate (1/1000) and t is the time window (75 years to the year 2100).
MATH DANGER ZONE ENDS
This means the odds of a VEI-7 or greater eruption before 2100 is 7%. Counterintuitively, it doesn’t matter how long it’s been since the last eruption. Odds are the same whether the last blowout was a hundred or a thousand years ago.
So Señor Stoffel is doing some kind of plain and fancy statistical tapdancing to get a value more than twice that of traditional math. To get to his 16% chance by the year 2100 number, the recurrence interval of massive eruptions would have to be 430 years, and there’s no evidence of that.
(Curiously, and perhaps not coincidentally, the chance of a massive eruption by the year 2200 is indeed 16% … but I digress.)
Now, about this “end of mankind” bit—don’t bother searching the Nature study. You’ll find plenty of detail on mineralogy and melt percentages, reams of electron microscope scans, and lots of caution about inferring timeframes. You will not, under peer review or USGS letterhead, find predictions about humanity’s extinction. What the data actually say is that Yellowstone’s timetable is wildly non-periodic, there are no clear cycles, and the statistical sample is, by any reasonable standard, too tiny for fortune-telling.
The actual hazard of monster eruptions somewhere on the globe? Real, yes. And it has been for the last nine thousand years.
Visibly increasing? No. The hazards have been the same over the entire nine millennia, and our ability to deal with such events has never been better.
Likely to bring about the “end of mankind”? Well, the last eight such events didn’t even begin to end mankind. They brought a few years of bad weather, sometimes very bad early on and close to the eruption. But not many people died worldwide.
So I’m gonna put my money on “No chance in hell” that the next one ends mankind.
Worth prepping your doomsday bunker over a panicky headline about eruptions written by a man with a PhD. thesis titled “Spatio-temporal variations of rockfall activity into forests – results from tree-ring and tree analysis”?
Maybe not this quarter.
So next time you see “Scientists Warn” above a picture of a bison grazing a steamy caldera, remember: It’s never the geologists issuing the press-release countdown to Armageddon. The real science, as usual, is in the fine print—buried under three layers of model assumptions, and almost always ending with some version of, “we simply don’t know when.”
My very best to everyone, life is good.
w.
As I May Have Mentioned Before: When you comment, I implore you to quote the exact words you are discussing. I’m SO tired of people saying something like “You are entirely wrong about volcanoes” or the like, and I have no idea if they’re talking to me or someone else, and if so, about what.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Nice article Willis. Real geologist here…whenever I read the doomsayers mouthing off about some geologic event I shake my head. Geologic time tends to temper any handwringing about current events…and predictions about the future.
Why Poisson?
The Poisson distribution is often used in volcanology and other natural hazard studies because it models the number of independent random events (like eruptions) that occur in a fixed time interval, under a few assumptions:
Randomness of occurrence
Eruptions are assumed to occur randomly in time rather than in cycles or predictable patterns.
Each eruption is independent of the last (the fact that one happened today doesn’t make tomorrow’s more or less likely).
Constant average rate
Over long time spans, a volcano tends to have a roughly stable average eruption rate. For example, one eruption every 30 years on average.
Memoryless property (exponential waiting times)
The Poisson process implies that the waiting time between eruptions follows an exponential distribution, which has the property that the probability of an eruption occurring soon does not depend on how long it’s been since the last one.
This matches the idea that volcanoes don’t “store up” eruption probability the longer they’re quiet — at least in the simplest statistical model.
Mathematical convenience
The Poisson model is simple, requires only one parameter (the mean eruption rate λ), and gives straightforward probabilities for eruption counts or waiting times.
I think he might have used the Rule of 72. Confused? Possibly. 😉
Stoffel utterly failed in his attempt to scare the public with his prediction because he didn’t even mention the increased CO2 coming from such an eruption, nor implicate CO2 in global warming from his concatenation of half-truths. What was he thinking? He needs to understand that people can’t be stampeded into demanding that governments do something (anything!) unless he includes scary words like “acidic magma” or “acidification of basic basaltic magma to create acidic rhyolitic magma.” He blew his chance at making a name for himself.
The key for the non-scientific research community-educated folks to learn is this: ‘publication is a game scientists play’.
You have to publish to get on in academia. You don’t get tenure without it.
In order to publish, you have to have ‘novel, insightful, meaningful’ things to say.
Who defines what is ‘novel, insightful and meaningful’?
Editors and Editorial Boards (which are glorified old-boys’ clubs in case you wondered).
You absolutely DON’T get published saying: ‘We did some really honest, really detailed, really rigorous work looking into something of potential concern to the general public and the truth is: there’s not really that much to worry about right now’.
It’s a shame that you don’t, because that would make science research one hell of a lot more cost-effective.
Another lesson the general public should learn about scientists is this: ‘Never solve a problem in 3 years if you can get funded to solve it in 15 years’.
The result of that is endless rehashing of basically the same story, having spent another £10/50/200m of taxpayers’ money in the process.
So the net result of all that is that you expect scientists to publish a tiny morsel of novel findings very regularly whilst hyping it all up in the worst traditions of tabloid newspapers.
Especially in the titles (titles of publications are the scientific equivalent of tabloid headlines).
When was the last time you saw a scientific paper entitled ‘Yellowstone Caldera extremely unlikely to erupt in the next 75 years’, eh?
When was the last time you saw a headline saying ‘Ebola virus has killed under 10,000 people in the entire time we actually knew it existed’? Never.
Because if you did, all the scams to generate Ebola Virus vaccines, Zika virus vaccines SARS-CoV2 virus vaccines etc etc would be rejected by return of post. Not dangerous enough. Not worthy of taxpayer funds anywhere in the USA or Europe.
It’s not helpful to the science research trades unions to have editorials espousing ‘Volcanoes erupt every so often: DEAL WITH IT!!’
It’s also not helpful to say: ‘Virus vaccine scams are great for the molecular biology industry but an absolute waste of taxpayers’ money!’
No-one should expect scientists to write calm, sober publications: it would be like a gold trader saying ‘gold will make you less money than keeping your cash safe in a 5 year interest-bearing account’. A gold trader is going to hype gold to get you to invest. Just like Google tell you how much you can’t do without their products. Just like Wall Street plugs AI so they can make shedloads on their bets.
If you want the truth, you need a society that actually rewards truth telling.
It hasn’t existed in the Western world since the day I entered adult employment exactly 40 years ago.
Until it does, you need to read between the lines and expect lies to be spun out of professional self-interest.
The next massive volcano eruption will cause climate chaos — and we are unpreparedVolcanic activity will be experienced differently in a warmer world. Researchers need to understand these risks and how they could spiral.
Don’t understand this. Why should volcanic activity, more than anything else, be experienced differently if its warmer? Everything will be experienced differently in some sense of course, just as things are in summer and winter.
I am working on a piece for submission to peer review. It is headed ‘The Kings Gambit will be experienced differently in a warmer world’. It will go on to report that scientists warn that this could affect the Bishops Opening and Vienna Gambit, and some even fear that the much loved Scheveningen Variation could be under threat….
Let us play a game.
For the game, we accept as valid and correct the speculation that one of 16-20 super volcanoes will pop by 2100.
Question 1: What can we do? Probably nothing.
Question 2: If the world will end by 2100 due to the super volcano erupting, why bother with Net Zero, CO2 restrictions, etc. Why not lift people up to enjoy life to the fullest with the time remaining?
Debate on the conjecture has been done in other posts. No need to repeat.
Not to worry about Yellowstone, as I saw a documentary on TV that showed the project to destroy Earth, by the Greys, was discovered and put to an end. As a result, the Greys were forced to hand over their rights to other aliens.
This reminds me of statistical rainfall events. “it was a 1000 year flood! The worlds coming to an end.” It’s just a statistical number and you can have one 2-weeks in a row!
That headline reminded me of the Weekly World News.
I checked the Daily Galaxy website. It might be legit, but it definitely goes for WWN headlines.
Willis….you used “yklept” again….I had to look it up cuz I forgot what it meant since the last time I had to look it up when thou used it…
I’m pretty sure Yellowstone is part of the planet.
But having said that, the fact that we’re hearing this prediction from a non-vulcanologist means that we can safely ignore it.