When Earth Kept Freezing Itself Over and Over

OlaSeven hundred and seventeen million years ago the equator was buried under ice-Not a dusting-Sheets-the whole planet locked.

AmaraAnd stayed that way for fifty six million years; Which, okay, that number should break your brain a little.

OlaIt broke a lot of climate models, too. According to Harvard's PNAS study, the standard Snowball frameworks can't actually account for a freeze that long. They just don't.

AmaraSo get this; two very different research teams looked at the same problem and came back with completely different answers! That's where today gets interesting.

OlaRight; the University of Southampton team went to the Garvellach Islands of Scotland's west coast, analyzed two thousand six hundred individual rock layers, each one a single year of sediment, and what they found inside those layers?

AmaraSeasons, solar cycles, El Niño-like patterns during a global deep freeze.

OlaIn the Snowball, yeah.

AmaraI love that the planet is supposedly a giant ice cube and it's still out. Still out here doing its climate thing.

OlaAnd then Harvard comes in from a completely different angle-their model says distortion wasn't even one continuous freeze-it was repeated Snowball to hothouse cycles driven by a CO2 feedback tied to the Franklin Large Igneous Province.

AmaraSo the question becomes, which version of Snowball Earth are we actually dealing with-a frozen prison with a pulse or something that kept thawing and refreezing for fifty

Speaker 3million years?

AmaraFifty six million years!

OlaFollow the incentives, not the rhetoric, and today the incentives are in the rocks.

AmaraWe've got Scottish varves, Harvard models, and an argument about whether complex life could have survived any of this. Fault lines start now.

OlaPicture this! It's seven hundred seventeen million years ago. You're standing on the equator.

AmaraOkay, so get this: the equator.

OlaIce, as far as you can see; grounded ice sheets at sea level; the tropics look like Antarctica.

AmaraAt the equator?

OlaAt the equator. And here's why that matters: once ice covered enough surface, it reflected solar energy back into space, which cooled the planet, which grew more ice,

AmaraWow!

Olaa feedback loop with no obvious exit.

AmaraTHE ALBEDO SPIRAL

OlaExactly. Standard models said volcanic CO2 would eventually build up and melt it-fine, that checks out-but those same models said the whole thing should last. After a few million years, tops.

AmaraAnd here's the thing nobody talks about; it lasted fifty-six million years!

OlaRight! fifty-six million years! Harvard's SEAS published a study on this, and they put it plainly: the Sturtian Snowball ran from roughly seven hundred seventeen to six hundred sixty million years ago, far longer than any standard model could explain.

AmaraWait, wait, fifty-six million years; for context, that's longer than the entire The entire age of the dinosaurs.

OlaWay longer, and for decades nobody had the convincing answer for why the planet just stayed frozen.

AmaraSo the CO2 valve, volcanism pumping greenhouse gas into the atmosphere, why didn't it work?

OlaThat's the puzzle. The mechanism was real, the models weren't wrong about the physics. Something was suppressing the thaw, keeping CO2 levels low enough that the ice held. But what?

AmaraPlot twist! The planet that should have been dead may be the reason we're alive.

OlaPossibly; I mean, complex life came after this-the timing is not subtle.

AmaraNot subtle at all, so we've got a freeze that defies the math and it lasted longer than anything evolution should have survived.

OlaAnd here is where the new research gets interesting: a ScienceDaily piece out of the Southampton University found that the climate inside that freeze wasn't simply silent, the planet was doing something under all that ice.

AmaraWhich raises a completely different question, right?

OlaWhat was actually happening inside fifty six million years of deep freeze, and could that be exactly what kept early life from going extinct? So varves! Let me put my Research hat on for this one, because these rocks need a proper introduction.

AmaraPlease, because I keep saying Scottish rocks and that does not do them justice, right? So a varve is basically a layer of sediment deposited in one single year. Think of it like tree rings, but underwater and seven hundred million years old. According to Science Daily, the University of Southampton team Beam examined two thousand six hundred individual layers from the Port Askaig formation on the Garvellach Islands off Scotland's west coast; each layer is one year.

OlaTwo thousand six hundred years of climate data locked in rock.

AmaraAnd here's where it gets good: the statistical analysis of those layer thicknesses turned up four distinct repeating cycles: annual seasons, El Niño-like oscillations every four to four and a half years, longer solar cycles, and centennial scale variation spanning roughly one hundred thirty to one hundred fifty years.

Speaker 4Wow!

Speaker 5Wow.

AmaraThese are patterns we recognize from today!

OlaWait! so they're finding modern climate fingerprints inside the Snowball Earth!

AmaraExactly! Professor Thomas Gernon called it-and I'm quoting-"the full suite of climate rhythms we know from today, all operating during a Snowball Earth.

OlaOK, but here's what I keep turning over: the Port Askaig formation, these specific rocks. How do we know they're from the middle of the glaciation and not just a transitional window at the very start or end? Because that distinction matters enormously.

AmaraOh, that's the real question, and the honest answer is that it's genuinely contested. The sediment signatures look like deep glacial conditions, not a deglacial event, but the paper is clear that Gernon himself frames it as the exception rather than the rule.

Speaker 6Mm-hmm.

OlaMm-hmm.

AmaraThe background state of Snowball Earth was deeply frozen; what the rocks are capturing is probably a short-lived disturbance, maybe thousands of years inside a multi million year freeze.

OlaSo a brief window, not the long thaw.

AmaraA brief window with a very loud climate signal.

OlaAnd the mechanism? Because nice cycles and the rock layers don't explain themselves.

AmaraAnd this is where the models come in. Southampton's climate simulations showed that a fully ice sealed ocean suppresses nearly all oscillations, but if roughly fifteen percent of the ocean surface stayed ice free, atmosphere ocean interactions could restart and suddenly you get those exact patterns the rocks recorded.

Olafifteen per cent.--that's not a lot of open water.

AmaraIt's a sliver, a tropical water belt maybe, a slush ball state. Not vast open seas, just enough for the climate system to find its rhythm again.

OlaFollow the incentives, not the rhetoric. The climate system doesn't need permission to oscillate, it just needs a crack in the ice.

AmaraA crack in the ice-that's literally true!

OlaI know; but here's what that crack raises. If climate was pulsing under the ice, that means the Frisk probably wasn't a static slab for fifty six million years straight; something was moving and that opens a really big question about mechanism, about why it kept cycling rather than just thawing out completely.

AmaraAnd that question has a very specific answer from a different lab. A 2026 PNAS study from Harvard is arguing the planet didn't freeze once; it froze and thawed repeatedly.

OlaA limit cycle, right. That's where the Scottish rocks hand off to something much stranger. But flip the Scottish picture on its head for a second, because the Harvard team is making a much bolder claim.

AmaraRight, so here's what Charlotte Minsky and her co authors published in PNAS last month: according to Harvard CS, the Sturtian wasn't one long, continuous freeze at all.

OlaThey're saying the planet cycled, fully frozen snowball conditions, then ice free hothouse intervals. Then back again, over and over.

AmaraHa! plot twist! The most extreme glaciation in Earth's history was also secretly a yo-yo.

OlaThat's one way to put it. The mechanism they propose starts with the Franklin Large Igneous Province, massive volcanic region in what's now northern Canada, erupted around 717 million years ago.

AmaraAnd volcanoes usually mean warming, right? CO2 going up. Going up.

OlaNormally, yes, but this is the counterintuitive part: when fresh basalt is exposed at the surface it weathers, rain hits rock,

AmaraWow.

Olasilicate minerals react, and that reaction pulls CO2 out of the atmosphere.

AmaraSo the Franklin eruption basically set a CO2 trap?

OlaExactly; according to the Harvard SEAS write up, the weathering of Franklin basalt drew down atmospheric CO2 enough to tip Earth into full glaciation-ice covers the continents, weathering stops because everything's frozen.

AmaraAnd then?

OlaVolcanoes keep going, CO2 slowly rebuilds, planet thaws, fresh basalt is exposed again, weathering kicks back in, CO2 crashes, freeze again.

AmaraWait, wait, wait!--so the same volcanic province that started the whole thing kept restarting it?

OlaThat's the limit cycle model. The Harvard team calls it a limit cycle, and it runs for the duration of the Sturtian. Fifty six million years of that loop.

AmaraOkay, so here's where I want to pull out my research hat, because the oxygen angle is the part that really got me.

OlaSay more!

AmaraSo, in a single unbroken freeze, the biosphere basically shuts down-no photosynthesis in the open ocean, no oxygen production. According to SciNews's coverage of the PNAS paper, Minsky's model predicts that those repeated thaw intervals are actually what kept

Speaker 3Antarctica's life going.

Amarakept atmospheric oxygen stable.

OlaAnd that matters enormously, because aerobic life needs oxygen: you can't pause that for fifty-six million years, and expect complex life to survive on the other side.

AmaraMinsky said it directly; this could help explain how aerobic life persisted through such an extreme interval.

OlaSo I want to push on this a little, because this is a coupled box model: it's not a direct sedimentary record of hothouse intervals, it's a mathematical framework that's consistent with the evidence.

AmaraHmm, there.

OlaThe real question is what an interglacial would actually look like in the rock record: what are we expecting to find? Carbonate layers? Organic matter spikes? Because right now the model says the hothouse intervals happened, but do we have field evidence?

AmaraAnd that's where this gets interesting, Ola, because two completely different methods, the Scottish Varve record and the Harvard Carbon cycle model, are are pointing in the same direction without being designed to agree.

OlaAh! that's interesting; I'm not sure it's settled.

AmaraNo, definitely not settled; but the convergence is worth something.

OlaIt is: follow the incentives here; if both studies hold up, you're not looking at a static frozen prison, you're looking at a pressure cooker, and that distinction has real consequences for what came next. Specifically, for what kind of life made it through?

AmaraWhich is exactly the argument we need to have, because the Ediacaran fauna-complex, multicellular animals-show up in the fossil record right after the Sturtian ends, and the question of why is suddenly a lot more interesting.

OlaYeah, that's the thread we're pulling on next. So here's the thing: two studies, two different methods pointing in a similar direction. That's interesting. But let me push on the Scottish rocks for a second.

AmaraOkay, push.

OlaThe Port Askaig Formation, the Garvellach Islands.

AmaraMm-hmm.

OlaThat outcrop is famous partly because it records the transition into the Sturtian, from a warm, tropical environment. That's the selling point.

AmaraRight.

OlaSo when the Southampton team reads climate cycles in those... Those two thousand six hundred varve layers-I want to know, are we inside the full glaciation, or are we sampling rocks from the edge, from a period when conditions were still, let's say, warming up to the deep freeze?

AmaraThat's a fair challenge.

Speaker 4Hm!

AmaraBut the Southampton researchers describe those deposits as some of the best preserved Snowball Earth rocks anywhere in the world. You don't get two thousand six hundred annual layers of glacial sediment at

Speaker 3the edge of the ice sheet.

AmaraAbridgment at the margin.

OlaThat's not nothing. I'll grant you that; but a box model from Harvard and a Scottish cliff face are still pretty far from confirmation of repeated hiatus intervals.

AmaraAha! Okay! here's where the biology drags the physics back to earth.

OlaYeah.

AmaraSomething survived fifty six million years of this! Aerobic eukaryotes, the ancestors of every complex animal that ever lived, were already present before the Sturtian started! PARTED.

OlaAnd they made it through.

AmaraThey made it through!" Science News reported on research showing multicellular animal ancestors had to be around by at least six hundred thirty five million years ago, maybe as early as seven hundred fifty one million years ago, so they were there before the freeze and alive after it.

Speaker 4Wow!

AmaraExplain that with a completely static zero productivity prison!

OlaI mean ice free oases, refugia. The coastal margins-the biology doesn't rule that out either.

AmaraBut now you need two special pleadings-refugia for the biology, and a static climate to preserve the physics. The Harvard model and the Scottish record both push against the static assumption independently; that convergence has weight.

OlaFollow the incentives, not the rhetoric. The incentive here is a very clean story-dynamic phase drove evolution. I'm not saying it's wrong.

AmaraBut you're not saying it's right.

OlaI'm saying a box model is not the same as finding geochemical evidence of actual hothouse intervals in the carbonate record. That's the next test.

AmaraWhich the model predicts. The Harvard study says repeated thaw intervals should leave oxygen signatures, testable, falsifiable. That's what good hypotheses look like.

OlaOkay, I'll give you that. If sulphur and carbon isotope records from that period show repeated oxygenation pulses the limit cycle model gets a lot harder to dismiss.

AmaraAnd if they don't?

OlaThen we go back to the prism.

AmaraThat's a very bleak fallback.

OlaWelcome to seven hundred and seventeen million years ago. The question we're left with is pretty sharp though: not was it frozen, but how frozen-and did that texture of freeze versus thaw make the difference for complex life?

AmaraAnd the answer is sitting in rocks somewhere, we just have to read them right.

OlaSo here's what changes on Tuesday if the limit cycle model actually holds.

AmaraRight, because this isn't abstract—geologists have Fifty-six million years of rock sequences that were read one way.

OlaAnd every one of those sequences gets reopened: what looked like a single unbroken freeze gets reread as alternating glacial and interglacial deposits, non-glacial mudstones, intermittent carbonates inside the Sturtian... Sturtian and record—those were curiosities before; now they're potential frost signatures.

AmaraAnd the Harvard team says the clearest test is actually pretty specific: look for discrete glacial and interglacial cycles that are global and synchronous; high-precision geochronology targeted stratigraphy—that's the work.

OlaWhich is science doing what it should; the model makes a prediction, you go find out if the rock. The rocks agree.

AmaraAnd the oxygen piece is the sharpest edge of that test. The Harvard study notes that a truly frozen Snowball should have exhausted atmospheric oxygen within a few million years as volcanism kept reacting it away, but aerobic life made it through.

OlaSo either the physics is wrong or the freeze wasn't what we thought!

AmaraThe sulfur isotope record is where this gets decided. If there's no sulfur mass-independent fractionation signal, in the immediately post-glacial sediments that's consistent with oxygen not collapsing. The Brighter Side of News flagged this directly from the Harvard paper.

OlaSo decades of textbooks hinging on "the planet froze solid" may need a footnote.

AmaraA very large footnote.

OlaBut here's the question I keep coming back to: if the chaos is what kept aerobic life alive across fifty six million years- Then the most extreme climate catastrophe in Earth's history may be the reason complex life got its footing.

AmaraThe freeze didn't almost end us; it trained us.

OlaThat's the question now. Okay, so? Fifty six million years of ice!

AmaraMm hmm.

OlaThe planet locked in and somehow life found a way through.

AmaraThat's the part I keep coming back to-not just that snowball earth happened, but that the freeze wasn't static.

OlaYeah.

AmaraScottish rocks, ancient varves, climate cycles still ticking underneath all that ice.

OlaAnd the big question we left open, right?--whether those cycles reflect the deep freeze or just the edges of it. We didn't fully agree on that.

AmaraWe did not; but I think that's the honest answer. The science is still being argued.

OlaThe takeaway for me? Chaos might have been the engine, the pressure that made complex life possible.

AmaraIf this one cracked something open for you, subscribe and leave us a review. Got a theory we got wrong? Email us at hello at heymado dot com.

OlaWe want to hear it. Thanks for being here.

AmaraWe'll see you next time.