Why did glacial cycles intensify a million years ago?

Researchers find clues on the bed of the Atlantic OceanPeer-Reviewed Publication


Sticking to It

Something big happened to the planet about a million years ago. There was a major shift in the response of Earth’s climate system to variations in our orbit around the Sun. The shift is called the Mid-Pleistocene Transition. Before the MPT, cycles between glacial (colder) and interglacial (warmer) periods happened every 41,000 years. After the MPT, glacial periods became more intense—intense enough to form ice sheets in the Northern Hemisphere that lasted 100,000 years. This gave Earth the regular ice-age cycles that have persisted into human time.

Scientists have long puzzled over what triggered this. A likely reason would be a phenomenon called Milankovitch cycles—cyclic changes in Earth’s orbit and orientation toward the Sun that affect the amount of energy that Earth absorbs. This, scientists agree, has been the main natural driver of alternating warm and cold periods for millions of years. However, research has shown that the Milankovitch cycles did not undergo any kind of big change a million years ago, so something else likely was at work.

Coinciding with the MPT, a large system of ocean currents that helps move heat around the globe experienced a severe weakening. That system, which sends heat north through the Atlantic Ocean, is the Atlantic Meridional Overturning Circulation (AMOC). Was this slowdown related to the shift in glacial periods? If so, how and why? These have been open questions. A new paper published today in the journal Proceedings of the National Academy of Sciences proposes an answer.

The researchers analyzed cores of deep-sea sediments taken in the south and north Atlantic, where ancient deep waters passed by and left chemical clues. “What we found is the North Atlantic, right before this crash, was acting very differently than the rest of the basin,” said lead author Maayan Yehudai, who did the work as a PhD. student at Columbia University’s Lamont-Doherty Earth Observatory.

Prior to that oceanic circulation crash, ice sheets in the Northern Hemisphere began to stick to their bedrock more effectively. This caused glaciers to grow thicker than they had before. This in turn led to a greater global cooling than before, and disrupted the Atlantic heat conveyor belt. This led to both stronger ice ages and the ice-age cycle shift, says Yehudai.

The research supports a long-debated hypothesis that the gradual removal of accumulated slippery continental soils during previous ice ages allowed ice sheets to cling more tightly to the older, harder crystalline bedrock underneath, and grew thicker and more stable. The findings indicate that this growth and stabilization just before the weakening of the AMOC shaped the global climate.

“Our research addresses one of the biggest questions about the largest climate change we had since the onset of the ice ages,” said Yehudai. “It was one of the most substantial climate transitions and we don’t fully understand it. Our discovery pins the origin of this change to the Northern Hemisphere and the ice sheets that evolved there as driving this shift towards the climate patterns we observe today. This is a very important step toward understanding what caused it and where it came from. It highlights the importance of the North Atlantic region and ocean circulation for present and future climate change.”

The research was led also by Yehudai’s advisor, Lamont geochemist Steven Goldstein, along with Lamont graduate student Joohee Kim. Other collaborators included Karla Knudson, Louise Bolge and Alberto Malinverno of Lamont-Doherty; Leo Pena and Maria Jaume-Segui of the University of Barcelona; and Torsten Bickert of the University of Bremen. Yehudai is now at the Max Planck Institute for Chemistry.

*  * * * * * * *

Scientist contacts:
Maayan Yehudai  my2430@columbia.edu
Steven Goldstein  steveg@ldeo.columbia.edu


Proceedings of the National Academy of Sciences


Data/statistical analysis


Not applicable


Evidence for a Northern Hemisphere trigger of the 100,000-year glacial cyclicity



From EurekAlert!

3.8 12 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
November 9, 2021 6:08 am

The core idea looks a little thin.

John Tillman
Reply to  ResourceGuy
November 9, 2021 9:47 am

The 41,000-year tilt cycle still operates. As the Pleistocene wore on, it got colder, and the former deglaciations became interstadials, in which ice sheets didn’t entirely melt. So now glaciations last either two or three tilt cycles, for an average of about 100,000 years, with interglacials of varying lengths between them.

Mike Dubrasich
Reply to  John Tillman
November 9, 2021 11:32 am

Yes. Obliquity was (is) ubiquitous before and after the “transition”, which was more of an intensification. The entire global cooling phenomenon began ~40mya driven by continental drift toward the poles that restricted oceanic circulation. The “transition” was just more of the same. And it’s going to get worse unless the continents start drifting away from the poles, which will probably happen 50-100 million years from now so don’t hold your breath.

We need to warm the planet. Warmer is better. More CO2, please, or something that works.

John Tillman
Reply to  Mike Dubrasich
November 9, 2021 12:33 pm

The present Cenozoic Ice House began at the Eocene-Oligocene boundary, about 34 Ma, when deep oceanic channels opened between Antarctica and South America and Australia, forming the Southern Ocean. This led to the buildup of ice sheets on Antarctica.

Then, around 3 Ma, the Isthmus of Panama formed, altering tropical ocean currents, leading to the formation of continental ice sheets in the Northern Hemisphere as well, during the Pleistocene Epoch, beginning 2.58 Ma.

Mike Dubrasich
Reply to  John Tillman
November 9, 2021 4:28 pm

Thank you JT. I stand corrected. 34mya is a good number — although the cooling started ~50mya after the Eocene hyperthermals. For most of the Eocene temps trended downward, possibly as much as -6°C before the Oligocene thermal crash (-5°C) at ~34mya. The opening of Drakes Passage is a prime suspect. In any case, restrictions on oceanic circulation at the poles is the principal cause of global cooling on geologic time scales. The atmosphere is the mist above the water; it’s Earth’s oceans that store and distribute the sun’s heat.

John Tillman
Reply to  Mike Dubrasich
November 9, 2021 5:02 pm

Yes, the Eocene cooled after its peak, but ice sheets didn’t form until the Oligocene. Ice retreated somewhat during the Miocene, as the Drake Passage shoaled during passage of the Scotia Plate.

Last edited 2 months ago by John Tillman
Clyde Spencer
Reply to  Mike Dubrasich
November 10, 2021 8:05 am

The Oligocene was also a period of extensive volcanism.

Reply to  Mike Dubrasich
November 9, 2021 7:43 pm

“Something big happened to the planet about a million years ago.

Scientists have long puzzled over what triggered this.

However, research has shown that the Milankovitch cycles did not undergo any kind of big change a million years ago, so something else likely was at work.”

One gets the impression they didn’t look very hard or far enough back.

The first warning was “Something big happened to the planet about a million years ago”. Their use of waffle words to infer/insinuate more than for which they have evidence.

Clyde Spencer
Reply to  ATheoK
November 10, 2021 8:09 am

Most of them are chemists, not glaciologists or even geologists.

“A man has to know his limitations.”

Last edited 2 months ago by Clyde Spencer
Nicholas McGinley
November 9, 2021 6:12 am

I do not think they know what the word “discovery” means.

To bed B
Reply to  Nicholas McGinley
November 9, 2021 10:17 am

Poor scientists tend to think that their first original idea is a momentous occasion.

Krishna Gans
Reply to  Nicholas McGinley
November 9, 2021 10:41 am

Let’s start here:

comment image


Last edited 2 months ago by Krishna Gans
Clyde Spencer
Reply to  Krishna Gans
November 10, 2021 8:11 am

To go where no humanoid of diversity has gone before.

Pamela Matlack-Klein
November 9, 2021 6:21 am

Wait, what? Ice sheets are sticking to the ground so not melting? Did I read that right? It is my understanding that ice sheets are pretty powerful bulldozers, pushing everything under them along, scouring the land down to bedrock in some cases. I am skeptical about this claim of stickiness.

Reply to  Pamela Matlack-Klein
November 9, 2021 7:34 am

Pamela M-K: It is my understanding that ice sheets are pretty powerful bulldozers, pushing everything under them along, scouring the land down to bedrock in some cases.”

That’s my understanding, too.

Their slippery substrate hypothesis doesn’t quite add up to me. The change from ~41,000 year glacials to ~100,000 year glacials was rather abrupt.

If it was a “gradual removal of the continental soils” with each cycle, it seems to me that the glacial periods would get longer with each cycle as the glaciers wore away where they ended each time and advanced a little farther with each glaciation.

It seems they are describing something like a dam failure. Each heavy rain weakens the dam (41,000 year cycles), then there’s the final heavy rain that comes that causes the dam to fail and we get the 100,000 year cycles.

If that’s so, aren’t the 100,000 year cycles chipping away at a ‘new dam’, and there will be an abrupt change to say, 150,000 year glaciations?

Pamela, I’ll be watching this comment thread for the geologists’ contributions. I just have an uneasy “I don’t get it” feeling whereas they should have some factual arguments for or against the gradual wearing away hypothesis.

Reply to  H.R.
November 9, 2021 8:44 am

H.R, Pamela.

You looking at a kind of science that claims already as factual an accelerated SLR…
even while Barry already stopped it some years ago.

What you expecting.

Everything that happens, happens for real in the background noise, as far as these “guys” care.


Last edited 2 months ago by whiten
Joao Martins
Reply to  H.R.
November 9, 2021 9:23 am

I have the same gut feeling too.

Pamela Matlack-Klein
Reply to  H.R.
November 9, 2021 10:01 am

H.R., I know there are people who visit here who are very familiar with glaciation. Hopefully, one of them will set this straight.

Reply to  H.R.
November 9, 2021 8:03 pm

Depends upon what you mean by “gradual”.

Glaciers scouring the ground beneath them over a period of 100,000 years is rather abrupt geologically.

A mile or even hundreds of feet glacier thickness is a lot of weight scraping the ground.
It’s why moraines are so deep over an immense area. Ice bulldozed everything down to bedrock and in that bedrock, glaciers left their signatures.

Reply to  ATheoK
November 9, 2021 10:12 pm

I’ll add you to the ‘scrapes to bedrock’ column, ATheoK.

Maybe Feminist Glaciology can explain it better, eh?.

Clyde Spencer
Reply to  Pamela Matlack-Klein
November 9, 2021 10:20 am

More like dragging than pushing. The reason that people often think of a bulldozer is because of the terminal moraine in front. However, that is the result of melting accelerating in the zone of ablation, and the ground and shear moraines continue to move forward until the entrained loads are dumped where the ice melts.

alastair gray
Reply to  Clyde Spencer
November 9, 2021 12:05 pm

WEll Clyde When I see Mons Klint in Denmark where the thrust sheets of Cretaceous chalk was overthrusted over itself it is difficult not to see it as a bulldozer. I was blown away when I heard of it

Clyde Spencer
Reply to  alastair gray
November 10, 2021 8:30 am

If you are pushing a soft, relatively weak material such as chalk, I would expect crushing, shearing, and chaotic piling up of material at the contact with the ‘blade.” On the other hand, if you have a sheet of ice frozen to the deformable bedrock, you have all the force of the glacier spread out over the entire surface; it makes sense to me that the underlying rock will fail as a large, relatively coherent mass and exhibit thrusting and folding throughout. Think of what happens trying to push a piece of spaghetti through a hole versus pulling it through.

Reply to  Pamela Matlack-Klein
November 9, 2021 12:10 pm

Plus, all the good ice sheet are now sliding to doom faster and faster in order to drown New Your City. Is this kind of like scratching at fleas?

Reply to  AndyHce
November 9, 2021 12:12 pm

The edit function is broken. Trying to edit, I get no text to modify.

John Tillman
Reply to  Pamela Matlack-Klein
November 9, 2021 12:40 pm

The hypothesis has been around for at least 23 years:

Origin of the Middle Pleistocene Transition by ice sheet erosion of regolith


I suppose that loss of continental soil would show up in oceanic sediments.

Last edited 2 months ago by John Tillman
Clyde Spencer
Reply to  John Tillman
November 10, 2021 9:21 am

As to the thickness of the ice, it is largely a function of the rate of snowfall, which we know little about, and the duration of glaciation, which decreases from ~100K years to years ~41K.

The article states, “This caused glaciers to grow thicker than they had before. This in turn led to a greater global cooling than before, and disrupted the Atlantic heat conveyor belt.” Even assuming for the sake of the argument that their first statement is true, they do not provide a plausible mechanism for thicker ice sheets to cause “greater global cooling.” The surface area stays about the same for Kansan (100K) and Illinoian (41K).

I haven’t found any estimates of the ice thickness of the North American stages, which would support their thesis that the Nebraskan stage should have been thicker than the Wisconsinan stage.

Clyde Spencer
Reply to  Clyde Spencer
November 10, 2021 10:11 am

comment image

Reply to  Pamela Matlack-Klein
November 9, 2021 1:44 pm

Scrape off the soil, and you have an extremely inhomogeneous bedrock. Some perhaps easier to glide across than soil, some rather difficult. Polished granite would make up a fair bit of it….watch a curling match to see how polished granite slides over ice.
A very poorly supported assumption.

John Tillman
Reply to  Doug
November 9, 2021 2:22 pm


Oldest exposed rock on Earth, thanks to the repeated spread of Laurentide ice sheets (or Innuitian).

Reply to  John Tillman
November 9, 2021 2:47 pm

John do u mean Keewatin not Innuitian? I thought the Labrador, Keewatin & Baffin ice sheets combined to form the Laurentide ice sheet. Perhaps yet another name has changed 🙁

John Tillman
Reply to  stewartpid
November 9, 2021 2:53 pm

The Innuitian Ice Sheet was distinct from the components of the Laurentide.
comment image

Last edited 2 months ago by John Tillman
John Tillman
Reply to  stewartpid
November 9, 2021 3:13 pm

The Keewatin probably excavated the Acasta Gneiss.

Reply to  Pamela Matlack-Klein
November 9, 2021 7:57 pm
  • Glaciers cause pressure. Under pressure ice flows, if necessary around obstructions.
  • Glaciers also insulate the Earth from the atmosphere.
  • Glaciers melt where their ice meets the Earth, lubricating the glacier’s movement.

Geologists know which direction the ice traveled. Debris, like rocks, pebbles, sand are forced across rock surfaces scarring the surface with deep longitudinal scratches.

  • Ice freezing to solid rock happens if the ice falls as freezing rain and then a cold front passes through.
  • Not conditions similar to beneath glaciers on their paths.
  • Even so, ice frozen to rocks quickly separates from the rock when the sun shines and warms the rock.
  • Meaning, freezing rain covered rocks only lasts in deep shade.

How much did these researchers get paid for such ‘glaciers stuck fast to rock’ silliness?

Pamela Matlack-Klein
Reply to  ATheoK
November 10, 2021 2:06 am

They probably go paid a lot more than the work was worth! One wonders if they even went out in the field to look around before proclaiming their “discovery” of sticky ice….

Clyde Spencer
Reply to  ATheoK
November 10, 2021 9:28 am

Glaciers obviously freeze to the bedrock. There are numerous examples of blocks being plucked, broken along joints in the rock. Most commonly, these examples are found on the downstream side of Roches moutonnées.


Reply to  Clyde Spencer
November 11, 2021 7:17 pm

From your link:

“The rough and craggy down-ice (leeward) side is formed by plucking or quarrying, an erosional process initiated when ice melts slightly by pressure and seeps into cracks in the rock. When the water refreezes, the rock becomes attached to the glacier. But as the glacier continues its forward progress it subjects the stone to frost shattering, ripping pieces away from the rock formation.”

The claim that “plucking from the lee side” is left unsupported, not proven.

Nor does the existence of a few “Roches moutonnées” suffice as evidence that is the only process.

Pleistocene glaciation, beginning about 2.6 million years ago, has shaped the surface topography of the Great Lakes basin.

Multiple repeated advances of continental glaciers, up to one mile thick and originating from the north, sculpted the surface of the bedrock.

The glaciers carved out the basins that are now occupied by the Great Lakes.

The less competent rocks tend to be more easily scoured and result in valleys.

This is especially true for Lake Superior which mimics the shape of the Midcontinent rift (Figure 2).

The center of the rift was filled with less competent sedimentary rocks, such as siltstone and fine-grained sandstone, whereas on the margin are the more competent dipping lava flow or to the north older Archean basement.

Thus, by preferentially scouring out the less competent rocks, the glaciers left behind a horseshoe shaped Lake Superior.”

“The youngest of the several episodes to have impacted the Great Lakes basin is termed the Wisconsinan episode which lasted from about 55,000 to 10,000 years ago.

The maximum advance occurred about 20,000 years ago (Figure 5).

During this advance, unconsolidated glacial debris left after retreat of previous glacial episodes was easily eroded and removed, hence little remains in areas north of the maximum.”

“Prior to glaciation, Phanerozoic rocks covered the entire Great Lakes basin rather than only being present in the southern part (Figure 4); they were removed from the northern area by glacial action.

Rocks of the latest Mesozoic and Cenozoic may have been removed from the Great Lakes basin too.

Thus, the glaciers played an important role in determining the bedrock available for exploitation by humans.

Besides its role in topography, the unconsolidated debris left behind by retreat of the Wisconsinan glacial episode provided abundant sand and gravel used throughout the region.”

Scoured, sculpted, scraped are operative words.

The evidences of glacial action in a mountain region are in general of three kinds—grooved and polished rock surfaces, characteristic topographic forms, and deposits of ice-borne rock débris.

Of these three kinds of evidence the first would naturally suggest itself, because of its highly distinctive nature, as the easiest to recognize and therefore the most valuable for the identification of ancient glacier paths.

Glaciers, indeed, literally grind, tool, and polish their beds with the rock débris that is frozen in their bottom layers.

The large angular blocks produce grooves a fraction of an inch in depth; the smaller fragments produce fine parallel “striae”; and the sand and mud scour and polish the rock until it fairly gleams.

The resulting striated and glassy rock surfaces are familiar to all who have visited intensely glaciated mountain areas. In the upper Yosemite region and the adjoining parts of the High Sierra such surfaces are particularly plentiful and extensive; there one may walk on “glacier polish” for considerable distances.”

rock debris that is frozen in their bottom layers” sounds a bit like “plucking”, but they’re not dropped as the glacier scours the ground with them.

Last edited 2 months ago by ATheoK
Clyde Spencer
Reply to  ATheoK
November 11, 2021 9:24 pm

The large angular blocks produce grooves a fraction of an inch in depth; …

Lateral moraines of alpine glaciers are composed of rocks that have cascaded down the sides of the adjacent mountains. Where do the “large angular blocks” found in continental glaciers come from if not from plucking?

I’m not convinced that the Great Lakes were carved from bedrock when most of the basement is coherent, flat-lying Paleozoic and Precambrian rocks. It seems more reasonable that the basins are a combination of structural features and water-carved courses filled with unconsolidated muds and sands.

The resulting striated and glassy rock surfaces are familiar to all who have visited intensely glaciated mountain areas.

Yes, I have seen them personally. Mountains are different in that the pre-glaciation soils are thin to non-existent, the slopes are high causing higher ice velocities, and the ice is often channeled into preexisting stream valleys. I think that the processes are somewhat different for alpine and continental glaciers because of the ground slope. By analogy, if you put out a block of Silly Putty, it expands by spreading and rolling over small obstacles. Whereas, if you put that Silly Putty on a steep slope it will slide down.

Last edited 2 months ago by Clyde Spencer
Gordon A. Dressler
Reply to  Pamela Matlack-Klein
November 10, 2021 9:46 am

Regarding the claim of glacial-size ice sheets “sticking” to underlying rocks beds (of any arbitrary mineral type):

“At -10 deg C an average shear strength of 600 kPa was determined. Special measures taken to reduce stress concentrations at the load application points resulted in an increase of the average shear strength to 1100 kPa.”-— source: https://www.researchgate.net/publication/44071396_On_Measuring_the_shear_strength_of_ice

To put that in perspective, 1100 kilopascals is equivalent to 160 pounds per square inch (psi), which is a extremely low shear strength compared to other solids. You may have noticed how frequently ice cubes fracture (crack), rather than popping out cleanly, when one twists a plastic ice tray to “loosen” the ice that naturally has some amount of adhesion between the ice and the plastic.

More perspective:
— soft lead has a shear strength of 20,000 to 30,000 kPa, or approximately 23 times that of the upper limit for ice
— granite has a shear strength of about 15,700 kPa, or approximately 14 times that of the upper limit for ice
— wood (southern pine) has a shear strength of about 9,800 kPa, or approximately 9 times that of the upper limit for ice

So, in summary, even if one assumed there was infinitely high adhesion of ice to rock across the surface-of-contact interface, at only a short distance above that interface the glacial ice would experience sufficient shear forces (> 160 psi shear) to crack the ice so as to allow the overlaying ice sheet to slide over that portion of ice asserted to be “sticking” to the rock.

On an engineering basis and considering the basic mechanical properties of ice, the claims cited by the EARTH INSTITUTE AT COLUMBIA UNIVERSITY researchers, per the above article, make no sense.

Pamela Matlack-Klein
Reply to  Gordon A. Dressler
November 11, 2021 1:10 am

Rhodes Fairbridge must be spinning in his grave over what now passes for science at Lamont-Doherty.

Peter Wells
November 9, 2021 6:21 am

Having studied this business since 2006, and having read “Climate Change in Prehistory” by Burroughs, among other stuff, their observations sound correct to me. All that land mass around the North Pole seriously obstructs worldwide oceanic circulation.

Leo Smith
Reply to  Peter Wells
November 9, 2021 7:12 am

That I will give you, but stickiness? nah!

November 9, 2021 6:33 am

Isthmus of Panama?

Reply to  guidoLaMoto
November 9, 2021 6:43 am

That was more like 3.5 million years ago.

John Tillman
Reply to  MarkW
November 9, 2021 1:50 pm

Probably a bit later than that, but the inter-American seaway had been shoaling for millions of years: Ten million years ago, Panama was still a megalodon shark nursery:


By about 3.0 Ma, the seaway was effectively closed. Maybe a chain of islands and shoals.


The main surge of North and South American animals passing each other through Panama occurred about 2.7 Ma.

The seaway briefly reopened around 1.8 Ma, but not enough to reorder major oceanic recurculation.

On the outer Barcoo
Reply to  guidoLaMoto
November 9, 2021 10:32 am

The isthmus didn’t develop overnight and the exchange of water between the Atlantic and Pacific was a gradual affair … a million years is a relatively short time in geologic terms.The simultaneous start of the current ice age ~2.5 my ago dovetails with major hominin evolution, viz., the appearance of early (tool-making) homo in East Africa.

Mike Dubrasich
Reply to  On the outer Barcoo
November 9, 2021 11:10 am

The onset of the Pleistocene also dovetails with the major hominid technology of fire-making. It’s no coincidence that global cooling led to the evolution of an animal that can make and control fire. Heating up our ice-bound planet is our special purpose and gift from Darwin.

John Tillman
Reply to  Mike Dubrasich
November 9, 2021 1:11 pm

H. habilis-grade humans, from about 2.3 Ma, probably hadn’t yet really controlled fire and almost certainly couldn’t make it. Matbe they used natural fires and tried to maintain them or even moved burning material around, but the evidence isn’t there yet.

At present, the earliest possible evidence of control of fire dates from 1.7 to 2.0 Ma, ie late H. habilis or early H. erectus-grade Homo. The evidence is better for 1.0 Ma.

Neanderthals and pre-modern H. sapiens made fire with flint and pyrite, from 300 Ka or earlier. Anatomically modern humans used fungus as tinder.

Last edited 2 months ago by John Tillman
November 9, 2021 6:42 am

Our research addresses one of the biggest questions

I do like the way the say “addresses” rather than “answers”.

Peta of Newark
November 9, 2021 6:46 am

Good grief – glaciers stick to soil?. What are these folks smoking, eating drinking
Just for starters, what made the Great Plains? ##

It’s very simple..
A very expansive northern forest was slowly dying – you know why – inexorable inevitable soil erosion.
The trees didn’t just ‘suddenly die’ – they got fewer farther between, grew more slowly, succumbed tp peats diseases and especially with all those things, caught fire much more easily.
And with them went the litter they’d built up underneath themselves and with that went the water it retained.
When the water went, so did the heat energy it retained (wintertime) and heat it ‘trapped’ (summertime) = The climate lost its ‘inertia’ – weather lurched/raced from one extreme to the next but with every lurch, the ice crept further south.
And when it did, sea-level slowly fell and then BANG – the ice blocked up the Bering Strait

That was the tipping point, that is what stopped the Atlantic flow -water could get into the Arctic Ocean but not back out.
Thereafter the ice had free reign.

Thus, be careful of cutting trees and making deserts – There Be Colde Playces
Be careful of a warming troposphere – that heat is what was under the trees, cooling as required and warming otherwise

## On the off-chance anyone doesn’t know, it was/is the soil that the Canadian (Boreal) Forest made during the previous (to this one) interglacial

Last edited 2 months ago by Peta of Newark
Rich Davis
Reply to  Peta of Newark
November 9, 2021 5:29 pm

Ah Peta! Just as for some there is nothing CO2 can’t do, in your mind, it appears that there is no problem without soil erosion being the culprit. Well, soil erosion or carbohydrates I guess, innit?

Clyde Spencer
Reply to  Peta of Newark
November 10, 2021 9:35 am

Much of the Great Plains is composed of wind blown ash from Yellowstone and loess blown from the glacial outwash plains at the end of the various glacial stages.

Homes were made from the sod of the Great Plains! More importantly, soil saturated with water will freeze into dirty blocks of ice that actually has higher strength than pure ice.

November 9, 2021 6:47 am

So, when did glaciers discover super glue and glue themselves to the substrate?

Abolition Man
Reply to  fretslider
November 9, 2021 9:32 am

Now we know where XR got the inspiration for their shenanigans; and I thought that they weren’t serious science junkies!

Rick W Kargaard
November 9, 2021 6:50 am

Grease the glaciers to stop the ice growth

November 9, 2021 7:05 am

A PhD thesis. Yawn.

M Courtney
November 9, 2021 7:24 am

It doesn’t seem unreasonable to suggest that friction between ice and loose soil would be different to friction between ice and bedrock.

Clyde Spencer
Reply to  M Courtney
November 9, 2021 10:23 am

Except the soil isn’t loose. It will be dirty ice.

On the outer Barcoo
Reply to  M Courtney
November 9, 2021 10:33 am

Even if the ice is kilometres thick?

M Courtney
Reply to  On the outer Barcoo
November 9, 2021 3:18 pm

In both cases, yes.
Vibrations cause fissures that allow movement, briefly. But no-one is claiming that glaciers ever move at more than a glacial place.
Vibrations do not crack bedrock.

Clyde Spencer
Reply to  On the outer Barcoo
November 10, 2021 9:37 am

Non sequitur!

R Taylor
November 9, 2021 7:25 am

From Columbia in NYC, and not a mention of CO2! Academic repression might be lessening.

Reply to  R Taylor
November 9, 2021 9:40 am

I was expecting it. A very pleasant surprise. But, sticky ice?

November 9, 2021 7:47 am

Perhaps the weight of the ice sheets depressed the continental crust—uplifting the ocean crust—reducing the slope.

November 9, 2021 7:54 am

“After the Mid-Pleistocene Transition at 1 Ma,….lower global sea levels encouraged grounding line advance…A positive feedback loop was established, with cooling leading to ice sheet growth and sea level lowering, which encouraged further ice sheet growth and cooling.”

November 9, 2021 8:16 am

Just looking at this as they lay it out I don’t see the correlation. It would seem an outside factor, such as asteroid/cometary material impacts, is a more likely cause than the ground suddenly becoming “sticky”. Or perhaps a period of substantially lower solar output followed by a gradual return to a more normal level. Glaciers remaining in one spot seems a bit of a stretch as to their having grown larger than they previously are thought to have been. Wouldn’t they have grown more from increased snow accumulation combined with a decades long period of cooler temps? The ground being less or more “sticky” just does not seem very likely.

Reply to  2hotel9
November 9, 2021 12:20 pm

They do say the soil was gradually removed through multiple ice sheet cycles until there wasn’t much soil left where the ice sheets formed. The moveable soil no longer existed and the ice formed on the bedrock, which did not move as the soil did, thus holding the ice sheets in place better.

Reply to  AndyHce
November 9, 2021 6:36 pm

Still not buying it. Where there is research being done where glaciers have receded there is soil, remains of trees and other vegetation. Hell, travel through US and Canada where glaciers covered the ground and SHAZAM there be soil. In Iceland, Greenland, Switzerland, Sweden, Norway etc etc where glaciers have receded we are finding evidence of human habitation and agriculture. Glaciers grow from accumulation of snow and ice on top, not from the bottom and since they are vastly heavy they don’t move very fast, even on dirt. Add to that glaciers would have increased after a drop in temps, not cause a drop in temps. Really just not feelin’ it.

Clyde Spencer
Reply to  2hotel9
November 10, 2021 9:43 am

Yes, the idea that glaciers act like giant bull dozers is a common misconception. They are more like pieces of coarse sandpaper. Glaciers mimic streams and rivers with the highest velocity of movement above the bedrock. For alpine glaciers, or outlets of continental glaciers, the friction with the side walls slow down movement and the highest velocity is in the center.

Reply to  Clyde Spencer
November 10, 2021 11:13 am

Friend of mine who does not comment online made the point yesterday that, in general, the motion of glaciers/icesheets is most restricted on the sides/edges not across the bottom of the structure. She is digging into this study after I sent it to her. Since she retired she has walked away from all the issues and devoted herself to her grandchildren, I occasionally send her things from here and other places and she gets suitably exercised enough to come by, sit in the kitchen with a cold beer and vent about the deplorable condition of the scientific field she spent so much of her life working in.

Clyde Spencer
Reply to  2hotel9
November 10, 2021 4:31 pm

That is true of alpine glaciers or the outlets of continental ice sheets. Once the continental ice sheets get thick enough to top most of the mountains, there is effectively no edge friction.

Reply to  Clyde Spencer
November 11, 2021 4:47 am

That would be a whole lotta ice! Only ever been on one glacier, in Alaska, and the ridges on each side were several hundred to a thousand feet high. Makes a person rather humble and insignificant seeing that scale.

Clyde Spencer
Reply to  2hotel9
November 11, 2021 9:30 pm

It is pretty clear from field evidence that most of the mountains in New England were over-topped by ice, with one notable exception being Mount Monadnock.

Reply to  Clyde Spencer
November 13, 2021 3:51 am

First time I saw evidence of the power of glaciers was in Wisconsin, where my stepfather grew up south of Fond Du Lac. People who are all heated up about the “climate” and “environment” need to get out into the real world and grasp just how small and insignificant they actually are.

November 9, 2021 8:25 am

If it was a change in the AMOC, probably the best place to look is the Arctic Ocean. If the currents in and out of there changed for one reason or another and the Arctic sea ice grew or waned depending, the fresh water increasing into the North Atlantic could influence the AMOC profoundly.

If the present day Beaufort Gyre lets loose, most models suggest a big change to the AMOC, mostly resulting in a much colder Europe. The only other variable would be a weakening of the warm water coming north in the Gulf Stream, which would take a big drop in the tropical North Atlantic water temperatures and unlikely.

Robert Stevenson
November 9, 2021 8:26 am

Clearly ice sheets are not sheets of glass ie pressure depresses freezing point and allows ice skaters to skate etc. Is it suggested that some anomaly kicks in at very high pressure causing stickiness?

Robert Leslie Stevenson
Reply to  Robert Stevenson
November 9, 2021 11:12 am

The melting point of ordinary ice falls on compression, until a value of -220 C is reached at 2040 atm. Further increase in pressure results in the transformation of ordinary ice into new modifications whose mp increases with pressure. At a pressure around 20,000 atm, liquid water freezes at about 100 C.

Robert Leslie Stevenson
Reply to  Robert Leslie Stevenson
November 10, 2021 5:26 am

Maybe the ice at the base is so cold that the pressure is not enough to produce liquid water, so the ice clings to the rock and refuses to budge.

Clyde Spencer
Reply to  Robert Leslie Stevenson
November 10, 2021 9:46 am

Unlikely. Ice acts as an insulator, impeding the geothermal heat from the crust. Thus, the thicker the ice, the higher the temperature at the base of the ice.

Robert Leslie Stevenson
Reply to  Clyde Spencer
November 12, 2021 10:01 am

Well spotted; I knew that having studied lava flows to demonstrate how low thermal conductivity leads to tunnel effects. The latter forming while the lava is still flowing freely. Molten slag from batch furnaces will take a long time for the slab to solidify once an insulating skin has formed due to low thermal conductivity. The analysis requiring the solution of a differential equation for transient conduction dt/dT=ad^2t/dx^2

Gordon A. Dressler
Reply to  Robert Leslie Stevenson
November 10, 2021 12:32 pm

Ice has a density about 92% that of water. Water produces a bottom pressure of about 0.43 psi per foot of depth, so ice will produce about 0.40 psi per foot vertical depth. A mile-deep ice sheet (glacier) would therefore produce a bottom pressure “only” about 2,100 psi, or about 143 atmospheres.

Therefore, references to ice pressures in the range of 2,000 to 20,000 atmospheres have no relevance to discussions about the interface between glacial ice sheets and underlying rock.

Robert Leslie Stevenson
Reply to  Gordon A. Dressler
November 11, 2021 5:03 am

That’s what I thought; just posting the pressure info as an enlightener in the category ‘not a lot of people know that’. Altogether six different polymorphic forms of ice have been found, with six triple points

Reply to  Robert Leslie Stevenson
November 11, 2021 8:04 pm


Last edited 2 months ago by ATheoK
Joel O'Bryan(@joelobryan)
November 9, 2021 8:36 am

“ice sheets in the Northern Hemisphere began to stick to their bedrock more effectively. This caused glaciers to grow thicker than they had before. This in turn led to a greater global cooling than before, and disrupted the Atlantic heat conveyor belt. This led to both stronger ice ages and the ice-age cycle shift, says Yehudai.”

Because seafloor core samples told you that Mr Yehudai??? Extreme Over-intrepretation at work.

Joel O'Bryan(@joelobryan)
Reply to  Joel O'Bryan
November 9, 2021 8:47 am

My point is there is nothing in their core sample data to dictate a direction of causality.

Their imagined scenario is:
stickier ice –> thicker glaciers –> more cooling –> AMOC disruption

But it could equally be argued:
AMOC disruption –> more cooling –> thicker glaciers –> ice gets stickier.

Or it could be:
more cooling (from something else) –> AMOC disruption –> thicker glaciers –> ice gets stickier.

Physicist Richard Feynman once said, “The first principle is that you must not fool yourself, and you are the easiest person to fool.” One way we fool ourselves is by imagining we know more than we do; we think we are experts.

Right-Handed Shark
Reply to  Joel O'Bryan
November 9, 2021 12:23 pm

But this is climate science! Start with a conclusion and work backwards. No assumption too ridiculous.

Clyde Spencer
Reply to  Joel O'Bryan
November 10, 2021 9:48 am

Joel, unless they present evidence that the Wisconsinan stage ice was thicker than the stages preceding, they have no evidence to support any of the above permutations of events.

November 9, 2021 8:39 am

Did they test if those glaciers floated in water, and therefore made of wood, proving that . . .

bill rocks
November 9, 2021 8:40 am

Remember to apply the “method of multiple working hypotheses”.

Have the investigators considered that with increasing time and ice sheet accumulation, the land under the ice sheets was depressed creating sub-ice bowls such as now exists under the Greenland ice sheet? The ice in the bowl does not slide up hill.

Reply to  bill rocks
November 11, 2021 8:18 pm

With ice over a mile thick, where are these bowls?

November 9, 2021 8:53 am

Why did glacial periods intensify a million years ago?

Glacial periods did intensify.


Reply to  whiten
November 9, 2021 11:31 am

Correction, sorry…

Glacial periods did not, did not intensify.



November 9, 2021 9:03 am

I’m sure it was caused by human activity and it’s going to destroy the world.

Reply to  Sailorcurt
November 9, 2021 9:42 am

Due to iatrogenic CO2….

Reply to  Sailorcurt
November 9, 2021 11:08 am

Well, similar.

External natural condition.

Similar, very similar in context of implications, to the anthropogenic condition,
as per the merit of so many many, many many Hiroshimas going and still keep going in, in an accelerated modus operandi,
going into the oceans… still as we speak, with no any consideration of slowing down… unless the self destruction considered as a worthy sacrifice.


Bill Rocks
November 9, 2021 9:21 am

Have the investigators applied the “method of multiple working hypotheses”?

Was the following considered: As the ice sheets accumulated they began to depress the crust below by isostatic adjustment resulting in a bowl shape below the ice such as now exists beneath the Greenland ice sheet? The ice does not slide uphill out of the bowl.

I know little about ice but have traversed a few glaciers and crossed a bergschrund or two. On second thought, more than two.

Last edited 2 months ago by Bill Rocks
On the outer Barcoo
Reply to  Bill Rocks
November 9, 2021 10:40 am

Recommended reading: “Fire into Ice” by Vernon Frolick. Canadian exploration geologist Chuck Fipke single-handedly launched the Canadian diamond industry with his ‘insane’ theory that ice can travel up-hill. He was right.

Bill Rocks
Reply to  On the outer Barcoo
November 9, 2021 11:35 am

Can you provide a succinct explanation or description of ice sliding uphill?

The book does sound interesting and I will investigate.

Clyde Spencer
Reply to  Bill Rocks
November 10, 2021 10:02 am

Upward movement of ice happens all the time over irregular terrain. All that is necessary is for the ‘cryostatic head’ pressure be greater than the elevation of the hill. It can happen either by ice being pushed uphill, or if the depression is filled, the ice might just shear across the top of the hill.

Although you probably won’t find it in any textbooks, a glaciologist I worked with while in the Army coined the term “shear moraine” to characterize the layers of dirty ice that were deflected by the terrain, and instead of coming out at the base, came out at a considerable elevation at the snout. When the ice melted, it left the shear moraine material sitting on the snout, protecting the ice from direct melting from the sun.

There are more things in heaven and earth, Horatio,

Than are dreamt of in your philosophy.

November 9, 2021 10:48 am

Could’ve been the increasing chilling of the deep abysmal ocean deeps, which are now way below the avg Earth’s temperature. That would cause increasing chilling in ocean-upwelling areas.

Last edited 2 months ago by beng135
To bed B
November 9, 2021 11:10 am

It doesn’t seem even remotely plausible that this mechanism could lead to a sudden change. A comment above made a suggestion about glaciation causing a permanent depression, but even that happening at most glaciers in the same glacial period would hardly be Will of Occam’s first choice. A bigger deformation of the crust, say below Drakes passage, could have a million year affect all of a sudden on ocean currents.

The last glacial maximum was dry. The Congo was thought to be savannah back then. I suspect a strong feedback of cooler so drier then even cooler climate as albedo increases. Then a sudden change to ocean currents leading to heavy summer rains in glacial areas creating lakes. Now what could cause this? Glaciation of Drakes passage leading to greater NS circulation of ocean currents?

Keith Rowe
November 9, 2021 12:37 pm

The oceans got colder and colder, since Antartica got isolated and Panama closed. This was the change. There wasn’t enough energy in the oceans to do interglacials every 41,000 years. Now 100,000 years the oceans are now super cold, then warming the surface and stratifying “locking” heat in the oceans but with a cold deep oceans. When the “conveyor” starts up again with insane amounts of energy stored and cold waters go to the bottom of the ocean pushing up warm waters for a few thousand years until it runs out again and back to the cycle as the deep oceans fill up again. Dansgaard–Oeschger events are where there isn’t enough stored energy but for a few decades of warm waters. One can see it in Antarctic waters. https://journals.ametsoc.org/view/journals/clim/aop/jcliD200049/jcliD200049.xml

The thinning mixed layer says, it’s going to get colder.

William Astley
November 9, 2021 12:41 pm

The paper that assertions changes in ice sheet friction …. Is a major factor in climate change…. Is conceptually lost.

Explain the temperature changes in the last 20,000 years first. Climate ‘science’ has spent the last 20 years pushing a prediction of warming which cannot explain the temperature change in the last 20,000 years including the Younger Dryas abrupt climate change at 11,900 years ago.

The recent warming has been discovered to have been caused by a reduction in cloud cover and an increase in short wave radiation not an increase in long wave radiation. That finding disproves CAGW.

The warming which we just experienced (in the same regions of the planet) has happened before. See this graph for the temperature on the Greenland Ice sheet for the last 11,000 years.

There have been 9 Dansgaard–Oeschger (D-O events also called Bond events) during this interglacial period.

The cyclic D-O warming periods occur in both interglacial period and the glacial period. D-O periods always end with an abrupt drop in temperature.

Does the Current Global Warming Signal Reflect a Recurrent Natural Cycle.

“Public media in the U.S., including National Public Radio (NPR), were quick to recognize the significance of this discovery. The past natural warming events reported by Mulvaney et al. are similar in amplitude and duration to the present global warming signal, and yet the past warmings occurred before the industrial revolution and therefore were not caused by anthropogenic greenhouse gases.”


Timing of abrupt climate change: A precise clock by Stefan Rahmstorf
Many paleoclimatic data reveal a approx. 1,500 year cyclicity of unknown origin. A crucial question is how stable and regular this cycle is. An analysis of the GISP2 ice core record from Greenland reveals that abrupt climate events appear to be paced by a 1,470-year cycle with a period that is probably stable to within a few percent; with 95% confidence the period is maintained to better than 12% over at least 23 cycles. This highly precise clock points to an origin outside the Earth system (William: Solar changes (more complicated than just a change in the solar magnetic fields/sunspots) cause the warming and cooling); oscillatory modes within the Earth system can be expected to be far more irregular in period.

November 9, 2021 12:50 pm

The multi-precession cycles of glaciation began after the Panama Isthmus was formed around 2Mya. That disconnected the North Atlantic from the Pacific. I expect that would be a significant factor in ocean currents.

November 9, 2021 1:26 pm

Well, why leave out the influence of the Sun? Occasionally, our central star has a sneezing session, frequently coinciding with a Coronal Mass Ejection (CME), as happened in 2006 and doesn’t crank up its furnace to “normal”, whatever that is. There is nothing in that article that takes such things into account.

I don’t understand why no one takes our central star into account about such things. Not enough solar heat to do more than make a few breezes blow might very well account for prolonged cold periods, while increased solar activity may account for increased warming periods.

There’s other stuff, too, but that’s the gist of it.

Reply to  Sara
November 9, 2021 1:43 pm

Scrape off the soil, and you have an extremely inhomogeneous bedrock. Some perhaps easier to glide across than soil, some rather difficult. Polished granite would make up a fair bit of it….watch a curling match to see how polished granite slides over ice.

A very poorly supported assumption.

Smart Rock
November 9, 2021 3:39 pm

It’s a hypothesis. In the absence of actual observational data about the adhesion of ice to different substrates, that’s all it is.

November 9, 2021 3:40 pm

Why did the cycle change? Because the earth got colder. I’d have thought that was obvious.

Tom in Florida
November 9, 2021 3:43 pm

The most obvious place to look first is the obliquity cycle of 41,000 years and the eccentricity cycle of 100,000 years. Something caused the latter to dominate over the former.

John Tillman
Reply to  Tom in Florida
November 9, 2021 7:48 pm

Eccentricity has two main periodicities: one cycle with an average of ~100,000 years and a longer cycle with a periodicity of ~413,000 years.

The most important cycle is obliquity, ie 41,000 years.

Tom in Florida
Reply to  John Tillman
November 10, 2021 4:43 am

Doesn’t the record show when eccentricity is lowest in the 413,000 year periodicity cycle it causes an interglacial period that is much longer. We are at that cycle point now.

Mark BLR
Reply to  Tom in Florida
November 11, 2021 4:41 am

Something caused the latter to dominate over the former.

NB : Read the posts by “John Tillman” above (the first response to the very first post here) and “Ulric Lyons” below before reacting to this one.

I think it’s more a case of something (gradual long-term cooling passing some “threshold” ?) resulting in the 100kyr cycle starting to modulate whether a particular phase of the 41kyr cycle would “trigger” a glacial-to-interglacial transition … or not.

I had to dig out the “Global Benthic / SPECMAP” data from my “spreadsheets from my previous, broken, laptop archive” to check it against the EPICA (Dome-C) dataset, but my vague memory of the most recent interglacials having 80kyr / 120kyr spacing rather than “the regular [100kyr] ice-age cycles that have persisted into human time” mentioned at the start of the ATL article appears to be correct.

Last edited 2 months ago by Mark BLR
November 9, 2021 4:28 pm

geology and continental drift and all this is really interesting

don’t remember seeing any of this in the climate models

all to interested in the flawed Rawleigh CO2 experiments of the late 1800

Ed Fox
November 9, 2021 6:17 pm

Al Gore already solved this. CO2 fluctuates on a 100 kya, so ice must fluctuate on a 100 kya. Why are we spending money to study settled science?

November 9, 2021 9:39 pm

Because of everyone’s favourite reason.

November 9, 2021 9:47 pm

Here’s mine – a long term trend of deepening glaciation. Glacial-interglacial flicker between attractors is a sign of transition to glaciation that could in future become deep and of the quasi “snowball earth” variety.

Obliquity is the dominant Milankovitch driver. From 3-1 million years ago obliquity alone started interglacials. This one every ~40k years. But at the “mid Pleistocene revolution” or MPR this ceased to be the case and from then on interglacials required the combined maximal warming peak of all three cycles, eccentricity and precession as well as obliquity. Thus one every 100k years.

The next transition will be to permanent deep glaciation unbroken by interglacials. Antarctica is leading the world toward glacial inception now:


Last edited 2 months ago by Hatter Eggburn
Clyde Spencer
November 10, 2021 8:00 am

… the gradual removal of accumulated slippery continental soils during previous ice ages allowed ice sheets to cling more tightly to the older, harder crystalline bedrock underneath, and grew thicker and more stable.

Except that the soil bed-load would have been frozen with ice and behaved similarly to clean ice, as can be seen in the shear moraines in ice tunnels in Greenland.
Also, the interglacials are long enough to develop new soils between glaciations!

Another thing to consider is that thicker ice generally increases the probability of the basal ice/frozen ground-moraine to actually melt from pressure and geothermal heat. The melted ice decreases friction in the same manner as ice skates work.

Lastly, it isn’t just planar contact friction that retards the movement of continental glaciers. In rugged terrain, such as the Canadian glaciers encountered crossing New England, barriers such as individual mountains can divert the glacial paths around or over the obstacles. In the case of deep valleys oblique to the flow of ice, forward movement can be stopped until the ice is thick enough to over-top the mountain(s) containing the valley. Then the ice shears over the obstacle instead of moving over the ground. There are places in New England where soil and grus are preserved in pressure shadows, and striations indicate a chaotic, turbulent movement at the base. Similarly, it was recently reported that soil and entrained vegetation was discovered in the Greenland core drilling done by CRREL in the 1960s.

I think they should have had a glaciologist on the team!

Last edited 2 months ago by Clyde Spencer
November 10, 2021 3:35 pm

A glacier very close to the equator formed as the Himalayas rose, lifted by the subduction of the Indian tectonic plate under the Eurasian Plate:

The role of terrestrial plants in limiting atmospheric CO2 decline over the past 24 million years, Nature 460, 85-88 (2 July 2009) doi:10.1038/nature08133 Letter, concludes that grasslands are the result of the CO2 suffocation of trees. Plate tectonics were responsible for CO2-depleting mineral formation, especially, the uplift of the Himalayas. CO2 levels dropped to 200-250 ppm from 1000-1500 ppm as forests starved for CO2, gave way to prairies. https://www.readcube.com/articles/10.1038/nature08133 (emphasis added)

Ulric Lyons
November 10, 2021 6:55 pm

Inter-glacial frequency over the last 1.2 million years is still essentially constrained by obliquity cycles, so the question is why did the new sequence start skipping peaks in obliquity, or skipping two from the Eemian to the Holocene. And why did it get both colder and warmer than around 2 million years ago.

comment image

Reply to  Ulric Lyons
November 11, 2021 12:50 am

One explanation is that global temperatures track the advance and recession of the Karakoram and Himalayan glaciers near the equator. When temperatures rise, more snowfall precipitates over the region, adding to the size of the glaciers, presumably, thus increasing earth’s albedo (ceteris paribus, holding other things constant). The mountain ranges constantly surge, 1 cm per year, thus the glaciers have advanced over the 5 million year period.

Last edited 2 months ago by migueldelrio
E.J. Mohr
November 15, 2021 8:42 am

I want to know more about this:  ”

In 2015, a team led by John Higgins, a Princeton geochemist, excavated the record-setting core.
At first, the oldest ice seemed to contain startling levels of CO2, several times the 407 parts per million (ppm) we see today, says Yuzhen Yan, the Princeton geochemist who led the new study. Further analysis, however, revealed the bubbles had been contaminated by CO2 percolating from beneath the ice, likely released by microbes. That meant the team had to toss out data from many of the oldest samples—a reflection of their conscientiousness, says Bärbel Hönisch, a geochemist at Columbia University.”

%d bloggers like this: