A “New” Theory About the Formation of the Antarctic Ice Sheet

Guest post by David Middleton


One of the big mysteries in the scientific world is how the ice sheets of Antarctica formed so rapidly about 34 million years ago, at the boundary between the Eocene and Oligocene epochs.

There are 2 competing theories:

The first explanation is based on global climate change: Scientists have shown that atmospheric carbon dioxide levels declined steadily since the beginning of the Cenozoic Era, 66 million years ago. Once CO2 dropped below a critical threshold, cooler global temperatures allowed the ice sheets of Antarctica to form.

The second theory focuses on dramatic changes in the patterns of ocean circulation. The theory is that when the Drake Passage (which lies between the southern tip of South America and Antarctica) deepened dramatically about 35 million years ago, it triggered a complete reorganization in ocean circulation. The argument is that the increased separation of the Antarctic land mass from South America led to the creation of the powerful Antarctic Circumpolar Current which acted as a kind of water barrier and effectively blocked the warmer, less salty waters from the North Atlantic and Central Pacific from moving southwards towards the Antarctic land mass leading to the isolation of the Antarctic land mass and lowered temperatures which allowed the ice sheets to form.

No one has thought to link these two competing explanations before

A group of researchers, led by scientists in McGill University’s Dept. of Earth and Planetary Sciences now suggest that the best way to understand the creation of this phenomenon is, in fact, by linking the two explanations.

In a paper published on the subject in Nature Geoscience earlier this week they argue that:

  • The deepening of the Drake Passage resulted in a change in ocean circulation that resulted in warm waters being directed northwards in circulation patterns like those found in the Gulf Stream that currently warms northwestern Europe.
  • That this shift in ocean currents, as the warmer waters were forced northward, lead to an increase in rainfall, which resulted, beginning about 35 million years ago to reduced carbon dioxide levels in the atmosphere. Eventually, as the levels of carbon dioxide in the atmosphere dropped, as a result of a process known as silicate weathering (whereby silica-bearing rocks are slowly worn away by rainfall leading the carbon dioxide from the atmosphere to eventually becomes trapped in limestone) there was such a significant drop in CO2 in the atmosphere that it reached a threshold where ice sheets could form rapidly in Antarctica.


To read “Enhanced weathering and CO2 drawdown caused by latest Eocene strengthening of the Atlantic meridional overturning circulation,” by Geneviève Elsworth, et al in Nature Geoscience: http://www.nature.com/ngeo/journal/vaop/ncurrent/full/ngeo2888.html



McGill Newsroom

Meet the new theory
Same as the old theory

My apologies to Pete Townsend for borrowing his lyrics… But this doesn’t sound “new” to me.  The only thing that appears to be “new” is the competing “theory” that a drop in atmospheric CO2 triggered the rapid formation of the Antarctic ice sheet.  The tectonics-driven disruption of oceanic circulation has been the generally accepted theory for as long as I can remember.

The Earth’s climate rapidly cooled about 34 million years ago.  This cooling marks the transition from the Eocene to the Oligocene…


Cenozoic temperature anomaly. Older is to the right, x-axis is millions of years before present (1950 AD).  Earth’s climate rapidly cooled about 34 million years ago.

Whereas atmospheric CO2 appears to have remained elevated for about 2 million years after the sudden drop in temperature…


Cenozoic atmospheric CO2. Older is to the left, x-axis is in millions of calendar years AD (BC). The Mauna Loa instrumental record  is depicted by the solid black line at year 0.  Oligocene CO2 levels drop from about 800-1,000 ppmv to about 600 ppmv from 32.5 to 30 MYA.

My apologies for the different x-axes.  I made these graphs several years ago and don’t recall where I saved the Excel files.

Otherwise, the press release is very well-written.  The paper, which is behind a paywall, also looks to be very well done…

Enhanced weathering and CO2 drawdown caused by latest Eocene strengthening of the Atlantic meridional overturning circulation

Geneviève Elsworth, Eric Galbraith, Galen Halverson & Simon Yang

Nature Geoscience (2017) doi:10.1038/ngeo2888
Received 14 October 2016 |  Accepted 03 January 2017 | Published online 30 January 2017


On timescales significantly greater than 105 years, atmospheric pCO2 is controlled by the rate of mantle outgassing relative to the set-point of the silicate weathering feedback. The weathering set-point has been shown to depend on the distribution and characteristics of rocks exposed at the Earth’s surface, vegetation types and topography. Here we argue that large-scale climate impacts caused by changes in ocean circulation can also modify the weathering set-point and show evidence suggesting that this played a role in the establishment of the Antarctic ice sheet at the Eocene–Oligocene boundary. In our simulations, tectonic deepening of the Drake Passage causes freshening and stratification of the Southern Ocean, strengthening the Atlantic meridional overturning circulation and consequently raising temperatures and intensifying rainfall over land. These simulated changes are consistent with late Eocene tectonic reconstructions that show Drake Passage deepening, and with sediment records that reveal Southern Ocean stratification, the emergence of North Atlantic Deep Water, and a hemispherically asymmetric temperature change. These factors would have driven intensified silicate weathering and can thereby explain the drawdown of carbon dioxide that has been linked with Antarctic ice sheet growth. We suggest that this mechanism illustrates another way in which ocean–atmosphere climate dynamics can introduce nonlinear threshold behaviour through interaction with the geologic carbon cycle.

Nature Geoscience



117 thoughts on “A “New” Theory About the Formation of the Antarctic Ice Sheet

    • Since CO2 is a radiative gas, higher concentrations in CO2 would have cooled the planet. Only during the day does CO2 convert IR radiation to heat and back again. During the day, its effects are a wash, no change. BUT, during the night, radiatiive gases such as CO2 and water vapor, with no solar input, actively convert heat energy in the atmosphere to IR which is lost to space. How long will it take these clowns to realize that even at 6000 ppm, a trace gas does not drive the climate. Never has never will.

      • so, you are claiming that IR radiation checks to see if the sun is shining before it does its thing?

      • To higley7:

        I think is misunderstand what happens.

        C02 and water vapor (as radiative gases) both absorb heat and radiate it, However, what they mostly absorb from the lower levels is from warmer temperatures than the temperatures they are at when their radiation finally escapes to space and is not reabsorbed. Hence the net radiation to space is less than it would have been if there had been none of these gases present. Less radiation equals less cooling which means relative warming. Since these gases do not absorb much solar radiation, there is not much difference in this from day to night.

    • If it was me, I would examine the possibility that the Temperature in the Antarctic went below zero deg. C for a while, like tends to happen every year during southern winters.

      Seems to be caused by the sun staying below the horizon for a while. That would make some ice.


    • Yes, we staved off the next ice age for some thousands of years…

      but the point is in doing so we created a very rapid (geologically speaking) increase in temps and climate change which mankind will likely have trouble adapting to.

      • We have done no such thing. Man-made GHGs have had no measurable effect on global temperature. Nor will they at 600 ppm.

        Nothing has happened since 1950 or is happening now with climate to which we can’t adjust.

      • In any case, the next glacial phase isn’t due for at least 2000 years.

        Before then, there will be another, bigger LIA and a cooler than present Post-Modern Warm Period. Then the start of the next Big Chill.

      • No true scientist would say that considering the lack of QUALITY data both pre 1960 and proxy temp data. NONE. NOT ONE. Only those following a religious like agenda would go so far. Which reminds me…,

        ACTIVISTS MARCH, NOOOT SCIENTISTS. Data clearly shows Activism Prevents Objectivity.

      • giffiepoo, has, as is normal for it, spouted absolute nonsense.

        What did your home base do giffiepoo? Print out some more false claims?

        Check David Middleton’s graph above. The pitifully few CO2 ppm mankind contributes to the atmosphere is lost when viewing mankind’s CO2 lost in ocean, volcanic and tectonic outgassing.

        Mankind will have zero trouble adapting.
        Perhaps you should get your breezy Queen Nefertiti garb ready. One can’t be too hasty in dressing correctly.

  1. “Otherwise, the press release is very well-written”

    Unfortunately I can’t agree since it has a A-grade blooper in the third sentence:

    “Scientists have shown that atmospheric carbon dioxide levels declined steadily since the beginning of the Cenozoic Era, 66 million years ago”

    Actually all proxies shows that CO2 levels went up for 20 million years from the rather low latest Cretaceous level. That’s a third of the whole Cenozoic. Then it starts going down about 35-40 million years ago, at about the same time as the Antarctic freezes over completely for the first time.
    So what came first? Did lower CO2 level cause the freeze, and if so why didn’t the same thing happen 65 million years ago when CO2 levels were about as low?
    Or was the freeze-up caused by reorganized ocean and air circulation and the lower CO2 level due to the fact that the cold deep ocean absorbed more CO2? We do know that the deep oceans switched to an “icehouse mode” about this time due to the glaciation of Antarctica which started up the production of cold, dense Antarctic Deep Water.

    • I didn’t say it was flawless. At least it didn’t contain a bunch of alarmist nonsense. And, yep, CO2 rose during the early Cenozoic, peaking about 10 million years after the Paleocene-Eocene Thermal Maximum.

      • Nah.

        Instead they bluster and blunder about with silly speculations:

        “That this shift in ocean currents, as the warmer waters were forced northward, lead to an increase in rainfall, which resulted, beginning about 35 million years ago to reduced carbon dioxide levels in the atmosphere. Eventually, as the levels of carbon dioxide in the atmosphere dropped, as a result of a process known as silicate weathering (whereby silica-bearing rocks are slowly worn away by rainfall leading the carbon dioxide from the atmosphere to eventually becomes trapped in limestone)”

        A) Shift warmer waters northward? First they decide that the Antarctica circumpolar current became too strong, now something has shifted warmer waters northward. From where to where? Specious.

        B) Then to an increase in rainfall. Rain that somehow reduced atmospheric CO2.

        C) They decide that “silicate weathering” trapped the CO2 in limestones?
        I must say, that is a new one to me.

        Silicate weathering is dreadfully slow. Plus it involves the slow weathering process of breaking down and leaching igneous sourced rocks, eventually into clay.

        CO2’s sole involvement is providing a slightly acidic environment, without actual loss of carbon into the ground.
        A competing theory is that groundwater leaching, especially during tectonics when magma heated solutions are quite acidic, helping to break down minerals.

        Then that limestone finish… Limestone as in deposited compressed microscopic skeletons and carcasses from ocean populations of plankton.

        Also leaving us the belief that limestone formations will all be found under gigantic worn masses of quartz and feldspars, from the leaching of silicate minerals.

  2. Sounds to me like they just wanted to keep CO₂ as the main driver of temperature change, even when accepting that the opening of the Drake’s Passage was the trigger.

    • Sounds that way to me too. There are all sorts of problems with CO2 as a driver (lagging indicator of temperature change being the most prominent), but they keep trying to make the square peg fit in the round hole. A bigger hammer is all they really need.

  3. The problem with CO2 and temperature changes is that due to each one affecting the other it is impossible to work out a proper cause-effect relationship. The current fashion is to adopt by default the CO2 -> temperature every single time. This is likely to be wrong and lead to a lack of progress in our understanding of past climate changes.

    • It’s actually not such a big problem. Ice core samples consistently show CO2 as a lagging indicator–temperature change comes first. Al Gore got around this problem in An Inconvenient Truth by showing them side by side, rather than overlaid, making it virtually impossible for the audience to catch what is really a devastating problem with his argument.

      • It is actually still a big problem because due to the firning process of the ice, the age of the gas does not match the age of the snow, and the age difference depends on precipitation rates. When you say that the lag is 800 years, you are assuming ice and CO2 have the same age, and that assumption is incorrect.

        See for example:
        Pedro, J. B., Rasmussen, S. O., & van Ommen, T. D. (2012). Tightened constraints on the time-lag between Antarctic temperature and CO 2 during the last deglaciation. Climate of the Past, 8(4), 1213-1221.

        Brook, Edward J. “Leads and lags at the end of the last Ice Age.” Science 339.6123 (2013): 1042-1043.

        Not precisely a settled matter.

      • At the beginning of an ice-age the lag is several thousand years, so large that temperature and CO2 frequently go in opposite directions.
        However Javier is right that the lag between ice age and gas age is uncertain and variable. It does not only depend on precipitation, but also on to what extent there is exchange between the atmosphere and the gas in the snow layer (changes in barometric pressure means that there will be at least some “pumping”), which in turn depends on i. a. whether there are any melting events (even a thin layer of ice will seal the gas below).

      • Javier,

        The lag between the ages of the ice and the entrapped air bubbles is not related to the lag between CO2 and temperatures. These are two different lags.

      • @ David Middleton The range of the two lags is about an order of magnitude apart as well, firn ~70-100 years, CO2/Temp ~800-1000 years. Some firn can be longer if it is in a dry area with low accumulation of snow.

      • The ice-air age lag is highly variable and generally dictated by precipitation rates. This also dictates the resolution of the ice cores. Law Dome has a high precipitation rate and the lag time is only about 30 years. This gives it a resolution of at least 30 years, possibly even 10 years. The lag time can be several hundred years in cores from areas with low precipitation. These cores have very coarse resolution, but can “see” back further in time.

        Unsurprisingly, ice cores with higher resolution demonstrate higher CO2 levels…

        The temperature-CO2 lag is primarily a function of the warming and cooling of the southern oceans. It’s generally around 800 years (+/-).

      • David,

        “The lag between the ages of the ice and the entrapped air bubbles is not related to the lag between CO2 and temperatures. These are two different lags.”

        It is clearly related. As the CO2 is younger than the ice, and you plot CO2 versus ice age you introduce a dating error in the CO2 curve and therefore a proper lag between CO2 and temperatures cannot be calculated unless the age difference is accounted for.

        Skeptics like to say that science is not settled. This is true even for the things on which our argumentation rests. The CO2 lag is not a particularly strong point. A temperature increase increases CO2, and a CO2 increase increases temperatures. Even if the first process is much stronger than the second as an skeptic would believe, they can still move in near synchronicity.

      • Javier,

        I see your point. The ice-air age lag introduces a highly variable potential dating error into the ice core CO2, which may affect the lag between ΔT and ΔCO2.

      • Javier & David,
        The time lag between temperature and time recorded in the 18O/16O ratio and the time when CO2 is entrapped within ice bubbles varies because they measure different events. BUT, the time when atmospheric gas became entrapped in ice bubbles has in several cases been measured directly — by using isotopic fractionation of N2 and AR ratios produced by diffusion. This method still indicates CO2 lags, but not as strongly, with perhaps uncertainty overlapping with zero.

    • Javier,
      A ten million year lag in CO2 peaking after the PETM peak strongly suggests to me that CO2 is a primary effect rather than a cause, even if they are part of a feed-back loop.

    • Pablo, thanks.
      1) Temperature leads the CO2 concentration.
      2) No causative effect between increasing CO2 and temperature that increased almost a millennia before CO2 started increasing.
      3) Causative effect of (lagged) temperature increases causing CO2 increases possible?
      4) What mechanism could cause CO2 to rise about 800 years after temperature rises?
      5) Now I’m confused??
      6) Science isn’t settled because CO2 takes nearly a millennia to decide how to respond to increased temperature. What processes could cause an 800 year delay in CO2 before it follows temperature increases?
      7) Maybe there is no relation between temperature increases and CO2 increases that we can see because warmunnists only started this global warming scare a very short time ago by turning up a thermostat before a Congressional hearing ?
      8) It is all so uncertain and fraught with tricks and deception and time scales we cannot comprehend; let alone measure any causative relationship.
      9) Is it time to discard the temperature CO2 speculations, tricks, and seances and return to real science?

      • Leonard Lane says:
        February 1, 2017 at 9:10 am
        6) Science isn’t settled because CO2 takes nearly a millennia to decide how to respond to increased temperature. What processes could cause an 800 year delay in CO2 before it follows temperature increases?

        I am no specialist, so please take my explanations with a grain of salt.

        The short answer: the ocean water cycle could be responsible for this delay.

        The longer answer:
        There is more CO2 dissolved in the ocean as in the atmosphere, I think it is something like 40:1 or more.
        CO2 is less soluble in water with increased temperature. When the ocean warms it emanates CO2, when it cools it absorbs CO2, or put otherwise, more CO2 gets dissolved from the atmosphere and incorporated into the ocean with colder water.
        The ocean water cycle is somewhere about 800 years or more.

  4. I discussed how changing circulation caused the formation of the Antarctic ice sheet and simultaneously the cooling of the deep oceans in the essay Antarctic Refrigerator Effect


    The asymmetry between formation of Arctic ice caps about 30 million years after the Antarctic ice sheet formed, makes it doubtful dwindling CO2 was a driver of global cooling cooling. Changes in CO2 can be attributed to the evolution of diatoms and there ability transport CO2 to depth, and massive increases in diatoms due to greater upwelling due to circulation changes

  5. Ah yes… it’s another computer model based study.

    Enhanced weathering and CO2 drawdown caused by latest Eocene strengthening of the Atlantic meridional overturning circulation

    ” In our simulations, tectonic deepening of the Drake Passage causes freshening and stratification of the Southern Ocean, strengthening the Atlantic meridional overturning circulation and consequently raising temperatures and intensifying rainfall over land. These simulated changes are consistent with late Eocene tectonic reconstructions that show Drake Passage deepening, and with sediment records that reveal Southern Ocean stratification, the emergence of North Atlantic Deep Water, and a hemispherically asymmetric temperature change.”


  6. One of the ‘blips’ that triggered my B.S. meter at the RealClimate website many years ago was their explanation that almost undetectable increases in CO2 initiated the interglacial warming periods where then much larger increases CO2 followed temperature increases by 500 or so years. That, plus their auditing of comments lead me to look for websites that appeared logical and reasonable.

    • The epitome of circular reasoning…

      What does the lag of CO2 behind temperature in ice cores tell us about global warming? This is an issue that is often misunderstood in the public sphere and media, so it is worth spending some time to explain it and clarify it. At least three careful ice core studies have shown that CO2 starts to rise about 800 years (600-1000 years) after Antarctic temperature during glacial terminations. These terminations are pronounced warming periods that mark the ends of the ice ages that happen every 100,000 years or so. Does this prove that CO2 doesn’t cause global warming? The answer is no.

      The reason has to do with the fact that the warmings take about 5000 years to be complete. The lag is only 800 years. All that the lag shows is that CO2 did not cause the first 800 years of warming, out of the 5000 year trend. The other 4200 years of warming could in fact have been caused by CO2, as far as we can tell from this ice core data.


      • That would be plausible except that, at least in some of the available graphs, the concentrations of CO2 don’t fall from their peaks until some time after the temperature starts dropping.

      • CO2 plays a small role in warming, by retrarding radiative cooling. However, it clearly isn’t the primary driver of warming and cooling…

      • When faced with a misbehaving feedback control system, the first thing we do is break the feedback loop. With the feedback loop in place, there is usually zero chance of properly diagnosing the system … and that’s for a nice linear time-invariant (LTI) system. The global climate is not LTI.

        The passage you quote is one of the most gobsmackingly ignorant things I have had the misfortune to read in a while. Scientists are not engineers. Their education is missing some of the fundamentals that would keep them from saying stupid things.

      • Dividing a 5000 year warming arbitrarily into an 800 year and a 4200 year part is not a parsimonious argument. It is very unlikely to be based on truth.

      • That RealClimate piece is complete bollocks (I’m being polite—the actual words I would choose are far saltier than anything a sailor would use). No wonder I eviscerate any warmist who tries using it as a reliable source to “prove” AGW from CO2 forcing.

      • CO2 and temperature are closely correlated. On the balance of evidence bi-directional causality is likely.

        It is possible temperature has more effect on CO2 levels than the reverse. But it is also possible that the effect CO2 has on temperature is more rapid than warming’s effect on CO2.

        If the effect warming has on CO2 lags hundreds of years+or-, how much lag is there for the reverse effect? Essentially immediate or just long enough to allow for a bit of chaotic SST mixing and albedo change?

        In other words it may take natural cycles centuries to warm things up but once CO2 (and gasses) start acting as accelerants.

        So if humans inject these geologically significant quantities of CO2 into the system it is completely plausible we could kick-start a warming event

        Hmm, that’s just what we see, but I suppose it could all be just a coincidence.

      • It’s not really circular reasoning, it’s something even more ridiculous. To paraphrase the argument:

        1. “something” caused it to start to get warmer
        2. warming caused NH ice sheets to retreat, and CO2 was liberated from freshly exposed tundra
        3. CO2 then caused lots of warming
        4. “something” went back into its shell until the start of the next interglacial

        And you’re right, David, this is the low point of warmist “logic”. In one version I’ve read, the “something” was said to be increased insolation dependent on orbital cycles (actually, not bad try but why did that effect stop?). Ignoring the change in albedo as continental ice sheets melt is the Achilles Heel of this “reasoning”.

        Plus of course they need “something else” to initiate the subsequent cooling towards the next glacial period while atmospheric CO2 is still rising, until it peaks at a level much higher than was deemed capable of causing warming at the start of the interglacial.

      • If CO2 really did amplify by positive feedback every warming or cooling, then climate would oscillate monotonically all the time. It does not. Instead it meanders, in the manner of Lorenz (DNF 63) classic nonlinear simulation. Thus the basics of nonlinear chaos dynamics show CO2 not to be a significant climate driver. That of course is why the climate establishment pay merr lip service to chaos-nonlinearity theory while avoiding / ignoring all its clear implications.

      • That is Jeff Severinghaus in apology mode. The idea is a Milankovitch jumpstart to get CO2 going. CO2 would work equally on the up and down trends. If you take 5000 years as the uptick, the long downhill slide takes 90,000 years for a ~100kyr post MPT cycle.

        Jeff must explain why CO2 radiative forcing works so much more impressively on the way up.

      • The reason is co2 isn’t working.

        If you accept the idea that co2 is this force that is magnified and is 80-90% of the ice age temps then the issue becomes why isn’t it having such a profound impact now. The fact it isn’t having such a large impact now in a real sense disproves that co2 is the major force in the ice ages.

        I have argued now that the real force is gravitational forces of milankovic and pressure from ice plates on the earth cause periodic increase in release of heat from the mantle through fissures. This process may slowly return to interglacial levels of release as the earth restores to a more flat shape and ice plates melt. Whereby the earth cools and repeats the cycle. In this case co2 is far less important.

        If water temperature stays high for a while solubility would increase slowly and co2 would disgorge from the atmosphere slowly. It is likely when the earth is distorted that the first releasss are more intense and cause rapid outgassing of co2.

      • Since we don’t know why CO2 doesn’t work, the slate is clean for any and all new ideas.

        I believe there are very good spectroscopic reasons for CO2’s failures.

        I also believe that during the Cretaceous sea level high stand and the massive large igneous intrusions, something like what you describe was going on. Very interesting that all this happened during a period without geomagnetic reversals, no?

      • I am not that familiar with geomagnetic reversal events. So I can’t comment but if you look at those periods and the co2 record the correlation looks random. It does appear there is very good reason to beleive that co2 has limited impact which as you say leaves open other possibilities. When I said this to climate scientists the basic argument is : well that’s useless because we can’t build climate models with unknown theories so we use what we have.

        Fine. That’s perfectly fine but what it really means is that they admit the models are probably missing all kinds of things they don’t understand so don’t take them seriously except as a research tool.

        They never say that to the public. They never say you know these are just some random theories we are playing with to see how they correlate and don’t pay attention and there are large unknowns and things could change.

        No they say the science is settled. We understand everytging well enough and our models work great. Shut up.

        I’m so sick of their corrupt clearly politicized crap. I cannot wait to clean this up and make these charlatans go away and have real fun and interesting science in this area where all kinds of ideas proliferate to explain lots of different things instead of this politically correct obviously impossible science that doesn’t work at any level.

      • Indeed. In a sense CO2 is just a place holder in the models. They are correct that the models go bonkers without it. The trick is to figure out what is doing the work that CO2 is essentially a proxy for.

  7. Was the Antarctic continent 35 million years in the same position as now? Am I wrong in thinking that the landmass shifting onto the pole was the trigger?

  8. This is a fascinating and important article.

    The levels of CO2 are far less important to the temperature than the geographic changes stipulated.

    CO2 changes historically have more to do with changing solubility of CO2 in the ocean and plant life carbon sinks.

    The changing geographic features combined with mantle offgassing due to gravitational fluctuations of Milankovic cycles and other events is primarily responsible for the climate changes. This is my opinion now that we’ve seen evidence that mantle leakage may vary dramatically on many factors and combined with ocean changes probably play a major role in climate far exceeding climate.

  9. Folks are overlooking an obvious flaw in the “CO2-as-climate-driver” paradigm:

    Think of the sawtooth pattern of CO2 over the last 600,000 years/4 major glacial/interglacial cycles.

    The cooling begins when CO2 is rising to its highest level, and then temp and CO2 decline together until full glacial maximum conditions are achieved.

    Similarly, the warming begins when CO2 is declining to its lowest level, and then temp and CO2 rise rapidly to full interglacial conditions

    Whatever triggers the reversals of warming/cooling CANNOT BE CO2 because it is trending in the opposite direction at the time of reversal

    The lag between T changes and CO2 changes also demonstrates CO2 is not the driver of change, but this evidence is secondary to the rreversal evidence

    • I agree, Geology Jim. Furthermore, the Antarctic ice sheet began when plate tectonics pushed the continent onto the southern geographic pole and surrounded it with a cold ocean in which warm currents are steered and deflected away by the circumpolar current, leaving the polar region isolated from the bulk of oceanic and atmospheric convective heating. I daresay that CO2 had [Matt Foley=ON] JACK SQUAT [Matt Foley=OFF] to do with it.

    • If you assume that co2 can’t be the cause of the ice ages or even a large portion then the rise and fall of co2 has to do with solubility in water and plant and animal life breathing.

      Many other things change. During milankovic cycles extreme pressure is put on the orbit of the earth and distortion of the earths mantle occurs. The orbit of the earth changes to give less heat to the earth from the sun. The gravitational forces on the oblong shaped crust reach a maximum and has been shown to open large number of fissures in the ocean floor. Heat escapes and warms the ocean which causes a huge outgassing of co2. The glaciers which have reached a peak during the glacial maximum put pressure on the mantle as well enhancing the heat release.

      As the heat pours out the glaciers melt and the distortion caused by the mass concentration at the polls decreases, milankovic gravitational forces distorting the earths shape and orbit start to diminish. More heat hits the surface of the earth as the earth warms. More melting of glaciers and less pressure on the earths shape and orbit flatten and close fissures. The heat escaping into the oceans drops and the cycle starts again.

      Co2 has been shown to provide at most 20% of the heat needed for an interglacial or vice Versa to cool the earth.

      This theory backed up with recent articles has no need for 5:1 amplification of co2 to get ice age variations. Co2 is a minor player.

      We know at 600 million years ago, 60 million years ago and 30 and 3 million years ago geographic changes and changes In the mantle produced radical variations in stable temperature of the earth. Up until 600 million years ago the earth was a total iceball but then a massive underwater release of every validated with undersea exploration show that it was fissures that ultimately brought earth out of this iceball state. Now we see that 60 and 30 million years ago changes in the Antarctic allows further cooling which forms the ice pack on the Antarctic and cools the earth for millions of years.

      Co2 doesn’t explain these events. It runs counter to co2 at times. Then 3 million years ago the 50,000 year cycle of ice ages goes to 100,000 years suddenly. Possibly the closing of the isthmus of panama may have caused this. A minuscule geographic change changes the earth cycles of ice ages and doubles the time periods between them possibly by further separating the oceans. Again co2 not such a big factor.

  10. When you compare Greenland with Antarctica, what do you see? Large pieces of land surrounded by cold water and most of the land area is contained within either the Arctic circle or the Antarctic circle. It’s as clear as mud as to why permanent ice caps formed upon these two large land pieces surrounded by cold water. There should be no mystery as to why the ice cap on Antarctica formed. But rather, there should be the realization that it is the ice cap on Greenland and its very existence, signifies this present Ice Age the earth has been in for the past 2.6 million years. When the Greenland ice caps melts, and it won’t for a very long time, that is when this present Ice Age ends.

    • Greenland isn’t a pimple on the posterior of the Antarctic ice sheets. The WAIS alone contains most of the fresh water on earth’s surface.

      Technically, the present ice age (or ice house) began when ice sheets formed on Antarctica ~34 Ma and won’t end until those are gone.

      The Carboniferous-Permian ice age lasted about 100 million years. At 30 million years, max, the Ordovician-Silurian was much shorter. Who knows how long this one will persist.

      • When South America joined with North America, in the forming of the Isthmus of Panama, that cut off the two major oceans of the world at the tropical ocean level. That occurred 3 million years. It was not a coincidence that 3 million years ago, the ice cap on Greenland started to form. Then four hundred thousand years later, the earth entered into an Ice Age. There was no Ice Age occurring when the Antarctic ice cap formed 30 million years ago. This present Ice Age commenced shortly after the Greenland ice cap formed. That is simply generally understood and accepted climate history.

      • It is a fact, ie a scientific observation, that the Antarctic ice sheet first formed after the continent was separated by deep channels from South America and Australia.

        An ice age or ice house is when earth has ice sheets. A hot house interval is when it doesn’t. There was no ice age in the Mesozoic Era, although around the Jurassic-Cretaceous boundary glaciers and other ice did form. In the Paleozoic, there were the two which I mentioned, and earth entered one again in the Oligocene Epoch of the Cenozoic Era.

        The Pleistocene glaciations which occurred after the closure of the Inter-American Seaway are just the latest phase of the Cenozoic ice house. It was traditional to think of the Pleistocene as THE ice age, neglecting Antarctica, because of NH bias. Also, when the ice age was first recognized, no one knew that Antarctica was once forested, nor that continents move.

        I don’t know what you mean about 400 Ka. NH glaciations started about 2.6 Ma. Until around a million years ago, glacial phases lasted about 40K years, in synch with our planet’s axial tilt cycle. After that time, since at least 800 Ka, they’ve averaged closer to 100K years, more or less, ie two or three obliquity cycles each.

        It’s correct to speak of the Pleistocene glaciations, but the Antarctic ice sheets define the Cenozoic ice age.

      • I get it now. You think it took 400K years after the closure for the glacial cycle to start.

        No one knows precisely when the seaway became effectively closed, ie oceanic circulation changed as a result of the emergence of the isthmus. It apparently briefly reopened again to some extent around 1.8 Ma.

        The interchange of animals between the Americas is generally dated to around three Ma, but that’s approximate.

        The seaway grew progressively shallower for millions of years, going from a nursery for megalodon sharks to New World monkeys. At what point its shoaling started reinforcing the Gulf Stream hasn’t really been robustly determined yet.


        An ice cap probably formed on the tip of Greenland already in the Pliocene, thanks to more moisture.

      • GM and others,
        Something that hasn’t been mentioned is that there was extensive volcanism during the Oligocene, at least in the Sierra Nevada of North America. A lot of ash was thrown out, enough to basically bury the Eocene landscape. That undoubtedly had a cooling impact on the Northern Hemisphere.

  11. This paper will be widely glorified by the pro-warmers because it is has the word “weathering” in it. They love the sciency sounding words.

    CO2 did not fall until 2 million years after the event. Antarctica feezes over, Earth cools, then CO2 falls (noting that 2 million years is like Eons of time versus just a lag). Then just as Antarctica started warming up again at 27 million years ago, CO2 promptly fell again. Opposite directions again.

    • CO2 and Temps for the last 40 million years.

      And then temperature and geographic changes last 45 million years. Pay attention to the Antarctic glaciation sequence line about how the glaciers changed over the whole period (and note the lack of correlation to any CO2 changes above.)

  12. It is disturbing how grimly some people are clinging to the misguided perception that CO2 has a significant effect on climate. Compelling evidence over the totality of history is that CO2 in the presence of about 35 times as many H2O molecules has no significant effect on climate.

  13. Dan, I agree. If the CO2 reading were ~7000 ppm about 525 MYA, why didn’t
    the temperature run away?

  14. It must be comforting to be a ‘climate scientist’ using the ‘settled science’ to produce absolutely certain scenarios for the politicians to use for mass changes to civilisation.

    It’s no wonder they have to control the peer reviewed literature used by the IPCC so that they do not get confused by all of the studies quoted in the pages of WUWT with their opposite views, contrary data and multiple possible histories.

    In the 2 quoted graphs above, the Miocene has low CO2/high temps and vice versa (note that the time axes are reversed) and yet CO2 is supposedly the cause of the opposite effect (high CO2/high temp).

    How is a layman such as I to believe anything about this ‘science’ in order to vote for politicians who are in control of my way of living?

  15. I agree with Z. Jawrorowski on the problems with ice core readings.
    My findings, however, indicate that CO2 readings in more temperate
    climates are related to the altitude above the ground, wind speed at
    the time they were taken, time of year, and the richness of the topsoil
    over which they were taken.

    In Beck, 2007’s readings in Harvard Forrest, the seasonal and wind
    speed variations are demonstrated. The 29 meters above the surface
    readings with wind indicate that the reading were not of a well mixed
    atmosphere reading. The slower the wind speed, the higher the
    maximum CO2 reading, indicating that the readings were a more
    direct reading of the major source of the CO2.

    If he had read the CO2 at the top of the topsoil, the readings would
    have been much higher.



    • Not reading the whole stream, but this segment, the issue is topoclimatology, where the bottom one inch of the atmosphere at the millimeter level determines radiance with mm altitude, and CO2 flux due to microorganisms. It is also the wind speed at that height, which is nil. I could not imagine trying to use 29 meter data for soil geochemistry. You are actually talking the inter-granular CO2 value in loose soil.

  16. The youngest volcanics that cut off ocean circulation and reorganized it was the separation of the Atlantic and Pacific by volcanism along the Central America ring of fire. The Yucatan is all limestone, upthrust in that action.

  17. This is a fairly reasonable representation of the landmasses and the ocean current patterns just before the initial glaciation event happened (noting of course, that Antarctica is the unlucky planet and has been glaciated over at least 4 other times before the last initiation).

    From Bijl 2010.


    At this time, closer to about 37 million years ago, there was not enough separation from Australia and South America for the Antarctic Circumpolar Current to fully operate. Particularly, there are numerous little cratons between South America and Antarctica (which are still noticeable today on Google Maps for instance), that got jostled around and even rotated between the two continents so that there was just not enough flow for a big current like the ACC.

    What happened is that the ocean gyres from the Pacific and Atlantic and Indian oceans circulated right up the continent and kept it warm enough. Still frigid and dark in the winter, but warm enough in the summers that the snow melted and only the central mountain chains had big glaciers.

    Then 33.6 million years ago, enough separation happened and the Antarctic Circumpolar Current started up (driven by the winds of the polar vortex which operated even then) and suddenly, Antarctica is isolated from the rest of the planet in an extreme polar climate. In less than 100,000 years, the glaciers covered the whole continent right out to the continental shelves going by the sea floor cores.

    But between 27 million years ago and 14 million years ago, those pesky cratons between Antarctica and South America got jostled around again and interrupted the current once again. Antarctica promptly lost half its ice and didn’t really get it all back until the ice ages started up 2.7 million years ago. From 14 million years ago to the ice age start-up they grew at a slow rate but didn’t get all the way out onto the continental shelves where the ice-shelves are today.

  18. Gosh can’t they just run the climate models backwards and find out why? After all so many “climate scientists” tell us how well they do both forwards and backwards…and they don’t even need real data, a computer program is enough!

  19. “My apologies for the different x-axes. I made these graphs several years ago and don’t recall where I saved the Excel files.”

    Are you related to phil Jones?

  20. If there is a causal signal (e.g temperature) and a response signal (e.g. co2), with a significant delay, and with very little modification of its frequency components, then I do not see there can be much feedback from the response back to the cause. In that case, the feedback would become (1/1-GH) where G is the gain and H is the feedback. G(f) and H(f) would be very different for frequencies much smaller and much larger than the delay, and the two signals would become very distorted with respect to each other, but they are not, so there is very little H feedback. So this delay in CO2 levels with respect to temperature suggests to me that temperature has a very low CO2 sensitivity.

    • +1
      I was trying to say something similar with my earlier comment:

      If CO2 really did amplify by positive feedback every warming or cooling, then climate would oscillate monotonically all the time. It does not. Instead it meanders, in the manner of Lorenz (DNF 63) classic nonlinear simulation. Thus the basics of nonlinear chaos dynamics show CO2 not to be a significant climate driver. That of course is why the climate establishment pay merr lip service to chaos-nonlinearity theory while avoiding / ignoring all its clear implications.

  21. The CO2 versus temperature thing. Kind of reminds me of the guy who claimed that wherever he moved his eyes, Pamela Anderson’s boobs followed.

  22. Erm. The paper is well done – it is in English. I have no idea what the news release was written in, but certainly not quite English. I would be embarrassed by the very poor writing exhibited in that piece.

  23. I am interested in the Eocene/Oligocene freefall as a possible inflection point between the Neogene regime where CO2 and temperature are correlated, but CO2 is the slave of temperature; and the prior Phanerozoic regime where CO2 and temperature do not seem related at all.

    • They were related at times in the Paleozoic, Mesozoic and Cenozoic Eras. CO2 fell to near present levels during the Carboniferous-Permian glaciation, and were generally high during the warm Mesozoic and early Cenoszoic.

      In the latter case, there is a causation-correlation problem however. Volcanism associated with the breakup of Pangaea both warmed the seas and released copious quantities of CO2.

      • I suspect you are referring to Royer (2009) Geocarbsulf (cough) VOLC+Solar evolution +CO2 [radiative forcing only] for your conclusion that CO2 was low during the Carbo-Permian glaciation. I believe Scotese has graphics that agree, but I have never seen the sources.

        I mean, rully, Royer’s CO2 series is a litany of compounded uncertainties, adding sulfur, volcanic, our sun’s status as a main sequence star, and supposed but unproven radiative forcing from CO2; to the already low resolution direct CO2 proxy data.

        This is what I get with simple proxies:

    • Estimated average CO2 level over each period (with the mean sometimes masking large swings):

      Cambrian: 4500 ppm (7000 ppm early)
      Ordovician: 4200 ppm (including a glacial phase)
      Silurian: 4500 ppm
      Devonian: 2200 ppm (evolution of large land plants possibly drawing down CO2)
      Carboniferous: 800 ppm
      Permian: 900 ppm
      Triassic: 1750 ppm
      Jurassic: 1950 ppm
      Cretaceous: 1700 ppm
      Paleogene: 500 ppm
      Neogene: 280 ppm (almost certainly too low an estimate).

      The Miocene Epoch (23 to 5 Ma) was by far the longest of the Neogene Period, during which epoch CO2 averaged around 400 ppm. The Pliocene, Pleistocene and Holocene, with generally lower CO2, have lasted only about five million years.

      • Higher Mesozoic CO2 perhaps thanks to break up of Pangaea, although that didn’t real start until the Triassic/Jurassic boundary.

      • Aren’t we still in the Pleistocene era? I was under the impression that Holocene is used to describe our current interglacial, not a new geological era.

  24. Maybe it’s because cold always sinks to the bottom, and Antarctica sits at the “bottom” of the world.😁

Comments are closed.