Climate Science Glossary

Term Lookup

Enter a term in the search box to find its definition.


Use the controls in the far right panel to increase or decrease the number of terms automatically displayed (or to completely turn that feature off).

Term Lookup


All IPCC definitions taken from Climate Change 2007: The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Annex I, Glossary, pp. 941-954. Cambridge University Press.

Home Arguments Software Resources Comments The Consensus Project Translations About Donate

Twitter Facebook YouTube Pinterest

RSS Posts RSS Comments Email Subscribe

Climate's changed before
It's the sun
It's not bad
There is no consensus
It's cooling
Models are unreliable
Temp record is unreliable
Animals and plants can adapt
It hasn't warmed since 1998
Antarctica is gaining ice
View All Arguments...

Keep me logged in
New? Register here
Forgot your password?

Latest Posts


Why does CO2 lag temperature?

Posted on 9 January 2010 by John Cook

Over the last half million years, our climate has experienced long ice ages regularly punctuated by brief warm periods called interglacials. Atmospheric carbon dioxide closely matches the cycle, increasing by around 80 to 100 parts per million as Antarctic temperatures warm up to 10°C. However, when you look closer, CO2 actually lags temperature by around 1000 years. While this result was predicted two decades ago (Lorius 1990), it still surprises and confuses many. Does warming cause CO2 rise or the other way around? In actuality, the answer is both.

Milankovitch cycles: CO2 vs Temperature over past 400,000 years
Figure 1: Vostok ice core records for carbon dioxide concentration (Petit 2000) and temperature change (Barnola 2003).

Interglacials come along approximately every 100,000 years. This is called the Milankovitch cycle, brought on by changes in the Earth's orbit. There are three main changes to the earth's orbit. The shape of the Earth's orbit around the sun (eccentricity) varies between an ellipse to a more circular shape. The earth's axis is tilted relative to the sun at around 23°. This tilt oscillates between 22.5° and 24.5° (obliquity). As the earth spins around it's axis, the axis wobbles from pointing towards the North Star to pointing at the star Vega (precession).

Milankovitch cycles: orbital changes in eccentricity, precession and obliquity
Figure 2: The three main orbital variations. Eccentricity: changes in the shape of the Earth’s orbit.Obliquity: changes in the tilt of the Earth’s rotational axis. Precession: wobbles in the Earth’s rotational axis.

The combined effect of these orbital cycles cause long term changes in the amount of sunlight hitting the earth at different seasons, particularly at high latitudes. For example, around 18,000 years ago, there was an increase in the amount of sunlight hitting the Southern Hemisphere during the southern spring. This lead to retreating Antarctic sea ice and melting glaciers in the Southern Hemisphere.(Shemesh 2002). The ice loss had a positive feedback effect with less ice reflecting sunlight back into space (decreased albedo). This enhanced the warming.

As the Southern Ocean warms, the solubility of CO2 in water falls (Martin 2005). This causes the oceans to give up more CO2, emitting it into the atmosphere. The exact mechanism of how the deep ocean gives up its CO2 is not fully understood but believed to be related to vertical ocean mixing (Toggweiler 1999). The process takes around 800 to 1000 years, so CO2 levels are observed to rise around 1000 years after the initial warming (Monnin 2001, Mudelsee 2001).

The outgassing of CO2 from the ocean has several effects. The increased CO2 in the atmosphere amplifies the original warming. The relatively weak forcing from Milankovitch cycles is insufficient to cause the dramatic temperature change taking our climate out of an ice age (this period is called a deglaciation). However, the amplifying effect of CO2 is consistent with the observed warming.

CO2 from the Southern Ocean also mixes through the atmosphere, spreading the warming north (Cuffey 2001). Tropical marine sediments record warming in the tropics around 1000 years after Antarctic warming, around the same time as the CO2 rise (Stott 2007). Ice cores in Greenland find that warming in the Northern Hemisphere lags the Antarctic CO2 rise (Caillon 2003).

To claim that the CO2 lag disproves the warming effect of CO2 displays a lack of understanding of the processes that drive Milankovitch cycles. A review of the peer reviewed research into past periods of deglaciation tells us several things:

  • Deglaciation is not initiated by CO2 but by orbital cycles
  • CO2 amplifies the warming which cannot be explained by orbital cycles alone
  • CO2 spreads warming throughout the planet

Many thanks to Ari Jokimäki who has been tirelessly tracking down papers on Milankovitch cycles.

0 0

Bookmark and Share Printable Version  |  Link to this page | Repost this Article Repost This


Prev  1  2  

Comments 51 to 65 out of 65:

  1. Not sure if I'm too late to get a response, but I have another question regarding the lag. I understand that cause for the lag at the initiation of warming, but it appears that in every case, as the climate shifts back to a glacial period, the temperature drops while the CO2 is still near it's peak concentration, with the temperature starting to fall very abruptly before the CO2 even begins to decline. Is there any explanation for that?
    0 0
  2. exp137#51:

    Read through the original post, especially the mechanism given after Figure 2. The temperature change is initiated by orbital variation and reinforced by atmospheric gas concentrations.

    It's important to understand the time frame when you say "temperature starting to fall very abruptly". That is really not the case: glacial onsets are typically much slower than terminations (see Figure 1 and note the time scale). Terminations can take 100s-1000s of years; cooling to full glacial is much slower. The temperature graphs take on a 'sawtooth' shape.

    0 0
  3. muoncounter#52:

    I've read the article, but it focuses exclusively on the timing of the termination as opposed to the onset. I'm very comfortable with the reasoning behind the lag at the termination, but still curious about the lag at the onset. If you look at the last interglacial period (before the current one) in Fig. 1, the temperature has dropped rather significantly while the CO2 concentrations are still largely above 270ppm. While the cause of the glacial onset may indeed be orbital forcings, does this really mean that the imfluence of CO2 on the glacial-inerglacial cycle is essentially neglegible? Or is it that the high levels of CO2 essentially "soften the blow" of the orbital forcings, by slowing the onset of the next glacial period? Please don't assume I'm a skeptic or unfamiliar with the basics just because I'm asking questions, as I am currently a graduate student studying paleontology. This is an argument I've heard from skeptics in the past and I'd like to find a solid explanation, if one is known.
    0 0
  4. exp137 - if two of the forcings go negative while GHG remain stable, then why would you expect temperature not to drop?
    Explicitly, when onset occurs, direct insolation from sun is dropping in NH (cooling), while CO2 concentration are possibly still rising - but very slowly as there is a considerable lag. On the other hand, ice albedo is a rapid feedback and the two together drop NH temperatures, freeze the methane sources, cool northern oceans which in turn leads eventually to lower CO2 and SH cooling as well.
    0 0
  5. scaddenp: I don't expect the temperatures to not drop at all, it's just that in the case of glacial cycles there doesn't appear to be much input from CO2 at all. This is a frequently used case by climate change skeptics that CO2 is not a climatic driver (or a very weak one at best) but instead simply follows temperature. I'm familiar with other cases in which CO2 does appear to drive temperature, such as the PETM and the Ordovician glaciation, but I just would like to understand this case better. I guess what I'm really looking for is a model that shows the relative influences of both insolation and CO2 during the Pleistocene glaciations. Because in this case, at least, it's obvious that changing solar input is the primary driver, but the role of CO2 isn't very clear.
    0 0
  6. exp137,
    ...there doesn't appear to be much input from CO2 at all.
    No? What about the fact that the glacial onset takes tens of thousands of years longer than a termination, because the orbital forcings/change in albedo are fighting against rather than in concert with CO2 changes?

    Another factor, of course, is one of pure area... on a sphere (the earth), there is more area at a latitude closer to the equator than a pole, and the sunlight is more direct and seasonally constant towards the equator, with the upshot that albedo changes are stronger when the ice sheet is further south.

    This in turn means that when the ice begins to retreat (during a termination), the strength of change in albedo is greater and faster at the beginning of the change than at the end as compared to the reverse, the growth of ice from the pole southward, when the growth (in area) is slow at first with a greater affect near the end.

    A model for the relative influences cannot possibly be as simple as you'd like, but one problem is exactly that you are eyeballing graphs and using a rule-of-thumb approach to the problem. It's easy to come up with any inference that appeals to you with that approach.

    To go over the glacial onset... orbital forcings cause summers to be shorter and cooler. This leads to a very, very slow build up of the snow and ice extent in the northern hemisphere. Each year, the snow melts not quite as far back as it did the winter before.

    Each time this happens, the total amount of sunlight reflected back to space without warming the planet is greater.

    At first, the difference is small, because the change happens at a high latitude. The total change in area is small, and the angle of incidence and W/m2 is less because of how the light strikes the earth at the poles.

    As the ice sheets extend further south, the effect becomes more and more severe.

    As temperatures drop due to this forcing, the amount of CO2 (and H2O) in the atmosphere falls, which causes temperatures to drop even further.

    There is no inconsistency in this.
    0 0
  7. exp137, which CO2 does appear to drive temperature...
    This is a bad way to look at things. Talk of a "driver" is a trap, or rather a denier's canard.

    In a natural climate there are very few true forcings. Orbital forcings are one. Change in solar output is another. A massive injection of CO2 due to, for example, volcanic or anthropogenic activity are also true forcings. A massive injection of dimming aerosols due to volcanic activity or a huge asteroid strike is yet another.

    These can be true "drivers."

    Outside of that, almost all of the possible influencing factors are interlocked. Atmospheric H2O rises/falls quickly with temperature changes, and drags temperatures further up/down. CO2 rises/falls similarly but less quickly, also dragging temperatures further up/down. Changes in temperatures can contract/expand ice sheets, which reduce/expand albedo, which drag temperatures further up/down.

    Virtually everything in the normal ebb and flow of climate is an interlinked feedback. Actual drivers vary, but the definition of the "driver" is more complex than simply "is this a driver?" The true "driver" somehow changed due do an event outside of the normal climate system... a quirk in the slow change in orbit, a massive flurry of volcanic activity, who-knows-what in the sun, or... a fossil-fuel based civilization.

    Albedo (snow/ice) is never a "driver"... orbital forcings cause that change, although orbital forcings don't themselves directly influence temperature, but rather the effects on snow/ice extent do so.

    CO2 is a "driver" only when it is "un-naturally" changed (volcanoes, humans). Otherwise it is a feedback.

    But the reason for the change in no way changes the efficacy of the forcing. If a car runs you over, it doesn't matter if someone was behind the wheel driving it or not, it still has the same effect.
    0 0
  8. My understanding of the role of CO2 in Pleistocene glaciations is twofold.
    1/ One it is a slow feedback amplifying effect of solar (which would be too weak for the scale of temperature change by itself).
    2/ It is part of the mechanism by which changes in NH also causes warming/cooling in the SH.

    Note that albedo is also major player in the feedback the amplifies the solar change. The relative importance of the forcings are discussed in Ch 6 of the IPCC WG1 report.
    0 0
  9. exp137,

    I should clarify something. I rather inadvisedly said "fighting against rather than in concert with CO2 changes." That didn't come out as I intended, and is incorrect. CO2 is a feedback that enhances the temperature drop. The point I intended to make was rather that because CO2 levels are high, the high temperatures that result help to hold back the expansion of the ice sheets. At the same time, because of the logarithmic nature of CO2, a "draw down" of CO2 levels and temperatures is much harder to enact than an increase.

    One must get more CO2 out of the atmosphere, at the start of glaciation, to get the same change in temperature. As such the effects of CO2 changes at the initial stages of glaciation are small/sluggish, the opposite of the result of changes in CO2 at the start of a glacial termination.

    This makes it look like CO2 provides a bigger "kick" to a glacial termination (it does, at the start, as opposed to glaciation where that kick is harder at the end).
    0 0
  10. exp137,

    Okay, one more point. I'm actually really glad that you brought this up, because it highlights a very, very important aspect of CO2, and that is that because of the logarithmic nature (i.e. double CO2 to get the same incremental increase in temperature, so 2x = 1, 4x = 2, 8x = 3, and so on) the effects of increasing CO2 are more evident when you raise it from a low value, and less evident when you first drop it from a high value.

    This has huge implications on the idea of adding it, and somehow later trying to reduce it and return temperatures to normal.

    What this inevitably means is that it is easier to raise temperatures by increasing CO2 than it is to drop them by decreasing CO2.

    As such, the impact of CO2 at the start of a glaciation is far less evident than the impact on the start of a glacial termination.

    Also... it is going to be a whole lot easier for man to raise temperatures by increasing CO2 than to draw them down by extracting it. The bang from the first X gigatons added to the atmosphere is far greater than the bang from the first X gigatons somehow subtracted (assuming we can come up with a cost effective way of doing so).
    0 0
  11. Sphaerica - You have the makings of a post/rebuttal there. Maybe a few graphics to explain how it fits together. I only mention this because it is a legitimate area of confusion, and perhaps we haven't adequately explained the process.
    0 0
  12. Thanks for the responses everyone. This was seriously informative, and needed.

    I agree with Rob Painting here, that it would be great to put this together into a post to fill the gap left by focusing mainly on the lag on the warming end of the cycle. From what I've seen, every time you leave something not explicitly explained, a skeptic will use this as a point of attack against the entire concept.

    Again, thank you very much for your quick and detailed responses.
    0 0
  13. "every time you leave something not explicitly explained, a skeptic will use this as a point of attack against the entire concept"
    Those aren't skeptics. Those are fake-skeptics.
    0 0
  14. It turns out the mechanisms for glacial onset are still uncertain.

    Timmerman et al 2010:

    Compared to the rapid glacial terminations, the buildup of glacial ice sheets in the Northern Hemisphere took tens of thousands of years. During the buildup phase, the growing ice sheets were subject to major orbitally induced summer insolation changes, without experiencing complete disintegration. The reason for this behavior still remains elusive.

    But why another lag post? See the existing 'CO2 lags temperature' rebuttals; both basic and intermediate are quite good - they generated 330 comments.
    0 0
  15. Muon - Sphaerica's comments helped clear exp137's confusion, whereas our current rebuttals did not. There is almost always room for improvement, and that's no slight on all our current rebuttals.

    I don't think that the number of comments is of that much value either - sometimes specific topics are contentious because they decimate well-worn fake-skeptic myths and said fake-skeptics swarm to defend it. The more valuable rebuttals, to my mind, are those that so thoroughly discredit a myth it never really gets going in the first place But YMMV.
    0 0

Prev  1  2  

You need to be logged in to post a comment. Login via the left margin or if you're new, register here.

The Consensus Project Website



(free to republish)



The Scientific Guide to
Global Warming Skepticism

Smartphone Apps


© Copyright 2014 John Cook
Home | Links | Translations | About Us | Contact Us