How long does carbon dioxide remain in the atmosphere?

I have remarked before that ‘climate change’ is full of unknowns, about which people often make extraordinarily confident assertions. A correspondent reported one such the other day — he had been listening to a lecturer in the UK who said that it would take ‘a million years for silicate weathering to remove the extra CO2 being put into the atmosphere’. He (the listener) thought that an extraordinary claim, and asked for comment. I think it is extraordinary too, and even Wikipedia, which has a definite tendency towards AGW orthodoxy, hums and haws a bit, before saying that ‘[t]he atmospheric lifetime of CO2 is estimated of the order of 30–95 years’. That is plainly a lot less.

Why is it important? Well, the AGW hypothesis says that adding carbon dioxide to the atmosphere must increase global temperature, which must be bad for everything and everybody. Hidden in that assertion is the notion that once a carbon dioxide molecule is added, it just stays up there. But  moment’s thought indicates that since CO2 is the fuel for all plants, more CO2 must be good news to the plant world, which might use the added gas as a sort of dessert. How much do the plants take, anyway? There are lots of estimates, and no one seems to be really sure. Then there’s the ocean, which releases carbon dioxide as it warms.

But back to the thousand-years claim. I know where that comes from (more in a moment), but it stimulated Tom Segalstad in Norway to write a long email on the subject, which I am summarising here, because it provided me with both the history and the sources.

Before the establishment of the IPCC the conventional estimate of CO2 residence time was  accepted to be five years. When that short time-span didn’t seem to fit what the early models projected, Tom says, ‘[t]he IPCC next constructed an “artificial” residence time for atmospheric CO2 to fit their model, of 50-200 years (IPCC 1990, Table 1.1)’. By and large the IPCC has stuck to this rather generously wide estimate ever since.

One of the leading figures in the IPCC, Susan Solomon, has argued for a thousand-year residence time, but her paper starts with a model, and ignores observations. It is a good example of the ‘scary’ AGW paper that assumes that it is completely right, despite the lack of supporting evidence. Moreover, if carbon dioxide had a life of a thousand years in the atmosphere, would we still be able to make carbonated drinks? Would plants to able to use CO2 so efficiently in photosynthesis? Tom points out, in any case, that ‘the measured rise in the atmospheric CO2 level is just half of that expected from the amount of anthropogenic CO2 supplied to the atmosphere’. In short, something is chewing up the carbon dioxide, and that makes any estimate of a thousand-year residence somewhat mysterious, if not head-scratching.

It seems as though the IPCC has a static model of the atmosphere, in which carbon dioxide is in some sort of natural balance, with the gas moving in and out of the ocean, the atmosphere and the ecosystem — an equilibrium model. When humans add more CO2 the system goes out of balance, which is bad. Humans must stop doing that, so that the system can return to equilibrium. You’ve heard this before, you say. Yes, so have I.

Tom has his own papers on the subject, and one of them argues that carbon isotope analysis supports a five-year residence time (yes, perhaps that early estimate was a good one), and that all the human-produced fossil-fuel carbon dioxide is only 4 per cent of the whole, the remaining 96 per cent having come from what was there last time we looked, outgassing from volcanoes and the oceans. Do you wonder about the volcanoes? Well, about 1500 are known to have been active on land in the last 10,000 years, and there are thousands under the seas. CO2 comes out from major fractures and dormant volcanoes, too.

Another paper of his lists nearly 40 articles, all of them between the 1950s and the 1990s, that calculated the residence time of carbon dioxide in the atmosphere; the range was from about one year to seven years. But none of these, as we know, fits with the modelling carried out for the purposes of the IPCC reports, and they have been discarded.

Does it matter? The current lack of significant warming, more than 17 years in the case of the RSS dataset and of varying smaller lengths in all of the other global temperature datasets, tells us that CO2 cannot be the principal driver of temperature. What is more, the apparent greening of the globe that has occurred in the last decades suggests that more carbon dioxide in the atmosphere is good for plants, and by extension, for us — at least so far.

But I learned a lot from reading these papers, and considering the argument again. For the moment, it seems to me that we know much less than the orthodox think we do about what happens when CO2 is added to the atmosphere, and that the outcome may actually be a beneficial one.



  • David says:


    It is great that you took, “[b]ut a moment’s thought” however if you had taken a second moment you would realize that a

    ‘a million years for silicate weathering to remove the extra CO2 being put into the atmosphere’ refers to a net change in CO2.


    “[t]he atmospheric lifetime of CO2 estimate of 30–95 years refers to a rate of replacement.

    Now I don’t know if either of these two statements is correct, but in my view I don’t think your “insight” that the latter is “plainly a lot less” than the former, is in any way useful.

    Apples and oranges!

  • BoyfromTottenham says:

    Actually chaps, that reference to silicate weathering sounds like a typical bit of cunning pro-AGW misdirection. So yes, David, it is “apples and oranges”, but also a very clever distraction, because there are probably hundreds of ways that CO2 gets added and subtracted from the atmosphere, and by choosing to talk about one which “takes a million years”, the lecturer hopes the listener will ignore the rest, which includes of course plants which can and do absorb vast amounts of the trace gas CO2 almost instantaneously and continuously, and the oceans, etc., etc., etc.

    • David says:

      “Instantaneously? In the space of 250 years C02 concentration has nearly doubled to 400ppm, rising to its highest levels in 2.1 million years. Not much sign of re-absorption, yet.

  • John Morland says:

    I have heard of many claims of years CO2 remains in the atmosphere, anywhere between thousand to about 10 years. A million years sounds an extraordinary claim.

    If you look at the rise of CO2 levels from the Mona Loa observatory you will find the upward curve is not smooth, but bumpy. This is caused by deciduous trees sprouting new leaves in the northern spring; not only reducing the rate of increase, but can reduce the atmospheric CO2 concentration (albeit for only a short time). So in a few weeks, despite mankind poring millions of tonnes of CO2 a year, the CO2 concentration dips, then steeply rises again during the northern autumn and winter. That response does not look like a million years of CO2 molecules hanging around in the atmosphere, or even thousand years..

    Just think, I walk in a park breathe out CO2, within minutes many (if not most) of the CO2 molecules I exhale will be captured by the tree leaves above me or by the grass near me. Of course some CO2 molecules would not be captured and be blown above the trees and into the wider atmosphere where they may stay for some time.

    Perhaps a better way of describing this would be in terms of a half-life.

    • Mike O'Ceirin says:

      John I have not read an explain of the saw tooth (bumpy) look before guess I did not question it. One thing though if 3% of CO2 is anthropogenic do you know how what part this is of the variation? I suppose I am asking what percentage is the variation.

  • BoyfromTottenam says:

    David – yes, of course plants consume CO2 “instantaneously” – meaning minute by minute throughout the sunshine hours of each day – its how plants grow! By implying that CO2 is not absorbed in this way by plants you are either denying that there are multiple sources and sinks of CO2 as I state (which is simply the use of reductio ad absurdum) , or you are deliberately copying the misdirection of which I spoke. Which is it?

    • David says:

      Here is what you said
      “…which includes of course plants which can and do absorb vast amounts of the trace gas CO2 almost instantaneously and continuously, and the oceans, etc., etc., etc”
      Define “vast amounts”. Obviously even “vaster” amounts of CO2 remain unabsorbed, that’s why CO2 levels to have doubled in the last 250 years.
      How many beans makes six?

  • Robert Holmes says:

    The IPCC itself reports 4 years as the residence time.

    The are papers on this, for example Essinghigh, 2009;

    Potential Dependence of Global Warming on the Residence Time (RT) being 5-16 years in the “Atmosphere of Anthropogenically Sourced Carbon Dioxide”.

    and work by Humlum Stordahl & Solheim also points to a very short residence time; n+atmospheric+carbon+dioxide+and+global+temperature&btnG=&as_sdt=1%2C5 &as_sdtp=

  • David Ellard says:

    Hello Don,

    The Tom Segalstad paper is interesting and reflects the author’s background in geology but it nonetheless contains a basic misunderstanding.

    Segalstad is right to claim that the atmospheric residence timescale of CO2 is about 5 years. He rightly says that this result is confirmed by e.g. measurements of 14C isotopes after the atom bomb tests of the 20th century.

    However he is wrong to think that the same timescale applies when we consider how long it takes to remove an excess QUANTITY of CO2 that has been placed in the atmosphere.

    The reasons are very complicated but to try to summarise: There are two exchange mechanisms for atmospheric CO2, broadly (1) exchange with the oceans (2) exchange with the biosphere (photosynthesis and respiration).

    (1) There is 50 times as much CO2 in the oceans as in the atmosphere. All of this CO2 can exchange with the atmospheric CO2 but only 10% of it can take part in the Henry’s Law equilibration which determines the balance of MASS of CO2 between ocean and atmosphere.

    (2) The rate of photosynthesis in plants and algae is first order in the concentration of CO2 in the atmosphere but the photosynthetic rate is constrained by the amount of available CO2 only for about 1/3 of production.

    In fact the ‘reaction timescale’ for the AMOUNT of CO2 to equilibrate with the ocean and biosphere is probably about 20 years, which is longer than the atmospheric residence timescale.

    • donaitkin says:


      I’ll pass your comment on to Tom.

      • donaitkin says:

        And Tom (Segelstad) has responded as follows:

        ‘Thanks for sending me this comment.

        There are different kinds of CO2 molecules, which have different atomic weight, depending on what kind of isotopes of carbon (12C, 13C, 14C) and oxygen (16O, 17O, 18O) they contain.

        The reaction speed will be different for the different CO2 molecules.

        The most common CO2 molecule is the one containing one 12C- and two 16O-isotopes.

        This most common CO2 molecule is the one that reacts the fastest, and which is mainly involved in the biological processes (selecting the 12C isotope).

        Many different studies have established that the average lifetime (half-life) of atmospheric CO2 is about 5 years (see Segalstad 1998). This average includes different sub-processes with other life-times, and also includes different life-times for different layers of the atmosphere.

        The latter is treated by, for instance, Rohde (1992), who finds that the life-time for CO2 in the lowest 1 kilometer of the Earth’s atmosphere is approximately 1 hour, and approximately 1 month in the lower 10 kilometers of the Earth’s atmosphere. In the Earth’s oceans the CO2 life-time is approximately 10 hours in the upper 50 meters; and approximately 2.5 years for the upper 1 kilometer.

        All this shows that the circulation of CO2 is quite fast both in the atmosphere and in the ocean.

        Does this answer the comment?


        Rohde, H. 1992: Modeling biogeochemical cycles. In: Butcher, S.S., Charlson, R.J., Orians, G.H. & Wolfe, G.V. (Eds.): Global Biogeochemical Cycles. Academic Press, London, pp. 55-72.

        Segalstad, T.V. 1998: Carbon cycle modelling and the residence time of natural and anthropogenic atmospheric CO2: on the construction of the “Greenhouse Effect Global Warming” dogma. In: Bate, R. (Ed.): Global Warming: the Continuing Debate. ESEF, Cambridge, U.K. (ISBN 0952773422), pp. 184-219.

Leave a Reply

Your email address will not be published.