Why is the concentration of carbon dioxide in the atmosphere increasing?

To get a feel for this there is a need to set a temporal context. Homo sapiens appeared from about 200,000 years ago and it really does not matter about what the climate was prior to this time since it is highly unlikely that Homo sapiens had any impact on the climate whatsoever.   However, since ice core gas studies have been taken back to about 800,000 years before present (BP), it is useful to use this data to set a temporal context equal to this period since it reveals the natural cycles that perturb the Earth’s orbit prior to and during Homo sapiens history.

Whilst ice core studies have been proceeding for some time, a collaborative paper by Lüthi et al. published in Nature in 2008 will be used to establish the temporal context used in this blog. This ice core study pushed the temporal limit back to 800,000 years.  Other studies have been done more recently that are trying to get back to 1 million years ago.  However, for the purposes of this blog, the duration period used in the 2008 Nature paper is adequate.

The paper clearly shows that the carbon dioxide concentration has been cycling between about 180 and 280 ppm (parts-per-million) every ~40,000 years. This behaviour is caused by a cyclic wobble of the Earth’s orbit and is called the Milankovitch Cycle. So in a trough of the cycle the Earth is colder than at the peak of the cycle.  As the Earth wobbles away from the trough towards the peak it exposes a different part of its surface to the Sun which causes a shift in the energy balance. This change in the energy balance initiates a heating cycle, the increase in temperature takes a little time (a lag) to work but it eventually perturbs the carbon cycle, which causes more carbon dioxide to be released from oceans and Arctic and Antarctic regions and mountains etc., which in turn enhances the natural greenhouse effect by the carbon dioxide released which causes positive feedback, which further increases the carbon dioxide concentration and so on, until the wobble has finished. At this time things settle down and a new, carbon cycle equilibrium is established (~280 ppm) at the peak of the cycle until the wobble starts to head back to the trough (~180 ppm).  As described by NASA, the carbon cycle, via the greenhouse effect, is essentially the thermostat that controls the Earth’s average global tempertaure.

The Earth is currently at or near the peak of a Milankovitch cycle which means if natural processes had continued as per normal over the last 800,000 years we should be measuring carbon dioxide in the atmosphere at very close to 280 ppm.  This is not the case since it is now over 400 ppm.  Why is this so?

The first hint is to look a at historical measurements and it can be readily seen that the carbon dioxide concentration started to increase above 280 ppm near the start of the 19th century somewhere between 1770 and 1800.  A period which is historically associated with the start of the industrial revolution. Since carbon dioxide is a greenhouse gas (GHG), its natural attribute is to absorb infrared radiation and many people have used the positive correlation of the average global temperature increase with the increase in carbon dioxide to conclude that it is the increase in carbon dioxide that it is causing the average global temperature to increase.  This is a logical argument but it is supportive evidence and is not conclusive.

To understand that the increase in carbon dioxide is caused by fossil fuel combustion we need to consider other possible sources.  The most popular candidate is emissions from volcanoes. Volcanoes do produce carbon dioxide but according to the BGS (2005) and the USGS (2011) they only represent about 1% of the total emissions produced by humans. Also, if volcanoes were playing a role then the graph of the observed increase in carbon dioxide would be very spikey just after a large eruption, especially if a volcano produced more carbon dioxide than produced by humans in their entire history, and this is certainly not the case. Volcanic eruptions can produce small cooling effects which lasts no more than a few years. Remember we are now considering the last 200 years or so of the peak of a Milankovitch Cycle which should have carbon dioxide at 280 ppm, which was the case just prior to the start of the industrial revolution.

So, the next step is to examine the nature of carbon dioxide. It is made up of 1 atom of carbon and 2 atoms of oxygen.  Carbon occurs naturally as the isotopes carbon-12 (99%), carbon-13 (~1%) and as a trace of carbon-14.  The latter is made in the atmosphere by the action of external radiation (Cosmic Rays etc.) on nitrogen in the atmosphere.  Carbon-14 is radioactive and has a half life of 5730 years, in other words it decays. So, if you burn fresh carbon you get carbon dioxide which is composed of a mixture of carbon-12 dioxide, carbon-13 dioxide and carbon-14 dioxide.  If you put the latter in an isolated, closed jar for 5730 years you would only have half of it left.  This is why one can use carbon-14 for historical dating but it is limited since after 50,000 years or so it cannot be measured since there is not enough left to detect.

So, the other factor which is often overlooked is the ever decreasing carbon-14 to carbon-12 ratio in carbon dioxide in the atmosphere, even though the concentration of carbon dioxide is increasing. The carbon-14 content of fossil fuels has long ago been lost because fossil fuels were laid down 250+ million years ago and we know that carbon-14 only has a half-life of 5730 years which means fossil fuels contain NIL carbon-14. The only source of carbon dioxide that is large enough to produce the observed dilution or Suess Effect is the combustion of fossil fuels by humans. Recall, that volcanoes do produce carbon dioxide but their GHG emissions only represent about 1% of the total emissions produced by humans.

The same isotope analysis can also be applied to the observed decreasing carbon-13 to carbon-12 ratio. It is not that plants are lazy but they are energy sensitive and have a preference for fixing carbon-12 over carbon-13 since the former is not as heavy as the latter.  It is not a large difference but what it means is that if you were to compare the carbon-13 to carbon-12 ratio in air immediately surrounding the plant to that carbon fixed in the plant you would find that in the plant, the ratio in the plant would be less than in the air.  So, since this also applies to fossil fuels since they are made from plants, carbon dioxide formed from burning fossil fuels is deficient in carbon-13.

Measurements of the carbon-13 to carbon-12 ratio in air are also decreasing over time which is further evidence that the carbon dioxide in the air is being diluted by carbon dioxide formed as fossil fuels are burnt. There is no other source of carbon dioxide that can cause the observed dilution or, Suess Effect. More recent studies by the Scripps Research Institute and their international collaborators also show that the global oxygen concentration is also decreasing by a small amount equivalent to that used to form carbon dioxide by the combustion of fossil fuels. This decrease in oxygen can only occur as a result of combustion. It is not dangerous, but it is just that a small change in a large background it is now measurable with good precision by modern instrumentation and the findings supplement the carbon isotope studies.

So the logical conclusion that can be made from the known mass of emissions and the carbon isotope studies is that the increase in the atmospheric concentration of carbon dioxide observed ABOVE that expected for a Milankovitch maximum of about 280 ppm after the period 1770 to1800 along with the observed decrease in global oxygen is caused by the combustion of fossil fuels.

Retired from the CSIRO Energy in August 2014 where I was manager of the Smog Chamber Laboratory in Sydney. My research areas were smog and ozone formation as well as real world vehicle emission effects on the atmosphere. Now a casual academic and course convener at UNSW Sydney in the area of managing greenhouse gas emissions.


  1. Jane Hilton 1 month ago

    Thanks for this explanation Dennys. Very helpful and interesting!

  2. Jim Lawless 1 month ago

    For completeness it would be useful to also comment on the 12/13C ratio of volcanic emissions: I don’t have that information to hand but it will be quite different to that from fossil fuels.

  3. Author
    Dennys Angove 1 month ago

    Jim, fair comment: I have seen modelling results of volcanic carbon isotopic studies which in my mind was consistent with the conlcusion given in the blog. However, I can’t find the paper but when I do, which I will, I will cite it here because it is interesting. Given that volcanic emissions have been concluded (USGS and BGS) to be about 1% of human carbon dioxide emissions, and that the increase in carbon dioixde emissions have been measured to rise above ~280 ppm from 1800 or so and that the monthly rise is realtively smooth apart from the vegetation cycles I am comfortable with the conclusion. As always in science (and politics) there are opposing views and/or differences in view which is what makes it exciting. I am relying on conclusions reached by the USGS researcher Gerlach who published in Eos, Vol. 92, No. 24, 14 June 2011 and the report by the British Geological Survey published in 2005 titled “Volcanic Contributions to the Global Carbon”. The other evidence which strongly supports the conclusion in the blog is the measured decrease in oxygen concentration which can be obtained from http://scrippso2.ucsd.edu.

  4. Dennys Angove 4 weeks ago

    Jim, I still have not been able to find the paper I have previously referred to. However, this link is to a nice measurement study using newish techniques. The text is meaty but there is some worthwhile reference material therein. It is available on-line and the conclusion is understated given the extent of the study in my view but is helpful.

Leave a reply

Your email address will not be published.