Is the Rate of CO2 Growth Slowing or Speeding Up?


The rate of growth in carbon dioxide concentrations in the atmosphere has accelerated since the beginnings of the Keeling Curve.  The rate has gone from about 0.75 parts per million (ppm)/yr in 1959 to about 2.25ppm/yr today. The attached graphic is from the Intergovernmental Panel on Climate Change Fifth Assessment Report and is based on data gathered by the Scripps CO2 Group and NOAA. The lower image shows the growth rate versus time, which has consistently stayed in positive territory since the late 1950s. No surprise here, because Mauna Loa record shows a clear rise from year to year. But this plot also shows that the growth rate (while staying positive) does vary quite a bit from year to year, something that is less evident in the Mauna Loa record. These variations are due to a combination of natural and human factors.

29 thoughts on “Is the Rate of CO2 Growth Slowing or Speeding Up?”

    1. It would appear to me that the -rise-of-China hasn’t been reflected in this graphic: could it be the mixing of the atmosphere takes so long(say, over ten years or more?!?) to register such a phenomena as the economic-activity-rise-of-China?
      (Of course I looked at the derivative-sinusoidal-graph-type-thing at inset but lack the knowledge to work it out intelligibly.)(I will keep trying …)
      Of course, this is a great page for us mere citizens to ponder and reflect on the important things in life so THANKS FOR ALL THE EFFORT THAT GOES INTO PRODUCING IT!

      Apparently the oceans take a long time to mix so…

    2. Has anyone heard of an in situ ice analysis boring head in use when studing ice cores?

      Something like this?

      It would seem that rather than drill and try to recover ice cores, just analyze the water as it’s generated by a heating head on a analysis device.

      1. A response from Scripps geoscientist Jeff Severinghaus:

        The idea of analyzing the water for isotopes from melting a stick of ice is indeed attractive.
        It is being done in several labs currently, among them the McConnell lab at the Desert Research Institute,
        Reno, Nevada. Gases and 34 elements are measured too, all on the same meltwater stream.

        Doing this onboard a melting head that penetrates downward into the ice is harder. The problem is energy –
        it is hard to get enough energy down thousands of feet of wire to melt the ice. So-called Philberth Probes
        were tried extensively during the 1950s and 1960s as a possible alternative to ice coring, and most were
        unsuccessful. It seems the most successful recent versions of this have used a fiber optic cable with laser
        light carrying the energy, and the fiber is spooled out behind the probe as it melts its way down into the
        ice sheet. The water in the hole behind the probe can then freeze, eliminating the energy-consuming need
        to keep the whole hole unfrozen.

        A short-range version of this concept, called “IceMole”, has been successful but it can only go a few tens
        of meters.

        I hope this is helpful.

        Jeff Severinghaus
        Professor of Geosciences
        Scripps Institution of Oceanography

  1. For the last six weeks CO2 levels seem to have leveled at about 400 ppmv. The same thing looks to have happened for the same period last year at about 398 ppmv. Am I seeing something that isn’t there or does this happen every year? If it is a recurring phenomenon, what causes it?

    1. Hi Larry,

      Here’s a response from Ralph Keeling: This pattern does tend to occur most years. The rising part of the cycle is mostly due to release of CO2 from land ecosystems in the fall and spring, with the winter being somewhat dormant by comparison. We are now in the winter plateau.

      1. That’s interesting. So during this winter plateau we can see the more constant contribution of fossil fuel combustion as CO2 increases a couple of tenths of a percent or so over the last six weeks. I understand that may be over reading the graph, but I think I see a slight trend.

        Thanks for taking my question.

          1. I’ve noticed this “winter plateau” as well. It does not happen every year, but it was pronounced in 2011, 2014 and again this year. It isn’t really adequate to just dismiss it as somehow related to the fact that it’s winter in the northern hemisphere – winter came as usual in 2012 and 13 but there was no plateau.

            I think it is plausible that with increased irrigation and farming of what were previously desert areas in the southern hemisphere, we are beginning to see a southern summer CO2 pulldown that still cannot rival that of the north, but makes itself seen when the northern winter is especially cold, slowing decompositional CO2 generation.

            Any thoughts about that?

  2. It is interesting to see others noted the plateau too. I agree with Kirk that “just” a winter hiatus from the Northern hemisphere doesn’t do it. I was wondering myself, could this maybe relate to the weather in Australia? I remember 2011 was one of the wettest years on record. This could maybe drive the growth of grass in dry areas and probably take up a lot of CO2, couldn’t it?

  3. I have also been intrigued by this “winter plateau” in the ML data (so thanks to Larry for raising it), but have thus far failed to find a reasonable explanation in any publication. As Kirk says, in many years it is not very pronounced at all (essentially absent), but in others it is very obvious. The data for early 2008 are a particularly good example of the “slowdown” in growth rate in Jan-Mar, whereas winter 2010 shows no hint of a “slowdown”.

    Like Kirk, I find the explanantion given by Rob/Ralph to be in need of some additional back-up. Perhaps there is a published paper to which we can refer?

    Regarding the title of the post, it seems that the evidence is becoming stronger that the rate of growth has been more or less the same for 10+ years, i.e. it is neither slowing down or speeding up.

  4. My thought (and it is just a best guess) is that the expected variations in southern hemisphere irrigation is a reasonable factor, but would not be large enough to explain the late-winter CO2 plateau at stations like Mauna Loa.

    The plateau likely has a lot more to do with plant physiology and soil microbe physiology, and is thus related to temperature set-points that kick in or turn off physiological processes. I.e., at a certain temperature land-plants leaf out, and the photosynthetic process is jump-started (and CO2 is intensively sucked out of the air). Likewise, soil microbes have temperature set-points where they kick in or turn off the decomposition processes. So, by this logic, the annual upward inflection happens when the CO2 removal is partially shut down (by frost-destruction of leaves), yet soil temperatures are still elevated so that microbes are very actively converting plant mass into CO2. As things cool down further into Winter, at some point the microbes also go dormant, and the CO2 plateaus. Then, as soil temperatures rise enough, microbial process resumes, causing CO2 to resume a climb for a few weeks (the second, post-plateau climb, ahead of the leafing process). Once leafing erupts in earnest, the powerful benefits of land-plant photosynthesis kick in, and we see the annual dive of the Keeling Curve.

    Reasonably, given that summer heat/drought extremes also shut down leafy land-plant photosynthesisn (through dormancy, or even plant death), we can expect to see a mid/late-Summer plateau on the downward CO2 curve. As AGW worsens, we can expect to see this plateau intensify. Seemingly, in terms of how plants compete with one another, long-season crops/plants will be placed at a disadvantage relative to short-season crops/plants.

    My view on this derives from a lifetime of home-gardening and studying my local forest. Land plants are quite sensitive, demanding relatively narrow ranges for temperature and moisture. Variations through these ranges will toggle processes on/off; rapid variations will weaken and eventually kill.

  5. So is that it?! I was pleased to stumble upon this blog and was looking forward to some answers from experts in the field, but I have been a little disappointed so far.

    The original post about the annual growth rate has received no comments at all, apart my own rather unscientific observation. Given the critical importance of the relationship between growth of emissions and growth of amospheric CO2, I was hoping for a better insight into the apparent/possible divergence of trends.

    With regard to the mid-winter pause in monthly growth, I appreciate the hypotheses put forward by Kirk and Jeff, but the very fact that these are hypotheses suggests that the answer to the question – what is the cause – seems to be that we don’t know. Nothing wrong with this of course and it is always helpful to understand what is known and what isn’t, but since it has been occuring for many years, I am surprised that there does not seem to be an accepted explanation.

    Finally, are the two “pauses” related in any way? Is there a correlation between the presence/absence of the mid-winter pause and the annual growth for the same year? If anyone has looked at this, I would really like to see their results and conclusions.

    1. It is possible to download their data and plot it yourself if you want to examine previous years for a winter plateau. You can see them most years, but last year and this year seem more distinct even when you set the axis scales to the same increments as the current 2-year graph on this site.

      Interestingly it appears a step change of about 1 ppm took place on Feb. 21st. I’m curious if this new plateau at about 401 ppm continues very long. By past observation, CO2 should start increasing again about now.

  6. Larry, I have indeed downloaded and plotted much of the data, but I am looking for the insight of the experts who have been analysing this data for many years.

    Yes, I also noted the 1 ppm rise and I would be fairly confident that this is the end of the mid-winter pause for this year. The question is, therefore, why is it a week earlier than last year! OK, that was a bit facetious, but seriously, this is real data and if we want to understand the cause, we need to be looking for possible correlations.

  7. Hi Jim and Larry,

    In response to the recent thread, Ralph Keeling had the following comment:
    The winter “plateau” in Jan-Feb is clearly present many years, and also emerges if you average the data from many years together. The averaged plateau is not a complete hiatus, however, but rather a slowing in the seasonal rise rate.

    Although the plateau has not been highlighted in the scientific literature (to my knowledge), it is present also in simulations of the CO2 cycle at Mauna Loa that have been done using numerical models that take account of exchanges of atmospheric CO2 with land vegetation and soils. The plateau is therefore basically “understood” although you’d have to dig into the models a bit to fully flesh this out. The slowing down of CO2 release from soils due to mid-winter cold is clearly an important factor.

    So what causes the differences from year to year in this plateau? A major factor must be year-to-year changes in wind patterns. In northern winter, the CO2 concentrations in the Northern Hemisphere are generally higher to the north, and lower to the south. This “gradient” is mainly the result of the seasonal release of CO2 from northern land ecosystems, which causes CO2 to build up somewhat in the north. With this gradient, you expect a stronger plateau in a year in which the synoptic winds at Mauna Loa were anomalously from north in early January and anomalously from the south in late February. Conversely, you’d expect a weaker plateau in a year with opposite wind anomalies at these times. Take 2015, for example: On or around Jan 1st, there was a short-term upward excursion in CO2, while on or around Feb 18th there was a short-term downward excursion. These features must be caused by specific weather patterns that happened this year, but cannot be expected every year. If you remove these excursions from the record (by eye) you also largely eliminate the visual impression of a plateau. Other factors may also be at work, but the detailed weather patterns clearly have a big impact on how strong the plateau appears by eye in a given year.

  8. Rob, Thanks very much to you and Ralph for this response. I now have something to think about. I may be back, however!

  9. I wasn’t offering an explanation. I just mentioned the data availability for those who may not have known it is easily accessed.

    I too appreciate the additional feedback. I figured the effect involved wind patterns based on the earlier reply. It’s interesting to watch how it changes.

    Anytime I come across someone who has doubts about the anthropogenic source of increasing atmospheric CO2, I don’t argue with them. I just point them to this site.

  10. As Ralph noted about the plateaus, “…If you remove these excursions from the record (by eye) you also largely eliminate the visual impression of a plateau….” This is quite clear. The astonishingly regular cyclical pattern within the overall Keeling Curve is hard to miss, and I suspect was one of the biggest surprises for Ralph’s father, when the first data was collected in the mid-1950’s.

    One other hypothesis possibly worth some data analysis is whether there is a correlation in northern hemisphere Arctic weather extremes and the plateau. It seems logical that a pattern that more aggressively disperses Arctic air southward will greatly increase the land area that is frozen, and may thus be effectively ‘capped off’ (by snow, ice, or a frozen surface area) and unable to release normal gasses from soils below. Once the Arctic event ends, and the surface thaws, the accumulated gasses would be released and nudge the Keeling Curve upward, back on track.

  11. Very nice but One other hypothesis possibly worth some data analysis is whether there is a correlation in northern hemisphere Arctic weather extremes and the plateau. It seems logical that a pattern that more aggressively disperses Arctic air southward will greatly increase the land area that is frozen,

    1. By “northern hemisphere Arctic weather extremes” do you mean the melting of its ice?

      If so then a correlation, if there is one that can be proved , could be explained by phase change. When ice melts the energy goes into completion of phase change!

  12. This is an extremely useful graphic. How about putting the second derivative up there too?

  13. I still find it odd that no-one has commented on Rob’s original question. So, let’s try this: does anyone see, and can provide evidence for, any significant change in the rate of growth of atmospheric CO2 over the last decade? For a bonus point: is the rate of growth evident in the decreasing part of the annual cycle, or the increasing part, or both? Further, does the size of the “mid-winter pause” have any impact on the annual growth rate for that year and, finally, does the size of the annual growth rate have any relationship to any measure of temperature?

Comments are closed.