Watt about CO2 levels?

Anthony Watts has a new post at Watts Up With That (WUWT) called an illustration that CO2 won’t roast the Earth in a runaway tipping point. He shows the following figure which illustrates atmospheric CO2 concentrations in previous geological time periods. The argument seems to be that it’s been much higher than it is now, so the Earth can’t reach some kind of tipping point where it will undergo a runaway greenhouse process.

Carbon dioxide concentrations over geological time.  (credit : Anthony Watts, WUWT)

Carbon dioxide concentrations over geological time. (credit : Anthony Watts, WUWT)

Okay, so Anthony may actually have a point. A runaway process is unlikely. But, this is really a strawman argument. Noone’s really suggesting that the reason we need to act is to prevent some kind of runaway process. What people are suggesting is that if we continue to add CO2 to our atmosphere at the current rate, we will double atmospheric CO2 concentrations – relative to pre-industrial levels – in about 80 years time (maybe sooner if the rate increases). There is strong evidence to suggest that the Equilibrium Climate Sensitivity is close to 3oC. So, eventually (within the next few hundred years) the Earth will be 3oC warmer than it was in pre-industrial times. This is hotter than its been for all of human history and will almost certainly lead to damaging climate change. Of course, if we continue to add CO2 so that concentrations more than double, the warming will be even greater.

So, the figure that Anthony should probably have shown is the one below. This shows atmospheric CO2 concentrations and temperature for the past 500 Million years. What’s clear is that when the CO2 levels were much higher than they are today, the surface temperature was much higher than it is today (25oC rather than 15oC). Furthermore, hundreds of millions of years ago, solar insolation was about 6% less than it is today. This means that the equilibrium temperature would be about 4oC less than it is today (there would probably be an even bigger difference in surface temperature given that this will change the temperature gradient in the atmosphere).

Carbon dioxide concentrations and surface temperatures for the last 550 Million years (source : Planet Fossils of West Virginia.

Carbon dioxide concentrations and surface temperatures for the last 550 Million years (source : Planet Fossils of West Virginia).

So, recent pre-industrial CO2 concentrations were about 300 ppm. To get to levels seen hundreds of millions of years ago (say 5000 ppm) would require about 4 doublings. Given that solar insolation was about 6% less than today, this means the CO2 has to provide (through forcings and feedbacks) about 15 degrees of warming, compared to what it is providing today. Hence, about 3 – 4oC of warming per doubling. I appreciate that this is a simplistic calculation that ignores many subtleties and details, but (unless I’ve made some kind of silly mistake) the data presented by Anthony in his post is essentially entirely consistent with current estimates that climate sensitivity is about 3oC per doubling of atmospheric CO2.

So, yes atmospheric CO2 in the past has been much higher than it is today. Yes, it’s unlikely that we are risking some kind of runaway process. However, what this past history tells us is that CO2 concentrations play a crucial role in setting surface temperatures and this data also tells us that climate sensitivity is probably around 3oC per doubling. This means that if we continue as we are we will probably have locked in 3 degrees of warming by the mid- to late-21st century. So, if Anthony could just extend his analysis a little he could make a positive contribution to the debate. We could all agree that we’re heading for at least 3 degrees of warming (if we do nothing) and could then focus on the discussion we should be having which is, should we do something about this and, if so, what?

This entry was posted in Anthony Watts, Climate change, Global warming, Watts Up With That and tagged , , , , , . Bookmark the permalink.

30 Responses to Watt about CO2 levels?

  1. Rachel says:

    I wonder whether it’s worth explaining that the reason there appears to be little correlation between CO2 and temperature in the graph you have supplied above is because it ignores the influence of the sun. I’ve seen climate science critics use that graph as proof CO2 does not affect temperature.

    Anthony Watts should also really point out that at times past, when CO2 and temperature were much higher than today, there were no humans.

  2. I believe that that is exactly the reason. I should maybe have mentioned it. I’ve seen this shown in a video, but I’ve yet to find a good illustration that includes CO2, temperature, and solar insolation. I think that would be useful for rebutting the claim that the graph here proves that CO2 does not affect temperature.

    Indeed, the idea that it’s okay because it’s been higher in the past is ridiculous given that we’ve only been present for a tiny fraction of the Earth’s history. There was a point in the past when asteroids and comets were regularly bombarding the Earth. To be consistent, Anthony should probably suggest that we shouldn’t be concerned about an asteroid strike since it has happened before.

  3. Marco says:

    There’s a lot to say about runaway climate change, including that it is unlikely, but also that a comparison with e.g. the Cambrian is inappropriate. Solar activity was likely considerably smaller (a few %, 3-4 or so) during the Cambrian (solar luminosity increases by an estimated 8% per billion years), which amounts to several degrees difference. Thus, we’d be starting from a higher baseline. Also the land distribution was much different, which will move the baseline yet again due to differences in albedo and ocean currents.

    Again, runaway is unlikely, but not as unlikely as Anthony would like to make it by his false comparison.

  4. I’ve got one little nitpick. If you’re talking about the equilibrium temperature that’s in the context of a 100 year timespan. As that is what the IPCC, and most scientists, mean when they are talking about temperature sensitivity and how much a planet will warm.

    Of course this doesn’t mean temperatures won’t rise after that. But that’s a different climate sensitivity. It all has to do with which feedbacks respond when and how fast:

  5. That’s a good point and one I had considered making. The baseline would be higher and so if we ever did get to the point were CO2 concentrations were 5000 ppm, surface temperatures would likely be an average of 30oC, rather than 25oC. So, not likely but not as unlikely as Anthony would have people believe.

  6. When I said equilibrium I was really meaning the non-greenhouse value (or, maybe, the top-of-atmosphere value) rather than the final surface temperature (i.e., balancing incoming flux and outgoing flux only). I was just trying to make the point that a 6% drop in solar flux would reduce the effective radiative temperature of the Earth by about 4oC so you need higher CO2 concentrations in the distant past to produce the same global surface temperature as today – I probably didn’t make that clear enough. You’re quite right though that there are different feedbacks (fast and slow) that can influence the final equilibrium and how long it takes for the whole system to reach equilibrium.

  7. I’m glad you commented on this. I was upset that he wasn’t giving the whole story, and leaving one to make assumptions about what his graph showed.

  8. Thanks. I should probably acknowledge that it may have been one of the comments that you made on the WUWT post that prompted me to write this. Apologies for not giving due credit πŸ™‚

  9. KR says:

    Regarding the graph you are looking for, see Royer 2006 ( Royer graph, “CO2
    -forced climate thresholds during the Phanerozoic”
    ), where in Fig. 2 combined radiative forcing for CO2 and sun are graphed against temperature.

    This was discussed at some length over at SkS in 2010 ( http://www.skepticalscience.com/co2-higher-in-past-intermediate.htm ).

  10. That’s great, thanks. Very interesting. I have a feeling I have seen that before, but have read so much in the last few months that I’m starting to lose track πŸ™‚

  11. KR says:

    My apologies – Royer only graphs the radiative forcing, stating that the temperature records over the Phanerozoic are too spotty for comparison. What is shown, however, are periods of significant glaciation (such as our current ice age cycles and polar caps) – none of which occur at CO2 concentrations over 1000 ppm, only two CO2 doublings from pre-industrial levels.

    What graphs such as the one in your opening post most clearly show is that ‘single-cause’ arguments (so frequent on the ‘skeptic’ blogs) are silly. CO2 is a factor in previous temperatures – so are insolation, orbital precession, positions of continents, etc. You need to examine -all- of the causes to tease out the individual contributions, and only then can you make predictions on how conditions will change along with an individual forcing change.

  12. No need to apologise. It’s an interesting graph. You’re quite correct that we need to consider everything. If I look at the Royer graph, 400 Myr ago the net CO2 plus solar forcing (relative to pre-industrial times) was about 4 Wm-2. If CO2 was at 4000 ppm (compared to 280ppm in pre-industrial times) that suggests that the CO2 forcing was 5.35 ln(4000/280) = 14.2 Wm-2 higher than in pre-industrial times and solar forcing was about 10 Wm-2 lower (although I think Royer assumed a linear change from 94.5% of pre-industrial values so what he uses could be worked out quite straightforwardly). So what I presume you’re suggesting here is that the Royer graph doesn’t really illustrate all the feedbacks, which given that the CO2 forcing is more than 14 Wm-2 greater than pre-industrial times, will be quite substantial.

  13. KR says:

    I’ve seen the CO2/temp graph before in bad ‘skeptic’ arguments (in fact, the site you source it from is in fact a denial site, pushing multiple myths, including arguing from wildly uncertain stomata CO2 data contradicted by ice cores). It’s often been used to argue that CO2 has little effect on climate – when after examining forcings in context, including the ‘faint sun’ effect, it’s quite clear that CO2 has a -very- strong effect.

    It’s not that the graph doesn’t include all the feedbacks, it’s that it doesn’t include all the forcings.

    The rather limited range of temperatures seen over the past half billion years (15C total) is itself a strong argument for the existence of geological timescale feedback – weathering (reducing CO2 from volcanic inputs) and glaciation (reducing exposed rock/weathering, allowing CO2 concentration to increase), for example, as opposing feedbacks keeping temperatures in that range. Unfortunately, those are orders of magnitude slower than current anthropogenic changes, and won’t be saving us from from significant and unpleasant climate changes – changes which, while geologically brief, are long on the human scale.

  14. Yes, you’re quite right that it’s been used all over the place to make all sorts of spurious arguments against CO2 playing a role in climate.

    I was probably using the term feedbacks a little loosely. The graph only contains CO2 and Solar, so ignores everything else, which is what I was referring to. I’m still learning the terminology πŸ™‚

  15. Och, don’t worry about it! Just glad I wasn’t the only one who found the whole thing a bit suspect!

  16. Tom Curtis says:

    Wotts, I come late to the party to find that KR has stolen my thunder.

    For what it is worth, feedbacks are a consequence of changes in temperature rather than changes in forcings. They vary slightly from one forcing to another due to different temperature responses with regard to latitude, altitude and time (summer/winter; day/night), but to a first approximation they can be considered identical across all forcings. In any event, the near identical planck response to a given forcing means the feedbacks to a net 4.2 W/m^2 forcing will be approximately the same whether due to 2.2xCO2, or 14.3xCO2 plus -10 W/m^2 of solar forcing.

    I will add that Royer produced a full phanerozoic temperature reconstruction (along with four others) in 2004. Figure 2 of that paper is reproduced here.

    Finally, I dislike the graph produced by Scotese which you reproduced in the blog. In particular the temperature series is produced largely by assigning a temperature for warm periods, and a distinct temperature for cold periods. Certainly it is not based on proxies beyond geological evidence of the extent of ice or tropical species. Further, the temporal resolution of the temperature series is very bad. The end Ordivician glaciation, for example, only lasted around 1 million years, but is shown as lasting 13 million years. The graph may have represented a best estimate when Scotese constructed it, but it is a poor substitute for the results of Royer’s research using oxygen isotopes as proxies of temperature.

  17. Richard Smith says:

    I was surprised to see that you agree with Anthony Watts that a CO2 tipping point is not something we should be worrying about.

  18. Tom Curtis says:

    “CO2 tipping point” is a very vague term. There are several potential tipping points that could be triggered by rising CO2 that we should worry about. However, we need not worry that a runaway greenhouse effect will boil the oceans dry.

  19. Thanks, Tom. The graph may well be a poor one. I just happened to have it handy when I quickly wrote this post this morning.

    Thanks for the links to the other Royer work. It does seem much more detailed and Figure 2 is very useful. I suspect I’ve sometimes being using the term feedback in the wrong context. I was thinking more of other forcings I suspect. Given that if the net forcing due to CO2 and solar, relative to pre-industrial times, was only 4 Wm-2 400Myr ago, that alone would only produce about 1 degree of surface warming. There must therefore be other forcings that produce a net increase of 4 to 6 degrees.

  20. Richard Smith – maybe I’m wrong – but I presume you’re being a little sarcastic. Well, my goal isn’t simply to criticise what appears on Anthony Watts’s blog. It is to address and at times give credit where credit is due. Also, I don’t think that I’ve ever expressed the view that we should be concerned about a runaway greenhouse process. We should be concerned about how we can continue to survive as we are if the planet becomes much warmer (by at least a few degrees) than it has been throughout human history. That’s not the same as being concerned about a runaway process.

    As Tom points out though, there are other potential tipping points that we should be concerned about.

  21. Fragmeister says:

    I think part of the denier play book is to exaggerate the likely changes so they can be discredited. Allows an alarmist picture to be built up rather than a realistic one. Denying, remember, doesn’t have much to do with reality at all, otherwise there would be no deniers.

  22. Tom Curtis says:

    Wotts, it is not clear that there needs to be another substantial forcing. First, 3C/2xCO2 only represents the Charney (Equilibrium) Climate Sensitivity. It does not allow for slow feedbacks such as the melting of ice sheets, changes in vegetation, and the increased release of CO2 and methane into the atmosphere by natural processes as a result of increasing temperature. The Earth System Sensitivity, which does allow for those factors is estimated as being from 4 to 8 C per doubling of CO2, being closer to the lower value when there are no major ice sheets, and the higher value when there are. Because of the low temporal resolution of the temperature estimate (approx 10 million years) ESS is the relevant value and may account for the additional warming of approx 6-7 C relative to preindustrial values.

    Second, continental configurations can change the responsiveness of the Earth to forcing in at least two ways. First, continents near the poles encourage the build up of ice sheets, which cannot form when most continental masses are in the tropics (over the temperature ranges seen in the phanerozoic). Second, different arrangements of continents can significantly alter the latitudinal distribution of heat, which effects the temperature response from a given forcing.

  23. Tom, yes I agree. I think I’m again using the terminology badly. But maybe this is an opportunity to clarify something that has maybe puzzled me a little. Ignoring this specific situation, if CO2 concentrations double then the forcing increases by 3.7 Wm-2. This alone would lead to about a 1 degree increase in surface temperatures. To get a 3oC per doubling of CO2 then requires additional effects. My understanding is that these could be increased water vapour and increases in other greenhouse gases. These would be forcings. There could be clouds (which can both trap outgoing longwavelength radiation or reflect incoming sunlight) and aerosols. There can also be changes in ice coverage which would change the albedo – and I assume that this would be a feedback rather than a forcing. Have I got the basic picture right here?

    So, in this specific situation (i.e., ~ 400 Myr ago) we have reduced solar flux and forcing due to CO2 resulting in a net forcing, relative to pre-industrial times, of 4 Wm-2 and a net increase in temperature of 4 – 6oC. If I understand you correctly there will be additional forcings (in addition to the 4 Wm-2 from CO2 and solar) due to – for example – water vapour and other greenhouse gases plus other feedbacks such as changes in ice coverage, which would depend on the arrangement of the continents, and changes in vegetation. Some of these can operate on short timescales and others on longer timescales. So, the ECS could be 3 degrees per doubling but the ESS can be larger due to the longer timescale effects.

    I feel a little silly writing a blog about this and then having to ask basic questions, but then I remind myself of what goes on at WUWT and would rather do this than blinding going on regardless πŸ™‚

  24. Tom Curtis says:

    Wotts, it is unsurprising that it is a bit confusing. The most important distinction between a forcing and feedback in many cases relates to timescale. If the response time of the factor is sufficiently short relative to the time scales involved such that the factor can be treated as being controlled by temperature, then it is a feedback. If not, or if the factor is not significantly influenced by temperature then it is a forcing.

    Thus, changes in solar radiation, for which temperatures at the surface of the Earth are not a causal input, are always forcings. In contrast, changes in relative humidity, which equilibriate based on temperature within a matter of days (or at most weeks) can be treated as controlled by the surface temperature of the Earth and are a feedback at all time scales.

    At time scales of less than several centuries, and given a significant change in emissions (natural or anthropogenic) from a base rate, CO2 concentrations will not reach a relatively stable level based on surface temperature, and are a forcing at that time scale. If the new emission rate remains constant over several thousands of years, however, changes in CO2 concentration over that period (after the first few centuries or so) will largely be determined by surface temperature and are a feedback. Similarly, on short timescales (ie a few centuries or so) ice albedo from ice sheets are effectively independent of temperature and are a forcing, but over millenia they are definitely a feedback.

    To confuse things still further, you can treat some relatively stable factors as forcings over time scales at which they are technically feedbacks (ie, CO2 changes from LGM to pre-industrial) to determine the Charney Climate Sensitivity.

    Finally, over the course of the phanerozoic, and at the 1 – 10 million time resolution involved, changes in CO2 concentration are effectively governed by large scale changes in the rates of emission of draw down by weathering of rock, and are forcings. However, changes at intermediate timescales (such as the variations between glacials and interglacials) are governed by temperature and CO2 is a feedback at those time scales.

    I suspect I have not been entirely clear (given the lateness of the hour), nor answered all your questions. But (again because of the lateness of the hour), that’s the best I can do for now.

  25. Richard Smith says:

    Tom, you can help me out here. I thought the term CO2 “tipping point” was fairly straightforward. Man adds CO2 to the atmosphere. The warmer atmosphere leads to a warmer ocean and the warmer ocean passes CO2 and water vapor into the atmosphere and those added greenhouse gases cause even more warming. Some believe this could lead to a tipping point, perhaps within the next century, whereby the process “runs away” and the earth warms at a very rapid rate – a rate much higher than the current rate, whatever that might be. What other tipping points are there that we should be concerned about?

    Fragmeister, is Tom Curtis’ reference to the ocean boiling dry an example of the deniers playbook of creating an unrealistic strawman?

    Wattsupwiththatblog, is Watts’ reference to tipping points or runaway really a strawman? Is this not something that has been talked about extensively by the climate scientists?

  26. BBD says:

    That’s a nice summary of feedbacks and forcings definitions, Tom.

  27. Thanks, Tom. Glad you said it was confusing. I find myself going through stages of thinking I get it, and then realising I’ve missed something. As BBD said, a nice summary.

  28. Richard, I don’t know if calling Anthony’s reference a strawman is actually justified. I’ve been aware of aspects of this debate for quite some time, but have only been reading about it extensively for the last year or so. As far as I can tell, very few are actually suggesting that a runaway process is likely, but I can’t claim that this wasn’t the case at some point in the past. Whatever the truth is, I don’t think what people are concerned about now is a runaway process. What people are concerned about now is what we do about the possibility that we’ve locked in about 2 to 3 degrees of warming (more as we continue as we are). Maybe some think there’s nothing to be concerned about. Many, as far as I can tell, disagree.

  29. dana1981 says:

    “the data presented by Anthony in his post is essentially entirely consistent with current estimates that climate sensitivity is about 3oC per doubling of atmospheric CO2.”

    Indeed, 3Β°C equilibrium sensitivity estimates are derived from paleoclimate data, among other methods. Most recently, see the PALEOSENS results, discussed here in the ‘Considering the Full Body of Evidence’ section:

Comments are closed.