Watt about climate models running way too hot?

Bjorn Lomborg has a new Guest essay on Watts Up With That (WUWT) called climate models are running way too hot. The post says

Over the past 30 years, they are at least predicting 71% too much heat. Maybe 159%. (see graph)

The figure he’s referring to is below. Here’s problem number one. These are temperature anomalies and so are relative to an arbitrary baseline (in this case, adjusted to zero 1979-1982). The percentage difference that Bjorn calculates depends entirely on the choice of baseline. If it was set relative to 1990, it would be massive. If relative to 1950, much smaller. An entirely meaningless calculation. The 159% presumably refers to the difference between the models and the satellite data. Well, the model results are presumably for the surface, so what is he doing comparing this with tropospheric temperatures (I believe, at least, that these are lower troposphere measurements)? He’s also considering only the ensemble model mean (I believe) and ignoring that climate models are still consistent with recent surface warming at the few percent level.

Comparison between measured surface temperature anomalies, satellite temperature anomalies and model results (credit : Bjorn Lomborg, WUWT)

Comparison between measured surface temperature anomalies, satellite temperature anomalies and model results (credit : Bjorn Lomborg, WUWT)

Additionally, Bjorn says the current climate models are running way too hot. I actually don’t think they are (maybe someone who knows more could confirm). They may be currently over-estimating the level of surface warming, but I think that they have been quite good at predicting overall warming. The planet continues, according to ocean heat content data, to accrue energy at the rate of 1022 J per year. So, yes, surface warming is currently slower than expected, but overall warming is not.

Bjorn finishes his post by saying

Yes, there is a problem, no, it doesn’t look like the end of the world.Let’s fix global warming without the fear.

Indeed, it doesn’t look like the end of the world. Yes, let’s fix global warming. Let’s do so, however, without underestimating and mis-representing what we are likely to face.

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 climate models running way too hot?

  1. Lars Karlsson says:

    One should always be suspicious when somebody shows a single line for the models.

    Ed Hawkins, who unlike Lomborg is a real climate scientist, has some great posts at Climate Lab Book about the topic.

  2. BBD says:

    Here’s problem number one. These are temperature anomalies and so are relative to an arbitrary baseline (in this case, adjusted to zero 1979-1982). The percentage difference that Bjorn calculates depends entirely on the choice of baseline. If it was set relative to 1990, it would be massive. If relative to 1950, much smaller. An entirely meaningless calculation.

    And that’s exactly the same trick Christy used in his misleading graph comparing mid-tropospheric temperatures to CMIP5 runs. He included other tricks too, but there’s no need to go into those here. Just noting that this one to watch out for when contrarians start waving graphs around.

  3. It’s time like this when it’s hard to remain civil. How can someone like Christy or Lomborg claim that they didn’t realise that their calculation was completely meaningless? Of course, they could correct it when pointed out. Admittedly, I’ve just glanced through the comments on WUWT and noone (yet) has mentioned that the calculation has no merit.

  4. Yes, I have read some of Ed Hawkins’s post in the past. The one you link to is very good.

  5. Just a quick addition, as I’m not gonna bother to read this claptrap. I noticed one thing in the above figure right away (apart from the baseline issue). It shows the RCP2.6 ensemble scenario (rather than the much more appropriate RCP4.5 or 6.0 scenario) which tends to warm fastest after 2000 (until 2020) as aerosol forcing quickly diminishes. Hence, it simply misleads the audience. But what else do we expect …

  6. Indeed. There was much more that could have been said, but there’s only so much effort I can put into debunking this type of nonsense.

  7. As an aside. I watched a HuffPo report today which included Michael Mann and Bjorn Lomborg. Bjorn stated that a carbon tax should be set at $5 per ton of CO2. Unless I did my calculation incorrectly, this would result in me paying about $5 more per year to drive my car. Doesn’t sound like a particularly effective tax if the goal is to reduce carbon emissions.

  8. BBD says:

    That’s why there’s a comments section!

    You do fine. Never trouble yourself about that.

  9. Thanks. I see now that my comment may have seemed a little defensive, but that wasn’t what I was intending 🙂

  10. Martin says:

    A carbon tax is usually set as a Pigovian tax. So, if it is set at a certain number per tonne, then because this corresponds to the net present damage (cost) caused by that tonne emitted now. This seems, at first glance, like just paying for the damage done. But the thing is that a price signal is set, and through the invisible hand of the blah, blah… the amount of CO2 emitted is reduced in accordance with that pricing. So, the statement “Doesn’t sound like a particularly effective tax if the goal is to reduce carbon emissions.” does not make much sense in that context: it implies that the reduction target is set independently from the tax level (and that the tax is there merely to “effectively’ reach that target), when it’s actually on and the same thing. But one could argue that the Lomborg number is simply too low, even if one does not go into the catastrophe insurance business and such.

    This is all very confusing (at least for me), and I really recommend Weitzman’s classic “Prices vs. Quantities”:

    Click to access weitzman.pdf

  11. It is funny that the people who do not accept the temperature increase of 0.8 °C over a century, make such a fuss about a deviation of one or two tenth of a degree in a decade.

    I am confident about the large long-term increase and am almost surprised that climatology seems to be able to say something about that minute deviation (ocean heat content and ENSO seem to be the best current candidates).

    If we actually understand this minor deviation, and I will stay careful for some more years (the scientific literature moves slower as the blogs), that would be an impressive feat of climate science. While the estimates of the climate sensitivity and its uncertainty did not change much the last 20 years, which could suggest little progress, being able to explain such an minor deviation would show that climate science has made huge increases in the observation, modelling and understanding of the climate system. Impressive.

  12. Derek says:

    So the missing heat is hiding in the deep ocean? Yet the data to support this thesis comes from ocean buoys and requires measurements in hundredths of a degree of temperature change which is simply not possible. Moreover, ocean heat uptake has been slowing over the past eight years.

  13. The data partly comes from Argo floats. I don’t know enough about them to know precisely what they measure. It also comes from measurements of changes in sea level and there are direct satellite measurements of the TOA energy imbalance. The rate may have slowed slightly in the last 8 years or so but is still (according to the most recent studies) close to 1022 J per year.

  14. I shall have to read that, but all I meant by my statement was that if I was going to be paying $5 extra per year to drive my car it would be unlikely that I would consider changing my car to one that was more efficient (or that I would really take this into account when buying a new one). Maybe, more correctly, that such a cost alone would be unlikely to influence my decisions.

  15. What BBD said 😉 … plus what Victor said below!

    As a little aside, I got used to check peoples references (in case there is one) if they appear to merely repeat a series of well known myths. If I then only learn that the person in question has been spouting nonsense for several years (as is the case with Derek) … well, you know what comes. Sadly, some people are simply lost. Tragic when they are young, but luckily, most are not exactly young anymore. That’s what all my hope for this planet is based on.

    And dear wotts, please note that this side note is not meant to be an advise of any sorts. Everyone of us has to go through this process for oneself, as you’ve aptly pointed out in one of your recent postings. That was spot on! And believe it or not, it was as if I got sent through some sort of time machine upon reading. So once again, you are doing perfectly fine … just as BBD said ;). We wouldn’t comment if it weren’t for your highly appreciated debunking efforts.

  16. So we cannot measure the heat uptake of the deep ocean, but you do know that the derivative of the ocean heat uptake is negative? Could you explain this paradox?

  17. Tom Curtis says:

    Wotts, the direct satellite measurements of TSI are very accurate for changes from month to month, and year to year, but poorly calibrated as to absolute value. That creates difficulties in creating reconstructions of changes in TSI over multiple satellite missions. You can see the problem clearly in figure 1 here. The same is also true of satellite measurements of OLR. In consequence, while month to month, and year to year changes in the energy balance are well constrained (allowing for differences in multi-satellite composites), the absolute value is not well constrained by satellite measurements.

    Stevens et al, 2012 say:

    “For the decade considered, the average imbalance is 0.6 = 340.2 − 239.7 − 99.9 Wm–2 when these TOA fluxes are constrained to the best estimate ocean heat content (OHC) observations since 2005 (refs 13,14). This small imbalance is over two orders of magnitude smaller than the individual components that define it and smaller than the error of each individual flux. The combined uncertainty on the net TOA flux determined from CERES is ±4  Wm–2 (95% confidence) due largely to instrument calibration errors. Thus the sum of current satellite-derived fluxes cannot determine the net TOA radiation imbalance with the accuracy needed to track
    such small imbalances associated with forced climate change.
    Despite this limitation, changes in the CERES net flux have been shown to track the changes in OHC data. This suggests that the intrinsic precision of CERES is able to resolve the small imbalances
    on interannual timescales, thus providing a basis for constraining the balance of the measured radiation fluxes to time-varying changes in OHC.”

    (My emphasis)

    This issue had also been raised by Trenberth in his discussions of Space/atmosphere, and atmosphere/surface energy balances.

    I am not aware of any more recent measurements more tightly constraining the absolute error. Essentially, therefore, we know the value of the TOA energy imbalance by measurements of change in OHC rather than by satellite measurements – the satellites not yet being sufficiently accurate to constrain the value.

  18. Tom Curtis says:

    Derek, ARGO takes literally thousands of measurements of ocean temperatures. They are measured accurately to a value of plus or minus 0.005 C. Your claim that measurements of temperature changes of one hundredths of a degree “are not possible” simply ignores the fact that ARGO floats measure accurately to 5 thousandths of a degree. It is simply a factoid made up for your convenience (or more probably simply made up by somebody else and echoed by you).

    Further, in taking the mean temperature change of the ocean, we do not take individual measurements, but the mean of all the many thousands of measurements involved. Absent an actual time variant bias in the measurements, that means the accuracy of the final measurement relating to instrument error is much better than the error of individual instruments. I believe the standard error would be +/- 0.005/(n^0.5) where n is the number of individual observations involved. That brings the instrument related error down to 8 one hundred thousandths of a degree. (There is a larger error related to the fact that not all of the ocean is covered, which is quantified and allowed for in the reports.)

    What you should be asking yourself now is, why were you fed the claptrap you just spouted here? And why (if you consider yourself a skeptic) did you swallow it so gullibly?

  19. Thanks, Tom. Yes, I was aware that the error was large but wasn’t aware of the details. Very useful.

  20. Tom, as far as Mike Lockwoods opinion is concerned (and he is certainly one of the most prolific experts on that particular topic), he strongly favours the PMOD composite over any other estimate. I am just passing this information on without further judgement. I might add that I consider his an extremely credible voice on the issue at hand 😉

  21. Martin says:

    The German blog “Die Achse des Guten” (the German outlet for everything that has place between Glenn Beck and Bill O’Reilly; one contributor is Benny Peiser) has a nice summary about the “climate models all wrong” theme. The title is about “Small Sensation: Apparently, IPCC backpedals even more clearly.”


    Sources cited for all that stuff: WSJ, Daily Telegraph, Mail, The Australian.

    Apparently, IPCC lead authors are now attending crisis meetings “because of the new tone.” Why, exactly, they would do this, the author of the weblog does not say – but reading between the lines it is probably BECAUSE THEY ARE SCARED TO DEATH BECAUSE NOW EVERYBODY KNOWS ABOUT THEIR LIES TAKE THAT MISTER MICHAEL STALIN MANN.

    The reversal is so phenomenal because “scientists from Potsdam” (he probably refers to the PIK, which usually serves as point of reference for Communist Climatestalinazis in German skeptics circles – about what the NOAA is elsewere; and Schellnhuber is the local incarnation of Hansen, I guess, like the Horus of Nekhen compared to the nṯr Horus in Ancient Egypt?) “talked about four, five or even more degrees Celsius warming not so long ago”. Imagine that! Withouth any context they talked about that, for the sole purpose of scaring people, and because they hate markets, too.

    There is also the Himalaya glacier “false prognosis”, and all that stuff.

    It’s as if The Onion had written it.

  22. Kulke is a liar! Plain and simple. Amazing how he managed to further twist Rose’s already severely twisted non-statement from Myles Allen. Particularly grotesque in the face of Myles correction in the comment section under his article. Had a funny chat with Myles about all that today in the office … if Kulke only knew how many light-years and universes off the mark he is …

  23. Martin says:

    Of course he’s a liar. I guess one could attempt to defend him by noting that he could simply been deluded. But then, I am not sure if that is really a defense…

  24. Martin says:

    Upps, didn’t see the response. Well, a carbon tax is not a premium on fuel consumption. As the fuel production process (beginning with extraction) and its maintenance isn’t carbon free itself, I guess the base price would be higher, too. But even that is beside the point. Even if it does not change your fuel consumption behavior, it might still have an effect on coal heavy production/generation (think steel, electricity in Germany/China), the worst of it all (even without all the crap other than CO2), for example.

    Point is: supposed we have set a carbon tax, and it’s exactly right, why would it be important that the consumption of one particular product by a person of a certain income level be reduced? What’s the significance of this specific product? If it does not influence fuel consumption at your income level, uhm, so?

    In your last sentence, you acknowledge the crux: in an of itself, minimally higher fuel costs would not do much. But that’s not what a carbon tax is. And very low taxes might not even have much of an impact for the middle class. If that’s the case, then there is no reduction target that is unlikely to be met – if middle class does not feel the tax (again under the idealistic assumption that it’s exactly right), then because we simply do not cause much damage (though one would have to look at the tax incidence: if coal plant operators pay, who loses?). The Lomborg problem lies exactly here, I think: in the low damage valuation itself (I guess rosy, cherry-picked impact scenarios combined with a high discount rate, and no catastrophic risk considered, but whatever).

  25. Thanks for posting this. I’m putting together a list of contrarians making this bogus argument to rub it in their faces in 10 days when the IPCC report comes out and proves them wrong (which it will). Pielke Jr. made a similarly inept argument today (only plotting the multi-model mean and ignoring the envelope of model runs and uncertainty range).

    So much for these two being ‘honest brokers’ or, you know, competent at interpreting data.

  26. Thanks. It’s going to be a busy few weeks. I’ve just seen Pielke Jr’s post. Yes, very odd that he and Lomborg are regarded as ‘honest brokers’.

  27. Rachel says:

    You may have already seen but Bjorn Lomborg is tweeting this graph:

  28. Thanks, I’ve just responded to him to point out that if he resets the baseline to be 1999-2001 it will be 250% too hot. That should help make his case even stronger!

  29. Pingback: The self induced implosion of Dana Nuccitelli | Watts Up With That?

  30. Pingback: Fact mongering | …and Then There's Physics

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.