Arguments
New Skeptical Science study finds physics trumps economics, suggests global warming is man-made
Posted on 26 June 2014 by MarkR
A new paper by several Skeptical Science contributors (Richardson et al., 2014) in Quaternary International points out that a previous study (Chen et al., 2013) made mistakes in using economics techniques to calculate how much recent global warming is man-made. Most climate science studies, such as the 8 examples here and Stone et al. (2005), Stott et al. (2011), Hegerl et al. (2011), Santer et al. (2013) and Ribes and Terray (2013), conclude that man-made greenhouse gases are responsible for recent global warming. However, Chen et al. calculated that only 50.2% of warming from 1990-2010 was man-made.
It’s important to try new and different techniques to answer difficult scientific questions, and Chen et al. brought in the concept of ‘elasticity’ from economics which measures interactions in the economy. For example Andreyeva et al. (2010) use elasticity to calculate that if the price of soft drinks goes up by 1%, then the amount that people buy tends to go down by about 1%.
How climate reacts to carbon dioxide
Carbon dioxide (CO2) helps to heat Earth's surface, so if you have more CO2 in the atmosphere then temperatures should go up. Instead of looking at the total amount of CO2 in the air, Chen et al. look at how much is emitted in a given year. Their equation says that if the rate of emissions slows down, even though CO2 is still being added to the atmosphere, then they expect temperatures to cool down.
This is like putting a frozen pizza in the oven at 200 ºC (392 ºF), turning the heat down to 180 ºC (356 ºF) a few minutes later and expecting the pizza to re-freeze because the rate of heating has slowed down. It’s obviously not true, but this tricky detail was hidden behind equations.
There’s another major problem. If you’re measuring a physical property and doing it properly, then the result shouldn’t depend on how long you measure it for. I can use a pendulum to measure gravity for 5 minutes today, or I could do it for an hour tomorrow and I’d get the same result.
Chen et al. used data from 1960 to 1989 to calculate the ‘elasticity’ of temperature to CO2. If this is a real result then you should get about the same answer if you change the measurement period, but this isn’t what happens. If you just shift the dates by 10 years you can make the answer go down by 50%, or up by 100%. These changes are so big that it’s clear that this elasticity is not a real physical property and the result is not reliable.
Figure 1 This shows how the elasticity calculated by Chen et al. changes with the start year you choose for analysis. If it were a rigorous measurement, then the result should be a flat, horizontal line around the chosen start date. Chen et al. used the value in 1960, when the result is changing rapidly so the calculation is obviously not reliable, and small changes in which data are used can change the result greatly.
Why did Chen et al. think their result was good?
Chen et al. calculated a coefficient called r2 that tells you whether one thing is related to another, and found r2 = 0.57 for their data. However, just because two things correlate, it doesn't mean that one thing is truly related to the other. We a physics-based equation called the 'one-box model' (Held et al., 2010) to link the effects of greenhouse gases, air pollution, the Sun and volcanoes to global temperatures. We get r2 = 0.82 and our higher number means a stronger correlation. The one-box model also works over a longer period of time, uses physics and doesn’t predict crazy things like temperatures going down if you heat up Earth. We calculate that about 100% of the recent global warming is man-made although since we just looked at the correlation and don’t fully explore this we say that it’s better to rely on research that used much more complicated techniques to work this out.
Figure 2 The green line shows the predicted temperature response to changes in heating from greenhouse gases, the effects of some types of air pollution, volcanoes and the Sun. The blue line shows the measurements from the UK's HadCRUT4 near-global temperature record. The long-term changes largely match the expectations of the physics, although there are short-term squiggles that are generally related to changing winds and ocean patterns.
Conclusion
Chen et al. tried a new method to estimate how much of the recent global warming is man-made. They concluded that it’s about 50%, although the consensus result from most climate science studies is around 100%.
It’s great that Chen et al. tried out a new method as this is an important part of science and if different methods give the same answer then we can be confident in our results. ‘What’s causing global warming?’ is a hard question and most studies say that the answer is that humans have caused 100% since 1950.
Chen et al. applied a method from economics to calculate what's causing global warming. This was a good idea: the question 'What is causing global warming?' is an important one and if different methods give the same answer then we can be more confident about the result. However, if they disagree then we need to ask why. It's now clear that Chen et al. didn't go far enough in understanding the physics before doing their calculations. As a result, they predict that heating can make temperatures go down, and give completely different answers depending on the years considered. Our approach is physically reasonable and reaches the same conclusion as the overwhelming majority of attribution studies: recent global warming is man-made.
At first I thought the title was an Onion headline! Do we really have to prove that basic physics is a bit more reliable than economics when it comes to predicting global warming??!! Apparently so!
Huh? In no way does properly applied science "trump economics" by showing that a single, laughably bad misuse of statistical technique gives predictably bad results.
Can we please stop broadbrushing like this?
I agree with Lou, Mark. That's a right proper rebuttal, and congratulations most definitely, but the headline is a little loose.
"New Study Surprisingly Finds Physics More Useful Than Economics in Understanding Atmospheric Physics."
I think some dry humour was intended in the title and that it shouldn't be taken at face value.
In Figure 2 the unibox model is a bit below HadCRUT4 ~1990-1999 then it catches up but Cowtan & Way estimate ~0.09 higher at 2010 so if that turns out to be correct then the unibox model would stay a bit below Cowtan & Way krigified ~1990-2010 in which case I would muse about whether a slight warming effect started ~1990 that isn't in the unibox model. Also, I wonder whether coefficient r2 could sub for DNA in a paternity suit.
Good article...
might be a typo ?
Actually, the oven temperature analogy should be even worse... a decreasing rate of CO2 emissions isn't the equivalent of turning down the oven temperature... it is the same as decreasing the rate at which you are turning it up.
Their analysis also ignored ~98% of the warming to focus on atmospheric temperatures, assumed that temperature changes would occur instantly rather than time being required for heat to build up in the climate system and for feedbacks to play out, and (looking at the paper itself) they actually calculated wildly different 'elasticity' values for different parts of the planet... which in and of itself invalidates their entire hypothesis given the well mixed nature of CO2 in the atmosphere.
I was wondering if some actually useful results could be obtained if this sort of analysis was done properly (like the RSS satellite temperature data eventually coming out of the original wildly inaccurate UAH results), but there haven't been any instances of decreasing total atmospheric CO2 levels from which to calculate the 'elasticity' value. Further, if we succeed in switching to non-fossil fuels the correlation between economic activity and CO2 emissions will be broken... and thus decreasing atmospheric CO2 levels at that point couldn't be tested against economic activity either.
This is what happens when you try to mix psuedo-science (Economics) with real Science ..... You end up with the equivelent of an Alchemist trying to turn lead into gold ....
Anyone that has ever graded exams or homework assignments knows how much harder it is to figure out why someone screwed up than did something correct.
You deserve medals for debunking this dreck, including Beenstock et al.
We don't use methods that are intended to "learn" human-behavioral game theory trends (economics, finance) on problems that obey hard physical laws (earth sciences).
Once again, the trend in modern rightwing culture to assume economics (read 'free market capitalism') is bigger than government and even Physics. How did Chen forget the Physics in his AGW study? Simple: Physics can say nothing about a Physical problem once Economics has entered the ring. Put that together with this: Hank Paulson gave a wonderful argument for taking AGW seriously the other day. But what bothers is that Paulson is a banker, former CEO of Goldman Sachs, and former Treasury Secretary. So, to recap: after 120 years of Scientific warnings go unheeded, AGW is now 'official' in the eye's of our conservatives because a Banker says so?
Where is Lewis Carroll when you need him.
This whole concept of " ‘elasticity’ of temperature to CO2" is so counterintuitive that I simply dforgotten it overnight and must have reread this article to recall thisd morning what the fuzz is about.
In more broad context, it's amazing how the contrarians while seeking the alternative science explanation, come up with convoluted explanations defying or trying to workaround simple basic physics we've grown with since primary school. I remember enjoying the first climate science course by David Archer, that fit so nicely into and reminded me my long-rusted physics. Yet contrarian explanations, including this one here, or any others e.g. by Dick Lindzen or Judith Curry require strange mental excercises, including forgetting the world we are living in and grasping some abstract ideas like "stadium vawe" or "cosmic rays".
Actually, as a co-author of this comment paper, my impression of the original paper is completely different from the interpretation which is emerging here.
I don't have the impression that Chen et al are contrarians. They may be, and there may be extrinsic evidence to support that, but at this point my assumption is not.
I suspect this was a well intentioned but misguided attempt to apply a new method to testing the ideas of climate science. I suspect they though their results were supporting the consensus position.
I also think that, while wrong, the work is not completely without value. The elasticity calculation would have given a plausible result had they given it the right inputs. While it adds nothing to the science, it may have had some value in explaining the science in terms which would have been immediately familiar to economists.
Kevin C @12, certainly the paper would have been classified as endorsing AGW by the Cook et al (2013). It finds >50% of recent warming to be due to CO2 emissions.
Although this attempt to apply standard concepts from another discipline has not been successful, several major advances in science have been made by porting concepts and/or mathematical techniques from one discipline to another. The authors should not be faulted for that attempt in itself, but only for attempting to apply their technique to the wrong quantity (emissions rather than concentration). Applied to concentration, the results may or may not be usefull, but that can only be known after the attempt.
longjohn119, not all economics is pseudo-science. There are plenty of economists who take a proper 'scientific' approach to the field. Unfortunately, there are also entire branches of economics that do not... and they are often dominant in policy debates.
Indeed, one of the most disturbing things about the science denier movements (i.e. 'global warming deniers', 'evolution deniers', various carcinogen and pollutant deniers, et cetera) is the thought that they might one day succeed in corrupting a significant portion of the 'scientific' community to their nonsense the way economics has been corrupted. There are already large segments of the general populace who have bought into fake science. What happens when the children of some of those people grow up and become 'scientists' dedicated to advancing fiction the way many 'economists' currently do?
There hasn't been much sign of this so far. Just an ever diminishing circle of actual scientists with contrary beliefs that they cling to, and produce flawed research in support of, despite the evidence. That's inevitable in any field and has been common throughout the history of science. Still, all the elements are there for the corruption of economic science to spread to other fields.
My original training was in mathematical economics, which, at least in the modeling sense, is a statistical “science”. While economics itself, generally, sorry to some good people here, is not. Economists do not usually have the option of performing significant peer-reviewed experiments, or the need: the "successful" ones do not need to be validated, just repeated as necessary, while the experiments that fail usually must count millions of victims among their "results".
What we who were schooled in the dismal un-science can (hopefully) do is understand the efforts of the physical ones and relate the implications to economic phenomena and mechanisms. Then we can do our real job, which is to suggest useful policies to achieve desirable sociopolitical goals. The most important of which right now is ensuring the survival of our species along with as many others as we humans can manage not to kill off in the process.
Or, equally usefully, we can show the real and opportunity costs of avoiding unpalatable political and business actions based upon an utter ignorance of and disdain for science and its practitioners.
By all means, expose the scientific inadequacies of Mr. Chen et al, and have some fun at the expense of Mr. Paulson, and Mr. Tol, and even (the non-policy brother) Mr. Nordhaus when they miss the climate science mark. But please remember we are all inside the same greenhouse, and all equally committed to its safe preservation.
End of mini-rant.
Thanks to all who participate here, but very specially to those among you who give your time and considerable expertise to help make Mr. Cook’s vision such a magnificent educational resource. You have taught this ancient numbers guy much in the last five years.