Guest commentary from Barry Bickmore (repost)
The Wall Street Journal posted yet another op-ed by 16 scientists and engineers, which even include a few climate scientists(!!!). Here is the editor’s note to explain the context.
Editor’s Note: The authors of the following letter, listed below, are also the signatories of“No Need to Panic About Global Warming,” an op-ed that appeared in the Journal on January 27. This letter responds to criticisms of the op-ed made by Kevin Trenberth and 37 others in a letter published Feb. 1, and by Robert Byer of the American Physical Society in a letter published Feb. 6.
A relative sent me the article, asking for my thoughts on it. Here’s what I said in response.
Hi [Name Removed],
I don’t have time to do a full reply, but I’ll take apart a few of their main points.
- The WSJ authors’ main point is that if the data doesn’t conform to predictions, the theory is “falsified”. They claim to show that global mean temperature data hasn’t conformed to climate model predictions, and so the models are falsified.
- The WSJ authors say that, although something like 97% of actively publishing climate scientists agree that humans are causing “significant” global warming, there really is a lot of disagreement about how much humans contribute to the total. The 97% figure comes from a 2009 study by Doran and Zimmerman.
- The WSJ authors further imply that the “scientific establishment” is out to quash any dissent. So even if almost all the papers about climate change go along with the consensus, maybe that’s because the Evil Empire is keeping out those droves of contrarian scientists that exist… somewhere.
But let’s look at the graph. They have a temperature plot, which wiggles all over the place, and then they have 4 straight lines that are supposed to represent the model predictions. The line for the IPCC First Assessment Report is clearly way off, but back in 1990 the climate models didn’t include important things like ocean circulation, so that’s hardly surprising. The lines for the next 3 IPCC reports are very similar to one another, though. What the authors don’t tell you is that the lines they plot are really just the average long-term slopes of a bunch of different models. The individual models actually predict that the temperature will go up and down for a few years at a time, but the long-term slope (30 years or more) will be about what those straight lines say. Given that these lines are supposed to be average, long-term slopes, take a look at the temperature data and try to estimate whether the overall slope of the data is similar to the slopes of those three lines (from the 1995, 2001, and 2007 IPCC reports). If you were to calculate the slope of the data WITH error bars, the model predictions would very likely be in that range.
Comparison of the spread of actual IPCC projections (2007) with observations of annual mean temperatures
That brings up another point. All climate models include parameters that aren’t known precisely, so the model projections have to include that uncertainty to be meaningful. And yet, the WSJ authors don’t provide any error bars of any kind! The fact is that if they did so, you would clearly see that the global mean temperature has wiggled around inside those error bars, just like it was supposed to.
So before I go on, let me be blunt about these guys. They know about error bars. They know that it’s meaningless, in a “noisy” system like global climate, to compare projected long-term trends to just a few years of data. And yet, they did. Why? I’ll let you decide.
So they don’t like Doran and Zimmerman’s survey, and they would have liked more detailed questions. After all, D&Z asked respondents to say whether they thought humans were causing “significant” temperature change, and who’s to say what is “significant”? So is there no real consensus on the question of how much humans are contributing?
First, every single national/international scientific organization with expertise in this area and every single national academy of science, has issued a statement saying that humans are causing significant global warming, and we ought to do something about it. So they are saying that the human contribution is “significant” enough that we need to worry about it and can/should do something about it. This could not happen unless there was a VERY strong majority of experts. Here is a nice graphic to illustrate this point (H/T Adam Siegel).
But what if these statements are suppressing significant minority views–say 20%. We could do a literature survey and see what percentage of papers published question the consensus. Naomi Oreskes (a prominent science historian) did this in 2004 (see also her WaPo opinion column), surveying a random sample of 928 papers that showed up in a standard database with the search phrase “global climate change” during 1993-2003. Some of the papers didn’t really address the consensus, but many did explicitly or implicitly support it. She didn’t find a single one that went against the consensus. Now, obviously there were some contrarian papers published during that period, but I’ve done some of my own not-very-careful work on this question (using different search terms), and I estimate that during 1993-2003, less than 1% of the peer-reviewed scientific literature on climate change was contrarian.
Another study, published in the Proceedings of the National Academy of Sciences in 2010 (Anderegg et al, 2010), looked at the consensus question from a different angle. I’ll let you read it if you want.
Once again, the WSJ authors (at least the few that actually study climate for a living) know very well that they are a tiny minority. So why don’t they just admit that and try to convince people on the basis of evidence, rather than lack of consensus? Well, if their evidence is on par with the graph they produced, maybe their time is well spent trying to cloud the consensus issue.
The WSJ authors give a couple examples, both of which are ridiculous, but I have personal experience with the Remote Sensing article by Spencer and Braswell, so I’ll address that one. The fact is that Spencer and Braswell published a paper in which they made statistical claims about the difference between some data sets without actually calculating error bars, which is a big no-no, and if they had done the statistics, it would have shown that their conclusions could not be statistically supported. They also said they analyzed certain data, but then left some of it out of the Results that just happened to completely undercut their main claims. This is serious, serious stuff, and it’s no wonder Wolfgang Wagner resigned from his editorship–not because of political pressure, but because he didn’t want his fledgling journal to get a reputation for publishing any nonsense anybody sends in.[Ed. See this discussion]
The level of deception by the WSJ authors and others like them is absolutely astonishing to me.
Barry
PS. Here is a recent post at RealClimate that puts the nonsense about climate models being “falsified” in perspective. The fact is that they aren’t doing too badly, except that they severely UNDERestimate the Arctic sea ice melt rate.
Well, there are engineers and then there are engineers. Some are more open to received wisdom than others.
That presumes the audience is honestly applying suitable skills to the case and data that you bring. I would suggest that many “skeptical” engineers are not. Hopefully I’m not applying overkill via repetition, but for a recent and impressive case study go see Burt Rutan in action (including the comment thread) denying almost entirely the robustness of the critiques of his horribly flawed “anti-case” and almost the entirety of the scientific case, despite having been provided with large amounts of quality evidence and argument.
This is probably only true if you want detailed predictions. The basics of radiative equilibrium constrain the system in ways that are not difficult to understand – especially for engineers – and that have some fairly straightforward consequences. You can probably do a reasonable analysis of the risks of consequences inferred at that level of analysis without recourse to a detailed model. (Most “skeptics” don’t as far as I know…)
And then if you think the information at hand is too uncertain and/or the system is to complex (often considered to be two sides of the same coin) to be sure what’s going to happen, you should apply standard risk mitigation analysis: do you screw with atmospheric composition in ways that you know will have an effect and then hope like heck it turns out OK, or do you stay within an envelope where you know humanity and the ecosystem it relies upon for survival and prosperity have done well over the last few thousand years?
Uncertainty and complexity is not the friend of the “let’s screw with it and hope things will be fine” strategy.
Oh, and BTW and FWIW “I are also an engineer”.
A bit late, but I wanted to point out that the first signer, Claude Allègre, forgot to mention something. He is not only a former director of a science institute, he has integrated the political staff of the president-candidate Sarkozy.
But don’t forget, he is a “free thinker” “outside the politics” [/sarcasm]. Double standard to the extreme.
Remember engineers and IT people are users of science, not practitioners of science. From the outside science looks much more cut and dried than it is.
I’m more familiar with the blind spots of IT people than I am with those of engineers. IT people are thinking in terms of the single error which will propagate and bring everything down. The systems that they work with can have this. They aren’t used to thinking in terms of things that are only approximately right, of systems that are like jigsae puzzles where most pieces fit imperfectly but well enough. And they look for a single conclusive line of proof. Scientists look for multiple independent proofs. They think in terms of consilience. That is not in the intuitons of IT people. I can understand why so many of them can get fooled by denialists.
Dbostrom, I did not claim that engineers were myopic (that was MA Rodger. I feel it necessary to correct this misattribution so that I do not get in (more) trouble with my wife–who trained as an EE and then became an environmental scientist.
I think all this speculation of why engineers or IT types or physicists, etc. fall victim to crank theories misses the point. The real point is that one can obtain an advanced degree in science, engineering, IT, etc. without really having the foggiest idea of how science works. Hell, one can even have a fairly successful career in science without fully understanding why the techniques you are applying work. In some ways, this is both the strength and the weakness of science–the techniques of science are relatively easy to learn and apply, but understanding why they work is a deep epistemological exercise. There aren’t a lot of folks with the patience for such an exercise. When such an impatient scientist is also an arrogant one, then you get a recipe for idiocy every time he ventures out of his narrow area of expertise.
If anyone remembers Anteros’ comments upthread and was confused, SkS has posted on this. In sum, Anteros was either confused or deliberately misleading:
Wall Street Journal ’Skeptics’ Misrepresent the IPCC
http://www.skepticalscience.com/wsj-skeptics-misrepresent-ipcc.html
Ray: Dbostrom, I did not claim…
Sorry!
CO2 effects on climate are an impulse response only on geological time scales. “…20–35% of the CO2 remains in the atmosphere after equilibration with the ocean (2–20 centuries). Neutralization by CaCO3 draws the airborne fraction down further on timescales of 3 to 7 kyr.” As far as human society from an historical perspective, it’s a step response.
Anteros has appeared in comments over there, arguing much the same as here IIRC (and asserting that the authoritative FAR predictions are the textual descriptions and NOT the graphs, which is a curious position to take unless he has some supporting evidence).
And he’s still making a great deal of noise about prediction vs projection, and is now claiming that “the emissions scenario was the big error in 1990″(!) and so forth.
If I understand Anteros’ “argument” correctly, it is that the FAR made “predictions”, and the textual description of them trumps the graphs, and the trend over the century validly applies to each decade including the early ones, and the BAU scenario [despite not eventuating] is a valid choice to use when comparing “predictions” to subsequently observed temperatures [in fact, he seems to argue that it is the only scenario for which a “prediction” was made despite commenters providing quotes that refute this], and that all of that means the trend line depicted in the WSJ was an eminently reasonable choice (and under no circumstances was it the “highest” choice they could have made because the high sensitivity model went even higher under BAU).
Trouble is, if we use an emissions scenario close to what was realised the WSJ trend line was significantly higher than the highest choice on offer (i.e. the high sensitivity model), no matter whether you go by the text or the graphs.
> the IPCC says they are predictions.
When they did use that word, they used it to describe what nowadays are called scenarios.
The IPCC does, at 5-year intervals, report on the current state of the science.
What do you think they call that kind of thing nowadays?
Interesting exchange on the topic of engineers. I must add my (late) €0.02, because I’m currently doing my MSc thesis work in engineering fluid dynamics, a postgraduate degree of mechanical engineering at my university. And I also don’t recognise myself in the image of stubbornness that emerges now and then.
Since I’m Dutch, I think I should elaborate slightly on our education system, as there may be differences with the Anglophone world. Engineering degrees can be obtained at “HBO” level (tertiary vocational education/college) as a 4-year B-Eng curriculum, distinguished from university level, where it is a 3yr BSc + 2yr MSc programme. The former degree is more focused on direct application of the engineering sciences, places lighter requirements on prior high school education and is taught in classes with (often) mandatory attendance. The latter is inherently academic, and hence requires students to become deeply familiar with the theoretical backgrounds. Consequently, a much heavier demand is placed on mathematical knowledge, and students are advised to choose their high school curriculum accordingly.
So, one could say these are two different types of engineers, although it is possible to complete a university MSc degree with a HBO B-Eng degree, after following a 1-year deficiency (“pre-master”) track.
In other words, students who want to work in the field of engineering, but are less interested in where the technology comes from (research), will generally go for the B-Eng degree, and I think that is the biggest distinction between people’s mindsets (as Ray also points out above). Because, speaking for myself, the academic degree requires and trains you to constantly evaluate your assumptions; especially in fluid mechanics you can’t take much for granted. There is no solid ground in the form of generally valid rules of thumb, tables of data, etc. that you can fall back to when researching new technology. It must first hold up to the remorseless scientific method, you can’t skip any steps in the chain of evidence. It is hence a thorough exercise in humility, which usually expresses itself in the subtle and careful way scientists communicate. Never 100% sure. Or, the more one knows, the more one knows how little one knows. I have learned to be cautious about people who appear incredibly confident or treat scientific findings as Boolean values. This is, however, all too often the way these are presented in mainstream media.