Vemdalsklimat

Jag är nyligen hemkommen från en mycket givande konferens om klimatvetenskap och statistik i Vemdalen. Fyra världsledande forskare på området hade flugits in från utlandet: Caitlin Buck, Douglas Nychka, Andrew Parnell och Richard Smith. Deras föreläsningar, och diskussionerna däremellan, fokuserade, som så ofta i statistiska sammanhang, på (de ofta mycket stora) osäkerheterna i klimatologiska studier. Vi hann också med en del skidåkning mellan varven. Som ett lite lättsammare komplement till de vetenskapligt tunga inslagen erbjöds jag tillfälle att föreläsa om debatten med klimatskeptiker, och jag infogar mitt föreläsningsmanus nedan. Trogna UI-läsare kommer att känna det mesta av sakinnehållet, men disposition och språkdräkt är nya.

En personlig reflektion: När man som jag ägnat så mycket tid åt debatten med klimatskeptiker känns ett möte av det här slaget ljuvligt befriande. I den öppna vetenskapliga diskussionen utan politiska agendor behövde ingen oroa sig för att det i publiken skulle sitta någon mindre nogräknad figur på jakt efter smaskiga citat att lyfta ur sammanhanget och förvränga.

Vid sidan av mitt eget föredrag var det en tämligen liten del av konferensen som ägnades de vanliga diskussionspunkterna i debatten med klimatskeptiker (detta är ganska typiskt: medan klimatskeptikerna fastnar i enkla missförstånd går vetenskapen vidare mot lösandet av klimatvetenskapens verkliga problem). Som ett undantag som eventuellt kan intressera UI:s läsare kan nämnas en uppsats från 2007 med Nychka som medförfattare som kom upp till diskussion. Uppsatsen visar vad som händer om man på Michael Manns ursprungliga hockeyklubbsdata från slutet av 90-talet applicerar nya och vassare statistikare metoder än Mann själv mäktade med; resultaten visar att ryktena om hockeyklubbans död är, lindrigt talat, överdrivna.

* * *

Confronting the Climate Skeptics

Olle Häggström

41st Winter Conference in Statistics, Vemdalen, March 2010

I was drawn into public confrontation with climate skeptics in Sweden more or less by accident in 2008, and expected to be able to move on just as quickly. Things turned out otherwise. The debate has consumed a lot of my time, but it has also been rewarding by forcing me to learn a lot about climate science – an extremely interesting topic. Today I would like to report to you some of my experiences from the debate.

For the benefit of our foreign guests, let me begin by saying a few words about the situation in Sweden. It is much better than in the United States, where climate skeptics constitute a large enough minority in congress to have disastrous political influence. In the Sweden only some three or four members of Riksdagen (our national parliament) endorse a climate skeptic agenda. But it is not clear how stable this situation is. The climate skeptics have become increasingly visible on the opinion pages of newspapers and elsewhere, and recently Alf Svensson - well-known EU-parliamentarian and former leader of the Christian Democrats in Sweden - came out as a climate skeptic.

Let me apologize for the title of the talk. The term climate skeptic is not a good one. Especially not in spoken language, where it is cumbersome to distinguish between “skeptic” and skeptic, with and without quotation marks:

• A skeptic (without quotation marks) employs the kind of critical thinking that we all know as absolutely central to the scientific method:
  
  Are my data representative of the population I’m interested in? Are the data analysed by Jones et al representative of the population they claim conclusions about? Are the sources I’m using for my study reliable? Am I really entitled to reverse the order of summation and taking limits in my derivation of the expected value? Are the variations in humidity in my laboratory really small enough to be negligible? Can my experiment be replicated? Etc, etc.
  
  


• A “skeptic” (with quotation marks) dislikes some particular scientific theory (such as evolutionary biology or the theory of anthropogenic global warming) and is happy to pick up any argument or any one-liner that seems to fit his agenda (such as “what use is half an eye or half a wing?” or “considering how much more water vapor than carbon dioxide there is in the atmosphere it seems highly unlikely that carbon dioxide should matter much for the climate”) and then using them without stopping for a second to check if the arguments hold up to scrutiny.
  

Now, what word should we use instead? There is no really good word. In more informal circumstances I’ve sometimes used the term climate denialist, but that has the downside that it often triggers unpleasant discussions along the lines of “are you trying to compare us to Holocaust denialists?”, and if I then express my truthful opinion about the relevance of such a comparison, all chances of further communication go straight out the window. By the way, I’ve heard the following complaint twice from Sten Kaijser, an old friend, mathematician and nowadays a self-proclaimed climate skeptic:

Climate denialist is such a silly word! Nobody has ever denied the existence of climate. We prefer to call ourselves climate skeptics.

Well, professor Kaijser, does this mean you are skeptical about the existence of climate?

But all right, I will stick to climate skeptic for the rest of this talk, with quotation marks implicit and hopefully no risk of confusion.¹

*

Terminology aside, let me mention some recent trends in this debate.

• Over the last couple of years, there has been a tendency towards somewhat more sophisticated arguments among climate skeptics. Two years ago, they would typically rant about things like (a) how there is no reason to believe that the increase in atmospheric CO2 has anything to do with our burning of fossil fuels, or (b) how the greenhouse effect is either non-existent or independent of further increase in CO2 levels. Though such arguments can sometimes still be heard, an increasing number of climate skeptics have found (a) and (b) untenable and switched to a different tactic: accept anthropogenic CO2 increase, accept the greenhouse effect, but claim that the climate sensitivity of CO2, defined as the equilibrium increase in global average temperature resulting from a doubling of CO2 concentration², is much smaller than what mainstream science claims. The interval given by IPCC as likely, reflecting scientific consensus, is [2 °C, 4.5 °C] with a best estimate of 3 °C and the remark that values below 1.5 °C are “very unlikely”. Climate skeptics of this “new, sophisticated” kind claim that climate sensitivity is more like 1 °C (which is what you get theoretically using the Stefaan-Boltzmann law if you only take into account the greenhouse effect and ignore all feedbacks). This is an improvement over the cranky and unphysical denial of the greenhouse effect (it is not outright unphysical to think that the known positive feedbacks are counteracted by negative feedbacks of the same strength) but does require an extremely selective reading of the scientific evidence and literature.³
  


• On a much shorter time scale – just the last few months, after the email break-in known as ClimateGate, after the Copenhagen failure, and after the reports about errors in the IPCC 4th Assessment Report – it seems to me that the climate skeptics have gained a lot of ground in the debate. It’s not just that the climate skeptics themselves have become more visible. Also, and perhaps more significantly, mainstream journalists have increasingly begun to accept the climate skeptics’ take on climate science, phrasing issues in their language, etc. I will give an example of this below: the treatment of the BBC interview of February 7, 2010, with Phil Jones, director (temporarily stepped aside) of the Climate Research Unit at East Anglia University.
  


• Another extremely worrying trend is how groups of climate skeptics have begun to try to intimidate climate scientists using not only rhetoric but also with hooliganism and even crime, such as:
  
  
  
  • ClimateGate.
    
  • McIntyre’s⁴ FOIA (Freedom of Information Act) organized campaign against CRU at East Anglia, aimed at spamming and harassing CRU researchers with requests for data.
    
  • At a panel discussion with, among others, Sweden’s most popular climate skeptic blogger Maggie Thauersköld Crusell, in Stockholm in November 2009, she had managed to mobilize quite a crowd of supporters, who used coordinated scornful laughter (triggered e.g. by references to IPCC) as a means of intimidating the rest of us. The same crowd has employed the same extremely unpleasant tactic before, but this was the first time I experienced such a thing. (Needless to say, scornful laughter is not a scientific argument.)
    
  • Breaking news from climatesciencewatch.org on February 24, 2010: “Senator James Inhofe, ranking Republican on the Environment and Public Works Committee, has gone a step beyond promoting his long-notorious global warming denialist propaganda. He is now using the resources of the Senate committee to seek opportunities to criminalize the actions of 17 leading scientists who have been associated with the Intergovernmental Panel on Climate Change assessment reports. A report released by Inhofe’s staff on February 23 outlines this classic Joe McCarthyite witch-hunt: page after page of incorrect and misleading statements, a list of federal laws that allegedly may make scientists subject to prosecution by the U.S. Justice Department, and a list of names and affiliations of 17 “key players” in the “CRU Controversy” over stolen e-mails and their connections with IPCC reports.”
    

The really tragic aspect of these last two trends is that if the climate skeptics are successful in their campaigns, their description of climate science as a fortress defending unquestionable old dogmas may turn into a self-fulfilling prophecy. If climate science is constantly under such siege, there may be a real risk that the scientists themselves become guarded and refrain from discussing, e.g., uncertainties in the open and open-minded way we have seen at this conference.

*

Here’s an example of the kind of twisted news coverage that has become increasingly common lately. The Swedish radio news programme “Dagens eko” is normally pretty much the most respected and trustworthy news channel we have in Sweden. But listen to what they said on February 15, 2010:

The critical examination of the United Nations climate panel IPCC and their work continues, and now it is one of the leading characters who says in an interview with British BBC that he is no longer so sure about global warming. It is none other than the director of CRU at the University of East Anglia, professor Phil Jones, who says that there are no statistically significant proofs of global warming during the last 15 years.

And listen to the original interview:

Q: Do you agree that from 1995 to the present there has been no statistically-significant global warming?

A: Yes, but only just. I also calculated the trend for the period 1995 to 2009. This trend (0.12C per decade) is positive, but not significant at the 95% significance level. […] Achieving statistical significance in scientific terms is much more likely for longer periods, and much less likely for shorter periods.

[...]

Q: How confident are you that warming has taken place and that humans are mainly responsible?

A: I'm 100% confident that the climate has warmed. As to the second question, I would go along with IPCC Chapter 9 - there's evidence that most of the warming since the 1950s is due to human activity.

Of course it is bizarre to translate the non-significance of a single data series into doubt of the phenomenon at hand. It seems that Jones used the HadleyCRU data for global average temperature, which barely fails to give significance at the 95% level. The GISS series (Goddard Institute for Space Sciences) on the other hand, does give a statistically significant increase. But never mind. Besides the general error of focusing on the statistical significance in a single time series rather than on the total weight of available evidence⁵, there is in this case also the issue of whether the null hypothesis is relevant at all. When you push the linear regression button in your favorite statistical software package, the p-value it produces is based on a null hypothesis not only of a zero trend, but also of annual deviations from the mean being uncorrelated. This seems to me like the wrong null hypothesis to use, given the variations we have on biennial to decadal time scales such as the El Nino phenomenon and the sun spot cycle. Modelling this is much harder and requires deep understanding of climatology (i.e. cannot be done using statistical competence alone), because, e.g., any observed time series is consistent withe a null hypothesis of zero trend and sufficiently strongly correlated noise.

When you look at the BBC interview, Phil Jones seems to do pretty much everything right, scientifically, and yet the outcome, in terms of how he is represented in news and blogs around the world, is so bad. By giving straight honest to-the-point answers he exposes himself to cherrypicking, quote mining and distortion. Is it the case that if/when the media climate turns hostile, we (the scientists) are screwed whichever way we turn?

*

I am often asked about the psychology of climate skeptics. Do they really believe what they are saying or are they lying intentionally? I am somewhat divided as to whether it is a good idea to talk about this. On one hand, psychologizing the opponent takes attention away from the real issues, and carries connotations of Soviet mental institutions. On the other hand, it is a fun topic to speculate about, so what the heck, let me say a few words then:

• I know several of the leading Swedish climate skeptics more or less personally, and I find no reason to suspect that any of them don’t believe what they are saying about the climate threat being bogus and a result of junk science, etc.
  
  
  
  • Caveat: With some of them, in particular Lars Bern and Maggie Thauersköld Crusell, I suspect they realize that not every detail of what they say holds up to scrutiny. But this doesn’t matter (they say to themselves) because as long as you are right about the big picture the details don’t matter, and moreover if their opponents stretch the truth then it’s OK for them to do the same. Thauersköld Crusell has made some revealing slips on her blog:
    
    
    
    • About New York Times science journalist Andrew Revkin, September 24, 2009: “I like Revkin. He is not a climate skeptic per se, but more like a real skeptic, weighing arguments against each other.”
      
    • October 13, 2009: “Cherrypicking – we all know what it means. Picking out precisely those data that support our thesis. Don’t we all do it, more or less consciously? It would be incredibly stupid to defend one’s hypothesis by stating arguments to the contrary.”
      
    
    
  
  
• The Dunning-Kruger effect, according to Wikipedia, is defined as the cognitive bias in which "people reach erroneous conclusions and make unfortunate choices but their incompetence robs them of the metacognitive ability to realize it". I think this phenomenon is very widespread, and that pretty much everyone who becomes a climate skeptic begins by realizing something along the lines of “Hey, we went back and forth between ice ages and interglacials long before there were cars or coal power plants, so the climate scientists must be all wrong about what causes climate change”. After some time they typically progress towards somewhat more sophisticated arguments, but their basic shortcoming remains the same. The metacognitive ability I wish they had is to stop for a moment and think “Among thousands of climate scientists, could it really be that none of them have thought about this? Perhaps I should at least check this, and see what they have to say about it”. It seems to me a tad megalomanic to think that one can out-think an entire scientific discipline just like that. Or consider what Swedish journalist and TV personality Elisabet Höglund wrote in Aftonbladet on February 23, 2010, concerning the decrease in the number of weather stations in Siberia following the fall of communism:
  
  When temperature in no longer measured in cold and vast areas of the Earth, such as the huge Siberia, well, then this affects global temperature. If the low temperatures disappear from statistics, then the average temperature goes up!
  
  
• I fear that I and others can sometimes create climate skeptics out of people who are just luke-warm, ignorant, and not terribly interested in the issue, if we push them too hard. Consider, for instance, the meeting on climate and environmental statistics held by the Cramér Society last fall in Lund, and the report from the meeting written by Mikael Möller and Rolf Larsson and published in the Swedish Statistical Association newsletter Qvintensen. These are two highly esteemed colleagues who, however, as regards climate are precisely “luke-warm, ignorant, and not terribly interested”. Their ignorance is reflected in their report where they express regret over the fact that only one side of the scientific debate was represented at the meeting, they point out that there are in fact scientists who claim that human activities are not the only factor behind climate change, and they stress that there is controversy over what has come first: changes in temperature or in CO2 levels.⁶ As many of you have probably seen, I wrote a reply in the same issue of Qvintensen to correct their misunderstandings, but I did it with some agony and tried to be nice (something that does not always come naturally to me…) so as not to push them into the corner of climate skepticism.
  

Footnotes


1) After having ranted for a while about the sorry state of climate skepticism, I am sometimes asked whether there aren't any good climate skeptics, in the real (quotation mark-free) sense of the term. Well, there are plenty! From what we've seen so far at this conference, I'd say that all four main speakers (Caitlin Buck, Douglas Nychka, Andrew Parnell and Richard Smith) fit beautifully the notion of a climate skeptic, in the true quotation mark-free sense of the term.

2) What makes this a natural baseline quantity is that the greenhouse warming is approximately logarithmic in the CO2 concentration. This sounds nice, because it means that the more CO2 we add, the less is the effect of an extra ppm, and it is therefore no surprise that this is pretty much the only result in climate science that is never the target of the rhetoric of climate skeptics.

3) Sometimes a climate sensitivity of around 1.5 °C is obtained as follows. Global average temperature is plotted against CO2 concentration for two different years, say 1900 and 2009. Two points is enough to specify the logarithmic curve, and the climate sensitivity can immediately be read off as about 1.5 °C. However, this simplistic method ignores a number of factors that strongly influence the result, including climate inertia (not all of the warming corresponding to anthropogenic CO2 increase has happened yet) and aerosols (their cooling net effect – sunshield – hides much of the CO2-caused increase). There is much uncertainty about these quantities (something that of course does not warrant setting them equal to zero) but with reasonable values plugged into the calculation they tend to lead to climate sensitivities well within the IPCC range.
A favorite ploy by climate skeptics (which in particular retired Swedish oceanographer Gösta Walin likes to promote) is a variation of this calculation where they grant themselves a discount of a couple of tenths of a degree from the 20th century warming of about 0.7 °C, something that scales down the climate sensitivity proportionally, yielding about 1 °C. Why the discount? Because “even the IPCC agrees that there exist other factors causing climate change, beyond CO2 emissions”. This is beyond sloppiness…

4) Steve McIntyre, best known for his persistent harassment over several years of hockey stick pioneer Michael Mann.

5) See, e.g., The Cult of Statistical Significance: How the Standard Error Costs us Jobs, Justice and Lives by Stephen Ziliak and Deirdre McCloskey (2008), reviewed here.

6) In transitions from ice ages to interglacials and vice versa, temperature rather than CO2 is the prime mover. There is in fact no scientific controversy about this paleoclimatological result. I wrote about the interesting interplay between CO2 and temperature here.