Thursday, August 17, 2006

The IPCC

In a recent article that's getting lots of play in the blogosphere, John Fleck wrote:
So how should the public and policymakers, whipsawed by debate, sort out the competing claims? "That's easy," said Dessler, co-author of Cambridge University Press's "The Science and Politics of Global Climate Change." Dessler calls for the use of "expert assessments"- panels of specialists brought together to sort out and summarize scientific information for politicians, policymakers and the public. It's a common technique on all sorts of science-policy questions.

On climate change, a number of such reviews have been done, including work by the Intergovernmental Panel on Climate Change and the National Academy of Sciences. The panels' findings have been consistent and are reflected in the key finding in the IPCC's 2001 "Climate Change: The Science": Greenhouse gas emissions are altering Earth's climate.

Pielke Sr. has a different answer- listen to more diverse scientific voices. He thinks expert panels like the IPCC are inbred, representing a narrow focus on greenhouse gases. "The public is getting a very narrow view of the breadth of issues in climate science," he said.
There's a really interesting point here. I agree with Pielke Sr. that assessments have to include the range of opinions held by the majority of the scientific community. If half of the community says one thing and the other half says something else, then a good assessment will include both views. My sense is that the recent AGU assessment on hurricanes failed in this regard (see also this Prometheus post).

However, we have to be careful about including views that are "too diverse." The goal of a scientific assessment is to let policymakers know what the scientific community thinks on an issue. If 99.9% of the scientists think one thing (e.g., the Earth is warming), then should the assessment include a dissenting view held by 0.1% of the community? I would argue not. Policymakers are often unable to discern a 99.9% position from a 0.1% position, and they look to expert assessments to do just that. If they don't, then the assessment has essentially abdicated its responsibility to clarify the science for the policymakers.

Obviously, this is one of the hard choices that the authors of assessments have. They have to use their judgment about what positions are credible within the majority community and what positions are simply outliers. Writing assessments is a tricky business, but they play such a crucial role that scientific community has to continue to work hard to make them the best possible source of information for policymakers.

18 comments:

Tigre said...

while i am a conservative Republican, i do admit that global warming does exist and is an important issue. That's one of the things i don't agree with my party on. I just hope we can all come on board some day and do something about it. Unfortunately, the polarized political climate we currently live in will do all but bring us on board together.

Anonymous said...

Andrew, the difficulty with the NHC/Gray camp is that they aren't producing the science. Landsea has threatened that he's going to produce it, but hasn't done so (and even when he does, is there a chance it will find the additional *150* new strong hurricanes needed to overturn Webster et al?), Chan's (yet *another* former Gray student) critique didn't amount to anything, Michaels et al was the expected joke, Phil Klotzbach's paper gave too much weight to a historical pattern Kerry Emanuel had already noted, and Chelliah admitted over at RC that his and Bell's stuff doesn't make the needed case. This contrasts to what's getting to be a pretty big stack of papers on the AGW attribution side (and see the latest from Jim Elsner). From an outside perspective, the controversy appears to be nearly settled.

Anonymous said...

Andrew- Scientific assessments shouldn't summarize scientists' opinions. Opinion polls can do that. Scientific assessments should summarize the peer reviewed literature and do so comprehensively. This would mean acknowledging the diversity of views that appear in the legitimate, peer-reviewed literature. Selectively omitting peer reviewed science would represent a failure to assess the literature. The IPCC gets in trouble when it decides not to include or acknowledge certain peer-reviewed perspectives in its report, a point made not only by RP Sr. but also James Hansen. For the IPCC to say that certain peer reviewed science is "not credible" makes a string statement about the peer review process that produced the omitted science.

Anonymous said...

John Fleck said:

Steve - Re Landsea's recent contributions to the peer reviewed literature, I found Landsea et al, Can We Detect Trends in Extreme Tropical Cyclones?, Science 28 July 2006: Vol. 313. no. 5786, pp. 452 - 454, to be quite useful.

Andrew Dessler said...

Roger-

Thanks for your comments. I have to disagree, and I'll present the following scenario to illustrate my point. Let's consider Lindzen's iris hypothesis. Everyone now considers that wrong (except Lindzen and Gray). Even though it's in the peer-reviewed literature, I believe it would be wrong to give that idea any credence. If you were to insist that it be included, I suppose I could live with including a statement saying basically that no one believes it.

That's the job of the assessment --- go through the literature and determine what's been repeatedly verified and has been accepted by the community. Just because something is peer reviewed does not make it right, as you well know. It is the job of the assessment to make these distinctions, and if they don't, then they force policymakers to do so, which is a fundamental abdication of the scientific community's responsibility.

Regards.

Andrew Dessler said...

Steve-

I see your point. However, I would just respond that when I talk to scientists at meetings, no one is really all that certain that we can detect any increase in the strength of hurricanes (although most people agree that the signal is there, it's just really really hard to detangle from the noise). I felt the AGU statement did a poor job of communicating the squishiness of the conclusion. Perhaps as additional work is done, this will change.

Regards.

Anonymous said...

Dr. Dessler, as a layman and speaking as someone who would have to depend on the information and views put forth by the panels, I would agree with the 99.9% idea you mention if the .01% was something really off the deep end, but if there is responsible, professional dissent even from just a handful of professionals, I would want to hear it.

I think it would be advisable in terms of achieving public letigimacy for the majority view on the panel, for several reasons, but in the interests of brevity I will point out the most important:

The figure of the lone dissenter or tiny group of dissenters who eventually turn out to be right or even **just partly right** is deeply instilled in our national as well as our individual experience.

In the US "collective consciousness" the presumption is always in favor of the right of an individual --- especially a qualified individual --- to be heard above the crowd, and exclusion **always** suggests in the public mind an attempt to control the debate.

This can only have a negative impact on the attempt to legitimize the findings of such a panel.

Anonymous said...

Andrew- Presemuably Lindzen's work has been superceded in the literature. Science that has not been refuted in the literature should not be ignored. After all, isn't that what the literature is for?

Anonymous said...

Roger, you know perfectly well that plenty of flawed papers are just left to die a quiet citationless death. As well, plenty of papers are published because they pose ideas that are worth discussing even if very few people think they're right. Next, it seems very apparent that different journals have different publication standards. Given all of this, it makes no sense to treat all papers as equal unless they have been formally disproved. Even if that were so, there's a big distinction between that AGU assessment and a review article where such a standard might be arguably appropriate.

EliRabett said...

One simply observes where the noise is coming from and asks why.

Andrew Dessler said...

Roger-

We agree that science that has been refuted should not be included in an assessment. But the rub is what constitutes "refuted"? While most people would agree that the iris qualifies, Lindzen and a few friends would disagree. There is also the category of ideas that are not refuted, but for which (for any number of reasons) additional support has not appeared. In my view, it is up to the authors of the assessment to determine if an of these ideas are sufficiently supported to include in an assessment.

Tyler-

I see your point, but I still have a problem including the 0.1% viewpoint. While I agree with everything you say, there is an important downside --- that it gives that viewpoint more credibility than it deserves, and gives opponents of action an opportunity to needlessly debate science as a strategy to induce gridlock.

Regards.

Andrew Dessler said...

e-mailed from a reader:
I understand and appreciate the consensus on AGW and the CO2 role. However, am I correct in surmising that RPSnr's argument is not that CO2 is responsible for a good proportion of the 66-90% of the identified warming, but that CO2 is increasingly being identified as the main contributor to AGW (over other influences) only because researchers are tending to inappropriately apply the consensus as a form of occam's razor i.e. not as predicated by data analysis, but at the stage where their research question is being formulated, and therefore they are preventing themselves from investigating other influences (e.g. land use change) in a completely unbiased fashion?
If so, by how much do you feel, personally, that such a research approach could be restricting our broader understanding of the issue?

I'm not sure if I'm playing Devil's Advocate here or whether I'm really missing something.

>>>
my response: interesting question; however, I don't know RPSr. so I cannot speculate on how he arrived at his conclusions.

Anonymous said...

If the peer review process for science was 100% unbiased and without internal and external political or economic influences, I would tend to agree with your statement regarding how to select the viewpoints represented on a panel. However, we all know that within the scientific community there are groups and cliques that form around a given theory that will defend that theory ruthlessly, as their research funding is often highly dependent on maintaining the impression that their theory is the "mainstream" train of thought on the subject. Pick any field of study and look at the history of scientific study in that field since 1950, and you will see examples where new theories emerged and were immediately pounced on and the proponents were scorned by their colleagues as quacks with no data to back up their ideas. In many cases, those new theories turned out to be correct. Plate tectonics is one example from the field of geology and I am sure you have similar examples in your own fields of study. As recently as the mid to late 1960s, geology grad students at UT were discouraged from discussing volcanic features observed in West Texas field areas in their papers and presentations because of a particularly harsh public criticism that occurred at a national meeting during the plate tectonics/continental drift debate. Because their direct observations could not be justified under the "prevailing theory" derived from peer reviewed literature (continental drift) they were discouraged from discussing the problematic features in their own papers that would be subject to peer review. Thus for several years, geologists pursuing graduate degrees were forced to dance around the elephant in the middle of the room that they could clearly see (that west Texas had volcanism in its past) for fear of critical "peer review" that was based on protection of an incorrect theory.

I understand your concern about allowing every quack with an agenda and a theory to sit on a panel and muddy the waters around the eventual conclusions. But recognizing that the academic research world has its own political lines that are constantly being drawen, redrawn, and viciously defended will go a long way towards recognizing the flaws inherent in the way the current scientific peer review process reaches apparent consensus on an idea.

Anonymous said...

To respond briefly to John Fleck's earlier (fourth) comment:

John, I went back and carefully re-read that Landsea et al paper to make sure I wasn't missing anything. Essentially what it does is *threaten* to produce past TC reanalyses that will undermine the results of Webster et al, Emanuel etc.; other than that, sll it does is list some obvious anecdotes. The one place where it purports to apply those examples to generate a larger statistic, it does so in a very slippery way by using the basin with the smallest number of storms (so small as to be nearly meaningless in the context of a global analysis) that is also the most likely to have suffered from an undercount. In any case, the fact that there was an undercount prior to the '90s has been known in the TC science community for years. Indeed, the present reanalysis efforts were undertaken before the present controversy erupted. As Webster et al point out in their new BAMS paper, cancelling out their results would have necessarily involved mistaking a very large number of eastern hemisphere cat 4/5 storms as cat 1 storms. Well, maybe, but on the other hand a quick look at some satellite photos of cat 1s versus cat 4/5s causes me to question the likelihood of that. It's not as if the photos that Landsea's colleague is working from have been unavailable to others.

Another irony in this is that Landsea et al want to rely on past results from theory that the theoreticians have now said were an underestimate. In connection to that, how and why is it that the theoreticians wound up being the ones to do the needed TC meteorology studies? Something very strange is going on with that.

EliRabett said...

This has also broken out in Roger Pielke Jr.;s blog. He objects to the IPCC drafters of the AR4 wanting to have an impact on policy. Why the devil should they not want to and have a perceptible impact. They certainly know more about the issue than he does, and Roger definitely wants to have an impact. Is he better qualified than Pauchauri? Is the AR4 a drawing room exercise for retired climatologists to be filed and forgotten?

What Pileke is not admitting is that AR4 WGI report undoubtedly says that there is a major effect of green house gas emissions on climate, and WGII says that the effects will be not good and will get worse if no action is taken. That, of course, is a nono in Pielkeland. What Pielke is doing is preparing the ground for an attack on the AR4 based on his typical misdirection.

Roger should take Bismark to heart. The making of laws, policy and sausage have much in common. The prissy need not play. On the other hand, you can't play nice nice with a manipulator. You accept his misdefinitions and you are lost.

Anonymous said...

Dr. Dessler, yes, I see your point and am also cognizant of this.

And it occurs to me that an effective level of inclusion also depends on what is meant by the ".01%".

For example, if this refers not to broader areas but to disagreements over finer details, including the ".01%" could make the panel size unwieldly and really muddy up the water.

I am just pointing out that on the **broadest issues** there must be recognition of the general viewpoint of the dissenters or you'll have the proverbial elephant in the room. Even if it is just one panel member out of 5 who presents a summary of the four or five main lines of dissent, no one could then accuse the panel of ignoring the dissenters as a group, no matter what its size.

Anonymous said...

hey andrew,

i don't want to put words in rp sr.'s mouth, but i believe part of his issue is that there is a distinction in a scientific assessment between (a) answering the positive questions that frame the assessment and (b) choosing these framing questions in the first place.

if the ipcc was charged with assessing all the diverse human influences on the climate systems and their implications and relative importances, i believe roger sr. would be more satisfied. instead, the ipcc is focused (in practice at least) on assembling evidence to support the hypothesis that emissions of greenhouse gases and (to an extent) aerosols can explain the observed climate change over the last century+ (primarily using the metrics of global-mean surface air temp and sea level) and can plausibly be expected to create further changes in these metrics moving forward into the future.

i think this latter question is an important one, and i think ipcc does a good job of addressing it. however, the very fact of framing the ipcc around this question means that the authors, the process, and the report will view the scientific literature through this lens. if i wanted an assessment that focused on, say, vulnerability to regional-scale climate variability and change, and this change as a function of many different forcings (not just ghgs), i would need a different body producing a different report.

choosing which assessment (i.e., which framing questions) is the best use of everyone's time and resources goes back, of course, to the purpose of the assessment and the policy discussion it is meant to inform. one's feelings about the value of the assessment will likely reflect one's feelings about the value of the policy discussion that provide its raison d'etre.

Anonymous said...

this was very good and tought me a whole lot... thanks to all that wrote this....!