Climate Scientist Questions Consensus Process

IEEE Spectrum talks with climate panel veteran Michael Oppenheimer

5 min read

Members of the Intergovernmental Panel on Climate Change (IPCC) have a tricky task. After aggregating the most credible research on the causes and impacts of global climate change, the group must somehow package it into a report that policy-makers and the public can digest. In the process, sometimes they omit contentious and hotly debated items to get their main points across. But there is discontent among some climate scientists with a process that is so reliant on reaching consensus. After 20 years and four assessment reports, a few members of the committee are taking a critical and public look at how the panel represents uncertainty in predicting the magnitude of such changes as the rise in sea level. They detailed their concerns in the 14 September 2007 issue of Science .

One of those committee members, Michael Oppenheimer, a professor of geosciences and international affairs at Princeton University and a lead author of several IPCC reports, spoke with IEEE Spectrum reporter Morgen E. Peck on 11 September 2007. (This interview has been edited for content and clarity.)

How does the IPCC deal with uncertainty? What gets into the reports and what doesn’t?

The IPCC deals with uncertainty in a variety of ways. When dealing with the physical climate system it tends to rely upon using the models themselves that the predictions come from as the basis for assessing uncertainty.

One typical way it is done is to look at the full range of predictions from all the available models. That gives you an idea of what the possibilities are. That’s only one way to go about uncertainty assessment, and it’s not a comprehensive way because in some cases the models are, say, cousins—that is, they’re related to each other—and can have the same sort of bias within them. And in other cases there may be only a single model that’s the basis for an uncertainty assessment. You look over a range of possible parameters within that one model.

In order to get a full assessment of uncertainty—which is very difficult to do—in some cases it’s necessary to look at a broader range of information than is in the models. IPCC does do this; it’s just that in the end they are almost forced by the nature of their process to present numbers that are as precise as they can get them. And that sort of forces them back to looking at this internal view of what the models alone project.

As a result, has much information gone unreported?

There is full consensus on many, many points. It’s just that when it comes to certain areas where there is really less consensus, the IPCC has tended to shy away from them and not refer to them as part of the uncertainty range in the same breath that they would the numbers produced from models.

So, for instance, I just heard Bjørn Lomborg is back in the United States talking about sea level rise and how it’s no big deal. He focuses on the consensus number that IPCC published, which is roughly 7 inches to 2 feet [18 to 61 centimeters] for this century, based on model outcomes. But IPCC also, somewhere else in the report, said that the models are limited. And there are a lot of things ice-sheet models can’t do. So there’s a potential for a much higher sea level rise. But IPCC doesn’t say the two things at once. That leads to problems because, you know, the reporters and the government officials see the 7 inches to 2 feet and take it as the full story. And it isn’t.

What are they missing?

The West Antarctic ice sheet, for example, has a lot of uncertainty. The models are simply no good when it comes to assessing the West Antarctic ice sheet. They’re less than inadequate—let’s put it that way. IPCC, I think, was faced with a quandary. And rather than grapple with it by grabbing the bull by the horns and saying, ”Let’s have a full description here of the uncertainty—including the possibility that there could be a large response by the West Antarctic ice sheet,” [in the summaries for policy makers] they just chose to avoid the problem. I think that was a mistake, and that’s a sample of what happens when you have to basically fall back on what everybody agrees on. In consensus you will tend to leave out discussions of the parts that people don’t agree on. And those parts are very important, maybe the most important thing of all, because that’s where the big, big changes could be.

What would you do to more accurately represent uncertainties?

Well, in the case of ice sheets, for instance, there’s a lot of new information from paleoclimate data on sea level rise. IPCC does mention it, but unfortunately when it comes to producing the numbers they more or less push it to the side and they don’t present a full and rounded picture of the potential risks. You can’t extrapolate today’s changes into the future as a basis for assessing future sea level rise, but it should supplement what the models are saying. It should better inform the picture.

Something IPCC could also try is a process called expert elicitation. That’s where groups of scientists that are smaller than IPCC are put together in a room under controlled circumstances and derive their own judgment about what the range of probabilities are. And in some cases they produce results quite different from what the IPCC process does. I think all these approaches need to be looked at by IPCC.

If you begin including all the contested data, might not the report end up being more confusing and complicated than it is helpful?

That’s a good question and it’s a fair question. I think one reason IPCC handles things the way that it does is that it has thought in the past that simplicity was an important goal—in this case, not simplicity for the point of removing the meaning from things, but to simplify scientific complexity to the point where intelligent policy-makers could make rational decisions. I think, at this point, that we can afford to take a somewhat broader view. Have faith that policy-makers really get the big picture, and try to take them into some of the very important details, which is the stuff where the big changes could be hiding.

Are you at all concerned that if the IPCC began reporting on highly debated issues that climate change skeptics would seize the uncertainty as a way to discredit the panel?

I fully expect that to happen. But, you know, you can’t worry about that all the time. And I think the issue is mature enough that we don’t have to worry about it 24/7. Sure, people will use information for their own purposes. Sure, there’s a risk that some contrarian’s argument will catch fire. Sure, there are people around, leaders like U.S. Senator James M. Inhofe [R-Okla.] who aren’t going to be convinced, no matter what, and will use every fragment of information to argue in a particular way. But, if you let yourself be paralyzed by that you really stultify what’s an organic process, a process of learning. And if we’re not willing to take it where it leads us then I think we take a grave risk.

Do you think it might actually increase the credibility of the document for people to see that and say, ”Oh, look, they’re actually questioning themselves”?

Yes. That’s in fact what happened with this last IPCC report. You know, what we said in this paper has been said in scattered places before, and I think that’s a healthy sign. It’s a healthy sign the community is thinking with many brains and that it’s not marching in lockstep to some politically correct tune. And I think it’s a perfectly safe and healthy thing to do to encourage that kind of diverse discussion. I mean, look, these are scientists. You can’t force scientists into a mold. It’s going to break eventually. If we force people into a particular way of assessing uncertainty for too long it’ll make people uncomfortable with the whole process and they’ll give up. They’ll do something else next time and they won’t participate.

This article is for IEEE members only. Join IEEE to access our full archive.

Join the world’s largest professional organization devoted to engineering and applied sciences and get access to all of Spectrum’s articles, podcasts, and special reports. Learn more →

If you're already an IEEE member, please sign in to continue reading.

Membership includes:

  • Get unlimited access to IEEE Spectrum content
  • Follow your favorite topics to create a personalized feed of IEEE Spectrum content
  • Save Spectrum articles to read later
  • Network with other technology professionals
  • Establish a professional profile
  • Create a group to share and collaborate on projects
  • Discover IEEE events and activities
  • Join and participate in discussions