Academies recommend that IPCC make changes

The Intergovernmental Panel on Climate Change, an authoritative scientific organization set up in 1989 to assess climate science, took some heat today from a group that it commissioned to investigate its credibility. The oversight group uncovered procedural weaknesses that preclude IPCC from responding nimbly to events, it said — or from reliably identifying errors in its assessments.

In early March, the United Nations and the IPCC’s chairman requested that a group of the world’s science academies collectively review IPCC processes and procedures. On Aug. 30, this InterAcademy Council — or IAC — issued a 100-page report to IPCC that recommends a number of changes in how it conducts its business. For instance, the IAC called for greater transparency in the IPCC’s deliberations, term-limited appointments for the IPCC’s top eight officials and the development of a new top-tier committee that could make time-sensitive executive decisions on IPCC’s behalf (rather than waiting, as now, for annual meetings of a larger plenary group).

But what catalyzed the new IPCC review more than anything were challenges to the group’s credibility over a small passage referring to the melting rate of Himalayan glaciers. It said that “if the present [melt] rate continues, the likelihood of them disappearing by the year 2035 and perhaps sooner is very high.” The next sentence then appeared to contradict that, saying the total area of the glaciers in question could shrink from 500,000 square kilometers to 100,000 square kilometers — again by 2035.

Okay, so which is it?

That’s what one reviewer of the passage had asked. And was ignored. The reviewer also pointed out that without supporting references for this glacier-melt projection — references which were never cited — this drastic conclusion “should have been deleted.”

A second reviewer not only questioned the Himalayan glacier statements but also provided journal references that had come to different conclusions. “Had the [IPCC report’s] authors and/or Review Editor consulted the references,” the IAC report observes, “they would have found two peer-reviewed articles, which, at the very least, were more cautious about the disappearance of the Himalayan Glaciers.” Oops.

Harold Shapiro, president emeritus of Princeton University and IAC chairman, attributes the overly dramatic and conflicting statements on the Himalayan glaciers to the IPCC report’s authors “just not paying close enough attention to what the reviewers said.” He acknowledged that he could only speculate about what might have happened, “but it may be that at the time they were just overwhelmed by the number of comments and just didn’t take them all as seriously as they should.” In other words, Shapiro said, this was not a case of bias so much as negligence.

Chapters of the 2007 IPCC report received, on average, several thousand reviewer comments each; the whole draft report drew 90,000. That’s way too many to expect timely — and thoughtful — deliberation. IAC concludes that editors of IPCC’s massive reports reassessing the science of climate every few years should in future limit their written comments to just “the most significant review issues.” More minor challenges should get “abbreviated” responses, IAC recommends, and editors shouldn’t even bother responding to purely “editorial” comments (which seems to argue that they can simply relegate comments or criticisms based on opinion, bias or politics to the proverbial circular file).

Another interesting point Shapiro made focused on the types of problems that his group uncovered in a synthesis of IPCC’s last assessment that had been developed for policymakers. Two types of errors emerged.

One gave undo confidence to projections “for which there is very little evidence,” Shapiro noted. IAC now recommends that confidence assessments “should not be used to assign subjective probabilities to ill-defined outcomes” because studies have shown that people interpret ratings of how likely something is to occur very differently. Bottom line: Don’t say something is “likely” — versus, perhaps, “somewhat likely” — without offering more words to articulate the difference, particularly within the context of a specific scenario or projected outcome.

A second type of error in the policymaker’s cheat sheet was the assignment of high probabilities to useless projections. Shapiro offered one exaggerated (and hypothetical) example: that the IPCC has great certainty it will rain in New York at some point in 2011. Of course it will. But what people want to know is how much and when — i.e. will total precipitation run to just 10 percent of normal, or will all of the rain fall during the same month-long period, or perhaps not at all during the crop-growing season? Assigning certainty to when and how much rain will arrive is vastly more useful than merely that some will be some precipitation.

At a press briefing convened in New York City by the United Nations Environment Program, IPCC chairman R.K. Pachauri headed off most questions about what he thought of the new recommendations. As no one at IPCC has had a chance to look at the new document, he said, “I’m not able to comment on its findings.”

But one question he wished to deflect and couldn’t was whether he was willing to step down. The IAC has recommended that to keep creativity high, IPCC should retain its top officials for no longer than the time it takes to undertake one assessment of climate science. Pachauri led IPCC throughout the fourth assessment. So now that a fifth is underway, reporters asked him whether he is ready to move on?

Each time, the calm and phlegmatic Pachauri ducked the question. Implementing any recommendations that are ultimately adopted by IPCC— including term limits — is “a mission that I cannot shirk,” he said. Upon further prodding for a more explicit yea or nay, he added that “I’ve been elected to continue the fifth assessment process. But I’m a servant of the IPCC and of course I shall abide by any decision that the IPCC takes.”

I don’t know: Sounds to me like he’s itching to stay.