by Judith Curry
As the IPCC struggles with its inconvenient truth – the pause and the growing discrepancy between models and observations – the obvious question is: why is the IPCC just starting to grapple with this issue now, essentially two minutes before midnite of the release of the AR5?
Well I suspect that the short answer is that they didn’t think it was important and/or they thought they could get away with ignoring it; after all they are the experts and in charge of the ‘consensus.’ It seems that reviewers and policy makers have been clamoring for the IPCC to address this issue; I suspect that David Rose’s MSM assault on this issue, largely fueled by blogospheric analysis, has played a significant role here.
A few weeks ago, I ran a post Overestimated global warming over the past 20 years that referred to a recent paper published in Nature Climate Change by Fyfe et al. This paper starkly laid out the discrepancy between CMIP5 model projections and observations of global surface temperature change. This wasn’t exactly news to those of us who follow the skeptical blogosphere; we have seen similar analyses by John Christy (presented in his Congressional testimony) and the analysis of Ed Hawkins that was made famous by David Rose’s article.
My blog post on the Fyfe et al. paper triggered an email from Pat Michaels, who sent me a paper that he submitted in 2010 to Geophysical Research Letters, that did essentially the same analysis as Fyfe et al., albeit with the CMIP3 models.
Assessing the consistency between short-term global temperature trends in observations and climate model projects
Patrick J. Michaels, Paul C. Knappenberger, John R. Christy, Chad S. Herman, Lucia M. Liljegren, James D. Annan
Abstract. Assessing the consistency between short-term global temperature trends in observations and climate model projections is a challenging problem. While climate models capture many processes governing short-term climate fluctuations, they are not expected to simulate the specific timing of these somewhat random phenomena—the occurrence of which may impact the realized trend. Therefore, to assess model performance, we develop distributions of projected temperature trends from a collection of climate models running the IPCC A1B emissions scenario. We evaluate where observed trends of length 5 to 15 years fall within the distribution of model trends of the same length. We find that current trends lie near the lower limits of the model distributions, with cumulative probability-of-occurrence values typically between 5% and 20%, and probabilities below 5% not uncommon. Our results indicate cause for concern regarding the consistency between climate model projections and observed climate behavior under conditions of increasing anthropogenic greenhouse-gas emissions.
Drum roll . . . the paper was rejected. I read the paper (read it yourself), and I couldn’t see why it was rejected, particularly since it seems to be a pretty straightforward analysis that has been corroborated in subsequent published papers.
The rejection of this paper raised my watchdog hackles, and I asked to see the reviews. I suspected gatekeeping by the editor and bias against the skeptical authors by the editor and reviewers.
The author team includes prominent skeptics Pat Michaels and John Christy; lukewarmer blogger/analysts Lucia Liljegren, Chad Herman, Chip Knappenberger; and James Annan, a scientist critical of high-end climate sensitivity estimates, but not someone typically associated with climate change skeptics (and in fact, whose inclusion raised some eyebrows). Interestingly, James Annan was one of the two reviewers of Michaels’ original submission to GRL, and signed on as a co-author after helping them solve some issues noted in the first set of reviews. The second round of review consisted of 4 reviewers, none of which supported publication.
Well, it seems like ‘skeptical’ papers require a larger number of reviewers (2-3 is typical), especially after one of the original reviewers ‘defects’ and ends up as a coauthor on the paper. I’ve gone through the reviews and discussed them with Michaels and Knappenberger, and we’ve agreed on the following summary of the second round of reviews:
The primary concerns regarded the inclusion of error bars on the observed trends (which the authors pretty clearly have demonstrated was not required, and in fact would have been wrong to do so); misinterpretation of the statistics (although those comments too, seemed off the mark); that Michaels et al. did not take into account things that were absent in the model projections (such things as solar variability, indirect aerosol effect, stratospheric water vapor trends); and perhaps most universally, that Michaels “cause for concern” statement really seemed to rub the reviewers the wrong way.
My own personal reaction to the rather lengthy reviews (12 pages worth) is that all of the reviewers rejected the idea that IPCC model projections could be compared in such a way that led to the conclusion that indicate cause for concern regarding the consistency between climate model projections and observed climate behavior under conditions of increasing anthropogenic greenhouse-gas emissions.
For context, about 5 years ago the “pause” in the rise of the global average surface (and lower troposphere) temperatures was starting to get uncomfortably long, and stretchy efforts to reconcile the observed trend with climate model projections were starting to appear in the literature. The first was by Easterling and Wehner (2009) followed shortly thereafter by Knight et al. (2009). In both cases, the authors concluded that 10-yr periods with little to no warming were still perfectly consistent with climate model expectations when considering natural variability.
Three years later, it seems pretty obvious and widely acknowledged that climate models have been unable to correctly capture the earth’s surface temperature evolution over the past several decades. Lucia continues to do good work on this subject; head over to her blog for a technical discussion on this topic and the Michaels et al. paper.
And we see where ‘pause denial’ has led the IPCC, potentially to a crisis point in the AR5. It will be very interesting to see how this plays out in Stockholm next week.
JC message to James Annan: kudos, and thank you.