Editor-in-chief resigns as a new paper identifies errors in “fundamentally flawed” climate paper

Posted on September 6, 2011


Last Friday, Wolfgang Wagner of the journal Remote Sensing resigned as editor-in-chief. He took this extraordinary step because he felt that it was his responsibility that Remote Sensing published a “fundamentally flawed” climate paper by Roy Spencer and William D. Braswell, both of the University of Alabama – Huntsville (UAH). In response, Spencer wrote on his blog: “If some scientists would like to demonstrate in their own peer-reviewed paper where *anything* we wrote was incorrect, they should submit a paper for publication.” The first published response appeared this morning in the journal Geophysical Research Letters by Andrew Dessler of Texas A&M, and Dessler’s response points out multiple severe deficiencies in Spencer and Braswell’s paper titled “On the misdiagnosis of surface temperature feedbacks from variations in Earth’s radiant energy balance” (hereafter SB2011).

Wagner wrote an editorial in Remote Sensing to explain why he felt that SB2011 was “fundamentally flawed.” He wrote that “as a minimum requirement, [peer review is] supposed to be able to identify fundamental methodological errors or false claims,” but that SB2011 “is most likely problematic in both aspects and should therefore not have been published [emphasis added].” Wagner went on to point out that while he initially felt that SB2011’s controversial conclusion (that satellite data showed that clouds, not humans, were the dominant cause of climate disruption) deserved publication, he changed his mind because SB2011’s conclusions were not new and because Spencer and Braswell had ignored the challenges to those other conclusions when they wrote SB2011. Wagner wrote:

In other words, the problem I see with [SB2011] is not that it declared a minority view, but that it essentially ignored the scientific arguments of its opponents.

Barry Bickmore of Brigham Young University and John Nielsen-Gammon, Texas State Climatologist, both observed that Remote Sensing has a policy of allowing paper authors to suggest their own reviewers, a policy that can have serious consequences for a journal like Remote Sensing that may lack climate expertise. Bickmore noted that if a journal’s editors “don’t know who would be better, or that the suggested reviewers are all buddies of the authors, [the editors] might just ask the suggested reviewers.” Nielsen-Gammon speculated, “You don’t suppose the managing editor simply chose three referees from the list that Spencer and Braswell provided? Well, I do.”

Wagner wrote in his editorial that “the editorial team unintentionally selected three reviewers who probably share some climate sceptic notions of the authors.” Given the small number of “skeptics” among climate scientists, the probability that three randomly selected, qualified reviewers would have all turned out to be Spencer and Braswell’s brothers-in-arms is very, very small. In fact, recent data indicates that only 3% of climate scientists have serious concerns about humanity’s dominance regarding climate disruption, so the probability that three like-minded scientists would be selected at random is 0.033, or about once out of every 37,000 tries.

This is a problem with any journal accepting papers in areas with which the journal’s editors are not familiar. Peter Gleick, President of the Pacific Institute and currently on the editorial boards of three different journals (and previously on the editorial boards of two more), wrote in an email to S&R:

We regularly turn down journal articles, even good ones, that are not appropriate for the journal based on subject. Journals are often focused on a particular area of science, and we build up lists of reviewers competent in those areas. Receiving off-topic articles dilutes that focus and complicates reviewing.

Gleick is not on the editorial board of Remote Sensing and wrote that he “cannot judge whether the S&B article was suitable” for that journal. “Only its editors can speak to that.” In resigning, Wagner has done so.

While Wagner’s resignation as editor-in-chief casts a shadow of impropriety over SB2011, Dessler’s new paper titled “Cloud variations and the Earth’s energy budget” goes to the core of the paper’s scientific arguments and finds them deficient. First, Dessler finds that SB2011’s first equation appears to violate conservation of energy (“energy can never be created or destroyed”), one of the most fundamental laws of physics. The problem is that the way Spencer and Braswell wrote their equation permits the ocean to change temperature without additional energy, a situation that is analogous to a cup of coffee sitting on a desk top suddenly warming up for no reason.

The bulk of Dessler’s paper is devoted to the second error in SB2011, namely that the authors dramatically underestimated a critical ratio in their equation because they didn’t constrain their equation using measured data from the real world. SB2011 assumes that the ratio in question was about 0.5, but real data requires that the ratio be closer to 26, an error of about 50x. When the constrained ratio is used in the equations in SB2011, the paper’s results are suddenly right in line with the very models and papers that SB2011 was supposedly disproving.

Dessler also points out that Spencer and Braswell commit one of the cardinal sins of science, cherry-picking data. They wrote in SB2011 that they analyzed 14 models, but they only plotted the six models that agreed with their conclusions, ignored the models that agreed with the measured real world data in favor of the models that agreed with their conclusions, and neglected to mention that their analysis matched up well with three models that realistically model El Nino/La Nina cycles. Skeptical Science produced the following graph from Dessler’s Figure 2 that shows the cherry-picked information more clearly.

For comparison, below is SB2011 Figure 3a (which Dessler reproduced), from Barry Bickmore’s post on this subject:

As bad as all these errors are, Dessler saves the most serious for last. He points out that the analysis in SB2011 assumes that the sea surface temperature is independent of changes in other climatic effects such as El Nino/La Nina, clouds, and changes in global wind patterns. In reality, all of these effects affect each other.

It’s unclear whether Spencer and Braswell have attempted to corrupt the peer review process by using the new and inexpert journal Remote Sensing to publish a paper that, as Wagner wrote, “should never have been published.” Only Spencer and Braswell know for certain, and it would be a breach of ethics for Remote Sensing to publish the names of the anonymous reviewers without their permission. But what is clear is that Wagner felt that peer review had failed in this case and that Spencer and his allies are hypocrites for throwing about accusations of “pal review” as if they were stones thrown about glass houses. It is also clear from Dessler’s new paper that SB2011 has a significant number of major errors that cast into doubt every one of the paper’s conclusions and may cast into doubt the scientific integrity of Spencer and/or Braswell themselves.