Clouds Over Peer Review
Posted on 7 October 2011 by Captain Pithart
Guest post by Jörg Zimmermann
The opponents of the current state of science regarding climate change play on a wide range of instruments. On one end there are anti-scientific tirades in mass media and blogs, on the other the instruments of scientific work are used for their own purposes. Part of this is looking for ways to trick or circumvent scientific peer review, or to reframe scientific articles to fit an agenda. The scandal over the recent article by Roy Spencer and William Braswell in the recently founded scientific journal Remote Sensing belongs to this part of the spectrum of activity opposing the climate science consensus and delivers a déjà vu, weak science and an example for the impact of reframing scientific results in mass media and blogs.
The "AGW is no problem" crowd has failed in challenging the foundations of the state of science regarding climate change. No more can any scientist doubt without appearing foolish that the Earth is warming on a global scale, or that a "greenhouse effect" exists. Gerlich & Tscheuschner attempted this, but the topic was outside their area of expertise, and the polemical style of their contribution is so remote to a serious scientific discussion of the topic that it leads to doubts regarding their basic scientific competence. The interesting part of this case is that here too there was an attempt to play the system of peer review, which is supposed to work as a first plausibility check for scientific contributions.
Here are the common tricks used to sneak pseudoscience into the peer-reviewed literature:
- Publication in a professional journal outside the subject area.
- Utilizing a friendly or politically like-minded publisher or editor.
- Selection of like-minded reviewers, or just having negligent or over-challenged reviewers
- Publication in a low-profile journal with low rejection rate - what matters is that the article can be paraded as peer-reviewed.
- vague discussion of the topic within the article, which allows for serious interpretation, but also for subsequent public delivery of overhyped interpretations.
- Correct presentation of a detail, with the possibility of illegitimate generalization in the public debate; also hyping insignificant deviations as total refutation.
- Papers with data that were chosen to produce a desired result, or manipulative use of theory, model and data.
Gerlich & Tscheuschner for example published in a journal outside the subject area (1), apparently had friendly editors there (2) and a friendly and negligent reviewer (3) and thus were able to publish implausible and polemic assumptions as a peer-reviewed article. Had they been right, the article had evoked a dramatical revolution within climate science, and G&T were candidates for a Nobel Prize. Quite to the contrary though, meteorologists and climatologists found the absurd work not even worthy of a rebuttal.
Essex, McKitrick and Andresen (2007) also had success with utilizing a journal with a different thematic focus. They alleged that there is no global temperature, a nonsensical statement that would have had no chance to go with a journal of meteorology, climatology or geophysics. So at least point (1) applied here, with suspicion of (2) and (3), for who should be knowledgeable about the finer details of meteorological data series and methods in a journal on non-equilibrium thermodynamics?
McLean, de Freitas and Carter (2009) on the other hand were able to publish in the fitting and respectable Journal of Geophysical Research, but their thesis of climate change being only the result of a sequence of ENSO cycles was so error-ridden and nonsensical that one must speculate about point (3). Foster et al. (2009) demonstrated that their handling of data was manipulative (point (7)).
Lindzen and Choi
Besides the blatant nonsense there are other possibilities to plant seeds of doubt. The most important method here is to play down climate sensitivity (contradicting a large body of work pointing in the other direction) and to explain global warming with alternative theories. Usually additional forcings due to cloud cover are brought up, which by some mysterious mechanism are driven by cosmic influences, solar activity or unexplained internal variability of the coupled system of the atmosphere and oceans. But here too the discussion runs for over twenty years now, so that the wiggle space is shrinking for opposing views. Three approaches from this field could be witnessed in the recent past, each one pronounced the final nail in the coffin of Anthropogenic Climate Change Theory (final nails were driven in on a weekly basis for the last two decades, which made about as much of an impact on climate change denialists as the rather weak showing of Armageddon did to Jehova's Witnesses). There was Lindzen & Choi, On the determination of climate feedbacks on ERBE data, Geophysical Research Letters, 36, L16705 (2009), using arbitrarily selected periods of tropical oceanic weather to state proof for a negative feedback on global warming from cloud cover. By choosing slightly different periods, the opposite conclusion can be drawn. Lindzen & Choi (2009) furthermore displays significant misinterpretations of the tropical climate system and of the utilization of climate models. So this falls under point (7) in the above list, but here at least experts are needed to dissect the flawed work. Trenberth, Fasullo, O’Dell und Wong, Relationships between tropical sea surface temperature and top-of-atmosphere radiation, GRL 37, L03702, (2010) not only refute Lindzen & Choi (2009), but also simplifying models in general that implicate negative feedback on global warming due to clouds.
Displeased with the refutation of their paper, Lindzen and Choi tried to publish a sequel in the high-profile journal PNAS, playing on Richard Lindzen's membership at the publishing body, the National Academy of Science (Stoat has background information). Point (3) was tried here, when Lindzen and Choi chose two objectors to the scientific standard model: Will Happer, a physicist that never published on climate topics, and Ming-Dah Chou, Lindzen's co-worker in establishing the Iris hypothesis, in effect reviewing his own theory. When PNAS sent the work to four of their regular reviewers, this resulted in four rejections, in part due to grave errors, in part due to the paper not addressing the critique of their former paper by Trenberth et al. (2010). The latter was probably also the reason for the paper not getting into Geophysical Research Letters as a direct answer to Trenberth et al. (2010), as originally planned.
Real conspiracy theorists can be identified by immunizing themselves at such a point using crank theories. According to Lindzen, his work is flawless, and the unanimous rejection can only be explained by an IPCC-controlled conspiracy, which uses peer review to deflect all legitimate criticism. This excuse resides roughly on the same level as "dog ate my homework"; for there are ample works critical of the IPCC stance that made it into the peer-reviewed literature, and it is easy enough to read the peer reviews and find out about the weaknesses of Lindzen & Choi's paper. An amended version was finally published in the Asia-Pacific Journal of Atmospheric Sciences, making Lindzen's allegation of an IPCC conspiracy blocking his work seem quite paranoid. Another question is the scientific quality of the article. Andrew Dessler explained how the notion that clouds could drive cooling or warming of tropical oceans, instead of warming or cooling of the oceans driving cloud formation, is not compatible with the data at hand. And at the same time thus addresses another publication.
Spencer and Braswell (2011)
Lindzen & Choi was only the prelude to the Spencer & Braswell (2011) scandal. They, too, insisted on clouds working as a strongly negative feedback on global warming. One would assume that Spencer and Braswell necessarily had to mention Lindzen & Choi (2009) and Trenberth et al. (2010), for the latter had very definitely shown that a negative cloud feedback could not be demonstrated using selected episodes and simplified models. However, this was not the case. Furthermore, Spencer & Braswell's statements are very strong in their conclusions (refutation of IPCC stance), while using rather weak data and simplified models (further details here and here). This already hints to points (5) and (7) in above list, as demonstrated by denialist blogs welcoming the study as proof for a climate sensitivity much lower than accepted by the IPCC, thus ruling out predominantly human-caused global warming. The refuted work of Lindzen & Choi (2009) was promoted in the same way. Spencer & Braswell (2011) being published in Remote Sensing, which does not normally touch on the topics of climate data or climate science, was another alarm signal (point (1) above applies). The scandal erupted when Wolfgang Wagner, the journal's editor-in-chief, resigned. In his statement he came to the conclusion that the paper should never have been published due to the above mentioned flaws, which have been discussed in detail elsewhere. RealClimate gives a synopsis of Lindzen & Choi and Spencer & Braswell, relegating to Dessler (2011), which demonstrates in detail how the two papers run contrary to the reality of temperature flows above the oceans.
Wagner also mentioned the problem of failed peer review, as the three reviewers seemed to have been of the same conviction as Spencer, thus fulfilling point (3) in the above list: the peer review was not really independent; an objective reviewer would likely have asked why Trenberth et al. (2010) was not mentioned. A critical reviewer would furthermore likely not have seen the point of publishing the re-incarnation of a refuted hypothesis. That Spencer & Braswell did not address responses to their hypothesis that were known to them might be more damning than their flawed analysis, the vague results of which were subsequently overhyped.
Spencer reacted like Lindzen: the dismissive reaction of the professional world, or the resignation of the editor-in-chief with his statements elevating the Spencer & Braswell (2011) publication to a scandal are not seen by Spencer as a hint that there could be problems with his hypothesis. Rather he defiantly declares that only one error could be attributed to him in his entire scientific career, and that the peer-reviewed journals were in cahoots with the IPCC to keep critical works out. The fault was not with him, but with almost all other scientists in the field. With this statement Spencer did not quite succeed in driving the final nail into the coffin of the theory of man-made global warming, but at least managed to bury the last remains of his own scientific respectability. As a bonus, he delivered a déjà vu: Hans von Storch and half of the other editors of the journal Climate Research resigned in 2003 after one of their editors (Chris de Freitas) had rubber-stamped a botched paper by Willie Soon and Sallie Baliunas. Then, too, the peer review process had been undermined (points (2) and (3)), then, too, the journal's reputation had suffered, then, too, the denialist scene paraded Soon and Baliunas as witnesses that laid bare the global warming hoax. Soon & Baliunas (2003) described the climate change of the last millenium as the result of changes in solar activity by cobbling together different proxy data for precipitation and temperature so as to create the impression that there were times in the last thousand years that were notably warmer than the late 20th century. The flaws in the work were so grave that they were described as "ludicrous" or "absurd" by specialists.
An IPCC conspiracy?
Some questions remain: shouldn't the opponents of the state of climate science take issue with their witnesses to the prosecution (such as Lindzen, Spencer, Svensmark, Soon, Douglass, the Idso family, McKitrick) having great problems to publish their dissenting interpretations in reputable journals? Shouldn't it give them pause that scientists in line with the accepted state of science publish more, are cited more, and produce works more relevant to the topic? Shouldn't it make them think that the road to publication for papers supporting their view most often involves friendly editors or reviewers, journals concerned with a different subject area, or even fringe publications like Energy & Environment (which as a journal of sociology publishes a plethora of denialist contributions that purport to deal with topics in natural sciences, with an editor-in-chief that publicly states her mission as setting up an opposite view to the IPCC)? Shouldn't it strike them as odd that for at least two decades new publications are declared to falsify the theory of climate change, but that the professional world never takes heed of these?
If one does not come to the conclusion that the fraction unconvinced of AGW has long since grown into an esoteric circle, similar to the opponents of evolution, the theory of relativity or vaccination, to the adherents of homeopathy, anthroposophists, or various conspiracionists, one has to assume that scientists worldwide are in their majority participants in a great conspiracy to suppress truth and to erect an IPCC One World Government.
Reading between Lindzen's and Spencer's lines one gets the impression that at least regarding this conspiracy theory they are drawing near:
"...there is a bust-gut effort going on to make sure that either (1) no scientific papers get published which could get in the way of the IPCC’s politically-motivated goals, or (2) any critical papers that DO get published are discredited with any and all means available." (Roy Spencer, September 5, 2011)