by Anastassia Makarieva
A quick overview of the case (relevant documents are below). The authors (Makarieva, Gorshkov, Sheil, Li) submit a manuscript to the Geophysical Research Letters (GRL) where they show that precipitation mass sink in the equatorial region produces sufficiently large pressure gradients to drive Hadley cell. The Editor writes in the first decision letter that the manuscript can be publishable provided the authors respond to the concerns of the second referee. The first referee is strongly positive and finds the paper important. The second referee writes 138 words:
This paper is interesting but somewhat confused. Without dwelling on the details of calculating the effect of precipitation on surface pressure, the attempt to relate such changes to large scale motion systems seems to ignore the fact that they are referring to balanced flows. Thus, pressure gradients per se do not drive the Hadley circulation. Rather, the momentum balance leads to circulations that maintain pressure gradients consistent with geostrophy. The actual way the pressure gradients are produced is essentially irrelevant though it would be potentially interesting if the mechanism involved precipitation. The fact that Held and Hou do not provide a quantitatively realistic description of the Hadley circulation is almost certainly due to the omission of eddies (Robinson, 2006, JAS, Walker and Schneider, 2004, GRL, 2006, JAS) and the assumption of equinoxial conditions (Lindzen and Hou, 1988, JAS).
The Editor points out that it is very important to attend to these comments and emphasizes that how pressure gradients are produced is irrelevant for the Hadley cell. They warn the authors that their response and the revised manuscript will be shown to referee No. 2 as well as to a third referee, if needed.
The authors are encouraged by the fact that one referee is excitied, another one finds the paper interesting. They draft a detailed reply, where they clarify that without a pressure gradient the Hadley cell would stall and that while the conventional paradigm considers differential heating as the cause of this gradient, they offer a new driver (precipitation mass sink).
The authors submit the revised version of the manuscript. The second decision letter from the Editor rejects the manuscript. It is accompanied by a review of Referee No. 1 (who continues to characterize the paper as a breakthrough) and a review of Referee 3. No reaction from Referee 2 to whom the authors replied.
The third referee finds the paper erroneous and recommends rejection. The authors find that the comments of the referee are in error and send the Editor their response to the comments of Referee 3. The Editor declines to consider the comments because “Manuscripts are accepted only when all the reviewers are in agreement about publication. In your case, two of the reviewers (#2 and #3), both of whom are very well-respected, were strongly against publication.”
It later turned out (see letter of the Editor-in-Chief) that Referee #2, who had found the paper interesting, refused to see the revision.
The authors do not feel that the justifications given by the Editor are valid. When the corresponding author receives a questionnaire from the ACP publishers, we express our concerns, in particular:
A journal that would employ the policy of publishing papers if only ALL referees agree will certainly be losing a good proportion of ground-breaking results submitted to it. We have never heard of such policies in any mainstream journals. If it is indeed the official policy of GRL, this should be publicly announced such that the authors know that the fate of their submissions is ultimately decided by arithmetic vote of the referees rather than by scientific logic.
…
The Editor mentioned that both referees were ‘very well-respected’. This is highly disturbing. This confirms our concerns that the decision regarding our paper was based not on the quality of scientific arguments, but on the reputation of who delivered these arguments. As a strategy, it makes it possible to accept the arguments of well-respected scientists and ignore the arguments of less respected ones, making the category of ‘respect’ the rule of thumb in publication policies.
This letter prompted a response from the GRL Editor-in-Chief who informed us that in his opinion there were no problems with our submission. We clarified our concerns:
We do understand it is normal that scientists may disagree on controversial issues. It is also normal, as you would certainly agree, that if the authors see that the referee is very seriously mistaken, they would have an opportunity to point that out to the Editor. Our concerns arose when the Editor motivated that his decision was final and that our response should be disregarded by stating that “Manuscripts are accepted only when all the reviewers are in agreement about publication. In your case, two of the reviewers (#2 and #3), both of whom are very well-respected, were strongly against publication.”
You say in your letter that the Editor “has as all latitude to make his/her decision based on his own reading of a manuscript — of course, editors’ decisions are usually strongly guided by all the reviews they obtain, as I am sure you understand since this forms the core of the peer-review process.” This contradicts the Editor’s words that the manuscripts are accepted only when all the reviewers are in agreement (which means that every reviewer has the right to veto any manuscript). Upon receiving your letter, I understand even less what was the basis for the decision to disregard our response to the erroneous comments of Referee 3.
You say that ‘GRL editors make a specific effort to find the best possible reviewers, with the strongest expertise in the fields covered by manuscripts’. In the view of this, we feel that our response to Referee 3 should have been considered by the Editor at face value, because another referee (Referee 1) was strongly in favor of the publication.
Neither do we understand the issue with Referee 2. Did the referee refuse to re-review our paper before seeing our response? If so, why did the Editor inform us he was planning to send our response to Referee 2? If the referee, who characterized the paper as interesting, refused to re-review our paper after seeing our response, how should this be interpreted? Can the reviewer be strongly against a paper without any explanations?
So please appreciate our position. Our response to Referee 2 was never mentioned. Our response to Referee 3 was ignored because the referee was well-respected. We did not understand this procedure. We were not disturbed and worried that our paper was rejected. We were disturbed and worried that, in our opinion, the scientific logic was sacrificed and scientific arguments were not taken into consideration losing force in the view of some others. When we were asked by AGU Publishers of our opinion of the peer-review process, we did express our concerns.
We did not receive any further feedbacks on our letter(s). Later we asked the journal if they had any reservations against us making this correspondence public. Having received no reply, we are now publishing this here.
1. Submitted manuscript
2. Decision letter #1 (including reviews)
3. Authors reply to decision letter #1
4. Revised manuscript
5. Decision letter #2
6. Comments of Referee 3
7. Authors reply to decision letter #2
8. Editor writes back
9. Peer review quality questionnaire from AGU Publishers
10. Authors express their concerns
11. GRL Editor-in-Chief replies
12. Authors clarify their concerns
13. Note about publicity
Stormy Science 