For a long period of time I have growing temptation to list in my CV one of my articles -- Shchepetkin, A. F. and J. C. McWilliams: The Regional Oceanic Modeling System (ROMS): A split-explicit, free-surface, topography-following-coordinate oceanic model. ... -- not as published in Ocean Modeling in 2005, but literally as ..., 2003 manuscript JCOMP-D-03-00102 rejected from The Journal of Computational Physics, and actually be proud of it. Here is why: If two anonymous reviewers say that it looks like a textbook, then it probably looks like a textbook... ...and if it looks like a textbook, then probably people will bother to read it, because there is a certain sense of lack of textbooks... ... and if at the end the reviewers do not name a particular textbook to which the manuscript is alike, then there probably no such textbook, so may be there is something new in it. ...at least I was told that it was reviewed by two "senior members of our community" and the rejection was upheld on appeal by JCP Chief Editor Gretar Tryggvason - what can be MORE HONORABLE? ...cited over 1600 times after it was published in another journal and still counting, it decisively overtook Chief Editor Gretar Tryggvason's own most ever cited article about front-tracking method published back in 1992. Що, наковтався пилюки, Шаблезубий? Rest in Peace with Glory and Pride my JCOMP-D-03-00102 !!! ...Oh no, wait, it does not want to: it is still... kind of ...alive!!! The story does not end here. Some people read it. Some people cited it. Some people cited it without reading it first. Some people even copy-pasted it without reading it first, and in the process of doing so attached a list of 20 co-authors including several respected senior faculty members... ...there is no reason to complain. This is, after all ...RECOGNITION! Lets appreciate the comical aspect of the whole situation: the copy-paste ended up being published in the same journal, The Journal of Computational Physics where the rejected manuscript was submitted into back in 2003 to begin with. Whiskey-Tango-Foxtrot, Chief Editor Gretar Tryggvason? How did is happen, Editor Piotr Smolarkiewicz? Нет, ну тут я просто охуe..аю. Слов цензурных не хватает выразить всю глубину моих чувств. Пипец! Я просто тащусь как удав по банке дуста. Как же так случилось, Дорогая Редакция? ...may be reviewers overslept? No, they did not. At least one of them sound alarm and identified all the copy-pasted pieces. Still it did not stop it: in fact, it was overslept during the broad daylight by Editor Piotr Smolarkiewicz. The following part contains three e-mails from Gretar Tryggvason related to this matter. The "Correction Note" published in JCP in 2009 to sort out this mess. The e-mails are: (1) Original e-mail for Gretar Tryggvason rendering the final rejection of 2003 JCP submission along with the original reviews upon which it was rejected. (2) Initial response of Gretar Tryggvason to the submission of the Correction Note in 2008. (3) Response of Gretar Tryggvason after receiving reviews to the Correction Note along with 4 reviews. The comical effect comes from juxtaposition of what is said by the same person about basically the same subject. ############################### THE FIRST #################################### Date: Mon, 1 Sep 2003 17:18:17 -0400 To: jcm@ucar.edu From: "Gretar Tryggvason" Subject: Re: appeal of a JCP review decision CC: jcp@elsevier.com, jcp@WPI.EDU Dear Drs. Shchepetkin and McWilliams Thank you for your message from 8/27/03 requesting re-reviewing of your manuscript (our number: JCOMP-D-03-00102). I have now reviewed both the manuscript and the reviews and I cannot see a reason for re-reviewing it. Both reviewers (who are senior members of what I believe is your community) have serious concerns and I feel that the associate editor could not have made any other decision than to decline the manuscript. You can, of course, always submit a reworked manuscript again as a new submission. If you elect do so, I would encourage you to include a letter explaining why you believe that it should be considered again and how you have changed it to deal with the objections of the original reviewers. I should also note that we will not send the manuscript out for reviews again unless the changes truly justify treating it as a new submission. Sincerely Gretar Tryggvason From: "Journal of Computational Physics" To: old_galaxy@yahoo.com Subject: JCOMP-D-03-00102 Date: Mon, 16 Jun 2003 07:17:14 -0400 Re: JCOMP-D-03-00102 Dear Dr. Shchepetkin, The Editorial Office has just received the reports on the paper entitled "The Regional Oceanic Modeling System: A Split-Explicit, Free-Surface, Topography-Following-Coordinate Oceanic Model". The reviewers comments are as follows: ************************************************************ Reviewer #1: The message of this paper is the time splitting between the barotropic and baroclinic modes in numerical ocean models is not exactly consistent. This is not an unaware problem because mode-splitting method has been used in both atmospheric and ocean modeling for many decades. Its basic strategy is to advance the fast barotropic mode in several small time-steps within a baroclinic time-step and then to couple the two modes by averaging the barotropic values. Such a method can save computational cost because the 3D baroclinic calculation is more costly than the 2D barotropic one. A price paid for such a saving is the possible splitting error. Clearly, the error depends on the splitting ratio. This work is proposed to modify the splitting method. A). We know there are a lot of errors in numerical models because they are approximations of continuous equations. It is always helpful reducing these errors to be as small as possible. However, the case made by the manuscript for a big operation is poorly justified. It argues, "Although this error is typically small for basin-scale configurations, it can become significant and cause non-physical behavior and even numerical instability in coastal regions." My understanding is that water is shallower and less stratified in coastal regions than in deep oceans. Model grid size is much smaller; therefore, the time-step for the coastal baroclinic part has to be small anyway. There is no need to have a large time-splitting ratio between the two modes and the splitting error should be smaller, than in the basin-scale models. B). The method is essentially to average the barotropic values by weighting coefficients, rather than by the simple mean or middle value used previously (no justification is given for why such a weighting is necessary). It should be noted that the modified method is extremely sophisticate. By weighted- averaging, several places are modified accordingly and extra cost is needed. It is claimed that the new method is better than previous ones, but no comparison is ever given. Most of the figures are comparisons of phase relationships of various linear schemes. They cannot be used to judge the weighting method. The sole purpose of mode splitting is to save computational cost. By using a more sophisticated averaging method, it might result in a longer time-step. What is the trade off between the longer time-step gained and the extra cost, probably implicit diffusion, introduced by the weighting and filtering? The manuscript fails to address the most important question. C). If this is an adaptation of Higdon and de Szoeke method to a terrain- following model, as indicated in the Conclusions, some comparisons should be given. The Monterey Bay example uses a splitting ratio 60. According to my calculations, using 15 is good enough. The barotropic time-step is still under the CFL conditions (grid size 3 km and time-step 960/15 sec). It reduces cost and avoids the expensive modification. The manuscript is too long, many unrelated materials which can be found from basic numerical books. I don't understand why such a small modification to an existing code is such a big deal. All users modify old codes. The title of the paper gives me an impression that a new ocean model is being developed. However, mode-splitting, free-surface, and topography-following models have been developed long time ago. The manuscript should focus on its innovative value if there is. It is actually a sophisticated twist of existing techniques. I cannot judge its practical value either because no comparison with previous methods is given. Reviewer #2: [no review from #2 at all for unknown reason] Reviewer #3: This paper is far too long (61p). It contains excursions into the design of terrain following coordinates, time-discretization schemes, filters, spline interpolants, all in addition to the main subject of split-explicit integration of barotropic/baroclinic modes in ocean models. The author appears to lose sight of the main purpose of a journal article, namely, to disseminate novel results clearly. I cannot recommend acceptance of this manuscript in its present form. I would recommend that the paper be broken up into (possibly several) more digestible papers, that can more clearly develop and explain the several contributions, and resubmitted. The study of time-discretization schemes is particularly interesting and deserves a separate and more complete treatment. Nevertheless, the absence of Brown-Campana pressure averaging that doubles the time step of LF and makes it as efficient as FB is paricularly glaring, and possibly skewes some of the conclusions presented. There, are numerous instances of inefficient and inconcise writing. It is not possible to list them all; for example, why give details of four options for the baroclinic scheme, and two options for the barotropic/baroclinic algorithm? Possibly because of the excessive length of the paper, there are essentially no computational results presented. Thus, it is not possible to evaluate the the accuracy and efficiency of the methods presented. ************************************************************ In view of these comments made the Associate Editor regrets that we are unable to publish your work in Journal of Computational Physics. Yours sincerely, Pien van Spijker on behalf of the Editors of Journal of Computational Physics Editorial Office Journal of Computational Physics Elsevier Sara Burgerhartstraat 25 1055 KV Amsterdam The Netherlands Tel: +31 20 485 2358/2360 Fax: +31 20 485 2370 E-mail: jcp@elsevier.com ############################### THE SECOND #################################### From: "Gretar Tryggvason" Date: Wednesday, July 9, 2008 12:33 PM To: old_galaxy@yahoo.com, alex@atmos.ucla.edu, jcm@atmos.ucla.edu Cc: "Gretar Tryggvason" Dear Dr. Shchepetkin, I am writing to you about your submission to the Journal of Computational Physics: JCOMP-D-08-00439: A Correction Note for ``Ocean Forecasting in Terrain-Following Coordinates: Formulation and Skill Assessment of the Regional Ocean Modeling System by Haidvogel et. al., J. Comp. Phys. 227, pp. 3595-3624 As you probably appreciate, this is an unusual situation. When the associate editor handling H2008 alerted me to the issue you raise, I tried to stop the publication of H2008 but as it was already being printed I could not do that. I have therefore taken some time to ponder exactly how to treat your submission. It seems to me that there are at least three possible ways to proceed: 1. I can send your manuscript out to reviewers as it is. I cannot imagine that they (or at least some of them) will not object to the way you discuss the circumstances leading to the publication of H2008 and the motivation for the present manuscript. If Haidvogel desires to publish a letter disputing your presentation of the events, I would not be able to deny him the ability to do so. As the editor of the Journal, I have obviously no desire to see the Journal become a forum for such an exchange. 2. You could change the introduction to expressing a more passive desire to clarify the presentation in H2008 before I send it out to reviewers, increasing the probability of favorable reviews and reducing the likely number of iterations. 3. If you are willing to consider changing manuscript (simply stating your desire to clarify the presentation in H2008) I would be willing to see if Haidvogel would accept not to challenge your note. If we can produce a version that was acceptable to him and you, I would also consider going through an accelerated or abbreviated review process. Let me know if you would consider either 2 or 3. If not, I will proceed to get reviews. I would be happy to discuss this with you in more detail on the phone. Sincerely Gretar Tryggvason Editor, Journal of Computational Physics --------------------------------------- Gretar Tryggvason, Professor and Head Department of Mechanical Engineering 100 Institute Road Worcester Polytechnic Institute Worcester, MA 01609-2280 Tel.: (508) 831-5759; FAX: (508) 831-5680 www: http://www.me.wpi.edu/Tryggvason ############################### THE THIRD #################################### From: "Journal of Computational Physics" To: old_galaxy@yahoo.com Re: JCOMP-D-08-00439 Date: Wednesday, March 18, 2009 5:22 AM Dear Dr. Shchepetkin, The Editorial Office has received the decision on the paper entitled "A Correction Note for ``Ocean Forecasting in Terrain-Following Coordinates: Formulation and Skill Assessment of the Regional Ocean Modeling System by Haidvogel et. al., J. Comp. Phys. 227, pp. 3595-3624". The comments are as follows: ************************************************************ Dear Drs. Shchepetkin and McWilliams, I now have comments from four people on your Note. I have asked a large number of other people to give me comments but most have declined. Indeed, reviewer 2 and 4 below declined to send a formal review and instead sent me emails with their comments. They did, however, agree that I could share their comments with you. All the reviewers agree that Note should be published but they comment on the length and its "tone." You already made some changes before sending the last version to me but apparently that was not enough. Please revise the Note, taking into account the suggestions of the reviewers. I think a significant reducion in lenght is warranted and the "tone" could still be improved. See for example reviewer's 1 suggestion for a revised abstract. It has, I will readily admit, taken considerable time to get this far and I would really like to bring this issue to a conclusion. Thus, I am hoping that the revised version will be as close to acceptable as possible. Sincerely, Gretar Tryggvason Reviewer I: Review of: "A Correction Note for Ocean Forecasting in Terrain-Following Coordinates: Formulation and Skill Assessment of the Regional Ocean Modeling System by Haidvogel et. al., J. Comp. Phys. 227, pp. 3595-3624" by Shchepetkin and McWilliams submitted to Journal of Computational Physics, JCOMP-D-08-00439 The manuscript is submitted to provide corrective information for a previously published article in JCP by Haidvogel et al., (2008; referred to as H2008). Both authors of the submitted manuscript were listed as authors on H2008. The current manuscript is presented in an effort to provide information that rectifies several key aspects of H2008. The submitted manuscript does contain relevant information, and that material should be published. However, having said that, the manuscript requires a more scientifically directed tone, with some reorganization to provide a more coherent presentation. The authors can retain section 1, which is a list of specific corrections to H2008. But for the rest of the manuscript, I suggest that the authors take an approach that allows a more clear explanation of the time history of the development of the model, rather than focusing so much energy to identify what is incorrect in the H2008 paper. And please remove the interjecting attacks at others. Those attacks do not add merit to the manuscript. I tried to provide some specific suggestions below. To the editors: Perhaps this is a message to the Journals that all authors of a submitted manuscript should be required to provide a letter of acceptance before any article is published. This extra step could limit this type of event in the future. Detailed comments: 1) p. Abstract. As an example, to modify the tone of the manuscript, perhaps the abstract could be reworded as: "Although our names appear as co-authors in the above article (hereafter H2008), we were loosely aware of its existence. In our view the material presented in H2008 does not constitute a mathematically accurate statement about the hydrodynamic core of ROMS. We identify that a significant portion of H2008 (? 40% of its volume, or 10 pages) contains almost line-by-line copy from a previously published work, Shchepetkin & McWilliams (2005) (hereafter SM2005), but without some intervening material that explains the content. This creates a confusing situation and the purpose of this Commentary is to clarify and correct it, as well as to explain some of the algorithmic differences among ROMS versions now in use." 2) p. 12 footnote 13 - The commentary on the failed hostile takeover of the ROMS model via copyright registration is not appropriate for this journal. 3) p. 16 Acknowledgments: the commentary on the lack of acceptance by JCP of a previous submission is not appropriate. What is the intent of that information? 4) Is there really a need for 30 footnotes in a 'correction note?' The manuscript needs to be reformatted in a manner that retains the intent but looses the 'by the way, this also happened' approach. 5) As an example : Could section 2 just be titled "Evolution of perturbed vertical coordinate" ? With this new suggested heading, the tone would be modified from trying to show how the roms model is different from Song and Haidvogel, to a description of how a vertical coordinate system evolved over time. This would change the tone from being defensive to being instructive. 6) In the end, it is still unclear what reference(s) should be used by the ROMS community to cite the model. Reviewer 2: Dear Gretar, You have asked me to review JCOMP-D-08-00439, which is a submission entitled "A Correction Note for ``Ocean Forecasting in Terrain-Following Coordinates: Formulation and Skill Assessment of the Regional Ocean Modeling System by Haidvogel et. al., J. Comp. Phys. 227, pp. 3595-3624", by Dr. Alexander F. Shchepetkin. The problem is that this is not a regular article with new material that requires a review. It is more like a letter commenting on an existing article, or in this case, a correction note (albeit a very long one) that contains no new material and therefore requires an editorial decision rather than a review. In such a situation one would normally publish the comments together with a reply from the author(s) of the subject paper (ie, Haidvogel). An unfortunate aspect of this note is that it has an accusatory tone which may well antagonize the authors of the subject paper. Because this "note" is so long, the normal procedure may not be desirable. I would therefore suggest the following: 1) Remove the offensive language to make the note more objective, 2) Drastically shorten the note (2-3 pages) to just to list the inaccuracies or omissions of the subject paper with respect to the original paper of Schepetkin and McWilliams (2005), with references to the original paper for the corrections or omissions, where appropriate, instead of very long winded explanations, 3) Request a reply from Haidvogel, 4) Publish the note and reply together. Since this in not a review, I have chosen instead to send you this email explaining my position. Reviewer 3: The circumstances that prompted the writing of this note are astonishing. How can a paper get written without author consent? I certainly hope that JCP reviews its protocols for accepting papers from multiple authors. I also suggest that they entertain input from Haidvogel. He is a respected scientist, and certainly deserves some comment on the situation. But so far as evidenced from the present note from Shchepetkin and McWilliams, there is clearly something amiss when he publishes a paper without sending around proofs and reviews to ALL authors. What prompted such behaviour? Given the circumstances and the evidence presented, I feel totally sympathetic with Shchepetkin and McWilliams. They must set the record straight, and that is just what they aim in their manuscript. I generally feel they succeeded, and in particular are precise in identifying those areas where the Haidvogel etal. paper is wrong, confusing, or misleading. I recommend acceptance of this Note, and offer the following specific and general comments. ***specific: pg 4 line 31: reference should be Adcroft and Campin (2004) pg 4 line 31: H2008 (missing "H") pg 11: foonote needs a period. ***general: I suggest a summary or conclusion section is appropriate, even if brief. Absent such, the reader is left hanging. More significantly, this note should motivate THe authors to present the ocean modelling community with a far more thorough, and accurate, summary of the amazing suite of algorithms present in ROMS. The present note, as it is basically an extended and critical review of a published paper, serves only those who are already experts. What the community needs in addition is a pedagogical exposition of the ROMS computational algorithms, digested somewhat from the existing papers. And if I may offer some frank and critical words regarding which journal to publish this review in...it should NOT be JCP, given their darkened record with the H2008 review paper, which is the subject of the present note. Reviewer 4 Remarks on ``A Correction Note for `Ocean Forecasting in Terrain-Following Coordinates: Formulation and Skill Assessment of the Regional Ocean Modeling System' by Haidvogel et. al., J. Comp. Phys. 227, pp. 3595-3624'', by A. F. Shchepetkin and J. C. McWilliams. I was a referee for the paper by Haidvogel, et al, which is the subject of this Correction Note by Shchepetkin and McWilliams. My report is copied below. In that report, I documented several instances of verbatim copying of text from the 2005 paper of Shchepetkin and McWilliams. I did not hear anything from the JCP office afterwards. Much later, I happened to look at my "Completed Assignments" page under the "Reviewer Main Menu" at the JCP website, and I noticed that this paper had been accepted for publication. In response to my referee report, the author(s) had said the following: ``With respect to Reviewer 2: First, since the intent here is to reproduce the essential details of the Shchepetkin and McWilliams papes for the benefit of readers, I have not chosen to remove these summaries. In the spirit of complete disclosure, however, I have clearly stated that the material in section 3 is extracted from these papers, and have referred the reader to them for additional details. Given their seminal influence on the ROMS algorithmic core, Shchepetkin and McWilliams are now the only authors referred to by name in the revised manuscript. Second, the authorship list has been left unchanged. I feel that it is appropriate to recognize all who have contributed material, no matter how much in a relative sense. Lastly, the style of the references has been kept as it is, in conformance with JCP standard practice.'' I was rather surprised by this. I do not see the point of re-publishing text that has already been published elsewhere; instead, simply tell the reader to read certain sections of the other paper. I was surprised further to read the abstract of the Correction Note by Shchepetkin and McWilliams and learn that they were unaware of the review article until after it was published. To judge from this abstract, it is clear that Shchepetkin and McWilliams are unhappy with this situation, and I do not blame them. Because of a high volume of reviewing commitments right now, I do not have the time to make detailed comments about the Correction Note. However, my impression is that Shchepetkin and McWilliams are basically right in submitting this Note, and I believe that it should be published, possibly after any technical corrections that might be noticed by other referees. One specific comment is the following. In the Acknowledgments, Shchepetkin and McWilliams describe how their 2005 paper was initially submitted to JCP and ultimately published in Ocean Modelling. I do not think that this history is relevant to the present Correction Note, and I recommend that it be deleted. My earlier referee report on the paper by Haidvogel, et al, follows. ---------------------------------------------------------------------- Referee report for ``Regional Ocean Forecasting in Terrain-following Coordinates: Model Formulation and Skill Assessment, by D. Haidvogel et al. The purpose of this paper is to give a review of the Regional Oceanic Modeling System, a mature and widely-used code for ocean simulations. Such a review is timely, and the authors are among the principal developers and users of this code. The contents of Sections 3.1-3.6 have already been published in reference [7] (Shchepetkin and McWilliams 2005, hereafter referred to as SM). This is the most technical portion of the manuscript. In the following examples, much of the material from the earlier paper was copied word-for-word, except for changes in equation numbers, and the rest was copied with some deletions. Section 3.1 of the present manuscript is also on page 349 of SM; Section 3.2, except for the first paragraph, is on pages 350 and 352 of SM; Section 3.3 (three pages) is contained in Section 1.3 of SM; Section 3.4 is on pp. 374-375 of SM; Section 3.5 through the bottom of page 20 (five pages) is found on pp. 376-380 of SM, and the last paragraph of Section 3.5 is in paragraph (iv) on page 381 of SM; Section 3.6 is in Section 5 of SM; Figures 2 and 3 are from SM. I realize that the present manuscript is a review paper, but I do not agree with re- publishing the material quoted above. Instead, the authors should reduce Sections 3.1-3.6 to a short summary (e.g., a page or two) that points to the pertinent sections of SM. Shchepetkin and McWilliams are two of the twenty authors of the manuscript under review, and the above material accounts for about 12 of the 35 pages of text. Under these circumstances, is it appropriate to list twenty authors? An alternative is to reduce the author list to one or a few, and then credit the others via citations in the References and/or by listing them in a short Acknowledgments section at the end of the paper. In the text of the manuscript, the authors usually refer to items in the References by number only. It would be more informative to the reader to list the authors along with the numbers, especially since this is a review paper. ************************************************************ In view of these comments made the Associate Editor who guided your article, Professor Tryggvason, has decided that the paper can be reconsidered for publication after major revisions. Therefore we look forward to receiving the revised version of the paper together with a reply to the reports and a summary of the revisions made. If the revised version is submitted within three months of receipt of this e-mail, the manuscript will retain the original submission date. After three months, your paper might be treated as a new submission and may be sent to new reviewers. Yours sincerely, Jeanette Bakker on behalf of the Editors of Journal of Computational Physics Editorial-Production Department, Elsevier Radarweg 29 1043 NX Amsterdam The Netherlands Tel: +31 20 485 3426 Fax: +31 20 485 2521 E-mail: jcp@elsevier.com Journal of Computational Physics is available electronically at http://www.ElsevierPhysics.com/