Scientific workflows applied to the coupling of a continuum (Elmer v8.3) and a discrete element (HiDEM v1.0) ice dynamic model

Memon, Shahbaz; Vallot, Dorothée; Zwinger, Thomas; Åström, Jan; Neukirchen, Helmut; Riedel, Morris; Book, Matthias

doi:https://doi.org/10.5194/gmd-12-3001-2019

Articles | Volume 12, issue 7

https://doi.org/10.5194/gmd-12-3001-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/gmd-12-3001-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 12, issue 7

Development and technical paper

|

15 Jul 2019

Development and technical paper |

| 15 Jul 2019

Scientific workflows applied to the coupling of a continuum (Elmer v8.3) and a discrete element (HiDEM v1.0) ice dynamic model

Shahbaz Memon, Dorothée Vallot, Thomas Zwinger, Jan Åström, Helmut Neukirchen, Morris Riedel, and Matthias Book

Download

Final revised paper (published on 15 Jul 2019)
Preprint (discussion started on 20 Jul 2018)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Review of Scientific Workflows Applied to the Coupling of a Continuum (Elmer v8.3) and a Discrete Element (HiDEM v1.0) Ice Dynamic Model', Anonymous Referee #1, 16 Aug 2018
- AC1: 'Referee # 1 - Comments' Response', Shahbaz Memon, 27 Aug 2018
RC2: 'Review of „Scientific Workflows Applied to the Coupling of a Continuum (Elmer v8.3) and a Discrete Element (HiDEM v1.0) Ice Dynamic Model“ by S. Memon, D. Vallot, T. Zwinger, J Åström, H. Neukirchen, M. Riedel and M. Book.', Anonymous Referee #2, 18 Sep 2018
- AC2: 'Review RC2 Response', Shahbaz Memon, 21 Sep 2018

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Shahbaz Memon on behalf of the Authors (22 Sep 2018) Author's response Manuscript

ED: Referee Nomination & Report Request started (24 Sep 2018) by Alexander Robel

RR by Anonymous Referee #2 (09 Oct 2018)

RR by Anonymous Referee #1 (13 Oct 2018)

Suggestions for revision or reasons for rejection

I must say, I was surprised when I read the revised manuscript. As I could not find a pdf with track changes (please include this in your next submission!), I have performed a visual diff of the revised and the original manuscript based on line-breaks and found virtually no changes. This is not what I would expect from a revision of a manuscript that is in danger of being rejected. I was very close to resubmitting my original review, but I will try to be more constructive.

As the manuscript is virtually unchanged, it still suffers from serious problems, that prevent publication in its current form. For a revision, the authors need to address those problems and this will require rewriting large parts of the manuscript.

I still see a serious problem in the lack of focus and coherence of the manuscript. As it currently stands, this manuscript is a mixture of the description of the details of the ice sheet model setup of Valot et al. (2018) and the usefulness of UNICORE for running models. The latter should be fairly independent of the details of all software steps employed. The glaciological figures still only have minimal importance to the story of the manuscript (and Fig. 2 actually seems to be straight from Valot et al. (2018) – might want to reference that). The authors should re-consider the figures and their necessity for the main story. For the details of the glaciology we can read the TC paper, now that it is referenced. Then there is a bunch of technical details of the protocols employed inside the UNICORE infrastructure – some of these might be better placed in the UNICORE technical documentation.

The manuscript needs to focus on the improvements over the commonly accepted good practice, and not on the clean-up effort that was involved in this work. The concept any journal is to present improvements over the state-of-the-art. Of course I am able to fork a run without major effort on my production setups, and my shell scripts give me a decent idea on where they fail – in bash “set -evx” usually does that job. When you have guests, they don’t want to know how long it took you to tidy up your place, but what that beautiful piece of art on your wall is. Similarly, I could not care less about the problems of the setup before cleanup. So, please feel free to detail on where UNICORE provides an improvement over cleaned-up scripts, but not on the ways, bad scripts are trouble.

Finally, claims like “the user shouldn’t have to be able to handle shell script” or “users shouldn’t have to know their supercomputer / their queuing system” really do not help your case. Users who are not able to handle standard shell scripts and job headers are in deep trouble on supercomputers. However, there is no lack of support for learning these skills and quickly solving standard problems. These claims cost you substantial amounts of credibility. Please focus on real improvements your system can bring instead of creating and solving pseudo-problems.

Hide

ED: Reconsider after major revisions (17 Oct 2018) by Alexander Robel

AR by Shahbaz Memon on behalf of the Authors (24 Jan 2019) Manuscript

ED: Referee Nomination & Report Request started (28 Jan 2019) by Alexander Robel

RR by Anonymous Referee #1 (19 Feb 2019)

ED: Reconsider after major revisions (21 Feb 2019) by Alexander Robel

AR by Shahbaz Memon on behalf of the Authors (02 Apr 2019) Manuscript

ED: Referee Nomination & Report Request started (15 Apr 2019) by Alexander Robel

RR by Anonymous Referee #1 (10 May 2019)

ED: Publish subject to minor revisions (review by editor) (15 May 2019) by Alexander Robel

AR by Shahbaz Memon on behalf of the Authors (22 May 2019) Manuscript

ED: Publish subject to technical corrections (24 May 2019) by Alexander Robel

AR by Shahbaz Memon on behalf of the Authors (28 May 2019) Author's response Manuscript

Short summary

Scientific workflows enable complex scientific computational scenarios, which include data intensive scenarios, parametric executions, and interactive simulations. In this article, we applied the UNICORE workflow management system to automate a formerly hard-coded coupling of a glacier flow model and a calving model, which contain many tasks and dependencies, ranging from pre-processing and data management to repetitive executions on heterogeneous high-performance computing (HPC) resources.