Simulating the Early Holocene demise of the Laurentide Ice Sheet with BISICLES (public trunk revision 3298)

Matero, Ilkka S. O.; Gregoire, Lauren J.; Ivanovic, Ruza F.

doi:https://doi.org/10.5194/gmd-13-4555-2020

Articles | Volume 13, issue 9

https://doi.org/10.5194/gmd-13-4555-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/gmd-13-4555-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 13, issue 9

Development and technical paper

|

25 Sep 2020

Development and technical paper |

| 25 Sep 2020

Simulating the Early Holocene demise of the Laurentide Ice Sheet with BISICLES (public trunk revision 3298)

Ilkka S. O. Matero, Lauren J. Gregoire, and Ruza F. Ivanovic

Download

Final revised paper (published on 25 Sep 2020)
Preprint (discussion started on 06 Dec 2019)

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

SC1: 'Executive Editor Comment on gmd-2019-304', Astrid Kerkweg, 16 Dec 2019
- SC2: 'Question about authorship of the record containing archived model code', Ilkka Matero, 31 Jan 2020
RC1: 'Paper review', Michele Petrini, 16 Jan 2020
RC2: 'Review of Matero et al.', Alexander Robinson, 19 Jan 2020
AC1: 'Author comment in response to referee and short comments', Lauren Gregoire, 09 Mar 2020

Peer-review completion

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

AR by Lauren Gregoire on behalf of the Authors (24 Apr 2020) Author's response Manuscript

ED: Referee Nomination & Report Request started (06 May 2020) by Philippe Huybrechts

RR by Michele Petrini (08 May 2020)

RR by Alexander Robinson (16 May 2020)

Suggestions for revision or reasons for rejection

The manuscript as it is now is clearly improved over the original submission. The figures have been improved significantly and now help to explain the messages in the paper. The scientific quality is high and the experiments are interesting. There has also been a clear effort to show how the manuscript fits into the topics of GMD. I still feel some detail could be added to the technical aspects of the experiments, but overall, I would suggest publication with only minor changes.

The discussion of the mesh refinement experiments needs improvement. Figure 7 does a good job now of highlighting the differences between the two resolution simulations (standard and AMR_2). And I believe the sentence on P19L1-5 summarizes the conclusion of this permutation well based on what is seen in the figures. But the later discussion of this topic in Section 5 is inconsistent, particularly the sentence at P26L26-28: “Increasing the level of AMR in our simulations results in a divergent response in the timing of the saddle collapse (fig. 6a), suggesting that further refinement of the mesh could result in the ice saddle reaching flotation earlier, with knock-on effects on the dynamics of the collapse.” Therefore, I suggest revising the paragraph of P26L23-32 carefully.

Note that I do not see this lack of impact of very high-resolution modeling in this context as a negative point. To the contrary, Ω0 could already be considered unprecedented high resolution for simulating the Laurentide, and that should be emphasized in the text.

And, in fact, since there was little change between Ω0 and Ω2, it would have been neat to see how much computation you could save and still have viable results (eg running a simulation of Ω2 or even Ω3, but with the lowest order resolution at 20km or 40km). If you could save on computation, that would allow you to run ensembles with more members and arguably make a stronger case for accurately assessing the uncertainty. Adding such a test at this point in the review is not necessary, but it certainly would add value.

Related to the above, but this is minor: I noticed the comments about runtimes did not seem fully consistent. Please correct this and consider using a common metric, either model years per day or model years per hour, but not both:
P8: Ω2 = 396 model years per day = 16,5 model years per hour
P8: Ω3 = less than 10 model years per day
P17: Ω2 = ~50 model years per hour
P17: Ω3 = 1 model year per day = 0,04 model years per hour
It is definitely interesting information for the reader to understand what kind of computational undertaking these experiments entail.

As a last point about the mesh refinement (I promise!), I would like to see a few sentences about how the topography was generated for the different resolutions, as this was not clear. Was the topography generated at the highest resolution, and then aggregated to the lower resolutions as needed, or vice versa? This has implications as to how much detail is conserved in the high resolution meshes.

In the tables, I do not understand the column “Reference” – does it refer to sections of the paper? This should be explained in the caption.

Hide

ED: Publish subject to minor revisions (review by editor) (02 Jun 2020) by Philippe Huybrechts

AR by Lauren Gregoire on behalf of the Authors (10 Jun 2020) Author's response Manuscript

ED: Publish as is (04 Jul 2020) by Philippe Huybrechts

AR by Lauren Gregoire on behalf of the Authors (03 Aug 2020)

Short summary

The Northern Hemisphere cooled by several degrees for a century 8000 years ago due to the collapse of an ice sheet in North America that released large amounts of meltwater into the North Atlantic and slowed down its circulation. We numerically model the ice sheet to understand its evolution during this event. Our results match data thanks to good ice dynamics but depend mostly on surface melt and snowfall. Further work will help us understand how past and future ice melt affects climate.