Comparison of sea ice kinematics at different resolutions modeled with a grid hierarchy in the Community Earth System Model (version 1.2.1)

Xu, Shiming; Ma, Jialiang; Zhou, Lu; Zhang, Yan; Liu, Jiping; Wang, Bin

doi:https://doi.org/10.5194/gmd-14-603-2021

Articles | Volume 14, issue 1

https://doi.org/10.5194/gmd-14-603-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/gmd-14-603-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 14, issue 1

Model experiment description paper

|

29 Jan 2021

Model experiment description paper |

| 29 Jan 2021

Comparison of sea ice kinematics at different resolutions modeled with a grid hierarchy in the Community Earth System Model (version 1.2.1)

Shiming Xu, Jialiang Ma, Lu Zhou, Yan Zhang, Jiping Liu, and Bin Wang

Download

Final revised paper (published on 29 Jan 2021)
Supplement to the final revised paper
Preprint (discussion started on 02 Jun 2020)
Supplement to the preprint

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

RC1: 'Review of “Multi-Scale Sea Ice Kinematics Modeling with a Grid Hierarchy in Community Earth System Model (version 1.2.1)” by Xu et al. (2020)', Nils Hutter, 07 Jul 2020
- AC3: 'Reply to comments from Dr. Nils Hutter (referee #1)', Shiming Xu, 10 Sep 2020
RC2: 'Review of of "Multi-Scale Sea Ice Kinematics Modeling with a Grid Hierarchy in Community Earth System Model (version 1.2.1)"', Frederic Dupont, 10 Jul 2020
- AC1: 'Reply to comments from Dr. Frederic Dupont (referee #2)', Shiming Xu, 10 Sep 2020
RC3: 'Comments on Multi-Scale Sea ice Kinematics Modeling with a Grid Hierarchy in Community Earth System Model', Véronique Dansereau, 11 Jul 2020
- AC2: 'Reply to comments from Dr. Véronique Dansereau (referee #3)', Shiming Xu, 10 Sep 2020

Peer-review completion

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

AR by Shiming Xu on behalf of the Authors (01 Nov 2020) Author's response Manuscript

ED: Referee Nomination & Report Request started (10 Nov 2020) by Alexander Robel

RR by Nils Hutter (04 Dec 2020)

Suggestions for revision or reasons for rejection

Review of “Comparison of Sea Ice Kinematics at Different Resolutions Modeled with a Grid Hierarchy in Community Earth System Model (version 1.2.1)” from Xu et al. (2020)
Nils Hutter

The authors have done a good job in answering to all reviewer’s comments and revising the manuscript. Thank you for that. The focus of the paper remains as the original but the robustness of the presented analysis and the clarity of the manuscript have been improved significantly by the authors. Major changes include new simulations with the Hibler (1979) ice strength that lead to a more realistic ice distribution and the presented scaling analysis of sea ice deformation has been extended to a 3-month period. With the modifications the paper is a valuable contribution to the ongoing scientific discussion and I recommend the paper for publication after my minor comments are addressed:

General comments:

1. Scaling analysis: The authors perform a scaling analysis for three winter months (Dec to Feb) and include the new section 3.3 discussing the variability of the scaling exponents during that period, which I appreciate. Why do you not present these new results in Figure 9 (lower row) and Figure 10 (lower row) as a substitute of the analysis of three day intervals. As a reader, I would find this results more interesting. Nevertheless, with Figure 11 you showed that the scaling behaviour is persistent throughout the 3-month period, which answers my main concern raised in the first review.

2. Future work plans: In “Summary and discussion” you state at many places your future research plans (P25: L 9-10, L12-13, L20-23; P27: L1-6,… ). I understand that these plans are the logical next steps from the results presented in this study. However, I recommend to generalise these plans to recommendations to the community where further research is needed. Right now, some parts of the summary reads more as project proposal, which is inappropriate.

Specific comments:

P1 L3 “sea ice kinematics”
add small-scale. Basin-scale kinematics are also simulated at coarse resolution.

P2 L6-8 “and quasi-linear kinematic features by visual inspections (Kwok et al., 2008), including local deformation regions of sea ice failures and shearing.”
a verb missing in the sentence

P2 L29-30: “Purely Lagrangian models such as neXtSIM (Rampal et al., 2019) are potentially 30 free of the resolution issues for resolving small deformation features.”
This is not correct. First, neXtSIM shows deformation features as small as the resolution of the grid but it is not resolution independent. Second, the reason for this is likely to be the combination of the MEB rheology generating these LKFs and the Lagrangian advection scheme preserving them. The Lagrangian advection scheme is hardly the only reasons like it sounds in your sentence.

P2 L33 “(Hutter and Losch, 2020),”
This paper does not comment on potential violations of continuum assumption in high resolution simulations. The issue with deformation features in VP models is also in my opinion rather the missing memory of past deformation (as discussed in the cited paper) than the continuum assumption.

P7 L10 “(Lipscomb et al., 2007; Hibler, 1979)”
I recommend the following to clearly state which ice strength definition you are using: ridging/rafting scheme (Lipscomb et al., 2007) and ice strength model (Hibler, 1979)

P12 L32: “deformation caused by numerics”
What is this? Do you mean noise generated by the EVP solver? Please be more specific.

P8 L33: “As a consequence”
The effective resolution is coarser than the grid resolution because a few grid cells are needed to represent steep gradients in the drift fields, not due to numeric noise. At least in converged solutions. Please clarify.

P20 Figure11: Please use the same limits for both y-axis. Since you are showing the running mean of the same data in the second y-axis, different limits are confusing and misleading.

P21 L28-29: “multi-fractal sea ice deformation is accurately modeled by all three resolutions”
Rephrase to: “the modelled sea ice deformation is characterised by multi-fractal scaling”. Since you did not compared to observed multi-fractal characteristics “accurately” is misleading.

P22 L8: is -> are

P22 L9-10: “Specifically, the process dependent scaling 10 properties were found for representative days for sea ice drift and AO index.”
Unclear what process dependent properties are. Please rephrase and clarify.

P25 L27-28: “Based on a traditional implementation of EVP in our model, we have witnessed asymptotic and converging behavior of modeled kinematics by the increasing EVP subcycle count.”
Two times the same sentence, please remove one

Hide

RR by Frederic Dupont (07 Dec 2020)

Suggestions for revision or reasons for rejection

General comments:

This is the revised manuscript accompanied by responses to the reviews of the original manuscript. The authors present the results of one component of a new earth system, namely the sea-ice. They first introduced a new grid generator for a tripolar grid. Then, they show initial results of the sea-ice component under normal-year-forcing (a kind of climatological forcing with synoptic scales sampled from a specific year) and different resolutions, including multi-scale analysis of the sea-ice deformation. Finally, they show the sensitivity of the dynamics of the model to the grid resolution and a long-debated parameter in the dynamics, namely the number of subcycling used in solving the viscous plastic equations.

The manuscript has improved in general but the English needs to be polished once more. The experiments are well described. The results are well presented and analyzed using existing diagnostic tools. More context for what the authors are aiming at would help however the reader as well.

1-I understand that the motivation in section 3.2 was to claim spatial multi-fractality in CICE which was originally disproved by Girard et al. (2009) => need to refer and discuss their results. Moreover, since the deformation field has not converged yet with respect to the NDTE parameter for TS015 and TS005 (ndte=240 in this section), I am curious to know the sensitivity to the parameter. In passing, I note that the sign of beta varies from negative to positive without clear explanations...

2-About the grid generation, I am grateful that the authors gave their rationale for developing their own generator, however this is done only in their response and is not reflected in the text. The reader is missing that context and is still left in the black. Notably the reference to the pioneered work of others is left to the appendix which leaves the wrong impression that the authors claim for the own the technique in the main part of the text. In their response, when comparing to the tri-polar ORCA grids, the authors imply that POP/CICE use a different and unique "U-fold", which I believe is a misconception. The POP manual (top page 46 of https://www.cesm.ucar.edu/models/cesm1.2/pop2/doc/sci/POPRefManual.pdf) does acknowledges that their tripolar grid is a courtesy of Madec (i.e. the ORCA grid). The only differences between ORCA and POP grids are in fact just technical (ORCA grids include a redundant row of points along the western, eastern and northern edges; ORCA grids accommodate both, what they call, F- or T-folds whereas POP grids exclusively rely on the F-folding technique, which the authors refer to "U-fold"). POP Manual also refers to Murray (1996) given below.

3-The authors in their response state that they found that their sea-ice state was more realistic with a change of ice strength formulation. I would refer them to Ungermann et al. (2017) where there is a discussion on the subject of Hibler (1979) vs Rothrock (1975).

4-"EVP rheology": Careful with using EVP as it were a different rheology in itself. The rheology remains VP (the constitutive laws) but EVP introduces artificially slow elastic waves for numerical (claimed) convergence. It is actually very akin to the pseudo-compressible method that was once in favour in computational fluid dynamics for solving for incompressible flows. Hunke use "EVP model", Koldunov et al. "EVP solution" and Bouillon et al. (2013) "EVP method".

5-ndte=960 does not seem to be enough in TS005 as we still see some blue in the CAA (fig.13), whereas, in the coarser resolution runs, the region is fully white. Same can be said for the central Arctic where it takes an increasing ndte to get the equivalent pattern in TS045, TS015 and TS005 respectively (.e.g the pattern of (TS005, ndte=240) is equivalent to (TS015,ndte=120), and that of (TS015, ndte=960) is in between to (TS045, ndte=240,960)). Same can be said from Fig.12. By the way, I like very much the colour scale in fig. 13, much easier to see patterns.

6-ndte is not indicated in fig.7-8-9. There is only one mention in the text that ndte=240 is used for all comparison of the ice deformation and scaling analysis until the next section. I would appreciate a reminder in the captions for fast browsing readers like me who like to go and forth through a paper and its figures.

7-Missing Figure S4?

8-Please include drags in Table B1.

9-Appendix B: should you take into account grid deformation in your calculation of deformation?

minor comments:

why TS005 on year 31 (or end of year 30) does not match that of TS015 in fig.4?

Abstract, line 5: I would "the" before "Community Earth System Model"
Abstract, line 8. "In specific" does not sound English
page 13, line 17-21: are still talking about fig.9? The term "cumulative" has been dropped here.
page 13, line 23 one instance of "CDF" remains (the others have been converted to "cumulative PDF") I don't think that the figure 9 has been corrected in that regard.
page 13, line 24-35: "Also, since we witness flatter cumulative PDF slopes in scaled datasets, we expect the “real” tails of cumulative PDFs at 2.4 km flatter than the modeling result from TS005, which have the slope of -1.0 for Dec. 20th, and -0.5 for Feb. 6th." Honestly, I don't understand the sentence anymore.
page 13, line 25 "we expect the “real” tails of cumulative PDFs at 2.4 km flatter than..." missing "to be" before "flatter"
page 13, line 27 "This is due to the temporal averaging on Eulerian grid points attenuates" misses a "which" before "attenuates"
page 14, line 24 "sea ice status" I think the authors meant "sea ice state"
page 14, line 25-36: the whole following sentence "Since the experiments in this study target at climatologlical sea ice states" could be phrased better (these experiment in this study aim at obtaining a climatologlically converged sea ice state?)
CORE instead of CORE2 in appendix B, line 21.

References:

Girard, Lucas, et al. "Evaluation of high‐resolution sea ice models on the basis of statistical and scaling properties of Arctic sea ice drift and deformation." Journal of Geophysical Research: Oceans 114.C8 (2009).

Ungermann, Mischa, et al. "Impact of the ice strength formulation on the performance of a sea ice thickness distribution model in the A rctic." Journal of Geophysical Research: Oceans 122.3 (2017): 2090-2107.

Murray, Ross J. "Explicit generation of orthogonal grids for ocean models." Journal of Computational Physics 126.2 (1996): 251-273

Hunke, Elizabeth C. "Viscous–plastic sea ice dynamics with the EVP model: Linearization issues." Journal of Computational Physics 170.1 (2001): 18-38.

Bouillon, Sylvain, et al. "The elastic–viscous–plastic method revisited." Ocean Modelling 71 (2013): 2-12.

Koldunov, Nikolay V., et al. "Fast EVP Solutions in a High‐Resolution Sea Ice Model." Journal of Advances in Modeling Earth Systems 11.5 (2019): 1269-1284.

Hide

ED: Publish subject to minor revisions (review by editor) (09 Dec 2020) by Alexander Robel

AR by Shiming Xu on behalf of the Authors (11 Dec 2020) Author's response Author's tracked changes Manuscript

ED: Publish as is (17 Dec 2020) by Alexander Robel

Short summary

A multi-resolution tripolar grid hierarchy is constructed and integrated in CESM (version 1.2.1). The resolution range includes 0.45, 0.15, and 0.05°. Based on atmospherically forced sea ice experiments, the model simulates reasonable sea ice kinematics and scaling properties. Landfast ice thickness can also be systematically shifted due to non-convergent solutions to an elastic–viscous–plastic (EVP) model. This work is a framework for multi-scale modeling of the ocean and sea ice with CESM.