the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 22 Nov 2021
Seasonal Sea Ice Prediction with the CICE Model and Positive Impact of CryoSat-2 Ice Thickness Initialization
Abstract. The Los Alamos sea ice model (CICE) is being tested in standalone mode for its suitability for seasonal time scale prediction. The prescribed atmospheric forcings to drive the model are from the NCEP Climate Forecast System Reanalysis (CFSR). A built-in mixed layer ocean model in CICE is used. Initial conditions for the sea ice and the mixed layer ocean in the control experiments are also from CFSR. The simulated sea ice extent in the Arctic in control experiments is generally in good agreement with observations in the warm season at all lead times up to 12 months, suggesting that CICE is able to provide useful ice edge information for seasonal prediction. However, the ice thickness forecast has a positive bias stemming from the initial conditions and often persists for more than a season, limiting the model’s seasonal forecast skill. In addition, thicker ice has a lower melting rate in the warm season, both at the bottom and top, contributing to this positive bias. When this bias is removed by initializing the model using ice thickness data from satellite observations while keeping all other initial fields unchanged, both simulated ice edge and thickness improve. This indicates the important role of ice thickness initialization in sea ice seasonal prediction.
This preprint has been withdrawn.
-
Withdrawal notice
This preprint has been withdrawn.
-
Preprint
(5668 KB)
Interactive discussion
Status: closed
-
RC1: 'Comment on tc-2021-353', Anonymous Referee #1, 11 Jan 2022
This manuscript analyzes impacts on sea ice thickness initialization on the simulation of sea ice extent and sea ice volume with the Los Alamos sea ice model (CICE) by comparing two sets of experiments initialized from the Climate Forecast System Reanalysis (CFSR) and CryoSat-2 satellite observations. The analysis of the experiments confirms results from earlier studies on initial sea ice thickness impacts on seasonal sea ice predictions. The manuscript is well structured and presentation of the results is reasonably clear. I would recommend acceptance of this manuscript for publication with a few minor revisions as listed below.
- Lines 30-32: Suggest adding Collow’s study to the citation. The work by Collow et al. (2015) is one of the earliest studies specifically on the need for improved sea ice thickness initial conditions. (Collow, T. W., W. Wang, A. Kumar, and J. Zhang, 2015: Improving Arctic sea ice prediction using PIOMAS initial sea ice thickness in a coupled ocean-atmosphere model. Mon. Wea. Rev., 143, 4618-4630. DOI: 10.1175/MWR-D-15-0097.1).
- Line 41: Suggest indicating that the UFS is to be the next NOAA’s operational coupled atmosphere-ocean-sea ice-land system for S2S predictions.
- Fig. 1: Reduce thickness of the curves so the differences can be seen more clearly.
- Lines 84-84; “The 12 once-per-month runs in 2014 are shown here as an example, as year to year variations are relatively small”. Does this mean the amplitude of interannual variations is smaller than that of model errors?
- Lines 87-89: “The Arctic SIE forecast matches observations better in the warm season than in the cold season at all lead times, and a positive SIE bias is seen in the cold season. When the SIV in the Arctic is higher than observations or reanalysis to begin with, this positive SIV bias often remains in the model throughout the forecast.”. Although initial sea ice thickness may have some impacts in the cold season, its impact is more significant during the melt season because it directly affects the melt rate. The larger SIE error in the cold season could be related to other factors such as model physics, atmospheric forcing, initial ocean state, and ocean dynamical processes.
- Lines 94-97 and Fig 2: To see the comparison between CTRL and Alt-Init more clearly, I suggest adding two panels to show the differences between CTRL and Alt-Init, one for SIE and the other for SIV. For SIE, it looks like the improvement in the summer melt season (Jun-Sep) is larger due to the use of better initial sea ice thickness.
- Fig. 7: I suggest making the curves 7c and 7d thinner.
- Lines 140-141: “Apparently, there are more bottom and top melt in the Beaufort, Chukchi and East Siberian Seas in the alt-init run than in the control run, …”. It looks to me the Alt-Init produces less top melt (blue colors) in Chukchi Sea and East Siberian Sea, and a large part of Beaufort Sea in the lower-left panel of Fig. 8.
Citation: https://doi.org/10.5194/tc-2021-353-RC1 -
AC1: 'Reply on RC1', Shan Sun, 21 Mar 2022
The comment was uploaded in the form of a supplement: https://tc.copernicus.org/preprints/tc-2021-353/tc-2021-353-AC1-supplement.pdf
-
RC2: 'Comment on tc-2021-353', Anonymous Referee #2, 09 Feb 2022
Review of "Seasonal Sea Ice Prediction with the CICE Model and Positive Impact of CryoSat-2 Ice Thickness Initialization" by Shan Sun and Amy Solomon.
This manuscript investigates the forecast skill of the sea-ice model CICE in stand-alone simulations. The analysis includes both the Arctic and Antarctic. The model CICE is shown to adequately reproduce the seasonality of the sea ice extent and volume, but with a larger difficulty in forecasting the winter and spring conditions. A particular interest is placed on the influence of the initial ice thickness on the forecast skill. Specifically, the use of the CFSR ice thicknesses, which over-estimates the ice thickness, is shown to decrease the forecast skills regardless of the initiation month. This is largely improved when ice thickness from Cryo-Sat2 is used instead. The simulations are presented as a baseline for future studies on the forecast skill of coupled models using CICE as the sea-ice component.
This manuscript is relatively well written, although many sentences are too long and confusingly constructed. The results are interesting and will interest many in sea-ice modelling community. Nonetheless, the manuscript suffers from an unclear problem statement in the introduction and from a tendency to describe figures without much in depth interpretations. This is especially important given that other studies have looked into the impact of ice thickness on forecast skills.
I believe this manuscript have potential for publication, but require major revisions.
Major points:
- There is a tendency (mostly in the introduction) in lumping too many ideas in complicated sentences. The text should be revised in that regard.
- The introduction lacks a problem statement and should be revised to clearly identify the scientific questions that the analysis aims to answer. A problem statement is vaguely formulated at L37-45 but mixed with the broader context. I believe that adding a paragraph devoted to the problem statement (mainly on the influence of the initial ice thickness on the forecast skills) would largely clarify the scope of the manuscript. In particular, it should clarify what information the stand-alone CICE simulations can bring that has not yet been documented.
- There is very little mention of the sea-ice dynamics (in the experiment setup, results and discussion) although it should largely affect the sea-ice extent, especially in a year-long simulation. The thermodynamics and dynamics contribution to the SIE is briefly investigated in Figure 7, although this analysis not clear (the methods are not described) and needs to be clarified.
- Much of the results are very descriptive and not thoroughly discussed. I believe that a comprehensive assessment needs to be include before publication. For instance, many statements are vague and general (e.g. the forecast skill is reduced in the alt-init simulations), despite the figures presenting much information. More in depth analysis of the results could include, for instance, explaining the differences in the SIC and SIV forecast skill patterns, and how it relates to the initial ice thickness.
Specific points:
L17-19: This sentence is confusingly constructed. Perhaps dividing it in smaller sentences would be clearer.
L20-22: This sentence is currently confusing as it covers too much while being too vague. For instance, what “forecast” and “predictability” are we talking about (weather conditions? Ocean? Climate?). The vague reference to the impact of sea ice conditions on teleconnections also needs expanding.
L23: It is not clear what we are talking about here. Weather?
L23: I am not sure that “leading” is the right verb... Perhaps “Driving”?
L29: The use of “In particular” is confusing here, as we were discussing the influence of sea ice on weather predictability, but now jump to sea ice forecasting.
L37-39: Do I understand that here, you validate the sea ice model component of a fully coupled ice-ocean-atm model, in a first step towards investigating how it feedbacks with the other components?
L37-45: The structure of this paragraph makes it difficult to understand the scope of the paper. It first indicates that the aim is to isolate feedback processes between coupled model components, then that it is to validate the sea-ice model used in NOAA UFS, in stand-alone simulations. However, I believe that the real goal here is to assess the influence of initial thicknesses on the ice predictability within CICE. This needs to be clarified.
L51-53: This sentence needs revisions.
L55: I suggest starting a new sentence after “experiment”.
Section 2: Some information on the dynamical component should be provided (e.g., I assume it is the standard EVP rheology and strength parameters in CICE ?).
L66-69: This long sentence could be improved.
Section 3: What is defined as a “reliable forecast”? This statement is made at various places throughout the result section, but sounds rather subjective.
L82: It should be specified here (not later) that Figure 1 only shows simulations from 2014.
L84: There is very significant inter-annual variability in Arctic (and Antarctic) sea ice extent, yet here you say that the inter-annual differences are small? This needs to be clarified. For instance, you show 2014, a year where the winter maximum was relatively small (~14.9 million km2) and the summer minimum remarkably large (5.0 million km2). It is possible that conclusions drawns from Figure 1 are not representative of different years, such as 2012.
L94: This result is very surprising to me, as the summer minimum is usually described as being more difficult to forecast and more dependent on the early summer meteorological conditions.
L91-102: Together, these results are confusing and should come with some analysis and explanation. The ice extent results seem to indicate that the model overestimates the ice growth, but that does not show in the SIV results. On the contrary, the SIV results seem to indicate biases in the spring and summer melt, but this does not show in the ice extent results. Why? Is this expected? Also, changing the initial thickness improves the overall skill but does not seem to change the temporal patterns. Does this implies that the thickness influences the error magnitude but not the predictability patterns?
L108: spatial distribution
Figure 2: A couple things that are concerning in these figures:
- Why is there a low concentration spots at the North Pole, but only for the top-middle and bottom-right panels, and of different size?
- What is that line of low concentration North of Franz Josef Land, running from the Laptev Sea to Svalbard? It is a very strange location and orientation for such an LKF, and it is seen both in April and October. More confusing, it is seen in the observations but only in October. Is it real or is it an artefact?
L117: On the initial thickness being the dominant source of error: how did you determine this dominance? There are no results on the contribution from other sources (dynamics mass balance, thermodynamics mass balances). This statement is also contradictory with the fact that the errors are large in the winter sea ice at all lead time: they are thus are likely dominated by other factors than initial ice thickness. The use of CryoSat also does not seem to change this pattern.
L122-123: This sentence is confusing and needs to be re-organised.
L129-130: I would edit to: “the SIE and SIV in the alt-init run in the fall ended up closer to NSIDC and PIOMAS than in the control run.”
L130-137: How are the dynamics and thermodynamics contribution measured? I find it somewhat surprising that the dynamics tendency is exclusively negative. Also unclear to me is how do you define the thermodynamics tendency in sea ice extent? The thermodynamics is usually defined by column physics, and not directly related to changes in area.
L138-144: How can the top melt be influenced by the ice thickness? It is more intuitive for the bottom melt, as the reduced thickness would also reduce the insulation, but it could also be mentioned.
L140: remove “apparently”.
Citation: https://doi.org/10.5194/tc-2021-353-RC2 -
AC2: 'Reply on RC2', Shan Sun, 21 Mar 2022
The comment was uploaded in the form of a supplement: https://tc.copernicus.org/preprints/tc-2021-353/tc-2021-353-AC2-supplement.pdf
-
RC3: 'Comment on tc-2021-353', Anonymous Referee #3, 18 Feb 2022
In this paper the authors analyse the seasonal prediction skill of a stand-alone CICE model forced/initialised using CFSR. The study is bipolar, although there is a much stronger focus on the Arctic. The authors show that initialising the model with observed sea ice thickness inferred from CryoSat-2 radar altimetry considerably improves the forecast skill, as has been shown previously for other models in other studies (as they correctly point out).
The manuscript is relatively well written and presented and the results will be of interest to the community. Therefore I think it worthy of publication in The Cryosphere.
However, the figures could do with a bit more attention in relation to figure captions and colourmaps. Furthermore, the study could be better motivated, and the discussion of the figures/results is often rather on the shallow side. I therefore recommend that this manuscript requires considerable revision before it is accepted for publication here.
## Particular points ##
A detailed list of comments can be found in the attached pdf document but I highlight here a few points that will particularly need addressing.
- the study needs to be a bit better motivated. The main motivation I can see for the study is lines 36-42 which states that fully coupled (AOIL) models are "considered the ultimate tool" (which incidentally would be considered an insult here!) for sea ice seasonal prediction but here the stand-alone model is used in order to "separate various feedbacks among the components of a fully coupled model". However, this separation of feedbacks is not done in the ensuing manuscript! It is also not mentioned anywhere (albeit a trivial point) that the stand-alone approach is much cheaper.
- there is no consideration of internal variability, which is a huge factor for sea ice and in polar regions generally, or significance. Many of the figures contain means of multiple years of model runs, which could also include error bars or shading to help understand the impact of internal variability (or at least inter-annual variability over the study period). Likewise hatching could be added to difference plots to try and portray to the reader how significant the changes are in relation to natural/chaotic differences.
- the CryoSat-2 data, and the way that it is used to initialise the model, are poorly described and so I am left wondering whether things have been done sensibly. There is no mention of what happens with thinner ice (for which CS-2 errors are near-infinite!) and no mention of what is done with the snow on top of the sea ice. Furthermore, it looks like they have not been very careful with their QC because the CryoSat-2 "pole hole" appears as open water in the sea ice concentration for their "alt-init" runs!
- >95% of the article is focussed on the Arctic but with approx. 4 sentences and a 1-panel figure on the Antarctic, which feels a bit orphaned within the bigger picture of this manuscript. I think the authors should drop the Antarctic and limit the scope of this study to focus on the Arctic only - particularly given that the impact of SIT initialisation cannot be evaluated there, which is actually the second half of the manuscript title!
- the results are often only described in a very shallow way without any mechanisms or processes being given. For example, the increased basal & top melting for the runs with thinner sea ice is not obvious and so the mechanisms should be talked about
- there is general confusion between 1D and 2D sea ice variables/quantities in the figures and accompanying text. For example, sea ice "extent", "area" and "concentration" seem to be used interchangeably and so are "thickness" and "volume"
- many of the titles, legends and colourmaps used in the figures are not intuitive for the reader. there are also some 'rainbow' colourmaps, which are also problematic for people who suffer from colour-blindness.
-
AC3: 'Reply on RC3', Shan Sun, 21 Mar 2022
The comment was uploaded in the form of a supplement: https://tc.copernicus.org/preprints/tc-2021-353/tc-2021-353-AC3-supplement.pdf
Interactive discussion
Status: closed
-
RC1: 'Comment on tc-2021-353', Anonymous Referee #1, 11 Jan 2022
This manuscript analyzes impacts on sea ice thickness initialization on the simulation of sea ice extent and sea ice volume with the Los Alamos sea ice model (CICE) by comparing two sets of experiments initialized from the Climate Forecast System Reanalysis (CFSR) and CryoSat-2 satellite observations. The analysis of the experiments confirms results from earlier studies on initial sea ice thickness impacts on seasonal sea ice predictions. The manuscript is well structured and presentation of the results is reasonably clear. I would recommend acceptance of this manuscript for publication with a few minor revisions as listed below.
- Lines 30-32: Suggest adding Collow’s study to the citation. The work by Collow et al. (2015) is one of the earliest studies specifically on the need for improved sea ice thickness initial conditions. (Collow, T. W., W. Wang, A. Kumar, and J. Zhang, 2015: Improving Arctic sea ice prediction using PIOMAS initial sea ice thickness in a coupled ocean-atmosphere model. Mon. Wea. Rev., 143, 4618-4630. DOI: 10.1175/MWR-D-15-0097.1).
- Line 41: Suggest indicating that the UFS is to be the next NOAA’s operational coupled atmosphere-ocean-sea ice-land system for S2S predictions.
- Fig. 1: Reduce thickness of the curves so the differences can be seen more clearly.
- Lines 84-84; “The 12 once-per-month runs in 2014 are shown here as an example, as year to year variations are relatively small”. Does this mean the amplitude of interannual variations is smaller than that of model errors?
- Lines 87-89: “The Arctic SIE forecast matches observations better in the warm season than in the cold season at all lead times, and a positive SIE bias is seen in the cold season. When the SIV in the Arctic is higher than observations or reanalysis to begin with, this positive SIV bias often remains in the model throughout the forecast.”. Although initial sea ice thickness may have some impacts in the cold season, its impact is more significant during the melt season because it directly affects the melt rate. The larger SIE error in the cold season could be related to other factors such as model physics, atmospheric forcing, initial ocean state, and ocean dynamical processes.
- Lines 94-97 and Fig 2: To see the comparison between CTRL and Alt-Init more clearly, I suggest adding two panels to show the differences between CTRL and Alt-Init, one for SIE and the other for SIV. For SIE, it looks like the improvement in the summer melt season (Jun-Sep) is larger due to the use of better initial sea ice thickness.
- Fig. 7: I suggest making the curves 7c and 7d thinner.
- Lines 140-141: “Apparently, there are more bottom and top melt in the Beaufort, Chukchi and East Siberian Seas in the alt-init run than in the control run, …”. It looks to me the Alt-Init produces less top melt (blue colors) in Chukchi Sea and East Siberian Sea, and a large part of Beaufort Sea in the lower-left panel of Fig. 8.
Citation: https://doi.org/10.5194/tc-2021-353-RC1 -
AC1: 'Reply on RC1', Shan Sun, 21 Mar 2022
The comment was uploaded in the form of a supplement: https://tc.copernicus.org/preprints/tc-2021-353/tc-2021-353-AC1-supplement.pdf
-
RC2: 'Comment on tc-2021-353', Anonymous Referee #2, 09 Feb 2022
Review of "Seasonal Sea Ice Prediction with the CICE Model and Positive Impact of CryoSat-2 Ice Thickness Initialization" by Shan Sun and Amy Solomon.
This manuscript investigates the forecast skill of the sea-ice model CICE in stand-alone simulations. The analysis includes both the Arctic and Antarctic. The model CICE is shown to adequately reproduce the seasonality of the sea ice extent and volume, but with a larger difficulty in forecasting the winter and spring conditions. A particular interest is placed on the influence of the initial ice thickness on the forecast skill. Specifically, the use of the CFSR ice thicknesses, which over-estimates the ice thickness, is shown to decrease the forecast skills regardless of the initiation month. This is largely improved when ice thickness from Cryo-Sat2 is used instead. The simulations are presented as a baseline for future studies on the forecast skill of coupled models using CICE as the sea-ice component.
This manuscript is relatively well written, although many sentences are too long and confusingly constructed. The results are interesting and will interest many in sea-ice modelling community. Nonetheless, the manuscript suffers from an unclear problem statement in the introduction and from a tendency to describe figures without much in depth interpretations. This is especially important given that other studies have looked into the impact of ice thickness on forecast skills.
I believe this manuscript have potential for publication, but require major revisions.
Major points:
- There is a tendency (mostly in the introduction) in lumping too many ideas in complicated sentences. The text should be revised in that regard.
- The introduction lacks a problem statement and should be revised to clearly identify the scientific questions that the analysis aims to answer. A problem statement is vaguely formulated at L37-45 but mixed with the broader context. I believe that adding a paragraph devoted to the problem statement (mainly on the influence of the initial ice thickness on the forecast skills) would largely clarify the scope of the manuscript. In particular, it should clarify what information the stand-alone CICE simulations can bring that has not yet been documented.
- There is very little mention of the sea-ice dynamics (in the experiment setup, results and discussion) although it should largely affect the sea-ice extent, especially in a year-long simulation. The thermodynamics and dynamics contribution to the SIE is briefly investigated in Figure 7, although this analysis not clear (the methods are not described) and needs to be clarified.
- Much of the results are very descriptive and not thoroughly discussed. I believe that a comprehensive assessment needs to be include before publication. For instance, many statements are vague and general (e.g. the forecast skill is reduced in the alt-init simulations), despite the figures presenting much information. More in depth analysis of the results could include, for instance, explaining the differences in the SIC and SIV forecast skill patterns, and how it relates to the initial ice thickness.
Specific points:
L17-19: This sentence is confusingly constructed. Perhaps dividing it in smaller sentences would be clearer.
L20-22: This sentence is currently confusing as it covers too much while being too vague. For instance, what “forecast” and “predictability” are we talking about (weather conditions? Ocean? Climate?). The vague reference to the impact of sea ice conditions on teleconnections also needs expanding.
L23: It is not clear what we are talking about here. Weather?
L23: I am not sure that “leading” is the right verb... Perhaps “Driving”?
L29: The use of “In particular” is confusing here, as we were discussing the influence of sea ice on weather predictability, but now jump to sea ice forecasting.
L37-39: Do I understand that here, you validate the sea ice model component of a fully coupled ice-ocean-atm model, in a first step towards investigating how it feedbacks with the other components?
L37-45: The structure of this paragraph makes it difficult to understand the scope of the paper. It first indicates that the aim is to isolate feedback processes between coupled model components, then that it is to validate the sea-ice model used in NOAA UFS, in stand-alone simulations. However, I believe that the real goal here is to assess the influence of initial thicknesses on the ice predictability within CICE. This needs to be clarified.
L51-53: This sentence needs revisions.
L55: I suggest starting a new sentence after “experiment”.
Section 2: Some information on the dynamical component should be provided (e.g., I assume it is the standard EVP rheology and strength parameters in CICE ?).
L66-69: This long sentence could be improved.
Section 3: What is defined as a “reliable forecast”? This statement is made at various places throughout the result section, but sounds rather subjective.
L82: It should be specified here (not later) that Figure 1 only shows simulations from 2014.
L84: There is very significant inter-annual variability in Arctic (and Antarctic) sea ice extent, yet here you say that the inter-annual differences are small? This needs to be clarified. For instance, you show 2014, a year where the winter maximum was relatively small (~14.9 million km2) and the summer minimum remarkably large (5.0 million km2). It is possible that conclusions drawns from Figure 1 are not representative of different years, such as 2012.
L94: This result is very surprising to me, as the summer minimum is usually described as being more difficult to forecast and more dependent on the early summer meteorological conditions.
L91-102: Together, these results are confusing and should come with some analysis and explanation. The ice extent results seem to indicate that the model overestimates the ice growth, but that does not show in the SIV results. On the contrary, the SIV results seem to indicate biases in the spring and summer melt, but this does not show in the ice extent results. Why? Is this expected? Also, changing the initial thickness improves the overall skill but does not seem to change the temporal patterns. Does this implies that the thickness influences the error magnitude but not the predictability patterns?
L108: spatial distribution
Figure 2: A couple things that are concerning in these figures:
- Why is there a low concentration spots at the North Pole, but only for the top-middle and bottom-right panels, and of different size?
- What is that line of low concentration North of Franz Josef Land, running from the Laptev Sea to Svalbard? It is a very strange location and orientation for such an LKF, and it is seen both in April and October. More confusing, it is seen in the observations but only in October. Is it real or is it an artefact?
L117: On the initial thickness being the dominant source of error: how did you determine this dominance? There are no results on the contribution from other sources (dynamics mass balance, thermodynamics mass balances). This statement is also contradictory with the fact that the errors are large in the winter sea ice at all lead time: they are thus are likely dominated by other factors than initial ice thickness. The use of CryoSat also does not seem to change this pattern.
L122-123: This sentence is confusing and needs to be re-organised.
L129-130: I would edit to: “the SIE and SIV in the alt-init run in the fall ended up closer to NSIDC and PIOMAS than in the control run.”
L130-137: How are the dynamics and thermodynamics contribution measured? I find it somewhat surprising that the dynamics tendency is exclusively negative. Also unclear to me is how do you define the thermodynamics tendency in sea ice extent? The thermodynamics is usually defined by column physics, and not directly related to changes in area.
L138-144: How can the top melt be influenced by the ice thickness? It is more intuitive for the bottom melt, as the reduced thickness would also reduce the insulation, but it could also be mentioned.
L140: remove “apparently”.
Citation: https://doi.org/10.5194/tc-2021-353-RC2 -
AC2: 'Reply on RC2', Shan Sun, 21 Mar 2022
The comment was uploaded in the form of a supplement: https://tc.copernicus.org/preprints/tc-2021-353/tc-2021-353-AC2-supplement.pdf
-
RC3: 'Comment on tc-2021-353', Anonymous Referee #3, 18 Feb 2022
In this paper the authors analyse the seasonal prediction skill of a stand-alone CICE model forced/initialised using CFSR. The study is bipolar, although there is a much stronger focus on the Arctic. The authors show that initialising the model with observed sea ice thickness inferred from CryoSat-2 radar altimetry considerably improves the forecast skill, as has been shown previously for other models in other studies (as they correctly point out).
The manuscript is relatively well written and presented and the results will be of interest to the community. Therefore I think it worthy of publication in The Cryosphere.
However, the figures could do with a bit more attention in relation to figure captions and colourmaps. Furthermore, the study could be better motivated, and the discussion of the figures/results is often rather on the shallow side. I therefore recommend that this manuscript requires considerable revision before it is accepted for publication here.
## Particular points ##
A detailed list of comments can be found in the attached pdf document but I highlight here a few points that will particularly need addressing.
- the study needs to be a bit better motivated. The main motivation I can see for the study is lines 36-42 which states that fully coupled (AOIL) models are "considered the ultimate tool" (which incidentally would be considered an insult here!) for sea ice seasonal prediction but here the stand-alone model is used in order to "separate various feedbacks among the components of a fully coupled model". However, this separation of feedbacks is not done in the ensuing manuscript! It is also not mentioned anywhere (albeit a trivial point) that the stand-alone approach is much cheaper.
- there is no consideration of internal variability, which is a huge factor for sea ice and in polar regions generally, or significance. Many of the figures contain means of multiple years of model runs, which could also include error bars or shading to help understand the impact of internal variability (or at least inter-annual variability over the study period). Likewise hatching could be added to difference plots to try and portray to the reader how significant the changes are in relation to natural/chaotic differences.
- the CryoSat-2 data, and the way that it is used to initialise the model, are poorly described and so I am left wondering whether things have been done sensibly. There is no mention of what happens with thinner ice (for which CS-2 errors are near-infinite!) and no mention of what is done with the snow on top of the sea ice. Furthermore, it looks like they have not been very careful with their QC because the CryoSat-2 "pole hole" appears as open water in the sea ice concentration for their "alt-init" runs!
- >95% of the article is focussed on the Arctic but with approx. 4 sentences and a 1-panel figure on the Antarctic, which feels a bit orphaned within the bigger picture of this manuscript. I think the authors should drop the Antarctic and limit the scope of this study to focus on the Arctic only - particularly given that the impact of SIT initialisation cannot be evaluated there, which is actually the second half of the manuscript title!
- the results are often only described in a very shallow way without any mechanisms or processes being given. For example, the increased basal & top melting for the runs with thinner sea ice is not obvious and so the mechanisms should be talked about
- there is general confusion between 1D and 2D sea ice variables/quantities in the figures and accompanying text. For example, sea ice "extent", "area" and "concentration" seem to be used interchangeably and so are "thickness" and "volume"
- many of the titles, legends and colourmaps used in the figures are not intuitive for the reader. there are also some 'rainbow' colourmaps, which are also problematic for people who suffer from colour-blindness.
-
AC3: 'Reply on RC3', Shan Sun, 21 Mar 2022
The comment was uploaded in the form of a supplement: https://tc.copernicus.org/preprints/tc-2021-353/tc-2021-353-AC3-supplement.pdf
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 655 | 244 | 41 | 940 | 30 | 30 |
- HTML: 655
- PDF: 244
- XML: 41
- Total: 940
- BibTeX: 30
- EndNote: 30
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1