Analysis of Sea Ice Timing and Navigability along the Arctic Northeast Passage from 2000 to 2019
Round 1
Reviewer 1 Report
The review comments by the reviewer have been addressed basically. The manuscript is now recommended for publication.
Author Response
Thank you very much for your publication comments. I have modified the results part.
Reviewer 2 Report
The authors of this manuscript have made major strides in expanding the description and improving the presentation of this manuscript. They have included substantial additions to the text, some which still needs clarification or modification. All the comments I have for this version are at the textual level, and many are minor grammatical notes.
Comments
Line 12: The full sentence on this line appears redundant. It could be cut to make space for suggestions below (if needed). If the strong adjective “crucial” is still desired, it could replace “apparent” in the line above.
Line 19-20: Assuming I’m reading this right, I think it would be clearer to say that the seasonal range of SIC is greater than the interannual variability from 2000-2019. The way this sentence is written, my initial interpretation would be that the change in seasonal sea ice concentration over 2000-2019 is greater than the change in annual sea ice concentration over 2000-2019, which is only true for some seasons.
Line 23-24: “the amount of multi-year ice” is vague. Saying “the high amount of multi-year ice” would be clearer. Similarly, describing temperature as “high” or “low” and the nature of the wind and sea ice motion (e.g., westerly, easterly, southerly, or northerly) would be helpful. The same modifications are desirable on Line 25 (temperature, wind and “amount of ice”), and Line 26 (wind). This should require one to two words added per concept, or eight to sixteen words total.
Line 70-71: The sentence starting “Citing this in both…” looks like advice from a reviewer, not something they intended to be copy/pasted into the manuscript.
Line 108: The “in September” at the end of this sentence is redundant and could be cut.
Line 109: Since this is not a hypothetical statement, I think “There can be” makes more sense than “There would be”.
Line 135-136: The University of Maine’s Climate Reanalyzer is only the plotting tool, so the dataset being plotted is important to note here. Based on the resolution, time span, and version listed by the authors, I suspect this is also ERA5 data, like the 2-m temperature. The text should be modified to clarify.
Line 166-167: The tense changes from present tense in the earlier part of the paragraph to future tense in this last sentence.
Line 171-172: The formula provided is only for calculating magnitude, not direction.
Line 196: Change “geographic” to “geography”.
Line 204: Change “gentle” to “gently” or “little”.
Line 288: Change “there is still a considerably” to “there was still a considerable”.
Line 289: Change “is relatively high” to “was relatively high”.
Line 301: Change “sparsely” to “sparse”.
Line 302-303: This statement is unclear. The authors say both that there is “a large amount” or unconnected crushed ice when sea ice concentration is 40-60%, but also say “which is relatively rare”. I’m not sure what “rare” is referring to – the presence of unconnected crushed ice? The occurrence of sea ice concentration conditions of 40-60%?
Line 315-316 & Figure 5: The distances cited in the text for the “ice segment” correspond to Sections 2 – 4 (e.g., Figure 1). The first value of 1131.71 nm does not, however, align with Figure 5. The first marker in Figure 5 precedes 1000 nm. Either the text or Figure 5 should be tweaked to match the following analysis. I think it’s the text that needs modification, because Figure 6’s x axis starts at 787 nm.
Table 2: I notice that only one vessel has something other than “Nordkapp” or “Bering Strait” listed for drive-in or departure – the Tianhui in 2018 has “Lianyungang” listed. This vessel also has the only duration greater than 12 days. My question is twofold: 1) Is “duration” the duration along the NEP or the total duration from port to port? 2) are those extra 7 days for the Tianhui in 2018 the 7 days necessary for the vessel to sail from the port at Lianyungang to the Bering Strait? If so, modifying that to the time from Bering Strait to Nordkapp would be better for analyzing the NEP.
Line 359: Change “cannot” to “could not”.
Line 361-362: In these lines the authors are using a comma to separate the thousands place, but they do not do so in Line 365-366
Line 401: It would be more helpful to specify that the temperature was “abnormally low by about 2°C” rather than just “abnormally about 2°C”.
Line 402: I agree that the Barents Sea had average temperatures in 2000, but it looks to me like the Laptev Sea is quite similar the Kara Sea in having negative anomalies for temperature. That’s also consistent with the wind descriptions above.
Line 416: Saying “these two years” instead of “the past two years” might be clearer.
Line 451: I believe the authors mean that the “temperature anomaly” is “obviously negative”, not the “temperature”. (In both the Barents and Kara Seas, there are temperatures over 0°C.)
Line 494: There is a rouge quotation mark at the end of this line.
Line 506: I believe this should say “northeast winds” instead of “northwest winds”. That is based on looking at the wind figures, but also the description of sea ice piling up on the east side of Severnaya Zemlya, which would require a wind with an easterly component.
Line 607-611: I have two comments for this section.
1) It is very short, which makes me think it might be better to incorporate into the current Conclusions section instead. It might therefore be better to label the last section “Dicussion and Conclusions”, but “Conclusions” would still suffice.
2) The bigger influence of downwelling radiation on sea ice is thermodynamic (melting) rather than dynamic, but here the authors seem to highlight instead that clouds and radiation fluxes are indirectly impacting sea ice motion by modifying the wind. Although cloud cover variability can have an influence on wind, the thermodynamic influence of clouds on sea ice is more important. Therefore, I’d tweak the language here to a) only emphasize the pressure gradient force in driving the wind field and b) separately calling out cloud cover as impacting radiation and therefore sea ice melt.
Line 654: Replace “he wind” with “the wind”.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
This paper addresses an important topic of Arctic ice conditions from the perspective of Arctic shipping. Despite of the originality and the obvious efforts in preparation of the raw data, the curent draft must be revised thoroughly to be meet the qualifciation of the scicentific journal. In particular:
- A new round of literature review is recommened to better understand the state-of-the art of this topic, with the emphasis from the perspective of martime transportation. Recent publications such as 1) Li, Z., Ringsberg, J. W., & Rita, F. (2020). A voyage planning tool for ships sailing between Europe and Asia via the Arctic. Ships and Offshore Structures, 1-10.; 2) Wang, Y., Liu, K., Zhang, R., Qian, L., & Shan, Y. (2021). Feasibility of the Northeast Passage: The role of vessel speed, route planning, and icebreaking assistance determined by sea-ice conditions for the container shipping market during 2020–2030. Transportation Research Part E: Logistics and Transportation Review, 149, 102235.
- There are many statement claimed or assumed without any references given. For example, 40% ice concentration is taken as the navigability, which is questionable. The importance of different ice-classes are generally ignored, and the role of icebreaker assistance was not mentioned at all.
- Some of the terms are wrong, for instance, “Velikitsky Strait” should read “Vilkitsky Strait”; “Eastern Siberian Sea” should be “East Siberian Sea”
Reviewer 2 Report
Overview
Understanding the factors that influence the navigability of the Northeast Passage can be impactful work for shipping activity. This paper highlights several years in the past two decades for which the Northeast Passage was closed or open for an especially long time. It explores how two general factors affected these years: a) summer weather (temperature anomalies and wind regimes) and b) sea ice preconditioning (the amount of multiyear/thick ice that is present along the route at the end of winter). Although some papers have delved into such issues in the past, this paper spends much more time on case studies that are particularly focused on the Vilkitsky Strait as a major chokepoint on the Northeast Passage. That is a valuable addition.
The pre-conditioning arguments are sensible to me, and the figures and interpretations algin well. Likewise, the relationships described between temperature anomalies and sea ice conditions follow logically. However, there are several instances for which the wind relationships described are unclear and/or inconsistent with what I see in the figures. Additionally, the wind vectors are difficult to see in some cases, so I used the University of Maine’s reanalysis plotter to compare results (https://climatereanalyzer.org/reanalysis/monthly_maps/). I sometimes (though not always) find myself in disagreement with the authors’ interpretations regarding the winds.
Major Points
- The issues with winds described above need to be resolved. Some of it might be resolved by improving the figure quality, but since what I saw on the Climate Reanalyzer doesn’t always match what the authors are describing, I think it likely that some of the interpretations should also be modified. (For example, there might be some cases when temperatures are higher than normal despite winds being northerly.)
- There are some aspects of the methods are not entirely clear, and that could be greatly helped by having a stand-alone methods section after the data section. Moving some of the statements from the results to such a methods section would help clarify some points.
- Especially in the abstract and conclusions, but also elsewhere, this manuscript would benefit from using more precise or specific language.
More specific examples of these points can be found in the below line-by-line comments.
Abstract & Conclusions
- Conclusion #1: See comment on lines 205-207. The data presented in this manuscript indicates they are the same length. This may just be a wording issue.
- Conclusion #2 and Abstract Lines 19-20: See comment on lines 148-149 as well. Perhaps there is some confusion caused by the wording, but I see plenty of interannual variation in Figure 2. More precise language and statistical tests (i.e., with a p-value) would be helpful for making statements about significance.
- Conclusions #4 & 6 and Abstract Lines 22-27: The statements here are sometimes vague. For example, in lines 24-26 and 608-610, it is not clear how temperature or wind impacted 2012, 2015, and 2019. Conclusion #5 is clearer because it specifically states how more multi-year ice than normal and anomalously low summer temperatures contributed to high sea ice concentration.
Geographic Terms: Often, this paper is describing the chokepoint in the Northeast Passage between Severnaya Zemlya (i.e., “Northland Islands”) and the Taymyr Peninsula, which I know as “Vilkitsky Strait”. The authors are using “Velikitsky Strait” instead. That might be worth double-checking. Moreover, the name of the islands seems to change throughout the manuscript. The authors use “Northland”, “Bordelia”, and “Bordi” all to refer to the archipelago north of Vilkitsky/Velikitsky Strait. At least, I think they’re all the same reference based on the map annotations. If not, more explanation of “Bordelia” and “Bordi” is needed. Finally, although “Northland Islands” is an accurate English version of the archipelago name, I typically to see “Severnaya Zemlya” in English-language journals. That might be better to use.
Figures
Several Figures: For any figure with the x axis as nautical miles (but especially figures 2-5), it would be helpful to have the sections labeled – e.g., at the top or bottom of each graph). Standardizing the sections like this might remove the need for some of the additional annotation with shapes.
Figure 7 and 10: The legends for these plots are difficult to read at this size. I think the color scheme for the left-hand plots is good and clear, but the rainbow color scheme for the right-hand plots makes the NEP line very difficult to make out. I also cannot read the vectors without zooming in a lot, and even then, I can clearly see only some. I started estimating direction from the SLP plots instead and then resorted to reproducing the plots with the University of Maine’s Climate Reanalyzer. To improve legibility I suggest
- Make the legends a little bigger
- Use a color palette for wind speed that does not clash with the NEP passage line (e.g., something monochrome).
- Make the wind vectors bigger and denser in the meridional direction but sparser in the zonal direction. Re-gridding the vectors to a polar grid before plotting may be the best option since the native grid has convergence at the pole.
Figures 8 and 11: I again find the legends small and difficult to read, but the rest is clear.
Figure 16, 18, and 20: The legends and wind vectors have the same issues as described for Figure 7 and 10.
Introduction
I think the introduction here is good, but there are two very recent papers that I suggest incorporating:
- Yu et al. (2021) notes the importance of wind direction on determining whether sea ice is advected into or away from Vilkitsky Strait. Citing this in both the introduction and results would be very beneficial.
- Chen et al. (2020) look at future scenarios in a modeling experiment, but they also focused on the Northeast Passage and (at times) Vilkitsky Strait.
Line 38: It’s not clear to me what “range” of Arctic sea ice means here. The difference between March and September Arctic-wide sea ice extent, perhaps? Regardless, clarifying in the main text would be helpful.
Line 81-92: This is a fairly long summary of results that is mostly a repeat from the abstract. I think it would be more helpful to see here what the authors goals and research questions were for this study rather than the results.
Study Area & Data
Line 103-104: Since the Laptev, Kara, and East Siberian Seas can be ice-free in September some years, this is an overstatement. Instead of saying “is covered by sea ice”, something like “may be covered by sea ice” would be better.
Line 117-122: Two notes here:
- Technically, the NSIDC data is a combined record of SSM/I and SSMIS (Special Sensor Microwave Imager/Sounder) for the period 2000-2019 – starting in January 2008, it’s SSMIS.
- There are two different algorithms from NASA for converting brightness temperatures into sea ice concentation. My guess is that the authors are using the NASA Team algorithm (https://nsidc.org/data/nsidc-0051), because that’s more common for Arctic research, but there is also the Bootstrap method (https://nsidc.org/data/nsidc-0079), which has different sensitivities (e.g., to melt ponds on sea ice). Including the URL/DOI would be helpful.
Line 118-119: I’m not sure why the study is limited to 2000-2019. I suspect that based on the 40% limit, this might be because there aren’t enough years with an open Northeast Passage before 2000. Regardless, I think articulating the reason here or in the intro would be best.
Line 128-130: The URL listed here does not go directly to a sea ice age dataset. I think the authors are referring to this dataset: https://nsidc.org/data/NSIDC-0611.
Comments Related to Wind & Temperature Anomaly Figures
These are the comments related to my major concern for this paper. I group them together because any revision might resolve multiple comments at once.
Line 347-349: I do not see this described relationship in the figures. The wind over the Laptev Sea is northerly but very weak. I can’t read the wind vectors for the other seas, and the SLP patterns in Figure 7a show the area of high pressure being centered over these seas. Although I would estimate northerly winds for the Chukchi Sea, I would not for the East Siberian Sea. Improving the wind vector presentation might help a lot here.
Line 349-353: Again, I agree on the Laptev Sea interpretation, but I am dubious about the other seas. In Figure 7d, it looks like northerly winds for the Kara Sea based on both the SLP pattern and the wind vectors I can make out. I can’t really read the wind vectors for the East Siberian and Barents Seas.
Line 381: The authors talk about north-westerly winds here impacting Velikitsky Strait, but earlier, they were discussing southerly winds in the Kara Sea. Figures 7b and 7d make me think the north-westerly direction is more appropriate for the entire Kara Sea, not just the Strait.
Line 399-409: For 2003, the winds are almost perfectly westerly throughout all four seas. It’s hard to see in the vector plots, but I know that type of SLP pattern. The interpretation of 2004 seems supported by the figures to me.
Line 413-415: The area through which the NEP runs in the Laptev and East Siberian Seas was not notably warmer or cooler than normal in 2003. The southeast sector of the Laptev Sea was warmer, but the southeast area of the East Siberian Sea was actually cooler than normal.
Line 416-418: I think Figure 11d shows the East Siberian Sea being colder than normal in 2004, not warmer than normal.
Line 491-492: In July-Sep 2012, the meridional winds at 850 hPa were predominantly northerly only in the East Siberian Sea – in the Barents Sea, the meridional winds were more southerly. Overall, winds were more zonal than meridional, though – mostly westerly. That is partly based on the provided figures and partly on what I re-produced here: https://climatereanalyzer.org/reanalysis/monthly_maps/.
Line 549-557: For 2019, the average SLP pattern over the Arctic Ocean is muted – no strong centers of action. The wind field (again, using both provided figures and https://climatereanalyzer.org/reanalysis/monthly_maps/) is highly variable, with weak winds overall. Therefore, wind-based arguments for 2019 are not compelling. That said, I can see northerly vectors in Figure 20b for the East Siberian Sea and southerly vectors for the Laptev Sea. That is consistent with the authors’ interpretation. But I also see northerly vectors over the Kara Sea. That is inconsistent with their interpretation.
Results (Except for Wind & Temperature Anomaly Figure Comments)
Line 148-149: The authors use the language here that “interannual difference” is “not significant”. However, there’s no statistical test being conducted, and I’m not sure what parameter they are describing as “not significant”. There certainly is year-to-year variability ranging over 20% at some points along the NEP. The earlier years in the record generally have the highest concentrations, suggesting there could be a trend. More specificity on the meaning of this statement would be helpful, and I suggest only using “significant” or “not significant” if associated with a p-value or confidence interval from a statistical test.
Line 163-165: Here, the authors are using the word “trend”, which normally in such research would mean a change over time. However, I’m not sure that this is what they mean. They also say “rate of variation is low” in 2015 and 2019, but it is not clear what that refers to either. The spatial variation along the NEP seems comparable for these years and 2012, so that seems unlikely to be what they mean. I suggest clarifying both points in the text.
Line 175: Again, the word “trend” is used, but not in a way I would expect. Perhaps the word “transect” would better describe variation along the NEP and “seasonality” would better describe differences amongst winter, spring, summer, and autumn.
Line 193 – 194: This line is unclear. It seems to say the NEP has fewer islands in spring than winter.
Line 205-207: The lengths for the melting and freezing periods are listed as Apr-Sep and Oct-Mar, respectively. Since those are both 6 months, it contradicts the statement that the latter is “slightly longer”.
Line 216-220: The statement that the NEP has sea ice “all year round” is misleading because the summer season includes July-September, and sea ice extent is substantially greater in July than September (e.g., Figure 4). I think Figure 3 does support saying that summer is the only season likely to be navigable for a wide range of vessels, so I would focus on that.
Line 224: “Linear change” suggests a regression analysis. “Transect” or “Spatial variability along the NEP” would be clearer language.
Table 1: Although using the same table, Shibata et al. (2015) used an upper limit of 30% sea ice concentration for “easy navigation”, not 40%, as used here. A recent paper by Yu et al. (2021) also used 30%, based in part on the work of Shibata et al. (2015). I think, then, that either 30% should be used or a different explanation for why 40% is better is necessary. To be clear, I think one is possible (e.g., Lei et al. (2015) used 50%), but just citing Shibata et al. (2015) is inconsistent since they went with 30%.
Table 2: It is not clear to me how the “actual” navigation is calculated. It might be based on the minimum duration along the NEP. It might be based on ship logs. It might be something else.
Line 322: I’m not sure of the meaning here.
Line 324: I assume that “summer” is July-September again (like in Figure 3), but it would be good to define “summer” here since June-August is also commonly used.
Line 334 & 480: “a certain impact” is vague for describing the La Niña. La Niña is also mentioned for 2012, but in neither case do the authors articulate why the La Niña matters to the NEP.
Line 360: There might be a missing word in “temperature of the sea” because it doesn’t say which seas. I’m guessing the authors mean East Siberian Sea, but I’m not sure.
Line 507-508: I don’t think “many storms” is the best characterization of the Parkinson and Comiso (2013) argument. In paragraph [12] they state: “Examination of wind fields from National Centers for Environmental Prediction (NCEP)/National Center for Atmospheric Research (NCAR) reanalysis data [Kalnay et al., 1996; Saha et al., 2010] reveals no major storms in the Arctic in June 2012 (see www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.surface.html). Storms are consequently rejected as the primary cause of the unusual June decay of the ice cover.”
There was one storm in August 2012 that made a big difference, but they also emphasize the importance of pre-conditioning (e.g., the thinner sea ice with a heavy fraction of melt ponds). I feel think that pre-conditioning argument could be helpful to the discussion here.
Line 515: Describing 2015 as reaching a “new low” for sea ice extent is deceiving since the September 2012 extent is still (as of May 2021) the lowest in the satellite record.
References
Some of the references (e.g, 5, 6, 21, 25) have the first author’s family name first whereas other references (e.g., 1, 3, 4, 24) have the first author’s family name second.
Works Cited in this Review
Chen, J., and Coauthors, 2020: Changes in sea ice and future accessibility along the Arctic Northeast Passage. Global Planet Change, 195, 103319, https://doi.org/10.1016/j.gloplacha.2020.103319.
Lei, R., H. Xie, J. Wang, M. Leppäranta, I. Jónsdóttir, and Z. Zhang, 2015: Changes in sea ice conditions along the Arctic Northeast Passage from 1979 to 2012. Cold Reg Sci Technol, 119, 132–144, https://doi.org/10.1016/j.coldregions.2015.08.004.
Miao Yu, Peng Lu, Zhiyuan Li, Zhijun Li, Qingkai Wang, Xiaowei Cao & Xiaodong Chen (2021) Sea ice conditions and navigability through the Northeast Passage in the past 40 years based on remote-sensing data, International Journal of Digital Earth, 14:5, 555-574, DOI: 10.1080/17538947.2020.1860144
Parkinson, C. L., and J. C. Comiso, 2013: On the 2012 record low Arctic sea ice cover: Combined impact of preconditioning and an August storm. Geophys Res Lett, 40, 1356–1361, https://doi.org/10.1002/grl.50349.
Shibata, H., K. Izumiyama, K. Tateyama, H. Enomoto, and S. Takahashi, 2013: Sea-ice coverage variability on the Northern Sea Routes, 1980–2011. Annals of Glaciology, 54, 139–148, https://doi.org/10.3189/2013aog62a123.
Reviewer 3 Report
Review of
Analysis of sea ice timing and navigability along the Arctic Northeast Passage from 2000 to 2019
by
Ji, Min, et al.
Summary:
This contribution aims to improve our understanding of navigability of the Northeast Passage (NEP) in the Arctic Ocean during years 2000 to 2019 using satellite passive microwave sea-ice concentration observations as the principal data source. Secondary data sources are sea-ice motion and age data sets and meteorological information provided by an atmospheric reanalysis. For the recent decade also sea-ice thickness observations are included. Special attention is paid to choke points en route that inhibit safe navigation along the NEP in its entirety. The analysis of the sea-ice concentration data is the most robust and convincing part of this contribution which comes to the conclusion that the likelihood for safe navigation along the NEP has overall increased over the two decades considered, and that two choke points exist - the central East Siberian Sea and the Vilkitsky Strait. Here safe navigation can be inhibited despite an otherwise open NEP. The more severe ice conditions encountered there during several years' summers are mainly caused by the combined action of the near surface atmospheric circulation, forcing sea-ice to drift into the NEP, and variations in the presence of older (than first-year ice) ice types in the vicinity of the NEP. This latter part of the contribution is the less convincing part of this contribution for reasons I lay out in my general comments.
I find the topic interesting. I also rate the data sets combined very useful and having great potential to unveal the observed variations in NEP opening. However, I have a number of serious concerns that pull me away from accepting this paper for publication.
I provide, in the following, my general concerns, a number of specific comments and a number of typos / editoral comments. I apologize for the overlap of specific and editoral comments; this is borne out of the generally difficult to read manuscript. I note that the general comments kind of summarize findings and issues I also wrote in the specific comments.
General comments:
GC1: The authors attempt to connect the observations along the NEP to other regions and processes on Earth like El Nino, extreme winters in North America and the like. I warmly recommend to avoid all of these in this technically oriented contribution. The statements made are often misleading, not to the point, and overly general, not being backed up by thoroughly worked out relationship between processes in involved. This is obviously an editoral comment.
GC2: The introduction but also the general layout of this contribution is not overly convincing when it comes to the processes or features along the NEP. Not considered are i) polynyas, ii) the flaw lead system, iii) fast ice coverage and duration, iv) melt-onset and freeze-up as derived from satellite passive microwave imagery. Here this contribution is neither complete with respect to embedding the work into what has been done nor is it complete with respect to discussing the influence of i) to iii) on the development of the NEP opening. Taking this general comment into account I rate a mandatory action.
GC3: The section about the observational data sets used is widely incomplete. A substantial number of additional information that is required is missing. Here I refer to the my specific comments. Mandatory to improve.
GC4: The choice of the atmospheric parameters is not well described. It confused me. I got the impression that the authors used the 850hPa air temperature and its anomaly to discuss processes at the surface such as melting.
In addition, I did not find the combination of air pressure and wind speed plus vectors particularly useful. The wind vector is a manifestation of the surface air pressure distribution; hence the information provided is redundant. Mandatory to improve.
GC5: What is completely missing is a description of the methodology. It is not clear how the average sea-ice concentrations along the NEP were derived. It is not clear how the sea-ice concentration data were geo-located with the NEP coordinates. Mandatory to improve. Nobody would be able to repeat the authors' analyses - as the paper is written currently.
GC6: I did not see an adequate motivation why the authors first looked at annual mean sea-ice concentrations, then seasonal mean sea-ice concentrations and then monthly mean sea-ice concentrations. The main conclusions they draw from the monthly data. Neither does - to my opinion - the analysis of the annual data reveal particularly useful information, nor is the conclusion new that during / at the end of summer the NEP is navigable best. My suggestions would be to focus on the monthly data rightaway. I cannot say this is mandatory to improve - I simply didn't see the added values of showing and discussing annual and seasonal mean sea-ice conditions under the main focus of this contribution.
GC7: The quality of the presentation / figures as well as the written text of section 4 requires substantial revisions. First of all, the paper would benefit from showing only a subset of the Arctic Ocean relevant for the processes along the NEP. That way you could improve the degree of detail - a very important point when it comes to the wind direction which is hardly to be seen. You could even plot these at 2.5 degree intervals. Getting rid of the surface pressure maps would aid in enlarging the panels of many figures, making the paper more easy to read and to follow for the reader. This is mandatory to improve.
Secondly, also the way the data are used to explain the conditions at the two choke points requires a more careful consideration of the time scales. It is not overly clear, why, e.g. summer-mean (July-September) wind and temperature fields are used together with a monthly drift field. It is not well taken into account that processes have a certain lag time and that sea-ice conditions in a particular region might be preconditioned by processes upstream in space or previously in time. Showing a sea-ice age map of March goes into the correct direction but why such maps are not shown for, e.g. August is not clear. I tend to say that to have a solid background to explain the NEP sea-ice condition variations towards the end of summer it would be very helpful to look into monthly maps of sea-ice concentration, sea-ice drift, sea-ice age and near surface wind vector from May through September. Ideally, you also take into account information about polynyas, fast ice coverage and the state of the flaw-lead system - all these pre-condition the sea-ice conditions along the NEP. This is mandatory to improve.
Thirdly, the way the results are written down is not overly concise and does often not use the common language / terms used. It is also very descriptive in nature with a number of repetitions which might be ok for such a technical paper but which could be streamlined such that you focus on the key elements of the processes / conditions involved. These are in my eyes: sea-ice drift, ice type, and sea-ice concentration distribution. Kind of mandatory as well because it would reduce the redundancy.
Specific comments:
L13/14: Suggest to begin a new sentence and write: "We use a sea-ice concentration threshold value of 40% to define the time window for navigating through the Arctic Northeast Passage (NEP)."
L15: Not clear what is meant with "year of abnormal navigation".
Please avoid beginning sentences with "And".
L19/20: What do you want the state in this first bullet? Interannualr variation of the sea-ice concentration as an annual average? What is the reference time window here?
"seasons are quite different" is not specific enough. Do you want to say that the inter-annual variation of the seasonal mean and/or the seasonal cycle of the sea-ice concentration has changed?
Lines 23-27: Air pressure and wind field are linked with each other. I suggest to get rid of the air pressure as a variable here.
L31-33: I am not sure that off-shore winds push sea-ice from the Arctic Ocean into the Vilkitsky Strait, because - in general - the Arctic Ocean is located north of the coastline and hence an offshore wind would tend to push sea ice away from the coast into the Arctic Ocean and not the other way round. I guess some reformulation would solve this discrepancy.
L37: I suggest to replace [2] by a more recent citation. In any avoid using a "The Cryosphere Discussion" paper if a final version in "The Cryosphere" is available.
L39: Consider replacing reference [4] by a more recent one. It is almost 15 years old.
L40: "melting season has extended" --> I am sure you find an adequate reference to back up this statement.
L41/42: I doubt "ground temperature" is the quantity you should refer to here. I guess it would be either surface temperature or near-surface air temperature.
I note that the reference [7] does not properly match with the statement it is used for. I guess the current IPCC report would be a better reference here.
L46: I am inclined to state there there are hundreds of papers which would be suitable as a reference here. Why do you use grey literature, i.e. a PhD study by using [8]? Please either delete the reference or add 2-3 more and an "e.g."
Adding an "e.g." in front of other citations further above might be a good idea as well because there are many papers dealing with the issues cited.
L76: Please spell out what SSM/I means. Please also add SSMIS because SSM/I data were superseeded by SSMIS data some time ago. The documentation of the data set used should reveal the date.
L77/78:
As suggested above: I would concentrate on wind and temperature and get rid of the near surface air pressure.
Avoid using "etc." Instead, specify which other parameters you used.
L81-92: This reads like part of the results section and is also already a repetition of some information given in the abstract. I suggest to delete these lines.
Instead, the introduction would benefit from a more specific description of the typical (already known) sea-ice conditions, better motivating the suite of data you are using. Information about polynyas / flaw leads and there year-to-year variability - promoting formation of ice free areas already in May, the fast ice extent and duration, the above-mentioned ice export out of Laptev Sea connected to likely thinner sea ice which melts out earlier in spring, the multiyear ice coverage north of your region of interest, and the studies dealing with melt onset and freeze-up date would greatly add to the introduction.
Also, a notion why you use satellite passive microwave data (you could have used ice charts from AARI or SAR imagery or visible/infrared imagery or scatterometer data ...) would be of advantage.
L102-108: "The middle section ... However ... became low. ... maximum in April" --> This reads quite complicated and I don't know what you intend with these sentences. Phrases such as "the sea ice ... experienced the melt process" or the "sea ice concentration became low" do not read fluently. "During Arctic summer sea ice concentrations decrease due to sea ice melt." could be one solution.
I suggest, however, to be more specific with respect to the ice cover in the sections of the NEP here. During winter (Nov.-April) the NEP is usually completely ice covered. Exceptions are the southern Chukchi Sea, and the Laptev and Kara Seas subject to the formation of - at times - quite persistent polynyas. These can trigger enhanced melt during spring. Melt-onset varies along the NEP (see a paper about melt onset you will dig out) but begins in May/June. Melt continues until September. Freeze-up usually commences in October and is typically rather quick with sea ice entering the NEP from the North but also from the South from along the Siberian coast. While the sea-ice area / extent are maximum in March I doubt that sea-ice concentration is at maximum in April. Please check and revise or delete.
But as I stated above ... it is not clear what your intention is here - whether you attempt to describe the climatological situation or similar. It might be good to back up this information by published (peer-reviewed) literature.
L117-122: This paragraph needs to be rewritten. You need to a) include SSMIS (because SSM/I sensors are not functional anymore), b) specify the data source and version (there are several sea-ice concentration products available), c) specify the limitations of its use / accuracy (land spill-over, reduced accuracy in summer e.g.), d) motivate why you prefer to use the NSIDC product over the novel and more innovative OSI-450/OSI-430-b climate data record of the sea ice concentration. In addition the final sentence "Compared with ..." appears not to be appropriate because i) the product based on SSM/I and SSMIS data would have 12.5 km grid resolution and ii) the 6.25 km product begins in 2002 and is based on AMSR-E data. Finally, please also provide the grid type (polarstereographic or EASE or other).
L124-128: Please motivate the usage of a the very coarse resolution NCEP reanalysis over finer resolved, more recent atmospheric reanalyses such as ERA5.
What is the grid resolution of this data?
Please specify your writing such that it is clear which parameters are surface parameters and which 850 hPa ones. Why do you include 850 hPa information? What is the motivation? Possibly the wind field is the 10m surface wind (u and v-components) and the temperature is the 2m air temperature ... ?
Did you get the anomalies from ESRL as well or did you compute these on your own? Since you used a quite specific base period of 2000-2019 I assumed the latter.
L128-136: Also the descriptions of all these data require more care and more details. Providing "arcticseaicenews" as the source for the ice age and motion data is not appropriate. Please provide the specific data set web pages for these two products. Please provide the time-period, temporal resolution and grid resolution plus the version of the data set used. Please make a note about on which primary data source the ice motion product relies on during summer (which is NCEP winds); please also provide the grid. It seems you used only data from August?
Please provide the time period for which you took the ice age data - as well as their grid resolution. You might also may want to add a note about their reliability (tendency to provide a too large fraction of old ice) and back up this information by an adequate reference (e.g. the paper by Korosov in The Cryosphere from 2018. Please motivate why you obtain a sea ice age map for the month of March. Instead of mentioning the tool used (ArcGIS) it would be more meaningful to state whether you averaged the 4 (?) maps of March and if so how. How did you treat a grid cell with an age changing over the course of March from 4 years to 1 year, for example?
Please be more specific with the data source for the sea-ice thickness data - I only see a general web page where I can get a diversity of data but not specifically the sea-ice thickness data. Please be more specific about the data them selves: gridded / non-gridded, spatial resolution, temporal resolution, time period used (you write March but not for which years).
L139-141: Please provide a motivation to use an annual average sea-ice concentration. Usually the information content of an annual average is not overwhelming.
In addition, your description of the methodology is poor. While Figure 1 provides a rough location of the NEP you now need to describe how you extracted sea-ice concentration values from the data set used. How did you co-locate the ship-track with the gridded sea-ice concentration product? Did you average over a certain number of grid cells left and right of the track? How did you treat eventually missing values? Did you chose an sea-ice concentration threshold?
I note in this context that you work with Nautical Miles for the NEP but the sea-ice concentration data come at an (so far not known) grid at km scale.
It is not sufficient to write that you used ArcGIS for this. You need to describe what you did.
L149 / Fig. 2: Since you refer to sections 1 through 5 in the context of the results shown in Fig. 2 I suggest to mark these sections in this Figure as well.
Are the horizontal dashed-dotted lines in Fig. 2 there by chance or do these have a specific meaning? Grid lines would have been put at regular intervals, therefore I am asking.
L171: I don't get this averaging of the remaining 13 years' sea-ice concentration data. Please be more specific and motivate your steps adequately.
L171/172 / Figure 3: "the average ... NEP" --> "the seasonal average sea-ice concentration along the NEP ... are calculated and shown in Fig. 3"
Even though you wrote something about the remaining 13 years the notion of "2002-2018" given in the legend could be misleading. If I were you, I would redo the analysis and compute the all-period average, i.e. 2000-2019 including the 7 selected years instead of wobbling around with averaging over these 13 years which overall are not that different from the selected 7.
L181/182: "sea ice density" --> sea-ice density is the weight of sea ice per volume. I guess you wanted to state something like "the black box highlights years 2000, 2001, and 2004 when the average sea-ice concentration in summer was above 50% in and around Vilkitsky Strait".
L193-196: "The NEP ... ablation of sea ice" --> I suggest to delete these explanatory statements. These are partly not correct and do not add useful knowledge here.
L196/197: "After the ... pole" --> I suggest to either delete this and stay focused at the NEP or write something like: "During summer the sea ice continues to retreat northwards, and ...". Reference to Fig. 3c should be made where you write about the NEP and not about the Arctic in general.
L202: Delete: ", which are at this time the least risky" at this point. It disturbs the flow of information because you have not made statements about the risk of navigation before (i.e. in winter and spring). Just keep that for a later stage in the paper.
L205-207: "Third, ..." --> This is a very global statement which is not backed up adequately by your Fig. 3 because it shows seasonal averages. Whether a value of 50% in spring originates from 100% until mid may and 0% afterwards until end of June it does not reveal. Therefore --> delete it. If you are after melt-onset and freeze-up times you need to dig into existing literature.
L216-220: Now tell me: What is new here? I don't see (yet) any added value compared to what is already known.
L223: You defined summer as July to September and put October into autumn further above. Hence "summer(from July to October)" is not consistent with your earlier definition. You could simply motivate inclusion of the first month of autumn by stating that the sea-ice cover minimum usually occurs in September and that it takes a while to grow the sea ice back.
L227: What you show here is not a time series chart. You show the monthly mean sea-ice concentration along the NEP for several years plus the average over 13 years (which I again suggest to change to all years).
Not sure I understand the 70% and the 80% threshold horizontal lines in July and October, respectively. These are also not noted in the caption.
L232-253: Several comments here:
a) Avoid to include descriptions about why sea-ice concentrations changed. This is going to be the topic of paper section 4 and can be completely omitted in section 3 for the sake of clarity and the sake of giving correct descriptions of the processes that occur. This applies, e.g., to "sea ice melting degrees" ... or ... "under the influence of ...
b) The sun rises but not the sea-ice concentration. It increases.
c) "the NEP was not open" --> this depends on your definition when the NEP is open, right? So far you have not presented any definition. Therefore, this needs to be rephrased.
d) "still significant anomalies" --> "significant" see one of my previous comments. "anomalies" are something you derive by subtracting an actual situation from the mean. Anomalies can have a positive and a negative sign. What you show are not anomalies. You could write that there "is still a considerably amount of sea ice"
e) "sea ice entered a freezing period" --> In October THE freezing period begins, hence use "the" instead of "a".
f) When it comes to commence of freeze-up in October you should take into account that although the sea-ice concentration increases quite a bit compared to September the sea ice is thin and hence more easy to ship through with an ice class for marginal ice zone for example. Another issue to consider is that sea-ice concentrations in October might be too low simply because the kind of algorithms used to compute the sea-ice concentration product underestimate the true sea-ice concentration in presence of thin ice (see e.g. Ivanova et al., 2015, The Cryosphere).
Finally: This is the third figure and hence the third attempt to discuss the sea-ice conditions along the NEP. While you try to develop the information logically, it isn't a surprise that the NEP has least sea ie during summer / beginning of autumn. Instead of working yourself (and the reader) through not overly new consideration of annual and seasonal mean sea-ice concentrations along the NEP I suggest to start rightaway with the presentation and interpretation of the monthly mean sea-ice concentration. This would be much more to the point.
L261-265: I don't understand what you do here. Over which sample the averaging ("average sea ice concentration") is carried out if you are working with daily data ("the first day", "3 consecutive days")? Please re-phrase accordingly.
L294-296: "For example ... early October" --> you need to state here that this agreement between your and the previous study refers to the actual opening dates given in Table 2
L301-308 / Table 3: I am puzzled by how meaningful it might be to have an average navigation time? Even if the average navigation time is 60 days ... a ship would not make it if one of the key locations is blocked. I - and possibly also the reader - would appreciate an explanation / motivation about why it does make sense to investigate the average period in addition to the actual one.
L342/343 / Fig. 7:
The panels of the figures are too small to obtain useful information from it. In line with earlier suggestions: remove the air pressure panels and only show the surface wind patterns which are the result of the surface air pressure distribution anyways. That way you can double the size of the remaining two panels and one can see where the wind blows. In the text you can of course mention that, e.g., northerly winds in the Vilkitsky Strait area in summer 2001 were caused by the combined action of a low pressure system in the Laptev Sea and high pressure over the Kara Sea; but you do not need to show this distribution.
Looking at the figures that come I would also suggest to not show the entire Arctic. It is a waste of space. I suggested to show a geographic area bound between 65N and 85N in latitude and between 45E and 195E. That way you can zoom in much better to the key locations. That way any wind vectors can be discriminated better (currently you plot them every 5 degree latitude which appears a bit coarse).
In addition, I suggest to highlight the choke points in the panels, i.e. those locations where the NEP remained blocked.
"flight segment"? I thought we are considering ships?
Figure 8: Panels a) and b) show summer mean 2m air temperatures? I doubt so. This would mean that, e.g., Stockholm had a summer mean air temperature of +6degC in 2000 and 2001. But perhaps ... these are 850 hPa temperatures. If this is the case, then the distribution and the values look ok ... however, then they cannot be used at all for the interpretation of sea ice melt. I am puzzled.
Also, "abnormal distribution" does not read well. Please use "anomaly" like you did in the data section.
L357-367 / Fig. 8:
a) After the data set description I am not sure panely c) and d) show 2m air-temperature anomalies as suggested here by your writing. Please check.
b) You need to be very careful with using causal relationships. The fact that the summer mean temperature of a particular year does not support melting overall is one way seeing it. It could be the other way round: sea ice that continued to be present inhibited any further warming and kept temperatures low.
c) air temperatures are not the sole reason for sea-ice melt. The net short- and long-wave radiation, their partitioning and the timing of their peak drive sea ice melt substantially as well - as does the snow cover.
d) Let's not debate about whether NCEP 2m-air temperature are reasonably accurate over sea ice during summer ... but doubts should be noted; or, in other words, I would not take an average 2m air-temperature of -3degC for granted. If I were you, then I would take a look at time series of local temperature observations of stations near the coast. These are possibly far more to the point.
L372/389 / Fig. 9:
While this analysis is worthwhile to do it is not carried out in a thorough enough way.
a) Why do you only use ice age maps of March? How do the respective maps look for July and August? The multi-year ice present in a certain area in March might well have moved away until September.
b) Why do you show a sea-ice motion map of August while you look at summer-mean values (July to September?) in case of the wind vector? This is not consistent and does not help in the interpretation. It is the full history of ice concentration and motion over the course of the summer which determines whether an area becomes ice free or not - in addition to melt. There is often also a time lag between forces / actions and the resulting sea-ice distribution.
c) I strongly advise you to take a look at peer-reviewed papers where similar processes were investigated and described to adopt the common language. This paragraph reads particularly difficult because it is full of formulations and statements that are not expressed adequately.
L391/392: Please avoid mentioning of weather phenomena somewhere else on Earth. Their causal link to NEP opening is not credible. This applies also to the previous sub-section (4.2.1) and subsequent subsections. Please stay focused on the NEP.
Figure 10 / L397: Same comment as for Fig. 7 in terms of the size.
L399-409: I kind of have the same comments as before with respect to your writing. It is partly confusing / not correct. How, for example, can "formed a large area of floating ice" be connected to "warm southerly wind"?
In addition, panels c) and d) are too small and coarse in resolution to adequately figure out the wind direction - even when zooming to 200%.
Please carefully check your statements about wind directions. While I agree that the Vilkitsky Strait area had a northerly wind component in summer 2003, the Laptev and East Siberian Sea had westerly winds.
I note that you argue already with temperatures here. Since you show these in Fig. 11 it might make sense to actually combine these, i.e. Fig. 10 (without pressure) and Fig. 11.
L413-420 / Fig. 11: I voiced my concerns about which temperature we see here already above. If this is the 850 hPa temperature - which I assume in the meantime - then it is not credible to discuss about ablation the way done in the manuscript.
L446-452 / Fig. 13: Instead of showing the pressure (is this the summer mean (July-September) or just September?) please show the wind speed and vectors along with the sea-ice drift vectors. What about the sea-ice age distribution in August 2002?
"It can be seen ... of the NEP" --> "Earlier termination of the actual navigation period in the NEP was caused by sea ice being advected by northerly to northwesterly winds (Fig. 13 a) associated with low pressure in the Laptev Sea (not shown) from the Arctic Ocean into the northeastern Kara Sea (Fig. 13 b), blocking the western entries to Vilkitsky Strait."
L458-465 / Fig. 14: Yes, in summer 2007 the surface air pressure distribution favoured an enhanced transpolar drift. No, your pressure map does not adequately explain why the NEP continued to be blocked near Vilkitsky Strait. In that summer, a series of low pressure systems moved into the Arctic Ocean from the Bering Sea / Sea of Okhotsk, traveling westward and pushing sea ice northwards / westwards. I suggest to show monthly wind speed and vector maps for July / August / September plus the monthly sea-ice drift and sea-ice age fields for June, July and August to adequately explain the situation that led to the late opening.
A side note: I suggest to add the information that this was the second lowest September sea-ice area within the satellite observation era.
L471-476 / Fig 15: Similarly to the previous case I suggest to illustrate the full series of events - simply because it is not understandable where the sea ice blocking the western entry to Vilkitsky Strait came from. Hence, I suggest to show wind speed / vector maps of June/July/August, and to show, for the same months maps of the sea-ice concentation, sea-ice age and sea-ice drift. I suspect that it is again the longer history of conditions that led to the late opening of NEP here.
L480: Please avoid to include teleconnections. It is misleading.
L483: "the third longest since it was opened ..." --> this reads as if the NEP was opened officially in 2000 ... please re-phrase.
L485/Figure 16: Same comments as I made for Fig. 7 and 10 as well as 8 and 11.
L487-495: Another paragraph to re-write once you have revised the Figure.
It doesn't matter whether the NEP was under predominantly low or high air pressure. What counts is the air pressure gradient and the direction of the resulting wind - as well as the cloud cover, which you omit completely in your manuscript - for good reason I think because it would have complicated your interpretations; still you should make a note later in the manuscript that you did not take cloud coverage into account and hence were not able to consider the effect of short- and/or longwave radiation flux anomalies.
There are no southerly winds in the Laptev and Kara Seas; wind speeds were quite low compared to any of the other years shown.
Figure 17: Here it becomes obvious that your choice of colors to denote different sections of the NEP is not optimal. An alternative solution would be to plot a chain of different symbols, i.e. plus signs for section 1, triangles for section 2, and so forth - all in black. I guess this would harmonize with all other figures shown as well - particularly if you accept my suggestion voiced in the context of Fig. 7 to reduce the area shown to certain latitude / longitude bounds.
L502-507: It seems you were tricked by your own rainbow color table. Please check your notion about sea-ice thickness values. I see for section 1: 2-2.5m, occasionally even 3m, section 2 (East Siberian Sea): around 2 m, section 3 (Laptev Sea) 0.5-1.0m(west) to 1.5-2m (east).
Note that coverage with valid sea-ice thickness data is smaller than the sea-ice age. Hence your comment about the "ice free state" in the Barents Sea should possibly be deleted.
L507-512: There are many more papers trying to explain the situation in 2012 than [26] - some opposing the statements made therein. Therefore and because again statements made in these sentences are not overly enlightening for the main focus of the paper, I suggest to delete this part.
L573-576: It is crucial that in this paragraph you also mention that you supported your conclusions by including information of sea-ice drift and age and for years after 2010 also thickness and by including atmospheric information such as near surface wind vector and near-surface air temperatures.
L579/580: "The closer ... concentration." --> Are you sure this statement holds after you have found that locations of the East Siberian Sea at relatively southerly latitudes and the Vilkitsky Strait are actually the main choke points of the NEP?
L581-594: I suggest to rewrite this whole paragraph by focussing rightaway on the monthly mean sea-ice concentrations.
L604-621: I suggest to condense the information presented in (4) to (6) to i) avoid repetition, ii) take into account that the wind field results from the air-pressure gradient (hence air pressure can be deleted), and iii) concentrate on 1-2 key processes that block the NEP. In the Eastern Siberian Sea this is most likely the presence of multi-year ice paired with the absence of strong winds (and hence sea-ice motion). In the Vilkitsky Strait region this is the persistence of sea-ice drift directions favoring blocking either the western or the eastern entries, additionally supported by the fraction of multi-year ice nearby.
Typos / Editoral remarks:
L28: Check writing of Vilkitsky Strait (throughout the paper)
L70: "believe" -->"suggest" or "conclude"
L97: Suggest to add "in the East" to be consistent with "Norwegian Sea in the West"
L149: I did not yet comment on the usage of "significant". But now I see that you use this term without having made any statistical tests about the significance. Therefore I suggest that you replace any usage of this term by either "substantial" or "considerable".
L157: Better: "Combining the results of Figure 2 with the geographic ..."
L158: Correct the unit. "meters" is wrong.
L161: "also small" --> "smaller"
L162: Which chart? Are you refering to Fig. 2?
L164-165: I suggest to avoid usage of the term "trend" in the context of this study. If you want to state that in more recent years the annual average sea-ice concentrations were comparably lower then it is fine to do this; but Fig. 2 does not adequately back up any statement given about trends.
L167: What is "the year of short navigation"? Please explain.
L179: "variation in sea ice concentration" --> "variation in the average sea-ice concentration in spring"
L183: "average sea ice concentration" --> "average seasonal sea-ice concentration"
L202: "a" --> "the"
L203: Delete "at the North Pole"
L204: You started refering to "Sections" rather than areas and I suggest to keep this notion.
L209-2011: Check grammar of this sentence please.
L298: "northward" --> ? Into the Arctic Ocean across the North Pole?
L302: "in 2008" --> "in 2000 and 2008"
L321: Novosibirsk is a town somewhere in southern western Siberia, Russia. What does it have to do with your navigation study.
L425-427: "There ... in summer" --> not clear, please re-write.
L423: "but the proportion was reduced" --> not clear, which proportion? reduced compared to what and when?
L429: "Arctic sea ice movement diagrams" --> these are maps, not diagrams. What is shown in these maps are "mean August sea-ice motion drift vectors". Please re-phrase accordingly.
L431: "the Kara Sea was subject to northwest wind control, with winds of about 4 m/s" --> "northwesterly winds of up to 4 m/s prevailed in the Kara Sea"
However: This is not to be seen from Fig. 12. This is something to be mentioned (if at all) in the context of Fig. 10.
L432-435: "This brought ... NEP." --> looooooongish complicated statement --> "Sea ice was advected by the prevailing northwesterly winds from the Arctic Ocean into the northeastern Kara Sea, accumulating to the west of Vilkitsky Strait and thereby blocking the NEP there."
L435: "Wind forces pushed ... Islands" --> "Sea ice was advected from the Laptev Sea to the eastern side of Vilkitsky Strait"
L500: "decreased sharply in 2012" --> reads as if the multiyear ice distribution stayed constant until 2011 and then dropped in 2012. --> "was substantially lower in 2012"
L501: "fragile seasonal ice" --> this "fragile" ice manages to survive summer melt favorably well in large parts of the Arctic Ocean. "seasonal" = first-year. Hence --> "was mostly first-year ice" ...
"which was easy to melt" --> "which usually melts completely earlier than thicker multiyear ice."
L514-516: "The sea ... km2)." --> not relevant --> delete.
L514: "2015 ... on record." Where? On Earth? On the moon? In the Arctic Ocean? Along the NEP? --> Please delete.
