the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 21 Sep 2021
Evolution of the Amundsen Sea Polynya, Antarctica, 2016–2021
Abstract. Polynyas are key sites of ice production during the winter and are important sites of biological activity and carbon sequestration during the summer. The Amundsen Sea Polynya (ASP) is the fourth largest Antarctic polynya, has recorded the highest primary productivity and lies in an embayment of key oceanographic significance. However, knowledge of its dynamics, and of sub-annual variations in its area and ice production, is limited. In this study we primarily utilize Sentinel-1 SAR imagery, sea ice concentration products and climate reanalysis data, along with bathymetric data, to analyze the ASP over the period November 2016–March 2021. Specifically, we analyze (i) qualitative changes in the ASP's characteristics and dynamics, and quantitative changes in (ii) summer polynya area, (iii) winter polynya area and ice production. From our analysis of SAR imagery we find that ice produced by the ASP becomes stuck in the vicinity of the polynya and sometimes flows back into the polynya, contributing to its closure and limiting further ice production. The polynya forms westward off a persistent chain of grounded icebergs that are located at the site of a bathymetric high. Grounded icebergs also influence the outflow of ice and facilitate the formation of a 'secondary polynya' at times. Additionally, unlike some polynyas, ice produced by the polynya flows westward after formation, along the coast and into the neighboring sea sector. During the summer and early winter, broader regional sea ice conditions can play an important role in the polynya. The polynya opens in all summers, but record-low sea ice conditions in 2016/17 cause it to become part of the open ocean. During the winter, an average of 78 % of ice production occurs in April–May and September–October, but large polynya events often associated with high winds can cause ice production throughout the winter. While passive microwave data or daily sea ice concentration products remain key for analyzing variations in polynya area and ice production, we find that the ability to directly observe and qualitatively analyze the polynya at a high temporal and spatial resolution with Sentinel-1 imagery provides important insights about the behavior of the polynya that are not possible with those datasets.
- Preprint
(15249 KB) - Metadata XML
-
Supplement
(5363 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on tc-2021-250', Anonymous Referee #1, 18 Oct 2021
This paper investigates the variability of Amundsen Sea polynya over the last five years (2016-20) using backscatter images from the Sentinel-1 SAR and sea-ice concentration (SIC) from the AMSR2 passive microwave radiometer. Unfortunately, this study has a severe problem in the definition of winter coastal polynya and the application of SAR data to polynya studies, as is described later. These shortcomings are fundamental to this study and are not something that would be remedied through the review process. Therefore, I recommend rejecting the paper.
This study defines coastal (latent heat) polynya as an open water area even in winter (L. 38). Specifically, a polynya (open water/low SIC) area is defined as a region where the SIC measured by the AMSR2 is less than 70%. This threshold is the one used by Parmiggiani (2006), Morelli & Parmiggiani, (2013), and Preußer et al. (2015), cited in this paper (L. 192). This threshold is based on a comparison between the SIC and the area of the coastal polynya. The polynya area was detected from brightness temperature observed by a passive microwave radiometer using the polynya signature simulation method (PSSM) by Markus and Burns (1995). They also compared with SIC, and the threshold of 75% was shown.
The question is whether the low SIC area of <70%, which is more extensive than the footprint size (>5 km) of the passive microwave radiometer, can appear in the Antarctic coastal region in winter when the weather conditions are very cold and windy. Under such conditions, even if an open water fraction does appear, it will soon freeze up and be covered entirely with new frazil (grease) ice, except for very close to shore where divergent ice motion is prominent. In SIC algorithms for passive microwave radiometer data, especially early ones, it is known that the SIC of thin (new) ice is underestimated. This is caused by the polarization ratio (PR) of the brightness temperature used in the algorithm: the PR value of new ice is similar to that of open water compared to that of first-year ice (Cavalieri et al., 1994). In addition, the PR value of landfast ice is similar to that of thin ice (Tamura et al., 2007). Fast ice develops in the ASP region (Fraser et al., 2020; 2021). Therefore, the low SIC region extending offshore during winter in Fig. 10 is speculated to the underestimation due to the influence of fast ice. That is, the SIC of 70 (75)% threshold has no physical meaning. These show that the SIC from passive microwave radiometer is underestimated in thin ice (coastal polynya) and fast ice areas. So what is the point of such an ambiguous parameter-based estimate of the polynya area and the ice production? The estimation of SIC by the newer ASI algorithm used in this study may improve the SIC estimation in thin ice areas to some extent. However, a comparison of SIC using the ASI algorithm in the Ross Ice Shelf polynya and the Mertz Glacier polynya in Antarctica with the PSSM polynya map clearly shows that coastal polynyas are covered by thin ice, not open water, in winter (Kern et al. 2007). Moreover, the SIC is underestimated in these regions. It is reasonable to assume that new (thin) ice with 100% SIC covers the winter coastal polynyas. This is supported by the surface temperature from infrared satellite images, the thermal ice thickness estimated from heat flux calculations using these images (e.g., Tamura et al., 2007), and the SAR images in this study. Based on these facts, the estimation of the winter coastal polynya area using SIC derived from passive microwave radiometer and sea-ice production there does not capture the reality and is meaningless. The heat insulation effect of sea ice decreases rapidly in the case of thin ice. Therefore, the more appropriate approach is to estimate the thermal thin ice thickness from brightness temperature observed by passive microwave radiometer, use it to determine the polynya area, and then estimate ice production from heat flux calculations (e.g., Tamura et al., 2008; Nihashi and Ohshima, 2015; Nihashi et al., 2017). Although these studies' estimates of polynya extent and production are cited (e.g., L. 123-128), the differences in the methods are not discussed at all.
Regarding the SAR data analysis in the first part of this study, the subsection title does indeed say “qualitative analysis” (L. 137; 335), but it is too qualitative. At L. 156-161, the authors describe the relationship between backscatter and ice (ocean) types. This would be generally true, but on what are these based? Quantitatively, which backscatter value corresponds to those surface types? There are no references at all. Furthermore, they did not compare with the SIC used in the second part of this study. At least, a comparison of SAR backscatter with surface temperature and ice thickness from MODIS infrared imagery, as shown in Nakata et al. (2019), will be needed.
In the “qualitative analysis” described in subsection 4.1 starting at L. 335, the SAR images are only shown in Figs. 2 and 3, and the story is developed based mostly on video S1. However, many of the SAR images in this video are missing from the central part of ASP. This leads to difficulties in following the text. It is also extraordinary to say, “Instead, ice formed by the ASP often remains in the ASP study area for months (Video S1, Fig. 3)” (L. 605) from a discussion which is based solely on SAR images. Why can this be said only from temporally sporadic SAR images? In reality, it is reasonable to think that the area of the thin ice area does not change, and the new ice formed there is either thermodynamically growing or advected and deformed at the edge of the polynya. Therefore, dynamical and thermodynamic analyses with an ice motion are essential to justify the author’s claim.
Again, the problems pointed out in this review are fundamental to this research and should not be improved through the review process. Therefore, I recommend rejecting the paper.
References:
- Markus, T. and B. A. Burns (1995), A method to estimate subpixel-scale coastal polynyas with satellite passive microwave data, J. Geophys. Res., 100, C3, 4473-4487.
- Cavalieri D. J. (1994), A microwave technique for mapping thin sea ice, J. Geophys. Res., 99, C6, 12,561-12,572.
- Tamura, T., K. I. Ohshima, T. Markus, D. J. Cavalieri, S. Nihashi, and N. Hirasawa (2007), Estimation of thin ice thickness and detection of fast ice from SSM/I data in the Antarctic Ocean, J. Atmos. Oceanic Technol., 24, 1757–1772.
- Fraser, A. D., R. A. Massom, K. I. Ohshima, S, Willmes, P. J. Kappes, J. Cartwright, and R. Porter-Smith (2020), High-resolution mapping of circum-Antarctic landfast sea ice distribution, 2000–2018, Earth System Science Data, 2987–2999, https://doi.org/10.5194/essd- 2020-99, 2020.
- Fraser, A. D., R. A. Massom, M. S. Handcock, P. R. Reid, K. I. Ohshima, M. N. Raphael, J. Cartwright, A. R. Klekociuk, Z.Wang. and R. Porter-Smith (2021), 18 year record of circum-Antarctic landfast sea ice distribution allows detailed baseline characterisation, reveals trends and variability, The Cryosphere, discussing.
- Kern, S.; G. Spreen, L. Kaleschke, S. de La Rosa, G. Heygster (2007), Polynya Signature Simulation Method polynya area in comparison to AMSR-E 89 GHz sea-ice concentrations in the Ross Sea and off the Ade Ìlie Coast, Antarctica, for 2002–05: first results, Annals of Glaciology, 46, 409-418.
- Nakata, K., K. I. Ohshima, S. Nihashi (2019), Estimation of thin ice thickness and discrimination of ice type from AMSR-E passive microwave data, IEEE TGRS, 57, 1, 263-276.
Citation: https://doi.org/10.5194/tc-2021-250-RC1 -
AC1: 'Reply on RC1', Grant Macdonald, 15 Nov 2021
Dear Reviewer 1,
Thank you for your feedback and the time you took to review our paper. We appreciate your criticisms, which will help us make the paper clearer if the editor gives us the opportunity to submit a revised version. Below we outline our initial responses to the points you made. We will provide more detail regarding specific changes if/when submitting a revised version of the paper.
The question is whether the low SIC area of <70%, which is more extensive than the footprint size (>5 km) of the passive microwave radiometer, can appear in the Antarctic coastal region in winter when the weather conditions are very cold and windy. Under such conditions, even if an open water fraction does appear, it will soon freeze up and be covered entirely with new frazil (grease) ice, except for very close to shore where divergent ice motion is prominent. In SIC algorithms for passive microwave radiometer data, especially early ones, it is known that the SIC of thin (new) ice is underestimated. This is caused by the polarization ratio (PR) of the brightness temperature used in the algorithm: the PR value of new ice is similar to that of open water compared to that of first-year ice (Cavalieri et al., 1994). In addition, the PR value of landfast ice is similar to that of thin ice (Tamura et al., 2007). Fast ice develops in the ASP region (Fraser et al., 2020; 2021). Therefore, the low SIC region extending offshore during winter in Fig. 10 is speculated to the underestimation due to the influence of fast ice. That is, the SIC of 70 (75)% threshold has no physical meaning. These show that the SIC from passive microwave radiometer is underestimated in thin ice (coastal polynya) and fast ice areas. So what is the point of such an ambiguous parameter-based estimate of the polynya area and the ice production? The estimation of SIC by the newer ASI algorithm used in this study may improve the SIC estimation in thin ice areas to some extent. However, a comparison of SIC using the ASI algorithm in the Ross Ice Shelf polynya and the Mertz Glacier polynya in Antarctica with the PSSM polynya map clearly shows that coastal polynyas are covered by thin ice, not open water, in winter (Kern et al. 2007). Moreover, the SIC is underestimated in these regions. It is reasonable to assume that new (thin) ice with 100% SIC covers the winter coastal polynyas. This is supported by the surface temperature from infrared satellite images, the thermal ice thickness estimated from heat flux calculations using these images (e.g., Tamura et al., 2007), and the SAR images in this study. Based on these facts, the estimation of the winter coastal polynya area using SIC derived from passive microwave radiometer and sea-ice production there does not capture the reality and is meaningless. The heat insulation effect of sea ice decreases rapidly in the case of thin ice. Therefore, the more appropriate approach is to estimate the thermal thin ice thickness from brightness temperature observed by passive microwave radiometer, use it to determine the polynya area, and then estimate ice production from heat flux calculations (e.g., Tamura et al., 2008; Nihashi and Ohshima, 2015; Nihashi et al., 2017). Although these studies' estimates of polynya extent and production are cited (e.g., L. 123-128), the differences in the methods are not discussed at all.
This comment is helpful because it shows we have not been clear enough with our definition of ‘open’ polynya area in winter. We will more clearly define this in a revised paper. We agree that open ocean areas in the polynya during winter are quickly covered by thin ice formation. Note that we define the transition between ‘winter’ and ‘summer’ periods for the polynya based on whether the ‘open’ areas seen in the SAR are exhibiting ice formation or not. By ‘open’ polynya area in winter we mean where existing ice has advected away, leaving an area that is then the site of active new ice production (new frazil ice).
By comparing our SAR imagery with the ASI SIC data we found that the SIC data did a good job of representing the ‘open’ area during winter. For example the 21-23 September 2020 polynya event shown in Fig. 2 and video S1 can be clearly seen in Video S2 (which displays SIC data). Variations in the SAR can be seen captured by the SIC data throughout Video S2. It is clear from this comparison that the ‘open’ areas we find through the winter are not variations in fast ice coverage – opening and closing occurs in a manner consistent with polynya events, and is visible in the SAR when matching images are available. In a revised version of the paper we will also include a supplemental figure directly comparing SAR images with SIC data using a 70% threshold.
We also note that ASI’s SIC data has been used in this way before in published work (Cheng et al., 2017; 2019) and we note that our mean annual estimate of ice production falls within the range of estimates by Tamura et al. (2008; 2016) and Nihashi et al. (2017).
In a revised version we will clarify the method used in the Tamura et al. and Nihashi et al. studies.
“Regarding the SAR data analysis in the first part of this study, the subsection title does indeed say “qualitative analysis” (L. 137; 335), but it is too qualitative. At L. 156-161, the authors describe the relationship between backscatter and ice (ocean) types. This would be generally true, but on what are these based? Quantitatively, which backscatter value corresponds to those surface types? There are no references at all. Furthermore, they did not compare with the SIC used in the second part of this study. At least, a comparison of SAR backscatter with surface temperature and ice thickness from MODIS infrared imagery, as shown in Nakata et al. (2019), will be needed.”
We agree that more information should be provided on this. The differences between the ice types are not only quantitative differences in backscatter but also textural. However we will include backscatter ranges, and an additional supplementary figure comparing the different ice types we identify that are important for our analysis. As mentioned above we will also include a supplementary figure comparing SAR imagery with SIC data (represented spatially).
We do not think that a comparison of SAR imagery with MODIS temperature and ice thickness data is necessary for qualitatively interpreting the imagery presented in this study, or would add significant information in the context of this study. MODIS data is not necessary to identify ‘open’ areas of ice production or polynya events, and would not aid in interpretation of other polynya dynamics (e.g. backfilling, obstruction).
“In the “qualitative analysis” described in subsection 4.1 starting at L. 335, the SAR images are only shown in Figs. 2 and 3, and the story is developed based mostly on video S1. However, many of the SAR images in this video are missing from the central part of ASP. This leads to difficulties in following the text. It is also extraordinary to say, “Instead, ice formed by the ASP often remains in the ASP study area for months (Video S1, Fig. 3)” (L. 605) from a discussion which is based solely on SAR images. Why can this be said only from temporally sporadic SAR images? In reality, it is reasonable to think that the area of the thin ice area does not change, and the new ice formed there is either thermodynamically growing or advected and deformed at the edge of the polynya. Therefore, dynamical and thermodynamic analyses with an ice motion are essential to justify the author’s claim.”
We make the claim “ice formed by the ASP often remains in the ASP study area for months” based on repeated and focused analysis of the SAR imagery. By visually analyzing the video and tracking changes in the ice it is evident that this is the case. Although temporal and spatial gaps in the video can be limiting, despite these gaps, by viewing the animated images it is possible to observe the flow and dynamics of the ice, including particular floes. While a dynamical and thermodynamical analysis of ice motion in the area would be valuable, we do not think it is within the scope of this study’s goal to analyze the evolution of the polynya, or necessary to make this claim.
Best regards,
Dr Grant Macdonald, on behalf of all authors.
Citation: https://doi.org/10.5194/tc-2021-250-AC1
-
RC2: 'Referee comment on tc-2021-250', Anonymous Referee #2, 02 Nov 2021
GENERAL COMMENTS
I think the manuscript's topic fits nicely within the scope of the journal. As far as I know, this manuscript is unique (i.e. original) in its detailed description of Amundsen Polynya, providing far more information than earlier publications. The study is also scientifically significant given the regional importance of the Amundsen Polynya, either in terms of sea ice production (4th in Antarctica) or for its role in phytoplankton production and atmospheric CO2 uptake. The methods and analyses were generally sound except in one particular analysis for which I proposed verifications and improvements (see "Specific comments: Major"). I thought the manuscript was structured in a straightforward/predictable manner and that the writing (and choice of references) was generally appropriate. Overall, I believe the manuscript has the potential to become a valuable addition to "The Cryosphere".
RECOMMENDATION
Unless my co-reviewers identify major flaws that I've missed, I think the manuscript will be acceptable for publication *after* moderate revisions are made (see specific comments below).
SPECIFIC COMMENTS: MAJOR
(1) Lines 521-522: "Daily mean wind speed at the polynya site and polynya area during the winter period has a weak but significant positive correlation"
Are you sure that these are the two variables (wind speed versus polynya area) you should be comparing together?
I would have tried to compare the winds against the *change* in polynya area, i.e. d(area)/dt.
A gust of winds would presumably correlate with a change in area ( d(area)/dt ), not the actual area.Another related comment is that you are using wind *speed* and ignoring the wind's direction.
Wouldn't the wind direction matter, e.g. in distinguishing between an opening (area increases) or a closing (area decreases) of the polynya?(2) Lines 529-531: "The mean wind direction throughout the ASP study area is approximately southerly. While this direction corresponds to the direction in which the polynya sometimes forms northward off the Dotson Ice Shelf, it does not correspond to the more typical westward formation off the iceberg chain."
Could you please expand on this in your "response to the reviewers", and demonstrate that the "southerly mean wind direction" isn't a plotting artefact or an error in the postprocessing of ERA5's winds? Although I haven't specifically worked with years 2016-2020, the area just north of the Getz and Dotson ice shelves usually shows mean winds blowing from the southeast or from east-southeast. In contrast, Figure 9 shows winds blowing from the south or from south-southeast.
A few thoughts and suggestions:
---A wind blowing from the south or from the south-southeast would imply that the `v' wind component is larger (in absolute terms) than the `u' wind component. Can you go back to the original ERA5 results (meaning before doing any sort of processing) and confirm that this is what you are truly seeing in ERA5's u,v wind components? (We are particularly interested in the region just north of Getz and Dotson ice shelf, i.e. the ASP).
---If you select one single location (latitude = something, longitude = something) where ERA5 winds are defined, please verify that the 'u' value and the the 'v' value (in the original ERA5 files) are quantitatively consistent with the arrow's direction shown at that particular location. If they don't quantitatively match (using the basic trigonometric relations), then there is an error in the plotting or the postprocessing.
---Since wind direction is important in Figure 9, shouldn't we add longitudes and latitudes to orient the reader who is not particularly familiar with the Amundsen Sea?
---Please clarify whether the statement about the "southerly" mean wind direction (Line 529) is specific to a time of the year, and if so, what time of the year exactly? This is a confusing thing, considering that Figure 8 is representative of April-October, Figure 9 is representative of November-December, and we just don't know about Line 529.SPECIFIC COMMENTS: MODERATE
(3) Lines 90-91: "...such as Antarctic bottom water formation and global thermohaline circulation..."
The Amundsen Sea doesn't produce Antarctic bottom water. You must be confusing the Amundsen with another location in Antarctica (maybe the Ross Sea?). You have to remove this passage because it simply doesn't apply to the Amundsen Sea.
(4) Lines 301-302: "Additionally, the total SIC for each day was calculated by calculating the sum of all percentage SIC values in the study region. These total SIC values should only be considered useful for analyzing relative changes in SIC"
By defining "total SIC" in this particular way, you are making it unecessarily difficult to compare your results with past/future studies, because the value obtained will be intimately tied to your grid resolution (the latter being an arbitrary choice).
For example, assume that SIC=100 over a certain geographical area, and that this area is covered by 4 grid points (with the resolution you've chosen). The "total SIC" for this area will be 400. Then, in two years from now, another researcher does a similar analysis and selects a resolution that's twice finer than the one you chose. The same geographical area is now covered by 16 grid points, and the "total SIC" is suddenly 1600.
I'd suggest normalizing the "total SIC" by the total number of grid points within the study region. This way, the number coming out of your analysis isn't so dependent on the arbitrary choice of resolution. Another benefit is that the numbers you'll obtain will be much smaller and less cumbersome. Right now, lines 581-583 are discussing numbers having 7 digits, and that's a bit awkward.
(5) Line 311: "Daily wind speed and direction at the site of the polynya, ERA5's hourly 'u' and 'v' wind products were processed for a region..."
The sentence makes it sound like these variables are peculiar to ERA5, but they really aren't. Please use the formal terminology to present what u,v are:
"Daily zonal (u) and meridional (v) components of the winds at a height of 10m were obtained from ERA5 and processed for a region..."
I also note a confusion between "monthly" (Line 308), "Daily" (Line 311), and "Hourly" (Line 313). This is terribly confusing for the reader. Please state exactly what you've downloaded from ERA5 (monthly? daily? hourly? It has to be *one* of the three), and then correct these sentences as necessary.
SPECIFIC COMMENTS: MINOR
(6) Line 50: Why is "Carbon Dioxide" capitalized? This looks unusual.
(7) Lines 107-108: "Westward coastal currents prevail in the area (St-Laurent et al. 2019)..."
"St-Laurent et al." is only a modeling study. I think it would be good to also cite an observational study with actual measurements of the coastal current: Kim et al. 2016, Estuarine, Coastal and Shelf Science, https://doi.org/10.1016/j.ecss.2016.08.004
(Same comment for Line 674 of the manuscript.)(8) Line 124: The acronym EOS isn't defined. Please define what EOS stands for, or remove the acronym.
(9) Figure 1a: To give the reader some context, please add to Fig.1a the climatological northward extent of the sea ice cover in Summer and in Winter. Such climatologies are available at, e.g., https://nsidc.org/data/NSIDC-0192/versions/3
Having this information will help the reader understand how you chose the extent of the red box. Without this information, the geographical extent of the red box appears arbitrary.(10) Line 224: "Following Cheng et al. (2017) the daily net heat flux, Q (in W/m2), of a pixel was estimated by..."
Shouldn't it be: "of a *ice-free* pixel"?
(11) Line 235: L_o = epsilon sigma T_0^4 (Equation 2)
Why use T_0 (the freezing point of seawater) in Equation 2? Since Equation 2 is the black body radiation of the ocean surface, then Equation 2 should use T_s (the temperature of the water surface), not T_0.
It's acceptable to assume, later on, that T_s is approximately equal to the freezing temperature, but Equation 2 should nevertheless be written as a function of T_s (not T_0) in order to make physical sense.
Similarly, line 241 should read:
T_s ~= T_0 = 273.15 - 0.0137...
rather than:
T_0 = T_s = 273.15 - 0.0137...(12) Lines 256-259: The symbol for pressure is "Pa", not "pa". Please correct throughout the paragraph.
(13) Line 274: Acronym GDAL isn't defined. Please define what GDAL stands for, or remove the acronym.
(14) Line 282: The letter "i" in "rho_i" should be a subscript (compare with Line 280 where it is correctly typed).
(15) Lines 289-290: "Caution should be used when interpreting the absolute numbers produced by the ice production model, particularly because the input data is modeled climate data..."
Are you referring to ERA5 as "modeled climate data"? ERA5 is an atmospheric reanalysis, which is very different from "modeled climate data". For one thing, an atmospheric reanalysis assimilates historical measurements, while a "climate model" doesn't.
(16) Lines 318-319: "wind direction was plotted using the 'matplotlib' function 'quiver'"
The results shouldn't depend on the choice of the graphics software. If this is relevant to the results, then explain in what ways it is relevant. If it's not relevant, then remove the passage.
(17) Figure 4,5,6: Can we make the curvers a bit thicker, so that their color is more apparent and makes it easier to distinguish one curve from another? Also, would it be possible to make these figure labels less pixelated, or larger so that the pixelation doesn't show as much?
(18) Line 575: "SIC for the broader ASP region on two days in 2017"
Please clarify why you picked these two specific days, and what they represent? Presumably the top figure represents some kind of minimum? And what about the bottom figure?
(19) Line 676: "(Koo et al., in review)."
You aren't supposed to cite studies that haven't been accepted for publication. Most journals will systematically remove such references at the copy-editing stage.
(20) There is something wrong in the caption of Figure S1:
"The figure covers background figure and area is the same as Fig 1b.".
Please correct as needed.Citation: https://doi.org/10.5194/tc-2021-250-RC2 -
AC2: 'Reply on RC2', Grant Macdonald, 16 Nov 2021
Dear Reviewer 2,
Thank you for your time, comments, and feedback that will undoubtedly improve this manuscript if we are invited to submit a revised version. Here we provide an initial response to your comments. If invited to submit a revised version, we will provide more details regarding changes made related to your comments.
SPECIFIC COMMENTS: MAJOR
(1) Lines 521-522: "Daily mean wind speed at the polynya site and polynya area during the winter period has a weak but significant positive correlation"
Are you sure that these are the two variables (wind speed versus polynya area) you should be comparing together?
I would have tried to compare the winds against the *change* in polynya area, i.e. d(area)/dt.
A gust of winds would presumably correlate with a change in area ( d(area)/dt ), not the actual area.
Another related comment is that you are using wind *speed* and ignoring the wind's direction.
Wouldn't the wind direction matter, e.g. in distinguishing between an opening (area increases) or a closing (area decreases) of the polynya?
We agree that change in polynya area would be a better variable to compare, we will amend this. We think it is valuable to include a comparison of changes in the polynya with just wind speed, as done in other studies, but agree wind direction is also important, so we will make reference to the daily wind direction shown in Video S3 when discussing changes in polynya area and its relationship with wind.
(2) Lines 529-531: "The mean wind direction throughout the ASP study area is approximately southerly. While this direction corresponds to the direction in which the polynya sometimes forms northward off the Dotson Ice Shelf, it does not correspond to the more typical westward formation off the iceberg chain."
Could you please expand on this in your "response to the reviewers", and demonstrate that the "southerly mean wind direction" isn't a plotting artefact or an error in the postprocessing of ERA5's winds? Although I haven't specifically worked with years 2016-2020, the area just north of the Getz and Dotson ice shelves usually shows mean winds blowing from the southeast or from east-southeast. In contrast, Figure 9 shows winds blowing from the south or from south-southeast.
A few thoughts and suggestions:
---A wind blowing from the south or from the south-southeast would imply that the `v' wind component is larger (in absolute terms) than the `u' wind component. Can you go back to the original ERA5 results (meaning before doing any sort of processing) and confirm that this is what you are truly seeing in ERA5's u,v wind components? (We are particularly interested in the region just north of Getz and Dotson ice shelf, i.e. the ASP).
We checked this and it seems to be correct, but we will check more thoroughly before submitting a revision. It would be more precise to say that we calculate the mean wind to come from the SSE in front of the Dotson, and SE to ESE in front of the Getz – we will amend the text to reflect this.
---If you select one single location (latitude = something, longitude = something) where ERA5 winds are defined, please verify that the 'u' value and the the 'v' value (in the original ERA5 files) are quantitatively consistent with the arrow's direction shown at that particular location. If they don't quantitatively match (using the basic trigonometric relations), then there is an error in the plotting or the postprocessing.
As above, it seems to be correct but we will more thoroughly check before submitting a revision.
---Since wind direction is important in Figure 9, shouldn't we add longitudes and latitudes to orient the reader who is not particularly familiar with the Amundsen Sea?
We will do this.
---Please clarify whether the statement about the "southerly" mean wind direction (Line 529) is specific to a time of the year, and if so, what time of the year exactly? This is a confusing thing, considering that Figure 8 is representative of April-October, Figure 9 is representative of November-December, and we just don't know about Line 529.”
The reference to ‘southerly’ on line 529 refers to the year-round mean. We will rephrase for clarity. Figure 9 is representative of the year-round mean, not November-December. Note it is November *2016* to December *2020*. However, we understand how this could confuse and we will adjust the caption to make clear it is the year-round mean over this whole period.
SPECIFIC COMMENTS: MODERATE
(3) Lines 90-91: "...such as Antarctic bottom water formation and global thermohaline circulation..."
The Amundsen Sea doesn't produce Antarctic bottom water. You must be confusing the Amundsen with another location in Antarctica (maybe the Ross Sea?). You have to remove this passage because it simply doesn't apply to the Amundsen Sea.
Thank you, we will remove this.
(4) Lines 301-302: "Additionally, the total SIC for each day was calculated by calculating the sum of all percentage SIC values in the study region. These total SIC values should only be considered useful for analyzing relative changes in SIC"
By defining "total SIC" in this particular way, you are making it unecessarily difficult to compare your results with past/future studies, because the value obtained will be intimately tied to your grid resolution (the latter being an arbitrary choice).
For example, assume that SIC=100 over a certain geographical area, and that this area is covered by 4 grid points (with the resolution you've chosen). The "total SIC" for this area will be 400. Then, in two years from now, another researcher does a similar analysis and selects a resolution that's twice finer than the one you chose. The same geographical area is now covered by 16 grid points, and the "total SIC" is suddenly 1600.
I'd suggest normalizing the "total SIC" by the total number of grid points within the study region. This way, the number coming out of your analysis isn't so dependent on the arbitrary choice of resolution. Another benefit is that the numbers you'll obtain will be much smaller and less cumbersome. Right now, lines 581-583 are discussing numbers having 7 digits, and that's a bit awkward.
Thank you, we agree that this was a poor choice and will instead normalize as suggested.
(5) Line 311: "Daily wind speed and direction at the site of the polynya, ERA5's hourly 'u' and 'v' wind products were processed for a region..."
The sentence makes it sound like these variables are peculiar to ERA5, but they really aren't. Please use the formal terminology to present what u,v are:
"Daily zonal (u) and meridional (v) components of the winds at a height of 10m were obtained from ERA5 and processed for a region..."
We will amend as suggested.
I also note a confusion between "monthly" (Line 308), "Daily" (Line 311), and "Hourly" (Line 313). This is terribly confusing for the reader. Please state exactly what you've downloaded from ERA5 (monthly? daily? hourly? It has to be *one* of the three), and then correct these sentences as necessary.
Daily wind speed was calculated from hourly ERA5 data, and we used the ERA5’s monthly product to analyze the overall mean over the study period (i.e. Fig. 9). We see how this is confusing and will rephrase for better clarity, or simply use the same hourly product to calculate the study period mean in order to avoid the unnecessarily confusing explanation.
SPECIFIC COMMENTS: MINOR
(6) Line 50: Why is "Carbon Dioxide" capitalized? This looks unusual.
We will change this.
(7) Lines 107-108: "Westward coastal currents prevail in the area (St-Laurent et al. 2019)..."
"St-Laurent et al." is only a modeling study. I think it would be good to also cite an observational study with actual measurements of the coastal current: Kim et al. 2016, Estuarine, Coastal and Shelf Science, https://doi.org/10.1016/j.ecss.2016.08.004
(Same comment for Line 674 of the manuscript.)
Thank you, we will do this.
(8) Line 124: The acronym EOS isn't defined. Please define what EOS stands for, or remove the acronym.
We will do this.
(9) Figure 1a: To give the reader some context, please add to Fig.1a the climatological northward extent of the sea ice cover in Summer and in Winter. Such climatologies are available at, e.g., https://nsidc.org/data/NSIDC-0192/versions/3
Having this information will help the reader understand how you chose the extent of the red box. Without this information, the geographical extent of the red box appears arbitrary.
Thank you, we will do this.
(10) Line 224: "Following Cheng et al. (2017) the daily net heat flux, Q (in W/m2), of a pixel was estimated by..."
Shouldn't it be: "of a *ice-free* pixel"?
Yes, it should be for an ‘open polynya’ pixel – we will amend this.
(11) Line 235: L_o = epsilon sigma T_0^4 (Equation 2)
Why use T_0 (the freezing point of seawater) in Equation 2? Since Equation 2 is the black body radiation of the ocean surface, then Equation 2 should use T_s (the temperature of the water surface), not T_0.
It's acceptable to assume, later on, that T_s is approximately equal to the freezing temperature, but Equation 2 should nevertheless be written as a function of T_s (not T_0) in order to make physical sense.
Similarly, line 241 should read:
T_s ~= T_0 = 273.15 - 0.0137...
rather than:
T_0 = T_s = 273.15 - 0.0137...
Thank you, we will make these changes.
(12) Lines 256-259: The symbol for pressure is "Pa", not "pa". Please correct throughout the paragraph.
Thank you for spotting this, we will amend it and double check for other similar errors.
(13) Line 274: Acronym GDAL isn't defined. Please define what GDAL stands for, or remove the acronym.
We will define the acronym (Geospatial Data Abstraction Library).
(14) Line 282: The letter "i" in "rho_i" should be a subscript (compare with Line 280 where it is correctly typed).
Thank you for spotting this, we will amend it.
(15) Lines 289-290: "Caution should be used when interpreting the absolute numbers produced by the ice production model, particularly because the input data is modeled climate data..."
Are you referring to ERA5 as "modeled climate data"? ERA5 is an atmospheric reanalysis, which is very different from "modeled climate data". For one thing, an atmospheric reanalysis assimilates historical measurements, while a "climate model" doesn't.
This does refer to ERA5. We will amend this incorrect wording.
(16) Lines 318-319: "wind direction was plotted using the 'matplotlib' function 'quiver'"
The results shouldn't depend on the choice of the graphics software. If this is relevant to the results, then explain in what ways it is relevant. If it's not relevant, then remove the passage.
We will remove the passage.
(17) Figure 4,5,6: Can we make the curvers a bit thicker, so that their color is more apparent and makes it easier to distinguish one curve from another? Also, would it be possible to make these figure labels less pixelated, or larger so that the pixelation doesn't show as much?
We will make the lines thicker and the labels larger.
(18) Line 575: "SIC for the broader ASP region on two days in 2017"
Please clarify why you picked these two specific days, and what they represent? Presumably the top figure represents some kind of minimum? And what about the bottom figure?
We will explain in the caption that
- panel (a) is presented to show an example of when the polynya had no ice boundary to the W/NW due to exceptionally low SIC in the region during the 2016/17 summer
- panel (b) is presented to show an example of the low SIC in the region, and narrow icepack bounding the polynya in the subsequent early winter.
(19) Line 676: "(Koo et al., in review)."
You aren't supposed to cite studies that haven't been accepted for publication. Most journals will systematically remove such references at the copy-editing stage.
We acknowledge this mistake, however the paper in question has now been published so we will retain it in the text.
(20) There is something wrong in the caption of Figure S1:
"The figure covers background figure and area is the same as Fig 1b.".
Please correct as needed.
Thank you, we will correct this. It should read "The background image and area is the same as in Fig 1b.".
Best regards,
Dr Grant Macdonald, on behalf of all authors.
Citation: https://doi.org/10.5194/tc-2021-250-AC2
-
AC2: 'Reply on RC2', Grant Macdonald, 16 Nov 2021
-
RC3: 'Comment on tc-2021-250', Anonymous Referee #3, 03 Dec 2021
I was asked by the editor to provide a 3rd review of the manuscript “Evolution of the Amundsen Sea Polynya, Antarctica, 2016-2021” by Macdonald et al. submitted to the Cryosphere 2021. Two reviews were available with one recommendation to reject and the other as a major revision.
The authors aim to investigate the evolution of the Amundsen Sea Polynya, Antarctica by the means of SAR images, AMSR2 ice concentration and ERA 5 reanalysis data. The study area is motivated by the highest primary productivity and key oceanographic significance. The time period 2016-2021 is very limited by the availability of Sentinel-1 SAR images. Specifically, the aims are to analyse qualitative changes in the characteristics, the summer polynya area and the winter polynya area and ice production.
The authors do not precisely identify the novel aspects of their investigation but I think it is mainly the use of Sentinel-1 data. This is in principle a good and interesting approach to investigate the polynya dynamics in the special Amundsen Sea environment. However, the limitations due to the sparse temporal sampling are not adequately discussed, and the knowledge gained is not presented in a comprehensible manner. Unfortunately, the quality of the supplementary video is only poor and does not support the claims. If the visual SAR video/image analysis were carried out by showing more example results, I would rate the present manuscript much better.
The scientific quality of the study is only poor. I agree with the fundamental concerns of reviewer #1. In addition to the concerns about neglecting the effect of thin ice thickness there is an inconsistency with the applied data sets. The high resolution AMSR2 ice concentration used for the manuscript is different from the low resolution OSI-SAF/OSTIA ice concentration used in the ERA5 reanalysis. This means that there are likely large inconsistencies for the calculation of heat fluxes when the low resolution data do not resolve the open water polynya. Unfortunately, there is no error assessment or sensitivity analysis.
I do not fully understand the study logic. I would have expected a comparison between the passive microwave and SAR derived polynya areas. The scalar wind speed is of very limited use to describe the polynya which can be seen in Fig. 8. The surprisingly low correlation indicates that the wind speed explains only very little of the polynya area variability. A more insightful approach would have been the usage of wind direction like outlined in Haarpaintner et al. (2001) who compared SAR-derived polynya width/area with model based results.
The literature regarding the impact of the Amundsen polynya is not well balanced, i.e. the impact for carbon sequestration is emphasized but this seems to be controversial (Lee et al., 2017). One interesting question is how to define a polynya during Summer, in particular when there is only surrounding open water.
In summary, I concur with review #1 and suggest to reject the manuscript but encourage the authors to resubmit their work after a fundamental revision. The aspect of the SAR image analysis would be very interesting when compared with the passive microwave data. The gain in knowledge has to be identified and the main hypotheses have to be worked out and supported by evidence.
Lee et al. (2017) https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2017GL074646
Haarpaintner et al. (2001) https://agupubs.onlinelibrary.wiley.com/doi/pdf/10.1029/1999JC000133
Citation: https://doi.org/10.5194/tc-2021-250-RC3
Status: closed
-
RC1: 'Comment on tc-2021-250', Anonymous Referee #1, 18 Oct 2021
This paper investigates the variability of Amundsen Sea polynya over the last five years (2016-20) using backscatter images from the Sentinel-1 SAR and sea-ice concentration (SIC) from the AMSR2 passive microwave radiometer. Unfortunately, this study has a severe problem in the definition of winter coastal polynya and the application of SAR data to polynya studies, as is described later. These shortcomings are fundamental to this study and are not something that would be remedied through the review process. Therefore, I recommend rejecting the paper.
This study defines coastal (latent heat) polynya as an open water area even in winter (L. 38). Specifically, a polynya (open water/low SIC) area is defined as a region where the SIC measured by the AMSR2 is less than 70%. This threshold is the one used by Parmiggiani (2006), Morelli & Parmiggiani, (2013), and Preußer et al. (2015), cited in this paper (L. 192). This threshold is based on a comparison between the SIC and the area of the coastal polynya. The polynya area was detected from brightness temperature observed by a passive microwave radiometer using the polynya signature simulation method (PSSM) by Markus and Burns (1995). They also compared with SIC, and the threshold of 75% was shown.
The question is whether the low SIC area of <70%, which is more extensive than the footprint size (>5 km) of the passive microwave radiometer, can appear in the Antarctic coastal region in winter when the weather conditions are very cold and windy. Under such conditions, even if an open water fraction does appear, it will soon freeze up and be covered entirely with new frazil (grease) ice, except for very close to shore where divergent ice motion is prominent. In SIC algorithms for passive microwave radiometer data, especially early ones, it is known that the SIC of thin (new) ice is underestimated. This is caused by the polarization ratio (PR) of the brightness temperature used in the algorithm: the PR value of new ice is similar to that of open water compared to that of first-year ice (Cavalieri et al., 1994). In addition, the PR value of landfast ice is similar to that of thin ice (Tamura et al., 2007). Fast ice develops in the ASP region (Fraser et al., 2020; 2021). Therefore, the low SIC region extending offshore during winter in Fig. 10 is speculated to the underestimation due to the influence of fast ice. That is, the SIC of 70 (75)% threshold has no physical meaning. These show that the SIC from passive microwave radiometer is underestimated in thin ice (coastal polynya) and fast ice areas. So what is the point of such an ambiguous parameter-based estimate of the polynya area and the ice production? The estimation of SIC by the newer ASI algorithm used in this study may improve the SIC estimation in thin ice areas to some extent. However, a comparison of SIC using the ASI algorithm in the Ross Ice Shelf polynya and the Mertz Glacier polynya in Antarctica with the PSSM polynya map clearly shows that coastal polynyas are covered by thin ice, not open water, in winter (Kern et al. 2007). Moreover, the SIC is underestimated in these regions. It is reasonable to assume that new (thin) ice with 100% SIC covers the winter coastal polynyas. This is supported by the surface temperature from infrared satellite images, the thermal ice thickness estimated from heat flux calculations using these images (e.g., Tamura et al., 2007), and the SAR images in this study. Based on these facts, the estimation of the winter coastal polynya area using SIC derived from passive microwave radiometer and sea-ice production there does not capture the reality and is meaningless. The heat insulation effect of sea ice decreases rapidly in the case of thin ice. Therefore, the more appropriate approach is to estimate the thermal thin ice thickness from brightness temperature observed by passive microwave radiometer, use it to determine the polynya area, and then estimate ice production from heat flux calculations (e.g., Tamura et al., 2008; Nihashi and Ohshima, 2015; Nihashi et al., 2017). Although these studies' estimates of polynya extent and production are cited (e.g., L. 123-128), the differences in the methods are not discussed at all.
Regarding the SAR data analysis in the first part of this study, the subsection title does indeed say “qualitative analysis” (L. 137; 335), but it is too qualitative. At L. 156-161, the authors describe the relationship between backscatter and ice (ocean) types. This would be generally true, but on what are these based? Quantitatively, which backscatter value corresponds to those surface types? There are no references at all. Furthermore, they did not compare with the SIC used in the second part of this study. At least, a comparison of SAR backscatter with surface temperature and ice thickness from MODIS infrared imagery, as shown in Nakata et al. (2019), will be needed.
In the “qualitative analysis” described in subsection 4.1 starting at L. 335, the SAR images are only shown in Figs. 2 and 3, and the story is developed based mostly on video S1. However, many of the SAR images in this video are missing from the central part of ASP. This leads to difficulties in following the text. It is also extraordinary to say, “Instead, ice formed by the ASP often remains in the ASP study area for months (Video S1, Fig. 3)” (L. 605) from a discussion which is based solely on SAR images. Why can this be said only from temporally sporadic SAR images? In reality, it is reasonable to think that the area of the thin ice area does not change, and the new ice formed there is either thermodynamically growing or advected and deformed at the edge of the polynya. Therefore, dynamical and thermodynamic analyses with an ice motion are essential to justify the author’s claim.
Again, the problems pointed out in this review are fundamental to this research and should not be improved through the review process. Therefore, I recommend rejecting the paper.
References:
- Markus, T. and B. A. Burns (1995), A method to estimate subpixel-scale coastal polynyas with satellite passive microwave data, J. Geophys. Res., 100, C3, 4473-4487.
- Cavalieri D. J. (1994), A microwave technique for mapping thin sea ice, J. Geophys. Res., 99, C6, 12,561-12,572.
- Tamura, T., K. I. Ohshima, T. Markus, D. J. Cavalieri, S. Nihashi, and N. Hirasawa (2007), Estimation of thin ice thickness and detection of fast ice from SSM/I data in the Antarctic Ocean, J. Atmos. Oceanic Technol., 24, 1757–1772.
- Fraser, A. D., R. A. Massom, K. I. Ohshima, S, Willmes, P. J. Kappes, J. Cartwright, and R. Porter-Smith (2020), High-resolution mapping of circum-Antarctic landfast sea ice distribution, 2000–2018, Earth System Science Data, 2987–2999, https://doi.org/10.5194/essd- 2020-99, 2020.
- Fraser, A. D., R. A. Massom, M. S. Handcock, P. R. Reid, K. I. Ohshima, M. N. Raphael, J. Cartwright, A. R. Klekociuk, Z.Wang. and R. Porter-Smith (2021), 18 year record of circum-Antarctic landfast sea ice distribution allows detailed baseline characterisation, reveals trends and variability, The Cryosphere, discussing.
- Kern, S.; G. Spreen, L. Kaleschke, S. de La Rosa, G. Heygster (2007), Polynya Signature Simulation Method polynya area in comparison to AMSR-E 89 GHz sea-ice concentrations in the Ross Sea and off the Ade Ìlie Coast, Antarctica, for 2002–05: first results, Annals of Glaciology, 46, 409-418.
- Nakata, K., K. I. Ohshima, S. Nihashi (2019), Estimation of thin ice thickness and discrimination of ice type from AMSR-E passive microwave data, IEEE TGRS, 57, 1, 263-276.
Citation: https://doi.org/10.5194/tc-2021-250-RC1 -
AC1: 'Reply on RC1', Grant Macdonald, 15 Nov 2021
Dear Reviewer 1,
Thank you for your feedback and the time you took to review our paper. We appreciate your criticisms, which will help us make the paper clearer if the editor gives us the opportunity to submit a revised version. Below we outline our initial responses to the points you made. We will provide more detail regarding specific changes if/when submitting a revised version of the paper.
The question is whether the low SIC area of <70%, which is more extensive than the footprint size (>5 km) of the passive microwave radiometer, can appear in the Antarctic coastal region in winter when the weather conditions are very cold and windy. Under such conditions, even if an open water fraction does appear, it will soon freeze up and be covered entirely with new frazil (grease) ice, except for very close to shore where divergent ice motion is prominent. In SIC algorithms for passive microwave radiometer data, especially early ones, it is known that the SIC of thin (new) ice is underestimated. This is caused by the polarization ratio (PR) of the brightness temperature used in the algorithm: the PR value of new ice is similar to that of open water compared to that of first-year ice (Cavalieri et al., 1994). In addition, the PR value of landfast ice is similar to that of thin ice (Tamura et al., 2007). Fast ice develops in the ASP region (Fraser et al., 2020; 2021). Therefore, the low SIC region extending offshore during winter in Fig. 10 is speculated to the underestimation due to the influence of fast ice. That is, the SIC of 70 (75)% threshold has no physical meaning. These show that the SIC from passive microwave radiometer is underestimated in thin ice (coastal polynya) and fast ice areas. So what is the point of such an ambiguous parameter-based estimate of the polynya area and the ice production? The estimation of SIC by the newer ASI algorithm used in this study may improve the SIC estimation in thin ice areas to some extent. However, a comparison of SIC using the ASI algorithm in the Ross Ice Shelf polynya and the Mertz Glacier polynya in Antarctica with the PSSM polynya map clearly shows that coastal polynyas are covered by thin ice, not open water, in winter (Kern et al. 2007). Moreover, the SIC is underestimated in these regions. It is reasonable to assume that new (thin) ice with 100% SIC covers the winter coastal polynyas. This is supported by the surface temperature from infrared satellite images, the thermal ice thickness estimated from heat flux calculations using these images (e.g., Tamura et al., 2007), and the SAR images in this study. Based on these facts, the estimation of the winter coastal polynya area using SIC derived from passive microwave radiometer and sea-ice production there does not capture the reality and is meaningless. The heat insulation effect of sea ice decreases rapidly in the case of thin ice. Therefore, the more appropriate approach is to estimate the thermal thin ice thickness from brightness temperature observed by passive microwave radiometer, use it to determine the polynya area, and then estimate ice production from heat flux calculations (e.g., Tamura et al., 2008; Nihashi and Ohshima, 2015; Nihashi et al., 2017). Although these studies' estimates of polynya extent and production are cited (e.g., L. 123-128), the differences in the methods are not discussed at all.
This comment is helpful because it shows we have not been clear enough with our definition of ‘open’ polynya area in winter. We will more clearly define this in a revised paper. We agree that open ocean areas in the polynya during winter are quickly covered by thin ice formation. Note that we define the transition between ‘winter’ and ‘summer’ periods for the polynya based on whether the ‘open’ areas seen in the SAR are exhibiting ice formation or not. By ‘open’ polynya area in winter we mean where existing ice has advected away, leaving an area that is then the site of active new ice production (new frazil ice).
By comparing our SAR imagery with the ASI SIC data we found that the SIC data did a good job of representing the ‘open’ area during winter. For example the 21-23 September 2020 polynya event shown in Fig. 2 and video S1 can be clearly seen in Video S2 (which displays SIC data). Variations in the SAR can be seen captured by the SIC data throughout Video S2. It is clear from this comparison that the ‘open’ areas we find through the winter are not variations in fast ice coverage – opening and closing occurs in a manner consistent with polynya events, and is visible in the SAR when matching images are available. In a revised version of the paper we will also include a supplemental figure directly comparing SAR images with SIC data using a 70% threshold.
We also note that ASI’s SIC data has been used in this way before in published work (Cheng et al., 2017; 2019) and we note that our mean annual estimate of ice production falls within the range of estimates by Tamura et al. (2008; 2016) and Nihashi et al. (2017).
In a revised version we will clarify the method used in the Tamura et al. and Nihashi et al. studies.
“Regarding the SAR data analysis in the first part of this study, the subsection title does indeed say “qualitative analysis” (L. 137; 335), but it is too qualitative. At L. 156-161, the authors describe the relationship between backscatter and ice (ocean) types. This would be generally true, but on what are these based? Quantitatively, which backscatter value corresponds to those surface types? There are no references at all. Furthermore, they did not compare with the SIC used in the second part of this study. At least, a comparison of SAR backscatter with surface temperature and ice thickness from MODIS infrared imagery, as shown in Nakata et al. (2019), will be needed.”
We agree that more information should be provided on this. The differences between the ice types are not only quantitative differences in backscatter but also textural. However we will include backscatter ranges, and an additional supplementary figure comparing the different ice types we identify that are important for our analysis. As mentioned above we will also include a supplementary figure comparing SAR imagery with SIC data (represented spatially).
We do not think that a comparison of SAR imagery with MODIS temperature and ice thickness data is necessary for qualitatively interpreting the imagery presented in this study, or would add significant information in the context of this study. MODIS data is not necessary to identify ‘open’ areas of ice production or polynya events, and would not aid in interpretation of other polynya dynamics (e.g. backfilling, obstruction).
“In the “qualitative analysis” described in subsection 4.1 starting at L. 335, the SAR images are only shown in Figs. 2 and 3, and the story is developed based mostly on video S1. However, many of the SAR images in this video are missing from the central part of ASP. This leads to difficulties in following the text. It is also extraordinary to say, “Instead, ice formed by the ASP often remains in the ASP study area for months (Video S1, Fig. 3)” (L. 605) from a discussion which is based solely on SAR images. Why can this be said only from temporally sporadic SAR images? In reality, it is reasonable to think that the area of the thin ice area does not change, and the new ice formed there is either thermodynamically growing or advected and deformed at the edge of the polynya. Therefore, dynamical and thermodynamic analyses with an ice motion are essential to justify the author’s claim.”
We make the claim “ice formed by the ASP often remains in the ASP study area for months” based on repeated and focused analysis of the SAR imagery. By visually analyzing the video and tracking changes in the ice it is evident that this is the case. Although temporal and spatial gaps in the video can be limiting, despite these gaps, by viewing the animated images it is possible to observe the flow and dynamics of the ice, including particular floes. While a dynamical and thermodynamical analysis of ice motion in the area would be valuable, we do not think it is within the scope of this study’s goal to analyze the evolution of the polynya, or necessary to make this claim.
Best regards,
Dr Grant Macdonald, on behalf of all authors.
Citation: https://doi.org/10.5194/tc-2021-250-AC1
-
RC2: 'Referee comment on tc-2021-250', Anonymous Referee #2, 02 Nov 2021
GENERAL COMMENTS
I think the manuscript's topic fits nicely within the scope of the journal. As far as I know, this manuscript is unique (i.e. original) in its detailed description of Amundsen Polynya, providing far more information than earlier publications. The study is also scientifically significant given the regional importance of the Amundsen Polynya, either in terms of sea ice production (4th in Antarctica) or for its role in phytoplankton production and atmospheric CO2 uptake. The methods and analyses were generally sound except in one particular analysis for which I proposed verifications and improvements (see "Specific comments: Major"). I thought the manuscript was structured in a straightforward/predictable manner and that the writing (and choice of references) was generally appropriate. Overall, I believe the manuscript has the potential to become a valuable addition to "The Cryosphere".
RECOMMENDATION
Unless my co-reviewers identify major flaws that I've missed, I think the manuscript will be acceptable for publication *after* moderate revisions are made (see specific comments below).
SPECIFIC COMMENTS: MAJOR
(1) Lines 521-522: "Daily mean wind speed at the polynya site and polynya area during the winter period has a weak but significant positive correlation"
Are you sure that these are the two variables (wind speed versus polynya area) you should be comparing together?
I would have tried to compare the winds against the *change* in polynya area, i.e. d(area)/dt.
A gust of winds would presumably correlate with a change in area ( d(area)/dt ), not the actual area.Another related comment is that you are using wind *speed* and ignoring the wind's direction.
Wouldn't the wind direction matter, e.g. in distinguishing between an opening (area increases) or a closing (area decreases) of the polynya?(2) Lines 529-531: "The mean wind direction throughout the ASP study area is approximately southerly. While this direction corresponds to the direction in which the polynya sometimes forms northward off the Dotson Ice Shelf, it does not correspond to the more typical westward formation off the iceberg chain."
Could you please expand on this in your "response to the reviewers", and demonstrate that the "southerly mean wind direction" isn't a plotting artefact or an error in the postprocessing of ERA5's winds? Although I haven't specifically worked with years 2016-2020, the area just north of the Getz and Dotson ice shelves usually shows mean winds blowing from the southeast or from east-southeast. In contrast, Figure 9 shows winds blowing from the south or from south-southeast.
A few thoughts and suggestions:
---A wind blowing from the south or from the south-southeast would imply that the `v' wind component is larger (in absolute terms) than the `u' wind component. Can you go back to the original ERA5 results (meaning before doing any sort of processing) and confirm that this is what you are truly seeing in ERA5's u,v wind components? (We are particularly interested in the region just north of Getz and Dotson ice shelf, i.e. the ASP).
---If you select one single location (latitude = something, longitude = something) where ERA5 winds are defined, please verify that the 'u' value and the the 'v' value (in the original ERA5 files) are quantitatively consistent with the arrow's direction shown at that particular location. If they don't quantitatively match (using the basic trigonometric relations), then there is an error in the plotting or the postprocessing.
---Since wind direction is important in Figure 9, shouldn't we add longitudes and latitudes to orient the reader who is not particularly familiar with the Amundsen Sea?
---Please clarify whether the statement about the "southerly" mean wind direction (Line 529) is specific to a time of the year, and if so, what time of the year exactly? This is a confusing thing, considering that Figure 8 is representative of April-October, Figure 9 is representative of November-December, and we just don't know about Line 529.SPECIFIC COMMENTS: MODERATE
(3) Lines 90-91: "...such as Antarctic bottom water formation and global thermohaline circulation..."
The Amundsen Sea doesn't produce Antarctic bottom water. You must be confusing the Amundsen with another location in Antarctica (maybe the Ross Sea?). You have to remove this passage because it simply doesn't apply to the Amundsen Sea.
(4) Lines 301-302: "Additionally, the total SIC for each day was calculated by calculating the sum of all percentage SIC values in the study region. These total SIC values should only be considered useful for analyzing relative changes in SIC"
By defining "total SIC" in this particular way, you are making it unecessarily difficult to compare your results with past/future studies, because the value obtained will be intimately tied to your grid resolution (the latter being an arbitrary choice).
For example, assume that SIC=100 over a certain geographical area, and that this area is covered by 4 grid points (with the resolution you've chosen). The "total SIC" for this area will be 400. Then, in two years from now, another researcher does a similar analysis and selects a resolution that's twice finer than the one you chose. The same geographical area is now covered by 16 grid points, and the "total SIC" is suddenly 1600.
I'd suggest normalizing the "total SIC" by the total number of grid points within the study region. This way, the number coming out of your analysis isn't so dependent on the arbitrary choice of resolution. Another benefit is that the numbers you'll obtain will be much smaller and less cumbersome. Right now, lines 581-583 are discussing numbers having 7 digits, and that's a bit awkward.
(5) Line 311: "Daily wind speed and direction at the site of the polynya, ERA5's hourly 'u' and 'v' wind products were processed for a region..."
The sentence makes it sound like these variables are peculiar to ERA5, but they really aren't. Please use the formal terminology to present what u,v are:
"Daily zonal (u) and meridional (v) components of the winds at a height of 10m were obtained from ERA5 and processed for a region..."
I also note a confusion between "monthly" (Line 308), "Daily" (Line 311), and "Hourly" (Line 313). This is terribly confusing for the reader. Please state exactly what you've downloaded from ERA5 (monthly? daily? hourly? It has to be *one* of the three), and then correct these sentences as necessary.
SPECIFIC COMMENTS: MINOR
(6) Line 50: Why is "Carbon Dioxide" capitalized? This looks unusual.
(7) Lines 107-108: "Westward coastal currents prevail in the area (St-Laurent et al. 2019)..."
"St-Laurent et al." is only a modeling study. I think it would be good to also cite an observational study with actual measurements of the coastal current: Kim et al. 2016, Estuarine, Coastal and Shelf Science, https://doi.org/10.1016/j.ecss.2016.08.004
(Same comment for Line 674 of the manuscript.)(8) Line 124: The acronym EOS isn't defined. Please define what EOS stands for, or remove the acronym.
(9) Figure 1a: To give the reader some context, please add to Fig.1a the climatological northward extent of the sea ice cover in Summer and in Winter. Such climatologies are available at, e.g., https://nsidc.org/data/NSIDC-0192/versions/3
Having this information will help the reader understand how you chose the extent of the red box. Without this information, the geographical extent of the red box appears arbitrary.(10) Line 224: "Following Cheng et al. (2017) the daily net heat flux, Q (in W/m2), of a pixel was estimated by..."
Shouldn't it be: "of a *ice-free* pixel"?
(11) Line 235: L_o = epsilon sigma T_0^4 (Equation 2)
Why use T_0 (the freezing point of seawater) in Equation 2? Since Equation 2 is the black body radiation of the ocean surface, then Equation 2 should use T_s (the temperature of the water surface), not T_0.
It's acceptable to assume, later on, that T_s is approximately equal to the freezing temperature, but Equation 2 should nevertheless be written as a function of T_s (not T_0) in order to make physical sense.
Similarly, line 241 should read:
T_s ~= T_0 = 273.15 - 0.0137...
rather than:
T_0 = T_s = 273.15 - 0.0137...(12) Lines 256-259: The symbol for pressure is "Pa", not "pa". Please correct throughout the paragraph.
(13) Line 274: Acronym GDAL isn't defined. Please define what GDAL stands for, or remove the acronym.
(14) Line 282: The letter "i" in "rho_i" should be a subscript (compare with Line 280 where it is correctly typed).
(15) Lines 289-290: "Caution should be used when interpreting the absolute numbers produced by the ice production model, particularly because the input data is modeled climate data..."
Are you referring to ERA5 as "modeled climate data"? ERA5 is an atmospheric reanalysis, which is very different from "modeled climate data". For one thing, an atmospheric reanalysis assimilates historical measurements, while a "climate model" doesn't.
(16) Lines 318-319: "wind direction was plotted using the 'matplotlib' function 'quiver'"
The results shouldn't depend on the choice of the graphics software. If this is relevant to the results, then explain in what ways it is relevant. If it's not relevant, then remove the passage.
(17) Figure 4,5,6: Can we make the curvers a bit thicker, so that their color is more apparent and makes it easier to distinguish one curve from another? Also, would it be possible to make these figure labels less pixelated, or larger so that the pixelation doesn't show as much?
(18) Line 575: "SIC for the broader ASP region on two days in 2017"
Please clarify why you picked these two specific days, and what they represent? Presumably the top figure represents some kind of minimum? And what about the bottom figure?
(19) Line 676: "(Koo et al., in review)."
You aren't supposed to cite studies that haven't been accepted for publication. Most journals will systematically remove such references at the copy-editing stage.
(20) There is something wrong in the caption of Figure S1:
"The figure covers background figure and area is the same as Fig 1b.".
Please correct as needed.Citation: https://doi.org/10.5194/tc-2021-250-RC2 -
AC2: 'Reply on RC2', Grant Macdonald, 16 Nov 2021
Dear Reviewer 2,
Thank you for your time, comments, and feedback that will undoubtedly improve this manuscript if we are invited to submit a revised version. Here we provide an initial response to your comments. If invited to submit a revised version, we will provide more details regarding changes made related to your comments.
SPECIFIC COMMENTS: MAJOR
(1) Lines 521-522: "Daily mean wind speed at the polynya site and polynya area during the winter period has a weak but significant positive correlation"
Are you sure that these are the two variables (wind speed versus polynya area) you should be comparing together?
I would have tried to compare the winds against the *change* in polynya area, i.e. d(area)/dt.
A gust of winds would presumably correlate with a change in area ( d(area)/dt ), not the actual area.
Another related comment is that you are using wind *speed* and ignoring the wind's direction.
Wouldn't the wind direction matter, e.g. in distinguishing between an opening (area increases) or a closing (area decreases) of the polynya?
We agree that change in polynya area would be a better variable to compare, we will amend this. We think it is valuable to include a comparison of changes in the polynya with just wind speed, as done in other studies, but agree wind direction is also important, so we will make reference to the daily wind direction shown in Video S3 when discussing changes in polynya area and its relationship with wind.
(2) Lines 529-531: "The mean wind direction throughout the ASP study area is approximately southerly. While this direction corresponds to the direction in which the polynya sometimes forms northward off the Dotson Ice Shelf, it does not correspond to the more typical westward formation off the iceberg chain."
Could you please expand on this in your "response to the reviewers", and demonstrate that the "southerly mean wind direction" isn't a plotting artefact or an error in the postprocessing of ERA5's winds? Although I haven't specifically worked with years 2016-2020, the area just north of the Getz and Dotson ice shelves usually shows mean winds blowing from the southeast or from east-southeast. In contrast, Figure 9 shows winds blowing from the south or from south-southeast.
A few thoughts and suggestions:
---A wind blowing from the south or from the south-southeast would imply that the `v' wind component is larger (in absolute terms) than the `u' wind component. Can you go back to the original ERA5 results (meaning before doing any sort of processing) and confirm that this is what you are truly seeing in ERA5's u,v wind components? (We are particularly interested in the region just north of Getz and Dotson ice shelf, i.e. the ASP).
We checked this and it seems to be correct, but we will check more thoroughly before submitting a revision. It would be more precise to say that we calculate the mean wind to come from the SSE in front of the Dotson, and SE to ESE in front of the Getz – we will amend the text to reflect this.
---If you select one single location (latitude = something, longitude = something) where ERA5 winds are defined, please verify that the 'u' value and the the 'v' value (in the original ERA5 files) are quantitatively consistent with the arrow's direction shown at that particular location. If they don't quantitatively match (using the basic trigonometric relations), then there is an error in the plotting or the postprocessing.
As above, it seems to be correct but we will more thoroughly check before submitting a revision.
---Since wind direction is important in Figure 9, shouldn't we add longitudes and latitudes to orient the reader who is not particularly familiar with the Amundsen Sea?
We will do this.
---Please clarify whether the statement about the "southerly" mean wind direction (Line 529) is specific to a time of the year, and if so, what time of the year exactly? This is a confusing thing, considering that Figure 8 is representative of April-October, Figure 9 is representative of November-December, and we just don't know about Line 529.”
The reference to ‘southerly’ on line 529 refers to the year-round mean. We will rephrase for clarity. Figure 9 is representative of the year-round mean, not November-December. Note it is November *2016* to December *2020*. However, we understand how this could confuse and we will adjust the caption to make clear it is the year-round mean over this whole period.
SPECIFIC COMMENTS: MODERATE
(3) Lines 90-91: "...such as Antarctic bottom water formation and global thermohaline circulation..."
The Amundsen Sea doesn't produce Antarctic bottom water. You must be confusing the Amundsen with another location in Antarctica (maybe the Ross Sea?). You have to remove this passage because it simply doesn't apply to the Amundsen Sea.
Thank you, we will remove this.
(4) Lines 301-302: "Additionally, the total SIC for each day was calculated by calculating the sum of all percentage SIC values in the study region. These total SIC values should only be considered useful for analyzing relative changes in SIC"
By defining "total SIC" in this particular way, you are making it unecessarily difficult to compare your results with past/future studies, because the value obtained will be intimately tied to your grid resolution (the latter being an arbitrary choice).
For example, assume that SIC=100 over a certain geographical area, and that this area is covered by 4 grid points (with the resolution you've chosen). The "total SIC" for this area will be 400. Then, in two years from now, another researcher does a similar analysis and selects a resolution that's twice finer than the one you chose. The same geographical area is now covered by 16 grid points, and the "total SIC" is suddenly 1600.
I'd suggest normalizing the "total SIC" by the total number of grid points within the study region. This way, the number coming out of your analysis isn't so dependent on the arbitrary choice of resolution. Another benefit is that the numbers you'll obtain will be much smaller and less cumbersome. Right now, lines 581-583 are discussing numbers having 7 digits, and that's a bit awkward.
Thank you, we agree that this was a poor choice and will instead normalize as suggested.
(5) Line 311: "Daily wind speed and direction at the site of the polynya, ERA5's hourly 'u' and 'v' wind products were processed for a region..."
The sentence makes it sound like these variables are peculiar to ERA5, but they really aren't. Please use the formal terminology to present what u,v are:
"Daily zonal (u) and meridional (v) components of the winds at a height of 10m were obtained from ERA5 and processed for a region..."
We will amend as suggested.
I also note a confusion between "monthly" (Line 308), "Daily" (Line 311), and "Hourly" (Line 313). This is terribly confusing for the reader. Please state exactly what you've downloaded from ERA5 (monthly? daily? hourly? It has to be *one* of the three), and then correct these sentences as necessary.
Daily wind speed was calculated from hourly ERA5 data, and we used the ERA5’s monthly product to analyze the overall mean over the study period (i.e. Fig. 9). We see how this is confusing and will rephrase for better clarity, or simply use the same hourly product to calculate the study period mean in order to avoid the unnecessarily confusing explanation.
SPECIFIC COMMENTS: MINOR
(6) Line 50: Why is "Carbon Dioxide" capitalized? This looks unusual.
We will change this.
(7) Lines 107-108: "Westward coastal currents prevail in the area (St-Laurent et al. 2019)..."
"St-Laurent et al." is only a modeling study. I think it would be good to also cite an observational study with actual measurements of the coastal current: Kim et al. 2016, Estuarine, Coastal and Shelf Science, https://doi.org/10.1016/j.ecss.2016.08.004
(Same comment for Line 674 of the manuscript.)
Thank you, we will do this.
(8) Line 124: The acronym EOS isn't defined. Please define what EOS stands for, or remove the acronym.
We will do this.
(9) Figure 1a: To give the reader some context, please add to Fig.1a the climatological northward extent of the sea ice cover in Summer and in Winter. Such climatologies are available at, e.g., https://nsidc.org/data/NSIDC-0192/versions/3
Having this information will help the reader understand how you chose the extent of the red box. Without this information, the geographical extent of the red box appears arbitrary.
Thank you, we will do this.
(10) Line 224: "Following Cheng et al. (2017) the daily net heat flux, Q (in W/m2), of a pixel was estimated by..."
Shouldn't it be: "of a *ice-free* pixel"?
Yes, it should be for an ‘open polynya’ pixel – we will amend this.
(11) Line 235: L_o = epsilon sigma T_0^4 (Equation 2)
Why use T_0 (the freezing point of seawater) in Equation 2? Since Equation 2 is the black body radiation of the ocean surface, then Equation 2 should use T_s (the temperature of the water surface), not T_0.
It's acceptable to assume, later on, that T_s is approximately equal to the freezing temperature, but Equation 2 should nevertheless be written as a function of T_s (not T_0) in order to make physical sense.
Similarly, line 241 should read:
T_s ~= T_0 = 273.15 - 0.0137...
rather than:
T_0 = T_s = 273.15 - 0.0137...
Thank you, we will make these changes.
(12) Lines 256-259: The symbol for pressure is "Pa", not "pa". Please correct throughout the paragraph.
Thank you for spotting this, we will amend it and double check for other similar errors.
(13) Line 274: Acronym GDAL isn't defined. Please define what GDAL stands for, or remove the acronym.
We will define the acronym (Geospatial Data Abstraction Library).
(14) Line 282: The letter "i" in "rho_i" should be a subscript (compare with Line 280 where it is correctly typed).
Thank you for spotting this, we will amend it.
(15) Lines 289-290: "Caution should be used when interpreting the absolute numbers produced by the ice production model, particularly because the input data is modeled climate data..."
Are you referring to ERA5 as "modeled climate data"? ERA5 is an atmospheric reanalysis, which is very different from "modeled climate data". For one thing, an atmospheric reanalysis assimilates historical measurements, while a "climate model" doesn't.
This does refer to ERA5. We will amend this incorrect wording.
(16) Lines 318-319: "wind direction was plotted using the 'matplotlib' function 'quiver'"
The results shouldn't depend on the choice of the graphics software. If this is relevant to the results, then explain in what ways it is relevant. If it's not relevant, then remove the passage.
We will remove the passage.
(17) Figure 4,5,6: Can we make the curvers a bit thicker, so that their color is more apparent and makes it easier to distinguish one curve from another? Also, would it be possible to make these figure labels less pixelated, or larger so that the pixelation doesn't show as much?
We will make the lines thicker and the labels larger.
(18) Line 575: "SIC for the broader ASP region on two days in 2017"
Please clarify why you picked these two specific days, and what they represent? Presumably the top figure represents some kind of minimum? And what about the bottom figure?
We will explain in the caption that
- panel (a) is presented to show an example of when the polynya had no ice boundary to the W/NW due to exceptionally low SIC in the region during the 2016/17 summer
- panel (b) is presented to show an example of the low SIC in the region, and narrow icepack bounding the polynya in the subsequent early winter.
(19) Line 676: "(Koo et al., in review)."
You aren't supposed to cite studies that haven't been accepted for publication. Most journals will systematically remove such references at the copy-editing stage.
We acknowledge this mistake, however the paper in question has now been published so we will retain it in the text.
(20) There is something wrong in the caption of Figure S1:
"The figure covers background figure and area is the same as Fig 1b.".
Please correct as needed.
Thank you, we will correct this. It should read "The background image and area is the same as in Fig 1b.".
Best regards,
Dr Grant Macdonald, on behalf of all authors.
Citation: https://doi.org/10.5194/tc-2021-250-AC2
-
AC2: 'Reply on RC2', Grant Macdonald, 16 Nov 2021
-
RC3: 'Comment on tc-2021-250', Anonymous Referee #3, 03 Dec 2021
I was asked by the editor to provide a 3rd review of the manuscript “Evolution of the Amundsen Sea Polynya, Antarctica, 2016-2021” by Macdonald et al. submitted to the Cryosphere 2021. Two reviews were available with one recommendation to reject and the other as a major revision.
The authors aim to investigate the evolution of the Amundsen Sea Polynya, Antarctica by the means of SAR images, AMSR2 ice concentration and ERA 5 reanalysis data. The study area is motivated by the highest primary productivity and key oceanographic significance. The time period 2016-2021 is very limited by the availability of Sentinel-1 SAR images. Specifically, the aims are to analyse qualitative changes in the characteristics, the summer polynya area and the winter polynya area and ice production.
The authors do not precisely identify the novel aspects of their investigation but I think it is mainly the use of Sentinel-1 data. This is in principle a good and interesting approach to investigate the polynya dynamics in the special Amundsen Sea environment. However, the limitations due to the sparse temporal sampling are not adequately discussed, and the knowledge gained is not presented in a comprehensible manner. Unfortunately, the quality of the supplementary video is only poor and does not support the claims. If the visual SAR video/image analysis were carried out by showing more example results, I would rate the present manuscript much better.
The scientific quality of the study is only poor. I agree with the fundamental concerns of reviewer #1. In addition to the concerns about neglecting the effect of thin ice thickness there is an inconsistency with the applied data sets. The high resolution AMSR2 ice concentration used for the manuscript is different from the low resolution OSI-SAF/OSTIA ice concentration used in the ERA5 reanalysis. This means that there are likely large inconsistencies for the calculation of heat fluxes when the low resolution data do not resolve the open water polynya. Unfortunately, there is no error assessment or sensitivity analysis.
I do not fully understand the study logic. I would have expected a comparison between the passive microwave and SAR derived polynya areas. The scalar wind speed is of very limited use to describe the polynya which can be seen in Fig. 8. The surprisingly low correlation indicates that the wind speed explains only very little of the polynya area variability. A more insightful approach would have been the usage of wind direction like outlined in Haarpaintner et al. (2001) who compared SAR-derived polynya width/area with model based results.
The literature regarding the impact of the Amundsen polynya is not well balanced, i.e. the impact for carbon sequestration is emphasized but this seems to be controversial (Lee et al., 2017). One interesting question is how to define a polynya during Summer, in particular when there is only surrounding open water.
In summary, I concur with review #1 and suggest to reject the manuscript but encourage the authors to resubmit their work after a fundamental revision. The aspect of the SAR image analysis would be very interesting when compared with the passive microwave data. The gain in knowledge has to be identified and the main hypotheses have to be worked out and supported by evidence.
Lee et al. (2017) https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2017GL074646
Haarpaintner et al. (2001) https://agupubs.onlinelibrary.wiley.com/doi/pdf/10.1029/1999JC000133
Citation: https://doi.org/10.5194/tc-2021-250-RC3
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,283 | 463 | 66 | 1,812 | 126 | 59 | 54 |
- HTML: 1,283
- PDF: 463
- XML: 66
- Total: 1,812
- Supplement: 126
- BibTeX: 59
- EndNote: 54
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1