the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 20 Mar 2024
Arctic Surface Snow Interactions with the Atmosphere: Spatio-Temporal Isotopic Variability During the MOSAiC Expedition
Abstract. The Arctic Ocean’s snow cover is crucial in moderating interactions between sea-ice and the atmosphere, yet fully grasping its isotopic composition and the processes shaping it presents substantial challenges. This study employs a unique dataset from the Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition to explore the complex interactions between deposition processes and post-depositional changes affecting snow on Arctic sea ice. By examining 911 individual snow isotope measurements collected over a full year, we identify a clear layering within the snowpack: the top layer, with lower δ18O values and higher d-excess values, indicates fresh meteoric snowfall, while the bottom layer, affected by the sea ice beneath, shows higher δ18O values and lower d- d-excess values. By integrating these discrete snow samples with continuous vapour isotope data, our research provides insight into interactions between snow and the atmosphere, as well as the processes that alter isotopic signatures within Arctic snow.
We observe a significant difference in δ18O values between snow and vapor during autumn, mainly due to delays in sampling after precipitation events, with d-excess ranges suggesting the impact of Atlantic moisture. Winter months exhibit sharp differences in δ18O and d-excess values, indicating kinetic fractionation amid extreme cold as the RV Polarstern traverses from the Siberian to the Atlantic sector of the Arctic Ocean. Conversely, summer months display a convergence in isotopic signatures, reflecting conditions favouring equilibrium fractionation, highlighted by increased air temperatures and humidity levels. While δ18O in vapour readily responds to changes in air temperature and humidity, surface snow δ18O is influenced more by subsequent processes such as sublimation and wind-driven redistribution. Sublimation, intensified by the snow’s prolonged surface residence and facilitated by the porosity of snow, plays a key role in isotopic enrichment. Wind-driven snow redistribution, occurring 67 % of the winter, led to a homogenised and depleted surface snow δ18O signal across the sea ice by spreading lower δ18O meteoric snow. This effect was especially pronounced in ridge snow profiles, where the top layers showed a uniform δ18O signal, in stark contrast to flat ice samples.
Furthermore, distinct isotopic patterns were detected along the MOSAiC expedition route from a region close to Samoylov Island to Fram Straight near Ny-Ålesund. Snow samples close to Samoylov Island exhibited notable seasonal δ18O variations, which were indicative of a continental climate. In contrast, samples from Ny-Ålesund displayed more consistent fluctuations, influenced by steady Atlantic moisture.
- Preprint
(3444 KB) - Metadata XML
-
Supplement
(351 KB) - BibTeX
- EndNote
Status: open (until 02 May 2024)
-
RC1: 'Comment on egusphere-2024-719', Lijun Tian, 21 Apr 2024
reply
General Comment:
The preprint, "Arctic Surface Snow Interactions with the Atmosphere: Spatio-Temporal Isotopic Variability During the MOSAiC Expedition," offers an in-depth analysis of the isotopic composition within Arctic snow and its dynamic relationship with the atmosphere. The research examined a robust dataset comprising 911 discrete snow isotope measurements, spanning the duration of an entire year throughout the MOSAiC expedition, complemented by continuous vapor isotope data. This research utilizes this unique dataset to explore the complex deposition processes and post-depositional changes affecting snow on Arctic sea ice. The study's findings are valuable for refining climate models and understanding the broader implications of Arctic climate change.
However, the discussion of the processes involved is too simple and elides much of the complexity involved, which could lead to overconfidence in the interpretation of isotopic signals. There needs to be a much clearer discussion of these processes and the potential errors they might introduce. Therefore, I recommend major revisions before it can be considered for publication.Specific Comments:
1. There is significant bias (δ18O-6.40‰; δ2H-36.4‰) between the samples from dataset 1 and dataset 2. You attributed this systematic offset to the evaporative fractionation effects during storage of dataset 2 samples. I have never encountered such a substantial isotope difference between labs, particularly when the samples were stored in a low-temperature, sealed environment. In my opinion, there might be an issue with the standards and/or the correction process in the WSL lab. You could verify this by sending samples with known isotopic values to the WSL lab for re-analysis.2. There is no information on "snow ice" in this manuscript. When the snowpack on sea ice is heavy and thick enough to depress the top surface below sea level, a slush layer or slurry is formed through the mixture of seawater or brine and snow at the base of the snow cover. The surface flooding and snow-ice formation might exert influence on the snow isotopic signals. You should provide freeboard information and give a brief discussion on that. Two refs on snow-ice formation:
https://doi.org/10.1002/2017JC012865
https://doi.org/10.1017/aog.2020.553. I did not see the boxplot for differences in d-excess values in Figure 2. You claimed "Winter months exhibit sharp differences in δ18O and d-excess values, indicating kinetic fractionation...". On the contrary, perhaps it is the equilibrium fractionation that makes the snow heavier than the water vapor (＞10‰ for δ18O). In the summer months, the convergence in isotopic signatures between snow and vapor might reflect enriched isotopic signals in snow due to kinetic fractionation. Please check the following refs on the sublimation process in summer:
https://doi.org/10.1029/2018JD030218
https://doi.org/10.5194/tc-2023-764. The snow and vapor isotopic signatures are particularly identical during July-Sept. 2020. You could expand on the mechanisms driving this phenomenon besides the vapor deposition, for example, the "supersaturation effect" might occur when the RH is high.
5. For the snow profile, you assumed that the lower layer with enriched δ18O and lower d-excess values was caused by snow interacting with the sea-ice surface through diffusion of sea-ice. Is it possible that the lower layer of the snow profile is snow-ice?
6. What are the control factors of the isotopic signals of δ18O and d-excess? Are the control factors identical or diverse? What kind of changes do the δ18O-δD line slopes have in different seasons? Two refs on the interpretation of isotopic signals:
https://doi.org/10.1029/2022JD037037
https://doi.org/10.1029/2018JC0137977. You mentioned the impact of wind-driven snow redistribution. It would be beneficial to include a more detailed analysis of this process and its broader implications for the isotopic homogeneity across the Arctic.
8. The authors should consider adding a section on the potential feedback loops that may exist between the observed isotopic changes and larger-scale atmospheric circulation patterns. You already have a discussion using isotope monitoring at Samoylov Island and Ny-Ålesund stations. Please consider using isotopic general circulation model and HYSPLIT trajectory model.
Technical Corrections:
1. In the Abstract, the isotopic terms "δ18O" and "d-excess" is mentioned but not initially defined.2. In the whole text, you could use a hyphen in "sea-ice" as an adjective; however, there are many unnecessary hyphens in the words as noun, such as "Arctic sea-ice".
3. There usually isn't a space before the per mil (‰).
4. The negative numbers should use the minus symbol instead of hyphen symbol.
5. Line 20- There is a typo of "d-excess" in the sentence "shows higher δ18O values and lower d- d-excess values."
6. Line 105- The "is" should be "are" in the sentence "Sublimation from snow into the atmosphere and deposition from the atmosphere onto the snow is controlled"
7. In Figure 4, the three boxes overlap together and it's rather hard to distinguish them. I suggest you could use violin plot or grouped boxplot. FYI: https://r-graph-gallery.com/265-grouped-boxplot-with-ggplot2.html
8. In "Reference" section, there are a few typos for the isotopic symbols. For example:
Akers, P. D., Kopec, B. G., Mattingly, K. S., Klein, E. S., Causey, D., and Welker, J. M.: Baffin Bay sea ice extent and synoptic moisture transport drive water vapor isotope ( δ <sup>18</sup>O, δ <sup>2</sup>H, and deuterium excess) variability in coastal northwest Greenland, Atmospheric Chemistry and Physics, 20, 13929-13955, 10.5194/acp-20-13929-2020, 2020.
Klein, E. S., Baltensperger, A. P., and Welker, J. M.: Complexity of Arctic Ocean water isotope (δ18O, δ2H) spatial and temporal patterns revealed with machine learning, Elementa: Science of the Anthropocene, 12, 2024.Citation: https://doi.org/10.5194/egusphere-2024-719-RC1 -
RC2: 'Comment on egusphere-2024-719', Anonymous Referee #2, 23 Apr 2024
reply
Review for „Arctic Surface Snow Interactions with the Atmosphere: SpatioTemporal Isotopic Variability During the MOSAiC Expedition“ by Moein Mellat et al.
General comments:
The authors present an exciting dataset from the underexplored central Arctic, a region that plays a key role in the climate system. The authors attempt to characterize the temporal and spatial (or snow type) variability of the isotopic composition of snow on sea ice but are limited by the snow sampling protocol. Nevertheless, some crude patterns are identified, and the dataset is compared to already published datasets of local isotopic water vapor and precipitation isotope data from two remote land-based stations. This comparison is motivated by the identification and deciphering of the drivers forming the snow’s isotopic composition. The manuscript is well-written, and the findings are presented in a clean way. With this paper, the authors contribute to the understanding of the polar water cycle by using the isotopic snow signal to identify major surface processes at play. I greatly support this type of study that explores isotope dynamics to improve polar process understanding. However, I am very skeptical about the quality of the presented dataset and thus must raise some major concerns:-
The presented isotopic snow dataset is a combination of two publicly available datasets, yet dataset 2 (DS2) has not been peer-reviewed yet. The uncertainty levels given (L. 152) for the DS2 samples measured at the WSL facility are unusually high and increased by a factor of 2 or higher compared to the state- of-the-art (Lis et al., 2008). The resulting uncertainty on d-excess is ~8‰ which makes interpretation difficult. The authors need to justify why they accept such high uncertainty levels.
-
The authors name evaporation/sublimation fractionation during storage as a reason for the alarmingly high identified bias between the DS2 sub-datasets and apply a linear bias correction. However, as the authors acknowledge themselves in L.80 kinetic fractionation during evaporation greatly influences d- excess. By definition, kinetic fractionation does not affect d18O and dD evenly. Thus, the original d-excess values cannot be “recreated” by applying a linear bias correction which might also explain why some d-excess values (of corrected!) samples are astonishingly low (< -30‰). Therefore, I have very low confidence in the presented d-excess values and am very skeptical about whether any d-excess interpretation/discussion is meaningful. The author group is aware of this problem as apparent from the statement in the data publication “We would like to emphasise that calculations of d-excess values for this dataset need interpreting carefully.” (Macfarlane et al., 2022), yet the issue is not discussed in this manuscript which is a major flaw.
-
L. 163: The bias correction applied to combine the datasets measured in two laboratories is described in one sentence and a reference is given to the online dataset where the same minimal explanation is given. This is clearly not enough to establish trust in the presented dataset correction method. If such extreme correction measures are necessary, the foundation for the correction method should be made accessible somewhere so the reader can judge the quality of the correction method and thus the quality of the data.
Thus, I strongly recommend to either:
- exclude the sub-dataset measured at WSL from the manuscript
- remeasure the samples at a different laboratory (to eliminate the possibility that it is a shortcoming of the laboratory instead of a storage problem)
- at the minimum, indicate clearly which of the presented data points are part of this problematic dataset in the figures and text
The manuscript is rather descriptive and lacks analysis that substantiates the claims being made: correlation is not necessarily equal to causality. Many relevant processes potentially influencing the snow isotope variability are named, but there is a clear need for more elaborate analysis to unambiguously identify responsible drivers including:
- Back trajectory analysis for the comparison to the land-based stations
- More statistical analysis testing for spatial and temporal trends
- The comparison of vapor vs. snow isotope variability by accounting for temperature-dependent equilibrium fractionation to substantiate the claim of non-equilibrium post-depositional processes as drivers for variability
- A discussion on melt influences on the isotopic composition of the snowI encourage the authors to reflect on the organization of the manuscript as the discussion section contains several figures presenting relevant data and statistical analysis which should be part of the results. The discussion could then detail the interpretation of these results and draw connections to other literature.
Suggestion: Based on my lack of trust in the combined dataset and the limited statistical analysis and substantiated interpretation I would recommend that the authors consider modifying this manuscript into a dataset paper submission (e.g. Copernicus ESSD journal) which would be a better fit to establish the credibility of the combined dataset, would offer the chance to outline the two land-based stations and the vapor dataset as a possible comparison, and still make this valuable dataset comprehensibly accessible to the community. Alternatively, a clear indication of the problematic data points needs to be made throughout the whole manuscript including the figures or these problematic data need to be excluded from the analyses entirely.
Specific comments:
L. 120: This last statement is not supported by the analysis that follows.
L. 181: Give specific data uncertainties for all datasets being used
L. 191: What does “Met City” stand for?
L. 196: Please indicate somewhere that the observation periods do not overlap
L. 211: The authors decide to discuss deviations from the median rather than deviations from the (weighted) mean. Please add a short statement of why this approach was chosen. Adding on to this, in the following paragraph “variability” in the dataset is discussed without defining what metric is chosen as the “variability” indicator. Please add.
L. 245: Exchange “typically” with a more specific word
Fig. 1: Try to find a better way to visualize the “subdivisions” you describe in L. 229 and maybe swap the color coding with the y-axis information to better visualize the variability. Add legend for marker shape.
L. 218: The following paragraph is confusing. It is not clear whether snow type or spatial variability is discussed. Consider adding an extra paragraph for “ridge” isotopes including the definition of ridges
L. 229: Without establishing the credibility of the d-excess dataset first, the following paragraph is difficult to interpret. Consider marking the problematic data points.
L. 278: Identify possible drivers for the observed “divergence” between vapor and snow by analyzing if vapor equilibrium fractionation can or cannot explain the snow behavior.
Fig 2: How is the precipitation data extracted from ERA5. The respective grid box of the moving ship?
L. 369: The Macfarlane et al., (2023) manuscript co-authored by the same authors is publicly available on a preprint server but has not been peer-reviewed, so interpretations from this manuscript should be addressed with the necessary precautions.
L. 418: Please elaborate or give details about which processes you expect to be responsible for a “more effective transfer of isotope signals” and provide evidence for the identified processes. An evidence-based credible process identification is missing in section 4.2 in general.
L. 426, L. 608: Please calculate this “disequilibrium” and show the analysis (Aemisegger et al., 2015; Wahl et al., 2024)
Section 4.3: Which snow sample values are being discussed in this section? Does “average daily snow” samples mean that all available samples were averaged? The comparison between snow and vapor isotopic composition (Fig 5b) should account for temperature since equilibrium fractionation is temperature-dependent, even if the effect for d18O is small. So if there is an exchange between snow and vapor, temperature probably has an effect.
L. 438: How was the 2mm aggregated precipitation level chosen? If it’s an arbitrary value, how robust is the analysis for different values, e.g. 1.5mm?
Section 4.4: For the comparison with continental locations a back trajectory analysis seems highly necessary to establish moisture pathways and general synoptic scale conditions.
L. 545: Exchange “sublimation” with “vapor deposition”. A negative latent heat flux is a deposition flux in this dataset.
L. 549: Commonly, snow sublimation leads to an enrichment in d18O and a decrease in d-excess (Hughes et al., 2021).
L. 560: The following interpretation of wind redistribution as the driver for surface snow depletion lacks evidence. See also (Wahl et al., 2024)
Fig 7 c: Are the relative humidity levels given referenced against saturation over ice or saturation over water? (Anderson, 1994)
L. 588: Combined vapor and snow measurements have been performed in the Arctic for at least 10 years (Steen-Larsen et al., 2014), so please rephrase “methodological advancement”.Technical corrections:
L. 18: Exchange “The Arctic Ocean’s snow cover” with “Snow on sea ice”
L. 24: delete “d-“
L. 27: «Vapour» or «Vapor»? consistency
L. 21 vs. L. 71: Consider defining what you mean by “deposition” to avoid confusion. Snowfall deposition or the phase transition between vapor and solid?
L. 129: Exchange “flow” with “floe”
Fig 1, caption: Correct plot number labeling
Fig 1: swap grey and white background coloring
L. 293-303: This is a repetition of information already given in the Data section. Please shorten.
Fig 4: The line for the median in the Surface Snow MOSAiC Box Plots is hardly visible L. 456: Add respective citation
Fig 5b: What are empty circles? Indicate the range of precipitation amounts. Have you applied a lower limit to account for the “drizzle effect” in modelled precipitation?
L. 535: delete “out”
L. 598: exchange “snow-sea-ice interactions” with “atmosphere-snow interactions”
L. 617: Please specify what you mean with this very general statementBibliography:
Aemisegger, F., Spiegel, J. K., Pfahl, S., Sodemann, H., Eugster, W., and Wernli, H.: Isotope meteorology of cold front passages: A case study combining observations and modeling, Geophysical Research Letters, 42, 5652–5660, https://doi.org/10.1002/2015GL063988, 2015.
Anderson, P. S.: A method for rescaling humidity sensors at temperatures well below freezing. Journal of Atmospheric and Oceanic Technology, Journal of Atmospheric and Oceanic Technology, 11, 1388+1397, 1994.
Hughes, A. G., Wahl, S., Jones, T. R., Zuhr, A., Hörhold, M., White, J. W. C., and Steen- Larsen, H. C.: The role of sublimation as a driver of climate signals in the water isotope content of surface snow: Laboratory and field experimental results, The Cryosphere, 15, 4949–4974, https://doi.org/10.5194/tc-15-4949-2021, 2021.
Lis, G., Wassenaar, L. I., and Hendry, M. J.: High-Precision Laser Spectroscopy D/H and 18 O/ 16 O Measurements of Microliter Natural Water Samples, Anal. Chem., 80, 287–293, https://doi.org/10.1021/ac701716q, 2008.
Macfarlane, A., Mellat, M., Dadic, R., Meyer, H., Werner, M., Brunello, C., Arndt, S., Krampe, D., and Schneebeli, M.: Ocean-sourced snow: An unaccounted process on Arctic sea ice, https://doi.org/10.21203/rs.3.rs-3572881/v1, 20 November 2023.
Macfarlane, A. R., Schneebeli, M., Dadic, R., Wagner, D. N., Arndt, S., Clemens-Sewall, D., Hämmerle, S., Hannula, H.-R., Jaggi, M., Kolabutin, N., Krampe, D., Lehning, M., Matero, I., Nicolaus, M., Oggier, M., Pirazzini, R., Polashenski, C., Raphael, I., Regnery, J., Shimanchuck, E., Smith, M. M., Tavri, A., Mellat, M., Meyer, H., Werner, M., and Brunello, C. F.: Snowpit stable isotope profiles during the MOSAiC expedition, https://doi.org/10.1594/PANGAEA.952556, 2022.
Steen-Larsen, H. C., Masson-Delmotte, V., Hirabayashi, M., Winkler, R., Satow, K., Prié, F., Bayou, N., Brun, E., Cuffey, K. M., Dahl-Jensen, D., Dumont, M., Guillevic, M., Kipfstuhl, S., Landais, A., Popp, T., Risi, C., Steffen, K., Stenni, B., and Sveinbjörnsdottír, A. E.: What controls the isotopic composition of Greenland surface snow?, Climate of the Past, 10, 377– 392, https://doi.org/10.5194/cp-10-377-2014, 2014.
Wahl, S., Walter, B., Aemisegger, F., Bianchi, L., and Lehning, M.: Identifying airborne snow metamorphism with stable water isotopes, The Cryosphere - Discussions, https://doi.org/10.5194/egusphere-2024-745, 8 April 2024.
Citation: https://doi.org/10.5194/egusphere-2024-719-RC2 -
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 229 | 38 | 13 | 280 | 20 | 6 | 5 |
- HTML: 229
- PDF: 38
- XML: 13
- Total: 280
- Supplement: 20
- BibTeX: 6
- EndNote: 5
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1