the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 08 Jul 2025
Multiproxy analyses of multiple firn cores from coastal Adélie Land covering the last 40 years
Abstract. Water stable isotope signals recorded in shallow firn cores are essential to constrain the variations of climate and atmospheric water cycle over the past decades to centuries. However, deposition and post-deposition effects add additional signal, often referred to as stratigraphic noise, to the isotopic signal. One way to reduce the local stratigraphic noise is to combine several firn cores at the same location.
Here, we study the water isotopic composition and chemical records from 9 firn cores (20 to 40 m depth) drilled in 2016 at 3 sites (D47, Stop5 and Stop0) with high accumulation rates (~200 mm w.eq ·yr-1) along a transect between the coast and the plateau in Adélie Land in Antarctica (100 to 385 km from the coastal station Dumont d’Urville). Each core covers at least the period from 1979 to 2016 and the high-resolution measurements permit to capture the seasonal variations in both chemical and isotopic records. At each site, similarities in the nssSO4 and δ18O variations between the different cores were used to combine the three isotopic records into a single stacked isotopic curve, thereby enhancing the signal-to-noise ratio. At two sites, we find a good agreement when comparing the water isotopic profiles recovered from the stacked records to those obtained as modeling output from virtual firn cores calculated using the two isotope-enabled atmospheric general circulation models, ECHAM6-wiso and LMDZ6iso over the period 1979–2016. At the very windy site of D47, building a coherent signal from the 3 individual cores is not possible because the isotopic and impurities signals are much more affected by stratigraphic noise. This study confirms that, even if the benefit of stacking is limited at very windy sites, combining several cores is of primary importance to faithfully reconstruct water isotope variability at one site and further investigate how much climate signal is actually preserved in coastal cores.
- Preprint
(2560 KB) - Metadata XML
-
Supplement
(1425 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-2863', Anonymous Referee #1, 02 Nov 2025
- AC1: 'Reply on RC1', Titouan Tcheng, 12 Dec 2025
-
RC2: 'Comment on egusphere-2025-2863', Ruth Mottram, 01 Dec 2025
REVIEW: Tcheng et al. Multiproxy analyses of multiple firncores from coastal Adélie Land covering the last 40 years
Note: This review was written by a research group following a thorough discussion of the paper within the group. This was an interesting paper that we have enjoyed reading and discussing and we find it a potentially valuable contribution to the community of Antarctic climate researchers. It is well written and presented and includes some valuable datasets.
The paper uses the stacking method of firn cores as a way to quantify and reduce uncertainties in firn core isotopic signal, called stratigraphic noise. For several isotopes, at three sites, with three cores at each site, they compare the individual isotopic signals to the stacked one, and to virtual firn cores compiled from isotope-enabled climate models. The authors show that there is greater resemblance between the stacked signal and the virtual firn core signals for two out of three sites. The individual records are quite similar for the first two sites, but present slight variations in depth and amplitude. The site where stacking is inefficient is known for high winds, meaning significant ablation and heterogeneous accumulation (surface roughness, recorded wind speed, modelled wind speed), causing significant differences in signal over very short distances, and generally making this site inappropriate to retrieve any climatic information. It is observed that stacking does erase extremes, including the warmest months. In general, stacking appears to be a reasonable method to increase the reliability in isotopic signals from firn cores.
We identify a few major issues in the paper and a few minor ones.
2 Major issues
2.1 1. Structure and justification of the paper
It is not totally clear at the start what the major scientific objective is and how it fits into previous research. It would be helpful to introduce the scientific ideas that provoke this study, the introduction section reads a little disjointed. The objectives do become clearer upon reading the discussion however, in particular section 4.2. We therefore suggest adding a little more detail to the abstract and the introduction around the overall aims, in particular the potential to identify climate/seasonal/weather signals from the firn cores.
We assume that another objective is to use these kind of datasets to evaluate general purpose GCMs as well as isotope enabled ones, and indeed this is a valuable dataset for the purpose, but it is not clear upon first reading of the paper and some discussion of this would also be useful.
2.2 2. Objectives
There are two listed objectives on lines 41-44. The first one focuses on improved isotopic record with respect to stratigraphic noise, the second is to expand that firn core knowledge to the recent climate. It would be clearer if you mention that these research questions can be answered by specifying stacking, but also they should be tied back to the wider points you make at the start about Antarctica’s role in the global climate system. The general importance of good isotopic records is clear from the intro, and stratigraphic noise is raised as a limiting issue (lines 39-40) but it feels a little disjointed.
2.3 3. Stacking
In this paper, you focus on stacking and comparison to the virtual firn cores. It would be nice to have a bit more background on the previous uses of firn core stacking, how it was developed and how obvious it is of a method (a very brief literature review found no recent papers), and it would be very beneficial to make it a key part of your research question. The definition of stacking is however a bit unclear, is it the application of the Palaeochrono software, which is described quite well? Or is it more complex than that? Reducing this ambiguity would be helpful for readers like ourselves who are knowledgeable on Antarctic climate, but not necessarily the specific isotopic methods applied in this paper. Similarly, I found the results rather interesting, but difficult to place in context. You find that stacking does improve things for 2/3 sites, is that consistent with other studies in the literature?
To help making your research question more relevant, it should be clear in the introduction how stacking was used before, and where stratigraphic noise was a limitation. Lines 454-463 clearly identify stratigraphic noise as a limiting issue, that one-core can be completely un-representative, and that stacking solves those issues. This would make a great ending to your introduction, leading directly into your research question. In addition, as mentioned in the previous paragraph, wind is a known issue (you cite Dallmayr et al. 2025b, line 414 on the effect of wind-induced high surface roughness). Even if this is not a completely novel idea, a very valuable finding would be whether stacking can improve the records in those locations. On a side note, we wonder if sharp changes in isotopic signal at the windy location could be used to reconstruct severe wind erosion events?
2.4 4. Models
It is not clear how you compare the 2 isotope enabled GCMs with the in situ data. They have two very different resolutions, with LMDZ-iso in particular being run rather coarsely. How did you sample or downscale this to the resolution of the in situ data? Does this have any impact on the results you find?
2.5 Extreme month ranking
The use of the top 12 warmest summers for each proxy is quite an arbitrary metric, quantitative alternatives exist. In Table 6, you compare the different proxies and look for agreement in predicting the top 12 warmest summers. This is a very arbitrary number, and a relatively arbitrary method too, where the 13th position is infinitely less important than the 12th, although there is no significant difference in those two rankings, but the difference between 1st and 12th is non-existent. The general intention of that table somewhat works, there is some agreement between different proxies in finding warmer summers. But we would encourage you to replace it or complement it with a more quantitative method, for example the Spearman’s rho, looking for resemblance in the ranking between the proxies.
3 Minor issues
- Lines 214-215: Is the Figure the focus of the sentence? Or is it the science that it carries? In other words, should it be ”X because of Y and Z (Figure A)” or ”Figure A shows that X,Y,Z”. We encourage you to make the science the focus of the sentence and not the figure.
- Figure 3: to a non-expert eye, there is very little resemblance between the δ18O and nssSO4 curves for different sites. The tie-points are used on Figure 2 between the repeated cores at the same site, and we understand that using such tie-points would have to be done with careful consideration for cores from different sites, but it would significantly help reading Figure 3. The significant increase in δ18O for D47 at 7400mm is not matched at Stop5 for example, is the Stop5 maximum offset left or right?
- Lines 400-401: you write that surface smoothing increases at windy sites, but your windy site has the least near-surface isotopic signal smoothing. You explain in the following sentence (line 403) that the multiple effects listed before compete, but it may be beneficial to be extra clear that the wind effects especially do not apply here.
- Lines 475-476: it is very common that ensemble means completely hides extremes. You may be able to connect your finding to other examples of extremes disappearing in simple ensemble means.
Citation: https://doi.org/10.5194/egusphere-2025-2863-RC2 - AC2: 'Reply on RC2', Titouan Tcheng, 12 Dec 2025
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 978 | 71 | 36 | 1,085 | 42 | 36 | 41 |
- HTML: 978
- PDF: 71
- XML: 36
- Total: 1,085
- Supplement: 42
- BibTeX: 36
- EndNote: 41
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The paper by Tcheng et al. is reporting an in-depth analysis of the isotopic composition of individual and stacked records from firn cores (a total of 9 cores) collected at three sites in a coastal area of Adelie Land in Antarctica. These sites are characterized by a relatively high snow accumulation rate allowing, through a continuous flow analysis method, to obtain high resolution d18O records along with nssSO4 profiles. These high-resolution records allow a very precise dating, at least for the two sites named Stop5 and Stop0. All the cores exhibit seasonal variations but at the D47 site the strong katabatic winds are strongly affecting the isotopic records. A stacking record, obtained for each of the three sites, permits to reduce the stratigraphic noise and is then compared to a Virtual Firn Core (VFC) calculated using two isotopic GCMs, ECHAM6 and LMDZ6, over a common period 1979-2016. The two less disturbed sites show consistent seasonal variability as the VFC records, allowing to observe common trends in d18O. This study, through a very detailed analysis if the high-res records, demonstrate the difficulty in obtaining a clear link between temperature and d18O, particularly in those sites that may be disturbed by redistribution processes by winds, as is the case for D47.
The paper is original and novel in the sense it uses a very detailed isotopic and chemical records for obtaining a climate reconstruction to be comparable to the climate data obtained from isotopic models. The data are well presented and discussed, and I have found the reading quite smooth although some parts could be reduced a little.
I recommend its publication after the authors have been considered to the following minor comments.
I have a general comment: why not attempting to calculate mean annual d18O values to be compared to VFC data and ERA5 precipitation weighted data? This could bring some information to eventual the d18O/T sensitivity at interannual scale. Has this been done or at least checked?
The title: I would suggest changing “covering the last 40 years” since it is misleading, they are not the last 40 years…
Page 2, line 1: “ … closely linked to …. ice mass loss …”
Page 2, line29: may you specify “long-term” how many years it is?
Page 7, line 147: may you add which is the period covered?
Page 10, line 232: all the nssSO4 profiles feature common patterns. However, If I am looking at figure 2 the sulphate records at Stop5 are not so similar….
Page 10, lines 248-249: the Pinatubo signal: I agree with the authors that it is quite ambiguous, but if I am looking at the figure S5 the signal is well evident at Stop0 (stacked profile).
Page 13, Table 3 and also in the text to be specified if the data regarding the snow accumulation from stake is referring to a mean value obtained from a stake farm or is one single stake value and which density values have been used and to which year is referred.
Page 18, figure 5: If I am looking at the records from Stop5 I found very different trends and patterns between staked records and VFC ones, around 1990 +/-3 years. May you comment on this? Perhaps are you referring at this point at page 19 lines 392-393? Add in the text.
Page 20, lines 411-413: may you explain better for the reader how the local roughness is calculated. There is an explanation in the table caption, but I would suggest you move it to the main text.
Page 20, lines 427-429: see my previous comment above (figure5).
Please, change in all the figures (text and supplementary) the X axis title and labels from mm to m.
Please check all the delta symbols in the text.
Please check all the table format.
In the Supplementary: the caption of figure S5 is referring to which core site? Stop0?
Is the code for calculating the VFC records free available?