the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 14 Apr 2025
Time series of the summertime diurnal variability in the atmospheric water vapour isotopic composition at Concordia station, East Antarctica
Abstract. Measurements of stable water isotopes in the atmospheric water vapour can be used to better understand the physical processes of the atmospheric water cycle. In polar regions, it is a key parameter to understand the link between the precipitation and snow isotopic compositions and interpret isotope climate records from ice cores. In this study we present a novel 2.5-month record of the atmospheric water vapour isotopic composition during the austral summer 2023–2024 at Concordia Station (East Antarctica), from two independently calibrated laser spectrometers (CRDS and OF-CEAS measurement techniques) which are optimised to measure in low humidity environments. We show that both instruments accurately measure the summertime diurnal variability in the water vapour 𝛿18O, 𝛿D, and d-excess when the water vapour mixing ratio is higher than 200 ppmv. We compare these measurements to the outputs of the isotope-enabled atmospheric general circulation model LMDZ6-iso and show that the model exhibits biases in both the mean water vapour isotopic composition and the amplitude of the diurnal cycle, consistent with previous studies. Hence, this study provides a novel dataset of the atmospheric water vapour isotopic composition on the Antarctic Plateau, which can be used to evaluate isotope-enabled atmospheric general circulation models. The dataset is available on the public repository Zenodo (https://doi.org/10.5281/zenodo.14569655, Landais et al., 2024b).
- Preprint
(5137 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on essd-2025-35', Anonymous Referee #1, 23 May 2025
The article presents a temporal series of the isotopic composition of water vapor during the austral summer 2023/2024 at Dome C, East Antarctica. The data have been obtained using two different laser spectrometers: a Picarro Cavity Ring-Down Spectrometer (CRDS) and an AP2E Optical-Feedback Cavity Enhanced Absorption Spectrometer (OF-CEAS). According to Lauwers et al. (2024), the low-humidity OF-CEAS analyzer, which was supposedly deployed in the field for this paper, should perform better than the CRDS at low humidity, but in this study it seems to give worse results.
A comparison between the measured and the LMDZ6-iso simulated isotopic composition of water vapor has also been carried out in this paper. The modeled and measured data show a good agreement for humidity and also show an overall agreement for d18O, dD and, to a certain extent, for deuterium excess, but LMDZ6-iso, while capturing the general variability, fails to obtain the correct isotopic values as well as the magnitude of the observed diurnal cycles.
Line 15 the measurement period is more than 3 months rather than 2.5 months
Line 46-47 change “such as encountered on the East Antarctic Plateau, is a technical challenge since most laser spectrometers are designed for measuring accurately within a range of humidities between 5,000 and 30,000 ppmv” with “such as those encountered on the East Antarctic Plateau, presents a technical challenge, as most laser spectrometers are designed for measuring accurately within a range of humidities between 5,000 and 30,000 ppmv”
Line 57-58 how is the commercially available Picarro laser spectrometer adapted for low humidity measurements?
Line 63 same as line 15
Line 91-91 I think it would be better to specify the Picarro model (L2130-i? L2140-i?) instead of the identifier of your instrument. The same for the AP2E instrument: is it a ProCeas?
Line 139 change “closest” with “close”
Line 140 Is it possible for the deuterium excess to be +90‰ (based on a d18O=-80‰ and a dD=-550‰)?
Line 160-161 Although you give an explanation in line 180-183, I am still wondering why you used two laboratory standards mostly outside the range of the measured water vapor isotopic values. Couldn’t you use, for example, the VSAEL standard with the FP5 standard?
Line 201-202 “i” is not subscript
Line 273 What model is the Picarro HIDS2319?
Line 284-285 change “Below 500 ppmv, both 𝛿18O and 𝛿D show a divergence with decreasing humidity levels, in the opposite direction as for both Picarro analysers” with “Below 500 ppmv, both 𝛿18O and 𝛿D diverge as humidity levels decrease, but in the opposite direction observed in both Picarro analysers”
Figure 4 d18Ohumcorr-d18Ohumcorr [‰] and dDhumcorr-dDhumcorr [‰] for the y axis is not very clear
Line 374-375 and 381-382 what does it mean that “both analysers capture the linearity between the true 𝛿18O value of the two laboratory standards”? A line passing through two point is always a linear equation
Figure 7 I can’t find “AP2E, raw” in panel a), which should be a blue dashed line; is it present in the graph? Was air temperature not available in the first and the last period of measurements? Where does the temperature come from? Is it AWS temperature or modeled temperature? You should specific it in the main text and in the figure caption
Line 455-456 there is a divergence between the two instruments between mid-February and mid-March. I know it’s due to the very low humidity which makes it hard for the laser spectrometers to correctly measure the isotopic composition, but how do you explain the different behavior of the Picarro and the AP2E laser spectrometers and which one is more reliable? I think this is an important point if you wish to measure the isotopic composition of water vapor in other seasons
Line 483 change “to correctly capture” with “in correctly capturing”
Line 485-486 change “Because of the large correction linked to the humidity-dependence on the 𝛿18O signal, even the 𝛿18O could be challenged” with “Due to the significant correction associated with the humidity dependence of the 𝛿18O signal, even the 𝛿18O measurement could be questioned”
Line 486-487 why you stopped the comparison at mid-February when you have data up to mid-March 2024? Is it because of the unreliability of the isotope data due to the very low humidity? I think it should be explained in the text
Line 490 change “show” with “shows”
Line 491 change “including for the amplitude of the observed diurnal cycle” with “including in terms of the amplitude of the observed diurnal cycle”
Line 495-496 change “the modelled 𝛿18O shows an overall positive bias during the period December to mid-February compared to the observations” with “the modelled 𝛿18O shows an overall positive bias during the entire period compared to the observations”
Line 538 change “during the summertime” with “during summertime”
Line 547-549 The higher deuterium excess in the measurements with respect to LMDZ6-iso could also be explained by sublimation
Line 556-557 change “might not be well representing the in-situ conditions” with “they may not accurately represent the in-situ conditions”
Line 570-571 change “Combining the observations of the water vapour isotopic composition” with “Combining observations of water vapour isotopic composition”
Lauwers, T., Fourré, E., Jossoud, O., Romanini, D., Prié, F., Nitti, G., Casado, M., Jaulin, K., Miltner, M., Farradèche, M., Masson-Delmotte, V., and Landais, A.: OF-CEAS laser spectroscopy to measure water isotopes in dry environments: example of application in Antarctica, https://doi.org/10.5194/egusphere-2024-2149, 15 August 2024.
Citation: https://doi.org/10.5194/essd-2025-35-RC1 -
RC2: 'Comment on essd-2025-35', Anonymous Referee #2, 08 Jun 2025
Review for Ollivier et al. (2025)
This manuscript documents a dataset of atmospheric water vapour isotope compositions observed during 6 Dec 2023 to 14 Feb 2024 at Concordia station over East Antarctica. The dataset represents a valuable effort of observations which require novel techniques to overcome harsh conditions at Dome C. The dataset is useful for evaluating isotope-enabled atmospheric models and investigating post-depositional modification of snow isotope compositions. The manuscript is well structured and clearly written. However, some improvements can be expected, as detailed below.
1, Line 1: Suggest removing “diurnal variability in the”, as the data are not “time series of variability”.
2, Line 13: “it is a key parameter”. What is a key parameter to “interpret isotope climate records from ice cores”? Isotope measurements from the ice cores? Please rephrase to be clearer.
3, Line 15: could you provide the exact start and end date of the record here?
4, Line 16: Please provide the full name of “CRDS and OF-CEAS” before using the abbreviation.
5, Line 19: “higher than 200 ppmv”. Based on Fig. 3, it seems the two laser spectrometers agree with each other up to lower limits in mixing ratios that are different for d18O and dD. Could you please provide quantitative estimates of the lower limits in mixing ratios that the two instruments agree? Is this agreement your criterion to decide whether the measurements are accurate? Could you also comment on the current valid humidity ranges for d18O and dD measurements separately?
6, Line 19: “output” is uncountable.
7, Line 21: “Hence…” Could you also mention the comparison of humidity, and then the added value of water isotopes for comparison? Based on the comparison, could you please briefly comment potential pathways to improve the modelled hydrological cycle?
8, Line 29: The deuterium excess is not defined to capture kinetic fractionation, or at least it is not only affected by kinetic fractionation. Even equilibrium fractionation can result in a slope other than 8 and affect the value of deuterium excess. You can refer to this paper: Dütsch, M., Pfahl, S., & Sodemann, H.(2017). The impact of nonequilibriumand equilibrium fractionation on twodifferent deuterium excess definitions.Journal of Geophysical Research:Atmospheres, 122, 12,732–12,746 https://doi.org/10.1002/2017JD027085
9, Line 35: As far as I understand, these studies show that snow isotopes can be modified post deposition, but how do these post-depositional processes affect the empirical relationships between isotopes and climate conditions? The empirical relationships using surface snow isotopes indeed already take the post-depositional processes into account, right?
10, Line 53: What is the uncertainty when the humidity is above 200 ppmv? And how much is it reduced in this dataset?
11, Line 60: “which permits” => "which permit". Could you please check through the manuscript to ensure no such mistakes? It can be easily done with tools like Grammarly.
12, Line 62: Just curious whether it is because people never tried or they failed to do it successfully.
13, Line 71: “The instrumental set-up … is installed”? Or just “The instrument”?
14, Line 72: “(+10°C)” => at a temperature of 10°C. Otherwise, it seems to be 10°C warmer than unheated ground.
15, Line 75: 1/4", is this diameter or radius in what unit?
16, Line 87: Could you please provide reference/website/product description source to the CRDS and OF-CEAS techniques for interested readers? From Fig. 3 it seems that at low humidity levels, AP2E measurements are closer (further) to the reference for d18O (dD) than HIDS2308. Do the authors have an explanation for the differences based on technique attributes?
17, Line 94: remove “led to”
18, Line 115: Which quantity do you use to recalculate relative humidity relative to ice, and how? Do you need temperature information and where is it from?
19, Line 121: Based on Fig. 2, it seems HMP155 measurements are larger than other two instruments on average, inferred based on the intercept. Could you please provide root mean squared error (RMSE), mean error (ME), and mean absolute error (MAE)? Do you think the systematic differences may result from different locations, heights, and inlets (one heated and how about the another one)? How much does these factors matter and how much does it matter if no humidity calibration is conducted? What is the difference between raw humidity measurements by the two instruments (RMSE, ME, MAE)?
20, Line 139: “closest” among what?
21, Line 142: How many calibrations did you conduct? Do you change the humidity level from 1100 to 50 ppmv for each calibration? Then how long does each humidity level last for the calibration and do you (need to) account for memory effects during consecutive humidity levels? In Fig. 3, it seems the calibration humidity is at some random levels and is different for two instruments, why?
22, Line 145: what does a reference humidity mean?
23, Line 176: “against VSMOW-SLAP”
24, Line 202: ‘any isotope species’ => ‘each isotope species’
25, Line 211: Do you need to take account of the coefficient 8 while estimating uncertainties in dxs?
26, Line 216: “referred to”
27, Line 221: provide the full name of LMDZ when you first mention it.
28, Line 235: how is the surface snow isotope composition over Antarctic ice sheet configured in LMDZ6-iso?
29, Line 248: As mentioned before, could you please provide RMSE, ME, and MAE between the independent analyzer and two instruments, as well as between the two instruments, ideally direct on the figure? There is an obvious cluster of outliers where two instruments indicate humidity larger than 200 ppmv and the independent humidity sensor gives values around 100 ppmv. Is this associated with large temperature deviations or different wind conditions? I would be interested to see a panel c where you plot Picarro vs. AP2E, which measure apple and apple. I am quite curious why the independent sensor is considered as ground-truth: is the technique more reliable, or are humidity sensors in Picarro and AP2E not optimized for humidity measurements, or humidity measurements may be biased by the heated sample line and the inlet?
30, Line 271: ‘humidity-isotope response curves’.
31, Line 278: “a much weaker divergence in 𝛿D”. Could you briefly discuss what leads to such improvements? And Line 280, it is surprising to me to see such big differences.
32: Line 290: Could you discuss why there is a tendency for Picarro to overestimate and AP2E to underestimate isotope ratios at low humidity levels? Could you quantitatively determine the applicable range of AP2E and Picarro for dD and d18O separately? So it can serve as an objective baseline for comparisons in future works.
33: Line 314: Could you please confirm the Eq. 2? The correction seems linearly depend on measured humidity.
34: Line 320: Could you simplify the legend and the figure? The hatch is nearly invisible when printed, do you really need to show the ±2 sigma? You can use one color for each instrument and one shape for each lab standard. Currently it is very confusing.
35, Line 330: What if you plot the humidity generated by LHLG against the deviations in d18O? Do you have records of room temperature?
36, Line 342: Do you mean a small spectral window? Could you provide the frequency range for d18O and dD? Could it also result in the different performance we saw in Fig 3? While for many analyses, people seem to pick either dD or d18O arbitrarily, here it seems that for observations, the relative uncertainty is smaller for dD and the application range is larger for dD than d18O.
37, Line 361: Could you provide RMSE here? How can you confirm that the vapour flux generated by LHLG has an isotope ratio as per the lab standard? In the legend, y=a±bx + c±d, not y=ax±b + c±d.
38, Line 382: Is it because AP2E is over-tuned for low dD values?
39, Line 410: Could you quantify the completeness of the dataset over a specific period?
40, Line 413: Could you provide the root mean squared difference, mean difference, R squared values, and mean absolute difference between raw and calibrated humidity, dD, d18O for each analyser, as well as the statistics between two analysers? The differences could be quite large if you include periods of low humidity. Could you determine a threshold to do the calculation and also as a recommendation on which range is more reliable (e.g. for model-data comparison)?
41, Line 487: Could you please provide statistics as mentioned before? Is it possible to compare temperature and provide it in the dataset as well?
42, Line 522: The anti-phase relationship may be an artefact due to the linear definition of dxs (see previous reference Dütsch et al. 2017).
43, Line 547: what are the isotope compositions of surface snow at the station in the model (and in reality)? How does it compare to vapour isotopes and if sublimation occurs with equilibrium fractionation, is the sublimation flux more depleted or enriched than the vapour fluxes?
Citation: https://doi.org/10.5194/essd-2025-35-RC2
Data sets
Isotopic composition of the atmospheric water vapor during the austral summer at Concordia Station, Dome C, East Antarctica (December 2023 – February 2024) Amaëlle Landais et al. https://doi.org/10.5281/zenodo.14569655
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 264 | 36 | 12 | 312 | 9 | 11 |
- HTML: 264
- PDF: 36
- XML: 12
- Total: 312
- BibTeX: 9
- EndNote: 11
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1