General comments

This paper reports the last ten years of precipitation isotope observations at Concordia Station, Antarctica. The data set is extremely valuable and should be published. The results are mainly based on correlation analysis and numerous comparisons with reanalysis data and isotopic models. On the other hand, I felt that the scientific novelty is unclear. Although the text is clearly written, the results of the correlation analysis are described in detail, and what the results mean is not discussed well. It should be clarified: "What do we learn from this observation?" I think seven figures in the main paper, 34 supplementary figures, and five tables are too much. The manuscript should be revised to reflect the following comments before publication.

Major comments

(1) L.110-115: I think the issue of the discrepancy between the three years of data (Stenni et al., 2016) and the Antarctic spatial slope is not fully discussed in the main text. The temporal d18O/T slope in this study is smaller than the spatial slope. The cause of this difference and its implications (especially its impact and implications on the recent controversy about temperature reconstruction of the Antarctic ice cores, e.g., Buizert et al., Science, 2021) need to be discussed.

(2) L155-160: Most precipitation isotopic ratios in this study are lower than that of SLAP. In other words, it has been extrapolated. Maybe future studies will be conducted to evaluate the effects of extrapolation, so please describe how many WSs were used (2 or more) and the isotope ratio of each WS.

(3) Fig. 1: The authors explained that the isotope ratios were measured only when there was enough snowfall to collect data. Specifically, what is the lowest accumulation (mm/day or more)? Based on Fig. 1, the number of isotope data varied considerably from year to year. In particular, 2010 and 2017 are large, and 2015 and 2017 are also somewhat large. Does this mean that the number of days of snowfall events varies significantly from year to year? Is this consistent with the number of days that snow was collected and the amount of snowfall in the reanalysis data?

(4) Regarding the discussion on the relationship between d-excess and d18O: Since the slope of D-18O is 6.6 (fig. 3), it is evident by simple mathematics that calculating d-excess with a slope of 8 would be inversely correlated with d18O (Figs. 1 and 5 and others). Relatedly, in Fig. 6, there is a large discrepancy between the model and observed d-excess, but this can also be understood because the slope of the model's D-18O is close to 8 (7.66). However, the authors briefly mention the possible decrease in MWL slope (L. 412-413). I think it is necessary to discuss this point more, as the authors repeatedly showed d-excess vs d18O relationship. Specifically, the logarithmic definition of d-excess (Uemura et al., Climate of the Past, 2012; Markle and Steig, Climate of the Past, 2022) has been proposed to alleviate this problem. At least the impact of the logarithmic definition on these results could be added to the discussion.

(5) If some of the Supplementary figures are to be deleted, the scatterplots of correlations (Fig. S16-23) would be a candidate.

Technical comments

L63 “the emprical d-T relathionship valid” -> “…is valid…”

L75: I think it is better to use some publication (white paper or perspective) instead of a URL as a citation.

L246 “Figure SI2” -> « Figure S2 ^»»

L260 ”may had led”-> “may have led”

L411 “explained with” -> “explained by”

L569  “worth to mention” -> “worth mentioning”

References:

Buizert et al., Science, 2021, https://doi.org/10.1126/science.abd2897

Markle and Steig, Climate of the Past, 2022, https://doi.org/10.5194/cp-18-1321-2022

Uemura et al., Climate of the Past, 2012, https://doi.org/10.5194/cp-8-1109-2012

General comments:

This manuscript presents a valuable dataset of isotopic compositions of daily precipitation at Concordia station for ten years from 2008 to 2017. The authors did detailed analysis of meteorological conditions, the water isotope data, and model-data comparisons. The dataset is useful to evaluate model performance, investigate climate controls on water isotopes in Antarctic precipitation, and quantify impacts of post-depositional processes on ice core records. However, the structure of the content can be more concise, and the added scientific values of this manuscript, especially compared to Stenni et al. (2016) that presents the first three-year data, are not very clear. Therefore, major revision is suggested before publication.

Major comments:

1, Titles of many subsections are too short to be informative. E.g. “2.2 Analytical ”, “3.5 Correlations”.

2, The authors stressed in many place that the dataset is ‘unprecedented’, “unique”, or “of extreme importance”. However, it is not clear what additional values do this manuscript bring to the research community compared to Stenni et al. (2016). For example, when this dataset is applied to evaluate model performance in Section 4, does it add more confidence in identifying model bias? Does it help to identify a direction to improve the model simulations or for further studies?

3, The authors presented analysis on both weighted and unweighted monthly or annual values. Can the authors elaborate which one is more suitable in which conditions? And why do the authors weight the variables using total precipitation from ERA5 (Line 296), rather than observed precipitation amount? If there is no corresponding precipitation in ERA5, how do the authors do the weighting?

4, The authors did extensive correlation analysis between different variables. However, different variables might be correlated because of their common correlation to another variable. For example, the correlation between deuterium excess and temperature discussed from line 388 might be related to their correlation with d18O. It a partial correlation analysis can confirm this point, what is the point of regression analysis between deuterium excess and temperature?

5, The structure and content of the conclusion section can be more concise. For example, the sentences starting from line 527 and line 533 can be shortened,  and the sentence starting from line 538 can be removed.

Minor comments:

Line 22: “AWS”, introduce the full form.

Line 24: 3.5 ‰/°C for δD/TAWS.

Line 46: “although occurring progressively over successive condensation events between the initial evaporation and the final deposition areas. ” This formulation is very strange. What is occurring?

Line 47: “Consequently, the different sensitivity of the empirical δ-T relationship in East Antarctic ice is generally poorly constrained with respect to other regions”. It is not clear what this sentence wants to express.

Line 143: “samples in the sealed bags”?

Line 268: “daily pattern” not “diel pattern”

Line 284: “during the days with collected samples” and “during the sampling days” are duplicated.

Line 485: Are the scatter plots in Fig.6 based on daily values? If the precipitation occurs on different days in simulations and observations, how are they matched together?

Line 537: “this could explain the occurrence of negative d-excess values in this season.” Do you have any supporting evidence for this statement?

Line 545: “mean monthly averages”.

Line 550: Is this based on annual data? Are ten annual data points enough to evaluate long-term trends?

Line 554: How can sublimation leads to lower slope in LMWL?

Line 559: “The high d-excess values found in winter, as well as its seasonal amplitude, are mostly due to the extremely low condensation temperature rather than to changes in moisture origin.” Do you have supporting evidence for this conclusion statement?