General comment

Stable water isotopes (δ¹⁸O) measure in polar regions like Antarctica are traditionally used to reconstruct past surface air temperature (SAT). However, this relationship is influenced by many parameters like surface elevation, air mass sources or sea surface conditions. Moreover, this relationship varies spatially and over time, and has not been investigates using historical simulations (1850 onward), yet. To tackle this issue, Goursaud Oger et al. investigated the SAT – δ¹⁸O relationship during the historical period in Antarctica using an ensemble of historical climate simulations (1851-2004) performed with the isotope-enabled HadCM3 general circulation model. The found strong SAT and precipitation temporal trends during this period, but only weak trends for δ¹⁸O, meaning no significant relationship between SAT and δ¹⁸O over one third of Antarctica. They conclude that the decoupling between δ¹⁸O and SAT occurs primarily because of the impact of autumnal sea ice loss during the simulated warming.

The analyses and idea are simple (in a good way) and well written, making the article easy to read and follow. To better quantify the influence of parameters other than temperature on stable water isotopic composition of ice in Antarctica is an important topic for paleoclimate community, and this article represents an additional valuable contribution to that discussion. This article is worthy for publication in CP after addressing the minor points detailed below, including more in-depth analyses of the circulation of air and moisture masses induced by changes in sea ice.

Major comments (but minor revisions)

• In terms of analyses, the warm – cold anomalies and the seasonal effects are very interesting and well investigated. On the other hand, in the abstract and the conclusion, the authors talk about the involved variations in moisture transport and air mass intrusions due to sea ice transport. I would expect deeper analyses on this aspect, and not just some general statements in the conclusion section. For example, the winds patterns are shown in Figure 4 but not cited in the text.
• In section 2.1, the model and the setup of simulations should described a little bit more. Six historical simulations have been performed. How the authors make these simulations a little different from each other? With different initial conditions? By changing the value of a parameter? It should be stated in this section. For the HadCM3 model, it should be stated from this section that this is a atmosphere-ocean coupled model. Also, parameterization relative to isotopes with sea ice should be described? How sea-ice - atmosphere exchanges are taken into account in the model for the isotopes? How ocean free vs. sea ice covered areas considered for the calculation of isotope concentration in surface water vapor? Is there any sublimation of snow on sea ice? With a fractionation effect? Or is there nothing specific coded for the isotope sides, meaning that isotopes are influenced by mainly by changes in air mass and moisture transport, only.
• The figures need to be improved. The font size of the titles, letters, equations, and labels are really too small in all the figures. The figure 2 needs big improvement. See for example the figures 2 and 5 in the recent paper from Servettaz et al. (https://doi.org/10.5194/tc-17-5373-2023). Moreover, it would help to note somewhere the name of the different regions of Antarctica, and to which colors they do correspond. In the current state of the paper, it is hard to follow which region is where for people not familiar with Antarctica geography.

Minor technical comments:

• Lines 5, 30, 58, 68, and others: I would replace “water stable isotopes” by “stable water isotopes”.
• 3^rd paragraph of the introduction: Cite also Servettaz et al., TC, 2023 (https://doi.org/10.5194/tc-17-5373-2023).
• Lines 46-47: “Antarctica2k” without space like in Stenni et al. (2017). Define A2k for Antarctica2k here, too.
• Lines 50-53: please add the study from Buizert et al., Science, 2021 (https://doi.org/10.1126/science.abd2897) et Cauquoin et al. CP, 2023 (https://doi.org/10.5194/cp-19-1275-2023), except if you want to focus on war m period only (in that case, some references like Werner et al. (2018) should be removed). If you choose the latter option, please precise at the beginning of the paragraph that you talk here only about warm periods.
• Line 92: see my comment about the definition of A2k above.
• Line 101: please describe briefly the ECHAM5-wiso simulation (AGCM at T106 resolution nudged to ERA40/ERA-Interim). And cite Werner et al. (2011, https://doi.org/10.1029/2011JD015681).
• Section 2.2: see 2^nd major comment.
• Line 115: which reanalysis data?
• Line 158: Medley and Thomas (2019)
• Section 3.2.2 and Figure 2: see my third major comment. For readers not used to Antarctica geography, it’s very hard to follow.
• Lines 170-171: sounds strange (strongest for one place and highest for another place)
• Line 179: Rephrase the beginning of the sentence to avoid the double use of “values”.
• Line 191: ECHAM5-wiso
• Line 2023 eastern Plateau or east Antarctic Plateau.
• Lines 226 and 227: remove the brackets for the two references.
• Lines 228-231: maybe the atmosphere-ocean coupling, including sublimation of snow on sea ice?
• Line 236: add a reference to Figure 5.
• Line 244: the largest (no capital T).
• Lines 244-246: say explicitly the months.
• Lines 253: there is no figure 5h.
• Line 278: It’s not ERA4 but ERA40 and ERA-Interim.
• Lines 279-282: see first major comment. I think more analyses of wind patterns (for example) to demonstrate these conclusions would improve the paper.
• Line 294: See second major comment. These specificities of HadCM3 should be said before in the paper.
• Line 295: higher (no capital H).
• Figure 1: increase the font size (for all figures), including for the equations, make the average and linear regression curves thicker, use something like a_all, a_recent, r_all, r_recent to differentiate the two equations in each plot.
• Figure 2: see 3^rd major comment.
• Figure 4: state in the legend that these anomalies relative to annual mean. The wind fields are visible. Put more space between the arrows and draw them thicker and bigger. There is a typo “>” in the last line of the legend.

Please see the attachment.