The paper presents a dataset describing the stable water isotope composition of approximately 3800 snow samples. Samples were collected at the deep drilling site of the East Greenland Ice Core Project between 14 May and 3 August 2018 along a 39 m long transect.

The dataset looks very interesting and unique in terms of detailed description, as well as for the proposed impact.

Three aspects related to the ESSD policy and requirements are major issues:

1) If this is a data description paper, ESSD states: "Although examples of data outcomes may prove necessary to demonstrate data quality, extensive interpretations of data – i.e. detailed analysis as an author might report in a research article – remain outside the scope of this data journal."

The paper matches the requested structure, but the section 2.5 introduces additional datasets complementary and outside to the presented dataset. It is ok to discuss them in the discussion section, mentioning them as references of published papers as authors have done for other datasets. Only the "high-resolution near-daily digital elevation models (DEMs) with a horizontal resolution of 1 cm and a sufficient accuracy for the purpose of this study", even published  in Zuhr et al., 2021. The other data are not included in the presented dataset, available in PANGAEA. 

 

2) ESSD data descriptions "should instead highlight and emphasize the quality, usability, and accessibility of the dataset, database, or other data product and should describe extensive carefully prepared metadata and file structures at the data repository."

In this case I looked at data on the PANGAEA repository and the usability is limited by the position description. Authors refer to a position number, a sampling distance and a sampling depth. They refer to a meter or submeter precision but no coordinates are available, even in the manuscript shows a coarse latitude and longitude positioning (only degree and minute for one location). The sampling depth is presented in the dataset in three different ways: as sampled depth interval, as average sample depth, and as a sample depth derived from digital elevation models. The first two parameters are redundant, thus the average depth from the surface and the sampled thickness complete the required metadata. The third parameter, expressed as percentage should be better described even in the paper. 

3) ESSD author obligations state "Fragmentation of research papers should be avoided. A scientist who has done extensive work on a system or group of related systems should organize publication so that each paper gives a complete account of a particular aspect of the general study. It is inappropriate for an author to submit manuscripts describing essentially the same research to more than one journal of primary publication.Fragmentation of research papers should be avoided. A scientist who has done extensive work on a system or group of related systems should organize publication so that each paper gives a complete account of a particular aspect of the general study."

Authors state that "high-resolution near-daily digital elevation models (DEMs) with a horizontal resolution of 1 cm and a sufficient accuracy for the purpose of this study (RMSE of 1.3 cm (Zuhr et al., 2021))." Are the same data included in the DEM derived parameter associated with the presented dataset? The interoperability between the two datasets should be the best practice.

Minor comments:

line 65 Whirl-Paks is a registered trademark, could it better to state the plastic bag in another way (high-purity sampling bag...?)

Figure 2 be consistent with Table 1: sampling modes "Normal" and "n" should be named in the same way

Table 1 declare the year in the caption.

Figure 3 The top panel with Latent Heat Flux should have the y axis without a colour since it refers both to daily and hourly values in the plot.

Check references Zuhr et al 2021 and 2023 are not differentiated even they are listed both in double.

The article presents an interesting study on water stable isotope variations in the upper snowpack nearby the EastGRIP drilling site in northeast Greenland. About 3800 snow samples were obtained during summer 2018 and analyzed for d18O and dD.

Other similar works were produced during last years on surface snow isotopic variability at this site.

The premise is that such a high temporal and spatial resolution in surface snow sampling will allow a better understanding of post-depositional processes.

Two high intensity period were chosen for a more frequent (daily) sampling, characterized also by a higher depth resolution of the samples, but in my opinion this, instead of providing useful data, creates a bit of confusion when compared to the rest of the dataset.

The difference in variations between different depths is higher for deuterium excess, where the upper layer shows a very different variability compared to the layers below, while for d18O the differences between those layers are not so pronounced.

The second sampling scheme, with less spatial positions but a more frequent sampling time and a higher vertical resolution, suggests that quick d-excess variations in surface snow are not captured on a 3-day sampling.

Line 2-4 I would use climatic parameters more than environmental

Line 28 change “as well as the stratigraphic features at a specific site” with “as well as by the stratigraphic features at a specific site”

Line 30-34 there is a bit of confusion in this part and I suggest a rephrase: the isotopic signal is imprinted during the formation of precipitation, after snow is deposited this signal can be significantly modified by post-depositional processes occurring through exchanges between snow and atmosphere or within the snow column (isotopic diffusion)

Line 42 change “contribute to pinning down the processes” with “has contributed in pinning down the processes”

Line 43 change “and which need a better quantification” with “and need a better quantification”

Line 63-64 moving the sampling position each time was certainly required in order not to resample the same point, but introduce a further variability which might not be solely due to the passing of time but also on the spatial variability given by sampling a different point

Line 68-70 I would rephrase as follow “The HIT 1 period, from 8 to 14 June 2018, followed a major snowfall event to study the signal imprint from the surface penetrating into the upper snowpack. During that period samples were collected only at ten sampling positions due to time constraints.”

Line 71-72 change “The second high-resolution sampling period (HIT 2) was from 24 to 30 July 2018 covering a snowfall-free period” with “The second high-resolution sampling period (HIT 2) was from 24 to 30 July 2018, covering a snowfall-free period”

Line 73 change “with the depth intervals 0 - 0.5 cm, …” with “with the following depth intervals: 0 – 0.5 cm, …”

Line 111 a root mean square deviation of 1.45‰ for d18O seems a bit too high

Figure 5 It seems that snow d18O changes more in the 1-4 and 4-10 cm layers than in the upper stratum: while the first centimeter of snow is almost entirely characterized by values equal or higher than -37‰, the lower layers are affected by a higher variability; how do you explain it? The sampling at 12-13 m of distance starts with d18O values well below -40‰ and ends up with values above -30‰ in both 1-4 and 4-10 cm, while it shows less variation in the first layer of snow; is it possible that the snowfall event prior to the HIT 1 period changed snow d18O for the 0-1 and 1-4 layers, but it seems to affect less the 4-10 cm layer. However, I do not observe significant d18O variations following this intense snowfall. HIT 2 lack of precipitation seems to affect more the surface layer than the other two: in the first centimeter there is an overall increase in d18O in all sampling points of this period. The last days of sampling are characterized by a minor d18O decrease in all sampling points in the 0-1 cm layer, while d18O values from 1-4 and 4-10 cm show an overall increase; how do you explain it?

Line 159 It is true when using a mean spatial d18O value, but when considering the single sampling points and according to the considerations I made above, I tend to disagree with this statement: at least for d18O, we see more (and faster) variability in the lower layers

Figure 6 could you use a lighter tone for temperature? It makes it difficult to correctly see the d18O dots and lines.

Figure 7 here it is more evident how d-excess is more sensitive to snow-atmosphere interaction and changes more frequently and more significantly in the upper layer. The period at the end of June is characterized by very d-excess values above 20, but only in the 0-1 cm stratum; these high values seem to have a low duration and decrease within few days, maybe following a snowfall event

Figure 8 for d-excess the spatial variability seems less important: the mean spatial values provide similar information to the single sampling points values presented in figure 7; this might suggest that the d-excess is more susceptible to atmospheric conditions, which are the same for the entire sampling place, than to snowfall accumulation and wind redistribution inhomogeneity

Figure 9 and 10 you have to specify in figure captions that you are showing mean spatial values

Line 226-228 you suggest that higher variability on the second sampling scheme is due to the atmosphere-snow exchange or wind distribution; why should that not be recorded when the snow is exposed for a longer period of time?