The authors introduce 6 years of meteorological, soil and, snow data collected at a tundra site in the Canadian high Arctic. The observed polygonal tundra environment located on Bylot Island, Nunavut is described as well-drained, void of standing vegetation, and typical of permafrost landscape. A thorough review of instrumentation, datasets, and validation at the site are presented prior to an assessment of data quality and correction of the temporal data where necessary. The authors highlight a lack of data available to drive snow physics models in herb (graminoid) tundra environments as motivation. Thin Arctic snowpack common to this area, and greater Arctic regions, consist of contrasting high-density winds slab and lower-density depth hoar elements. The predominate drivers of this composition, wind compaction and upward flux of water vapour, are clearly identified by the authors as a gap in the current generation snow physics models requiring new datasets to evaluate against and improve from.

The described work is a valuable contribution and should be published in ESSD. Public release of this dataset and its associated validation are a service to the snow, climate and meteorological communities where significant effort has been required to initiate and sustain these remote measurements. I have noted instances below where I felt minor clarifications were needed on corrections applied but in general, I felt the dataset was well documented. Where internet data resources were accessed, for example the ECCC station data, date of access should be noted in the references in case of future revision. Useful lessons and anecdotes on sensor deployment, calibration and adaptation punctuate the paper and I felt it could be more impactful if these were summarized in a concise discussion near the end. A discrete guide on how to setup an Arctic site is clearly beyond scope, but some of the more significant choices, for example the use of ERA5 over the CNR4 data, warrant clear recommendations. I would hope such as summary would be a catalyst for development of similar sites, amplifying the efforts here.

Minor comments with line numbers provided:

Lines 47, 68,73 and others – Where the authors are part of the narrative should the citations not be in-line instead of in parenthesis?

Line 149: What does significant mean in the contact of the spatial variability described here? It would be helpful to place this in the context of another study or even a metric of the locally observed variability to constrain the use of the word significant.

Line 165 Superscript is missing for ^3

Line 167: Add the total number of pits completed in the sentence or point towards a table to summarize.

Line 181: It is not clear if the correction applied was a constant or varied over time.

Line 198: This internet resource needs to be cited with the data of access.

Line 213: The gas station equation does not have its variables introduced in text with units.

Line 271 The Pond Inlet airport station has an observer. Were the observer comments ever used to evaluate selected precipitation phase threshold at +0.05C?

Line 279: The use of time lapses appears to be key here. Was this approach contrasted against an automated or empirical approach?

Line 307: It was previous stated that there was no small-scale relief at the tundra site in the context of hummock-tussock formations. Is this related to the wedge structures?

Line 395: This reference appears to be incomplete. Please expand on what type of correction was applied.

* general comments

The authors describe and provide an extensive set of in-situ meteorological, snow and soil data to force and evaluate snow schemes in a tundra environment. Such an extensive set of forcing and evaluation data, especially for snow, is unprecedented in the litterature in this type of environment. As the most complex snow models to date fail to represent tundra snow charateristics, despite their huge significance at the global scale and w/r to global warming, this paper and dataset are in my opinion an important contribution to snow science.

The site and the dataset are well described, the quality assessment of the data is thorough, and the corrections performed to the data are well explained and justified (see just some minor clarification needs in the specific comments).

I recommand the paper for publication in ESSD after these minor comments have been addressed.

* specific comments

- L51-52: "The explanation proposed (Domine et al., 2019;Domine et al., 2016b) is that Crocus and SNOWPACK were designed primarily for avalanche forecasting in the Alps, i.e. for mid latitude warm thick snowpacks while the Arctic features cold thin snowpacks." 

It would seem more faithful to me to cite also the other applications of these models besides avalanche forecasting (something like " were designed for avalanche forecasting and process-studies/other applications in the Alps")

- L58-59 : "This process is not simulated by Crocus or SNOWPACK, leading to erroneous outputs."

I think the attempts to partially account for this process in both models (Touzeau et al., 2018 and Jafari et al., 2020), should be mentioned here.

Touzeau, A., Landais, A., Morin, S., Arnaud, L., and Picard, G. (2018). Numerical experiments on vapor diffusion in polar snow and firn and its

impact on isotopes using the multi-layer energy balance model crocus in surfex v8.0. Geosci. Model Dev. 11, 2393–2418. doi: 10.5194/gmd-11-2393-

2018

Jafari M, Gouttevin I, Couttet M, Wever N, Michel A, Sharma V, Rossmann L, Maass N, Nicolaus M and Lehning M (2020) The Impact of Diffusive Water Vapor Transport on Snow Profiles in Deep and Shallow Snow Covers and on Sea Ice. Front. Earth Sci. 8:249.

doi: 10.3389/feart.2020.00249

- L181 : "Based on several years of simultaneous temperature measurements of the HCS2-S3-XT and CNR4 sensors, we corrected the CNR4 sensor values. We found that there was no bias between the two temperature measurements and a RMSD=0.784°C"

I am rejoining the comment from Referee #1 on this : Details on the correction (does it involve radiations ?) would be welcome as an help/ a reference for other people encountering similar problems at their sites

-L332 : "Figure 2 may also indicate a decrease in summer air temperature"

The statement is too vague, please clarify.

- L345: At several places like this one "However, all these criteria are not 100% certain, and there may be some data in the absence of snow." the honesty of the authors is much appreciated !

- L358 : "The deepest sensor was placed just above the frozen soil layer"

Maybe it would be more clear if you specify that this is the frozen layer at its summer position.

- L 395 : "Briefly, the amount of correction decreases with increasing snow density and is about 1.1 for dense wind slabs and 1.5 for soft depth hoar"

Is this a multiplicative factor ? Thanks in advance for clarifying this

- Dataset : As a suggestion for later updates of this dataset, a time-series flag for each meteorological variable indicating whether the data has been gapfilled, corrected, or is the raw measure, could be usefull and more precise/exhaustive than what is currently written in the paper (for wind speed for instance)

* technical corrections

- L 242: "showed a -2 W m-2 offset and a value of 2 W m-2" => the "-2" should be in exponent position



- Fig 6 : some of the legends would be better relocated (eg outside the graphs) for better readability



-L397 : "value at 2 cm in 2014-2015 has values around 0.02 W m -1 K -1".

Only one 'value' would be enough in the sentence.