MacFerrin et al. describe a dataset consisting of strain rates in shallow firn, 2 m air temperature, firn temperature, and surface height from eight sites across the Greenland Ice Sheet. The dataset consists of daily strain measurements and hourly meteorological measurements spanning 2013 – 2019. Multiple boreholes have been instrumented at each site allowing for users to assess viability and repeatability of strain measurements and examine compaction rates over different depth ranges.

Given the disagreement in firn models using both steady-state and transient modes, this dataset will provide important validation data spanning climate regimes. The dataset will provide opportunity for interesting and important future work examining firn compaction on seasonal timescales and investigating the effects of meltwater on firn compaction rates.

This paper is well-written and provides a concise but helpful literature review in addition to describing the dataset. I have a few general comments and questions along with some technical corrections that are detailed below, but I recommend this paper be published in ESSD after these are addressed. 

General Comments:

(1)

The authors provide a nice summary of the recent research on Greenland’s firn layer. I have one suggestion that may make their discussion clearer. The authors state four main reasons for studying the firn layer: (1) to refine estimates of mass balance using altimetry methods; (2) to estimate how much meltwater can be stored within the firn column and buffer future sea level rise; (3) to understand the development of near-surface ice slabs that block future meltwater percolation; and (4) to refine interpretations of climate records from ice cores.

In my opinion, numbers 2 and 3 essentially fall into the same category of understanding firn’s capacity to buffer future sea level rise. I would recommend phrasing this section as there being 3 main reasons for studying firn: (1) to use altimetry mass balance products; (2) to understand how much water can be stored in the firn column and buffer sea level rise – can all available pore space be filled or does the expansion of near-surface ice slabs block percolation and expand runoff zones? And (3) to improve interpretations of climate records contained in ice cores. Grouping the main reasons for studying firn in this way seems to make more intuitive sense to me and better links the impacts of ice slab formation on meltwater storage. Additionally, I would consider adding an additional phrase or sentence to expand on why firn structure is important for interpreting paleoclimate records. I would recommend just explicitly stating that knowing the ice-age/gas-age discrepancy allows us to accurately date past atmospheric conditions. This last reason just seemed a bit shorter than the others listed.

(2)

I am curious about the contribution of ice flow to vertical velocities that could be captured by the installed strainmeters and misinterpreted as firn compaction. If the ice column were undergoing longitudinal extension, the borehole could be shortening even in absence of firn compaction. I imagine that at sites like Summit, there likely is not much longitudinal compression, but with EastGRIP being located on an ice stream and there being local topographic variability at sites like Crawford Point, I am wondering how much ice dynamics could affect the calculated firn compaction rates. In studies that attempt to determine vertical strain at sites using phase-sensitive radar, some have attempted to separate out an ice dynamics component from compaction (e.g. Jenkins et al., 2006).

My two main questions arising from this curiosity are: (1) Would vertical strain resulting from ice flow be recorded by your strainmeter instrumentation? and (2) If so, are there existing data that users of your dataset could find to remove an estimated vertical strain component from ice dynamics to get a purely firn compaction component?

If the answer to the first question is no, then perhaps adding sentence or two explaining the concern and detailing that this instrumentation would not be affected by ice dynamics processes would help readers. If the answer is yes, then I think it would greatly help users of your dataset to point to existing measurements or potential models that could help resolve the influence of ice flow on the compaction rates that you present.

(3)

I am interested in knowing the distance between boreholes at each site. It may be helpful to include a sentence on line 111 saying that boreholes are spaced approximately xx m apart. If available, could each borehole’s coordinates be included in Table 2? It may make Figure 1 too busy, but potentially including insets for each site with the locations of each borehole may also be helpful.

This interests me because looking closely at Figure 5, it appears there are some slight differences in compaction rates between boreholes at the same site that are installed over approximately the same depth range. Specifically, I am looking at boreholes 13 and 16 at NASA – SE in the summer of 2015, boreholes 17 and 20 at Saddle, boreholes 30 and 33 at Summit (particularly in the summer of 2015), and boreholes 26 and 29 at EastGRIP. For future work, it would be interesting to examine discrepancies in compaction rates between closely-spaced boreholes instrumented over the same depth range, and knowing the length scale over which these differences occur would be useful.

(4)

Suspicious data readings that are not included in the preliminary analysis are mentioned on lines 210 – 211. Could a brief explanation as to why they are considered suspicious be provided here?

Minor Comments and Technical Corrections:

L18: Consider changing "snow temperature” to  "firn temperature"  or "snow and firn temperature" 

L42: "Adolf and Albert, 2014"  should be changed to "Adolph and Albert, 2014" 

L82: Does there need to be a dash here in "firn-density" ?

Throughout the manuscript, "SumUp" is written when I believe the correct capitalization is "SUMup" .

Throughout the manuscript text and figures, EastGRIP is written both as "EastGrip"  and "EastGRIP". I recommend that it be written as EastGRIP throughout the manuscript.

L256: Does there need to be a dash here in "firn-temperature" ?

Figure 3: The caption mentions "changes relative to installation depth plus 120 days", however, in line 195 you suggest removing the first month of data from analysis. In this figure, have approximately the first 4 months of data been removed? It may be helpful to state this in the text, perhaps just after mentioning that the first month of data has been removed but Figure 3 shows data with the first four months removed.

Figure 6: It appears that the firn temperatures at Summit are not captured by the scale bar on the right. I think the scale bar needs to be expanded to capture the low temperatures of firn at Summit. Additionally, the color scale used in this figure is slightly counterintuitive, as I often think of red colors as warmer temperatures. Perhaps flipping the scale may help readers interpret it better.

Table 4: I think the temperature difference column should be rounded to the same number of decimal places as the previous two columns. If there is not enough certainty to present measurements with two or more decimal places, it seems odd that the difference between the two is presented at a higher resolution than the measurements.

Reference included in review comment:

Jenkins, A., Corr, H. F., Nicholls, K. W., Stewart, C. L., & Doake, C. S. (2006). Interactions between ice and ocean observed with phase-sensitive radar near an Antarctic ice-shelf grounding line. Journal of Glaciology, 52(178), 325-346.

The dataset presented in this paper is novel and resulted from a large work effort. These data are therefore important and have high value to the scientific community. The manuscript is well-written, well-organized, and is comprehensive. I support publication of this paper in ESSD as a strong contribution to the literature. I have a small number of minor comments.



-The data are high quality and are accessible from a reputable data repository.

-Data ratings:

>significance: excellent

>uniqueness: excellent

>completeness: excellent

Comments



line 85: I believe the ‘is’ should be ‘are’ to match plurality.

Line 134: I was confused about numbers here. Understand that it was a lot of work for a failed installation, but it seems much more straightforward to just tell us about the 48 sensors with actual records.

Line 140: could report manufacture’s specs on strain of the wire. Over these short runs I would assume it’s a non-issue?

Line 145: if you excavated the PVC platforms, it would be nice to know how they settled with time: i.e., did they tilt heavily to one side?

Line 157: the approximate depth, or range of depths, of the snowpit installations would be good information to provide.

Section 3.2

-while the air temp appears to have been measured hourly, the firn temps were measured once per day? Clarification would be useful because some users may be interested in high time resolution firn temps.

-were the firn temperature holes backfilled (if so, how?) or left standing air filled? This is important information to include for future data users.

Line 170: ‘clearing chips’ will confuse some readers. “due to accumulated drill shavings on the bottom of the hole” …or some such.

Line 195-paragraph: appreciate this approach/section.

Figure 3:

-“…depth plus 120 days” is awkward wording that hung me up (L+T ?).

-The caption needs clarification of what the legend numbers/lines are …e.g., “legend is sensor number in Table X”.

Line 244: recognize that the scope of this paper is limited, but you mention a hypothesis that noise stems from interaction between meltwater and the borehole. This could use a bit more elaboration – meltwater dribbling down the hole somehow generates a noisy signal?

Figure 4: as w/Fig 3, the caption could use explanation of the legend numbers.

-Overall, an enjoyable read and high value contribution.

This manuscript introduces FirnCover, which is a unique dataset of firn compaction measurements taken between 2013 and 2019 at eight locations on the Greenland Ice Sheet. MacFerrin et al. detail the methods used to obtain the measurements, describe the format of the dataset, and include a preliminary analysis and interpretation of the data. Data from strainmeters, air temperature sensors, and firn temperature sensors at eight sites with different climates provide the opportunity to investigate variability in firn compaction based on climatic factors. Moreover, this dataset of in-situ observations from 50 strainmeters can act as a comparison for existing firn models, which the authors point out use different densification schemes and often disagree with one another. FirnCover is a novel, high-quality, and well-prepared dataset that will be a valuable contribution to the firn modeling and observation community.

Overall, this paper is very well-written and the dataset is of high quality. Included are two general comments as well as line-by-line technical corrections. Once these comments have been addressed, I recommend this manuscript be published in Earth System Science Data.

General Comments

1) Data overview and preliminary analysis - The data description and analysis (Section 5) is important to include in the manuscript since it describes and shows the actual data included in FirnCover. Since the goal of the paper is to present the dataset, finding a balance in the amount of preliminary analysis to include is difficult, and I think that toning back some of the analysis (or even including more of the raw data) could help reach this balance. This section could benefit from showing more of the raw data and focusing less on the analytical choices (e.g., averaging window, amount of smoothing, how much initial data to ignore), especially since the authors have included the full record of unsmoothed data in the publicly available dataset (a very good decision). For example, in Section 4, lines 192-198, the authors describe how the dataset includes the first month of measurements, but those data have been discarded for the analysis presented in the manuscript. Since users of the dataset will likely have to decide on how to filter these data for their own analysis, showing a figure of the full, unsmoothed record could be beneficial. This could even just be for one example record rather than all of them.

2) Interpretation – On a similar note, I would consider adjusting the amount of interpretation included in Section 5. There are a few instances where the authors state the reasons for higher firn density without referencing a figure or table (e.g., line 223-224), and it’s unclear whether these interpretations are based on the data itself or knowledge from prior studies. If based on the data, specifically referencing figures that support the interpretation will make the claim stronger. More specific examples regarding this comment have been listed in the line-by-line section below.

Minor Comments and Technical Corrections

Lines 53-58: It seems that this section might need a citation to better suppor these statements.

96: Consider expanding the description of these existing measurements and maybe even include them in Figure 1. If they are in regions or climates that are different from those in FirnCover, this would help amplify the need for FirnCover.

140: To what is the 2.032 m range referring? The lengths of the boreholes are several meters, so is this range the amount of length change that can be detected?

192: Wouldn’t the instrument settling depend on the snow surface density, such that the lower the snow density, the longer the settling time? If I’m interpreting this correctly, wouldn’t the one-month settling time of the high-density ice site be the minimum settling time among all sites? This seems to be mentioned later in lines 239-241. Consider combining this discussion or changing it to discuss how surface density affects settling rates, and the settling period is likely different for each site.

204: Explain why a period of two months is chosen for the running mean. Users of the dataset may be looking for guidance in processing/analysis, so a description of the authors' reasoning may help.

211: Why are these data suspicious? Consider elaborating on the suspicious nature or showing them in a figure

223: Change "this is due" to "this is likely due" and consider adding a citation or referencing a figure to back this claim.

Figure 3: The first line of the caption is confusing. The current wording makes it seem as though time is being added to depth. Is this record starting 120 days after the installation date? If so, why?

240: Consider changing "deformation" to "settling"

244: Is this really "noise" or this is a result of the densification processes mentioned? This is an area where I would be cautious of over-analyzing/interpreting for the purposes of this paper.

246: Earlier it was stated that KAN-U is completely ice (line 194), but here the authors write that it contains a 5-m ice slab. Consider updating either line 194 or 246 to reflect the same meaning.

249: Reference Figure 4 at the end of the sentence.

249-251: The wording in the sentence is confusing. Remove the "(resp. __)" and split into two distinct sentences.

Figure 4: Since there is dependency on borehole depth (line 237), could you also label the different lines by depth (in addition to instrument ID)? This could also apply to Figure 3.

257: Consider changing "compaction rates" to "surface height" since the referenced figures are showing surface height.

Figure 6: It seems these firn temperatures are interpolated; that should be mentioned in the caption. Also, it seems there was a different color bar used in the Summit plot. Consider flipping the color bar so warmer=redder, or consider using a perceptually uniform sequential color map (e.g., viridis) that has a larger range of colors and can visually capture small variations better.

268-270: Provide a clearer description of what two variables are being compared. Over what period is the average air temperature calculated? Is the firn temperature averaged over 10 m and over time, or is it the firn temperature taken at 10 m and averaged over time?

268-270: I wonder if this comparison would have more significance if examined seasonally rather than averaged over time. This is an area where I would consider being cautious with the interpretation since it seems to require a more in-depth analysis, which is not necessarily the focus of this manuscript.

284-287: Consider adding this information to either Table 1 or 2 so the reader can easily see what existing data are available at each FirnCover site.