A Call for More Snow Sampling

Round 1

Reviewer 1 Report

General

 

I like the idea of illustrating the issues in snow sampling methods using case studies, and the introduction outlines the reasons for the selection. However, it seems a little odd that the focus is much more heavily on snow depth than on SWE when the introduction is primarily aimed at the importance of SWE. (Incidentally, is it really true that the first measurements of SWE were taken as recently as the early 20^th century? It seems surprising to me.) It would also be good to outline the characteristics of each of the measurement methods to be employed so that the reader has some idea of what the technical issues in the comparisons are likely to be.

 

The main part of the paper itself is somewhat lacking in structure and detail, although there is evidently a great wealth of data behind it. I think the paper could be hugely improved by giving it a clearer structure based around spatial variability, perhaps by introducing the semivariogram tool earlier on. Because the discussion of individual examples is rather unclear, I don’t find the discussion to be a satisfactory development from them.

 

2.1 This would be easier to follow with a bit more explanation of measurements. A brief summary of the nature of the test site(s) would be very helpful. What was the precision and accuracy of the lidar measurements? (The resolution is mentioned as being ~ 1.5 m: is this the same as the footprint size, hence the sampling area, and what was the vertical resolution?) How were the probe measurements made? I assume with a magnetic compass and perhaps a tape measure to select and identify the measurements subsequent to the initial ones, but I don’t see how the ‘5% off’ calculation works or why the direction is specified as true North unless the subsequent measurements were all to the north of the marker. How were the lidar data registered to the probe measurements? How accurately were the probe orientations controlled when they were inserted into the snow? In fig 1, what is represented by 1(b)? I can’t see anything like as many as 684 points in this figure, so how does this square with the statement in the main text and the fig caption that that there are ‘up to 684’ depth probe measurements?

 

The semivariogram analyses (fig 2b) are potentially useful as a way of resolving the point made in (119-120) about the difficulty of making a direct comparison between lidar and probe measurements, but this analysis doesn’t really go anywhere. What does analysis of fig 2(a) reveal about the relationship between the two measurements beyond the mean differences reported in fig 1, and to what extent can the differences be understood in terms of the spatial variability illustrated by the semivariograms?

 

2.2 This section is currently rather unconvincing. Were the house measurements intended to test the variation of depth and SWE with spatial separation, or aspect relative to a wall, or both? The statement that ‘the wind blew from almost all ‘directions’ needs to be substantiated and ideally quantified to be convincing that the aspect had no bearing on the results, and were there any other factors – such as nearby trees or houses – that could also have invalidated the implicit assumption about isotropy here? The differences in measurements are asserted to be due to specific factors without actual evidence. I would hope to see some attempt to partition differences between measurements of snow depth into random and spatially-varying components. What was the purpose of the measurement with the snow roughness board? I think perhaps to give some basis for an argument about spatial variation in snowpack parameters based on solar-induced melting.

 

2.3 I must admit to being lost in this section. It needs some more explanation at the beginning so help the reader understand what is being tested here.

Specific points, typos etc. (by line no)

 

 

(37) model → models

(38) presents → present

(46) PMRS: spell out acronyms and abbreviations on first use

(52) ‘associated variability’ - spatial, or temporal, or both?

(55) primarily snow depth → primarily of snow depth

(58) for during → during

(74) are → area

(85) matching → coincident (?) would clarify the meaning

(86) design was → design involved (or something similar)

(89) implementation of → implementation by (If I have understood the meaning correctly)

(99) difference: which quantity has been subtracted from which?

(101) penetration into the soil: was the test area covered in soil?

(105) upper left: north-west??

(113) from are → are from

(113, 114) snow on → snow-on, snow off → snow-off

(116) The sentence ‘The mean difference...’ isn’t part of the caption to fig 1(b), I think.

(123) summarize → summarizes

(149) Here, (a) represent the matching… → (a) Corresponding…

(151) (b) is the snow … (b) Snow …

(155) snowpack can vary over multiple scale: I think this isn’t saying anything more than that the snow depth and SWE vary spatially, which point has already been made by fig 2(b). Or is timescale part of the consideration here too?

(205) references for the snow roughness board method seem to suggest that it is quite a recent idea. To the best of my knowledge, it was first described in 1998: Rees, W.G. (1998). A rapid method of measuring snow surface profiles. Journal of Glaciology 44 674-675.

(234) How were the solar shaded areas identified?

(245) essentially the same: please give numerical values to substantiate this.

(249) What are the green arrows in fig. 5?

(256) measure → measures

(257) exact same holes: what holes?

(258) depth probe as previous: what do you mean?

(259) as a function of → systematically with

(264) half of fig 6(b) is missing

(273) between what → between (I think: this sentence has lost its structure somewhat)

(274) location → locations

(279) Figure 2a → Figure 7a?

(279) but are somewhat consistent from year to year: meaning of this is not clear to me

(285) slower → more slowly

(287) expect → except

(288) Edge → The edge

(288) within several meters: not entirely sure what to read into this. If it was ‘within a few meters’ I’d read it as ‘not very far away’, but ‘several’ suggests ‘rather far’.

(291) varying materials → different materials (?)

(302) Fig 8: This is just a Google screen shot, and I’m not sure it adds anything to the manuscript.

(316) and this often changes over time: how is this a different statement from ‘can vary from year to year’?

(328) resolution of the terrain: presumably, spatial resolution of the representation of the terrain (or is this a statement about spatial variability of the terrain height itself?)

(331) land cover data are at a 30-m resolution: Why is this true? If I want a land cover map with a spatial resolution of 1 m I can make one using drone data.

(339) effort → efforts

(340) were → where

(356) This sentence has lost its structure somewhere along the way.

 

Author Response

I like the idea of illustrating the issues in snow sampling methods using case studies, and the introduction outlines the reasons for the selection. However, it seems a little odd that the focus is much more heavily on snow depth than on SWE when the introduction is primarily aimed at the importance of SWE. (Incidentally, is it really true that the first measurements of SWE were taken as recently as the early 20th century? It seems surprising to me.) It would also be good to outline the characteristics of each of the measurement methods to be employed so that the reader has some idea of what the technical issues in the comparisons are likely to be.

 

• The first sentence has been changed to present the link between SWE and runoff. More details are provided for the six different examples, specifically, three of snow depth only, two of depth, SWE, and density, and one of SWE only.
• A section has been added that explains the sampling and analysis methods used for the six examples

 

 

The main part of the paper itself is somewhat lacking in structure and detail, although there is evidently a great wealth of data behind it. I think the paper could be hugely improved by giving it a clearer structure based around spatial variability, perhaps by introducing the semivariogram tool earlier on. Because the discussion of individual examples is rather unclear, I don’t find the discussion to be a satisfactory development from them.

 

• A Methods section has been added prior to all the examples to explain the various sampling methods. This includes the comparison of data and an introduction to the semivariogram tool.

 

2.1 This would be easier to follow with a bit more explanation of measurements. A brief summary of the nature of the test site(s) would be very helpful.

• Some of this information is now presented in the new Methods section. More information has been added to address the specific questions.

 

What was the precision and accuracy of the lidar measurements? (The resolution is mentioned as being ~ 1.5 m: is this the same as the footprint size, hence the sampling area, and what was the vertical resolution?)

• More information has been added.

 

How were the probe measurements made? I assume with a magnetic compass and perhaps a tape measure to select and identify the measurements subsequent to the initial ones, but I don’t see how the ‘5% off’ calculation works or why the direction is specified as true North unless the subsequent measurements were all to the north of the marker.

• More information has been added and this has been rewritten. The depth probe was used to measure the distance from one point to another. The direction is to the north, east, south or west (added).

 

How were the lidar data registered to the probe measurements?

• This text has been changed. The lidar point closest to the snow depth measurement was chosen. The problem is that the depth probe measurement locations were reported and not actual, as per the previous comment, and they were reported to the nearest meter.

 

How accurately were the probe orientations controlled when they were inserted into the snow?

• It is assumed that the probes were inserted vertically. The problem of over-probing is also mentioned.

 

In fig 1, what is represented by 1(b)? I can’t see anything like as many as 684 points in this figure, so how does this square with the statement in the main text and the fig caption that that there are ‘up to 684’ depth probe measurements?

• More information is provided on the sampling distribution. This number has been changed to 680. Each 100x100 meter square has a starting depth measurement and at four additional depth measurements spaced 5, 10, 15, 20 or 25 meters apart in an “L.” In each quadrant there is one 100x100 meter square with 50 depth measurements.

 

The semivariogram analyses (fig 2b) are potentially useful as a way of resolving the point made in (119-120) about the difficulty of making a direct comparison between lidar and probe measurements, but this analysis doesn’t really go anywhere. What does analysis of fig 2(a) reveal about the relationship between the two measurements beyond the mean differences reported in fig 1, and to what extent can the differences be understood in terms of the spatial variability illustrated by the semivariograms?

• This section has been rewritten and reorganized. More details are given on the differences in spatial structure for the datasets.

 

2.2 This section is currently rather unconvincing. Were the house measurements intended to test the variation of depth and SWE with spatial separation, or aspect relative to a wall, or both? The statement that ‘the wind blew from almost all ‘directions’ needs to be substantiated and ideally quantified to be convincing that the aspect had no bearing on the results, and were there any other factors – such as nearby trees or houses – that could also have invalidated the implicit assumption about isotropy here? The differences in measurements are asserted to be due to specific factors without actual evidence. I would hope to see some attempt to partition differences between measurements of snow depth into random and spatially-varying components.

• “Two sets of measurements were taken around the authors house to assess spatial differences in accumulation and snowmelt, with the emphasis on quantifying the variability.” The point is the differences rather than a thorough investigation of the physics that causes the differences. For example, if I go out one side of my house I can get a different estimate than if I went out another side of the house. The difference at the two NWS sites is marked, and those sites are within meters of one another.
• A wind rose from the nearest station shows that the wind blew in most directions during the snowfall event. This has been added to Figure 3.

 

 

What was the purpose of the measurement with the snow roughness board? I think perhaps to give some basis for an argument about spatial variation in snowpack parameters based on solar-induced melting.

• The roughness board was used to obtain fine (sub-millimeter) resolution or continuous snow depth. This section has been rewritten.

 

2.3 I must admit to being lost in this section. It needs some more explanation at the beginning so help the reader understand what is being tested here.

• The introduction to this section is confusing. The nature of the data collected at the stations has been moved to earlier (section 2), and the focus is now on additional snow depth measurements.

 

 

Specific points, typos etc. (by line no)

 

(37) model → models

> changed

 

(38) presents → present

> changed

 

(46) PMRS: spell out acronyms and abbreviations on first use

> I added the full name of PRMS, VIC, and MODIS. I did not originally use those since they are known by their abbreviations.

 

(52) ‘associated variability’ - spatial, or temporal, or both?

> both

 

(55) primarily snow depth → primarily of snow depth

> changed

 

(58) for during → during

> changed

 

(74) are → area

> changed

 

(85) matching → coincident (?) would clarify the meaning

> changed

 

(86) design was → design involved (or something similar)

> changed

 

(89) implementation of → implementation by (If I have understood the meaning correctly)

> Yes, changed. This is “what the snow surveyor did in the field.”

 

(99) difference: which quantity has been subtracted from which?

> I have added “probe minus lidar snow depth”

 

(101) penetration into the soil: was the test area covered in soil?

> Yes, there is soil under the snow. The previous sentence has been changed to “due to penetration into the soil below the snow.” The next sentence has been modified to state that “[s]ince the average difference here is between -0.03 and 0.48 m, over-probing is likely not the cause of these differences.”

 

(105) upper left: north-west??

> changed here, and “lower left” is changed to “southeast”

 

(113) from are → are from

> changed

 

(113, 114) snow on → snow-on, snow off → snow-off

> changed

 

(116) The sentence ‘The mean difference...’ isn’t part of the caption to fig 1(b), I think.

> This is moved to before (b)

 

(123) summarize → summarizes

> changed to “summarized”

 

(149) Here, (a) represent the matching… → (a) Corresponding…

> changed

 

(151) (b) is the snow … (b) Snow …

> changed

 

(155) snowpack can vary over multiple scale: I think this isn’t saying anything more than that the snow depth and SWE vary spatially, which point has already been made by fig 2(b). Or is timescale part of the consideration here too?

> Yes, it does vary over time scales, but that is not the point here. The sentence has been changed to be a phrase that leads into the next sentence: “[s]ince the snowpack varies spatially over multiple scales [14; Blöschl 1999] …

 

(205) references for the snow roughness board method seem to suggest that it is quite a recent idea. To the best of my knowledge, it was first described in 1998: Rees, W.G. (1998). A rapid method of measuring snow surface profiles. Journal of Glaciology 44 674-675.

> Thank you for bringing this letter to my attention. I am not suggesting that this is a new method; I am just using the specific steps outlined by Fassnacht et al. [2009]. The text has been changed to “as per the method described in Fassnacht et al. [45; 2009]”

 

(234) How were the solar shaded areas identified?

> The text has been added stating that “[t]his was determined from a photograph taken the following day, illustrating shading on the snow using the technique of Fassnacht et al. [45; 2009].”

 

(245) essentially the same: please give numerical values to substantiate this.

> This has been included and then sentence also states “… only 1.2 cm different …”

 

(249) What are the green arrows in fig. 5?

> this is meant to represent the mean. The figure and the caption have been changed.

 

(256) measure → measures

> changed

 

(257) exact same holes: what holes?

> This has been rewritten to: “[t]he snowpack was melting, so the depth measurements taken on 2 June left holes in the snowpack that remained on 4 June, and thus depth was measured in the exact same location for the two dates.”

 

(258) depth probe as previous: what do you mean?

> this has been rewritten to explain that this is “the same depth probe used for the sampling presented in section 2.1 was used here.”

 

(259) as a function of → systematically with

> changed

 

(264) half of fig 6(b) is missing

> this is formatting from the journal. I have changed the figure

 

(273) between what → between (I think: this sentence has lost its structure somewhat)

> “what” has been removed

 

(274) location → locations

> changed

 

(279) Figure 2a → Figure 7a?

> changed here and the following instances throughout this paragraph

 

(279) but are somewhat consistent from year to year: meaning of this is not clear to me

> the sentence has been changed to: “the three pillows do not measure exactly the same SWE), yet the differences are consistent from year to year.”

 

(285) slower → more slowly

> changed

 

(287) expect → except

> changed

 

(288) Edge → The edge

> this sentence has been rewritten

 

(288) within several meters: not entirely sure what to read into this. If it was ‘within a few meters’ I’d read it as ‘not very far away’, but ‘several’ suggests ‘rather far’.

> a photograph (Figure 8b) has been added, and the specifics of the size of the pillow plus the distance apart is now included.

 

(291) varying materials → different materials (?)

> changed

 

(302) Fig 8: This is just a Google screen shot, and I’m not sure it adds anything to the manuscript.

> I have added a photo of the snow pillows. The forest surrounding the snow pillows is now shown in the Google image to illustrate that there is no differential shading on the snow pillows.

 

(316) and this often changes over time: how is this a different statement from ‘can vary from year to year’?

> There was a point when I wrote this, but don’t know what it is anyone. I deleted “and this often changes over time”

 

(328) resolution of the terrain: presumably, spatial resolution of the representation of the terrain (or is this a statement about spatial variability of the terrain height itself?)

> both spatial resolution of the data and variability within each pixel. The text has been changed

 

(331) land cover data are at a 30-m resolution: Why is this true? If I want a land cover map with a spatial resolution of 1 m I can make one using drone data.

> Most existing national or global datasets are at a 30-m resolution, and these are the ones that have been used. It is becoming more feasible, i.e., to use a drone to create finer resolution land cover data. The point is not the resolution, but the accuracy of on-the-ground/handheld GPS data. This sentence has been changed.

 

(339) effort → efforts

> changed

 

(340) were → where

> changed

 

(356) This sentence has lost its structure somewhere along the way.

> this sentence has been rewritten as: “[a] compromise is often necessary between having a random distribution of sampling locations and being efficient in moving across the terrain to collect more data.”

Reviewer 2 Report

The manuscript present a great and deep review and overview of the relevance of the snow sampling and their interest on  wide variety of fields. As a review it covers an important number of fields, and use all classic and new references. This work is well presented and the content is really interesting for the scientific community. It is a call to follow the research and investigation on snow hydrology due to its interest, as it has been clearly exposed by the author.

I do not see how to improve it, except small issues with some figures:

Figure 4 should be slightly wider to allow a well reading of the text inside the figure

Figure 6 c) is incomplete in the draft

So, I recommend to pubish it as it is after these short corrections.

Author Response

I think the reviewer for their comments on the figures. Below is how I changed them.

Figure 4 should be slightly wider to allow a well reading of the text inside the figure

> I have made the figure larger

 

Figure 6 c) is incomplete in the draft

> this is formatting from the journal. I have changed the figure

Reviewer 3 Report

In this article, examples of high-resolution snow measurement sampling are used to discuss the sampling strategy and recommend solutions for more representative sampling.

The article is overall rather confusing and it is not clear where the author wants to take us, with no real thread. The main objective of the article is not very clear, the paper is not well organized.  The author presents a succession of comparative studies of snow measurements without in-depth discussion to support findings.  
The figures are not of good quality and some are truncated (fig6).

The main topic of the article is of great interest but unfortunately more work is needed to clarify the objectives of the article and make its publication beneficial to the readers. 

Author Response

In this article, examples of high-resolution snow measurement sampling are used to discuss the sampling strategy and recommend solutions for more representative sampling.

 

The article is overall rather confusing and it is not clear where the author wants to take us, with no real thread. The main objective of the article is not very clear, the paper is not well organized.  The author presents a succession of comparative studies of snow measurements without in-depth discussion to support findings. 

> The end of the Introduction has been rewritten and the purpose of the paper has been explicitly stated. A Methods section has been added that lays out the methods used for data collection and evaluation of the data.

 

The figures are not of good quality and some are truncated (fig6).

> this is formatting from the journal. I have changed the figure

 

The main topic of the article is of great interest but unfortunately more work is needed to clarify the objectives of the article and make its publication beneficial to the readers.

> The end of the Introduction has been rewritten and the purpose of the paper has been explicitly stated.

Reviewer 4 Report

Dear Author, dear Editors,

I read the manuscript with great interest. I do have to say that there are not really any major remarks i am going to provide, even though i liked reading the manuscript and think it is highly relevant.

It is not your typical research paper, but makes a strong case for snow sampling, gives good recommendations and provides a comprehensive overview on approaches and the available literature.

I therefore recommend the manuscript for publication.

I would like to mention that the .pdf does not seem to be properly formatted, text and figures are somehow all on the right edge.

 

Best regards

Author Response

Here is the reviewer's comment:

I would like to mention that the .pdf does not seem to be properly formatted, text and figures are somehow all on the right edge.

> this is formatting from the journal. I have changed Figure 6 so it is completely on the page

Round 2

Reviewer 1 Report

The author has made some clarifications, but the manuscript is still quite poorly structured. It isn't always obvious why the reported measurements have been made (or indeed how, in some cases), and the discussion is not clearly linked to the foregoing presentation of experimental data.

 

S2.1. Maybe it is confusing the issue here to include material more relevant to terrestrial and spaceborne lidar than to the airborne systems that are the method to be considered in this section. Please also say something about the spatial sampling characteristics of LiDAR systems.

 

S2.3 Can you justify the choice of particular snow tube used in these experiments? Since the aim of the paper is to encourage more snow sampling, it would be useful to know what are the optimum characteristics for a measurement device. Also explain why snow surface roughness is a useful thing to measure – otherwise the use of a roughness board just seems like an unnecessarily laborious way of measuring depth. And explain what a ‘snow pillow is’ (this seems like a partial forward reference to the next section perhaps).

 

S2.4 This section refers to a specifically US system. I wonder if it is possible to broaden the scope by placing this measurement system in an international context – e.g. how do other countrie make automated snow measurements (where they do)?

 

S2.5 I feel this section lacks focus but could be sharpened into a discussion of the reasons for non-representativeness of snow depth (or SWE) measurements, preferably with the development of some mathematical framework into which subsequent results could be inserted. The qualitative description of some aspects of a semivariogram goes some way towards this, but I’d like to see some mention of random and systematic error. Why is fractal dimension mentioned here – is it useful to be able to measure it?

 

S3.1 Some important details are missing here I think. What does CLPX mean? Where is the study site located? How were the nine ISAs chosen (and how does the ‘North Park MSA’ relate to them)? Did they have vegetation cover? How were the lidar data acquired? The section is still quite confusingly presented, with rather unclear presentation of methods and no proper justification for some of the analysis performed on the data.

 

S3.2 Clarify what you are testing here. Is this about the spatial variability of snow measurements over distances of the order of metres, or a couple of kilometres? Or is it about the effect of nearby obstructions such as the house, fences, trees etc? It isn’t clear how these different potential effects can be disentangled from these measurements. What was the uncertainty in the measurements? Without knowing that, one can’t assess whether the various reported differences or similarities were significant. Also, how were the meteorological measurements acquired? And do they really justify the statement that wind blew from almost all directions? Fig 3d suggests a strong preponderance of wind from the north, which would I think be consistent with an increased accumulation on the south side of the house.

 

S3.3 Again, not really clear at the outset what the aim of this experiment will be. The mention of shading from the sun takes this reader by surprise – it wasn’t mentioned in the introduction, so why are we concerned with it now? How were the densities of the snow layers determined?

 

S4 Discussion. It isn’t always clear how the data described in S3 feed through to the discussion.

 

(72) ‘Light distance and radar’ → ‘Light detection and ranging’ (or variants – but certainly not ‘radar’)

(72) ‘optical portion’ More often in the near-infrared in fact

(82) ‘vertical resolution… millimeters’ This is a bit optimistic, especially for spaceborne systems.

(90) ‘error’ → ‘errors’

(94) ‘now’ → ‘not’

(108) Explain that ‘ds’ means depth of snowpack.

(109) ‘does snow’ → ‘does not’

(114) ‘form’ → ‘from’

(115) ‘Density is computed as… depth.’ Strictly, this statement cannot be true since the ratio is dimensionless and actually represents the ratio of the mean snow density to that of liquid water.

(128) ‘a second roughness was inserted’ - is there word missing here? These details in any case don’t belong in this section.

(153) ‘presents of snow accumulation’ - something has gone wrong with this sentence.

(160) ‘It is called the semivariance...’ The given reason is incorrect and doesn’t make sense.

(177) ‘The snow depth and vertical position were estimated...’ Do you mean by lidar, or by snow depth probe? I don’t understand what you mean by ‘vertical position’ - is height above some datum?

(192) ‘Buffalo Pass’, ‘Spring Creek’ - where are these?

(193) ‘nine-six’ what does this mean?

(204) ‘to horizontally measure’ - what does this mean?

(259) ‘Depth probe measurements are known...’ shouldn’t this statement be in section 2.2?

(278) ‘… but there was 10% difference...’ I am not sure I follow the implication of the word ‘but’ here.

(281) ‘...only 1% different.’ From what – one another?

(319) ‘… causing the snow to compact...’ How do you know that no other factors were involved in snowpack metamorphosis? E.g. what were the air temperature and wind conditions?

(365) ‘measures depth’ → ‘measurements of depth’ (?)

(401) delete second ‘the’ for logical consistency

(403) ‘forested openings’ - surely, not forested, otherwise the tree canopy would interfere with the measurements? I suppose you mean forest clearings?

(407) ‘shade differential’ - not sure I understand what you mean by this.

(408) ‘differences in snow pillow materials’. This is presumably part of the justification for performing this experiment, so it would be really helpful to tell the reader about these differences before, rather than after, presenting the results.

(463) ‘these can now be derived from lidar...’ Can land cover really be derived from lidar? Please give a reference for this.

Author Response

I thank this reviewer for thoroughly reading the paper and providing some good insight into restructuring the paper.

 

 

The author has made some clarifications, but the manuscript is still quite poorly structured. It isn't always obvious why the reported measurements have been made (or indeed how, in some cases), and the discussion is not clearly linked to the foregoing presentation of experimental data.

 

• The paper has been reorganized to first discuss issues related to sampling, then provide the general methods, followed by the specific methods for the examples presented. Methods sections have been removed from the Results. The Discussion now follows more in line with the Introduction and Methods.

 

S2.1. Maybe it is confusing the issue here to include material more relevant to terrestrial and spaceborne lidar than to the airborne systems that are the method to be considered in this section. Please also say something about the spatial sampling characteristics of LiDAR systems.

 

• There is only one sentence that includes reference to terrestrial and spaceborne lidar, and it is to put the resolution of the airborne data into context with the other two platforms. The resolution is provided for the dataset that is used.
• It is not clear what is meant with “say something about the spatial sampling characteristics of LiDAR systems”

 

S2.3 Can you justify the choice of particular snow tube used in these experiments? Since the aim of the paper is to encourage more snow sampling, it would be useful to know what are the optimum characteristics for a measurement device. Also explain why snow surface roughness is a useful thing to measure – otherwise the use of a roughness board just seems like an unnecessarily laborious way of measuring depth. And explain what a ‘snow pillow is’ (this seems like a partial forward reference to the next section perhaps).

 

• Two sentences were added at the beginning of this section to put the various snow tubes into context. Further, there was an error in the second sentence about the SnowMetrics tube, i.e., it should “not” be used for hard layers – this may have been a combination of mistyping and autocorrect. The SnowMetrics tube was used because it is designed for fresh snow sampling.
• The text now states: “Many different samplers have been created to extract a core and measure the mass [36; Farnes et al. 1983, 37; Dixon and Boon 2012, 38; López-Moreno et al. 2020]. Al-most any tube can be used for such sampling with proper care and acknowledging its limitations; this includes a stove pipe [39; Doesken and Judson, 1996] or a plastic drainage pipe [40; Fassnacht et al., 2010]. For sampling individual snowpack layers, a short metal tube sharpended at one end, often called a SIPRE (Snow, Ice, and Perma-frost Research Establishment) tube [41; US Army Corps 1956 or similar] is often used while for a shallow fresh, snow (ds < 30 cm), a plastic SnowMetrics snow tube <snowmetrics.com> can be used. Here, the latter was used to measure fresh snow.”

 

S2.4 This section refers to a specifically US system. I wonder if it is possible to broaden the scope by placing this measurement system in an international context – e.g. how do other countries make automated snow measurements (where they do)?

 

• Additional information has been added about operational snow sampling in other countries.

 

S2.5 I feel this section lacks focus but could be sharpened into a discussion of the reasons for non-representativeness of snow depth (or SWE) measurements, preferably with the development of some mathematical framework into which subsequent results could be inserted. The qualitative description of some aspects of a semivariogram goes some way towards this, but I’d like to see some mention of random and systematic error. Why is fractal dimension mentioned here – is it useful to be able to measure it?

 

• This section has been moved to the beginning of the Methods section. Random and systematic errors are now mentioned in this section.
• The fractal analysis sentence was removed. Although the fractal dimension can be used, it was not used here.

 

S3.1 Some important details are missing here I think. What does CLPX mean? Where is the study site located? How were the nine ISAs chosen (and how does the ‘North Park MSA’ relate to them)? Did they have vegetation cover? How were the lidar data acquired? The section is still quite confusingly presented, with rather unclear presentation of methods and no proper justification for some of the analysis performed on the data.

 

• The NASA CLPX project is referred to in section 2.2 (was section 2.1). The description of the methods is moved to section 2.2, so that this section can focus on the results.
• There was some vegetation cover in the six ISAs that were evaluated. These sites were chosen based on the experimental design of the NASA CLPX project – a citation is provided.

 

S3.2 Clarify what you are testing here. Is this about the spatial variability of snow measurements over distances of the order of metres, or a couple of kilometres? Or is it about the effect of nearby obstructions such as the house, fences, trees etc? It isn’t clear how these different potential effects can be disentangled from these measurements. What was the uncertainty in the measurements? Without knowing that, one can’t assess whether the various reported differences or similarities were significant. Also, how were the meteorological measurements acquired? And do they really justify the statement that wind blew from almost all directions? Fig 3d suggests a strong preponderance of wind from the north, which would I think be consistent with an increased accumulation on the south side of the house.

 

• The Methods have been moved out of this section. A more specific rationale for the experiment has been provided.
• I agree that the house can cause uncertainties. One of the points is that if we go outside and measure the snow it may not be representative, depending on where we go and what the meteorological conditions are.

 

S3.3 Again, not really clear at the outset what the aim of this experiment will be. The mention of shading from the sun takes this reader by surprise – it wasn’t mentioned in the introduction, so why are we concerned with it now? How were the densities of the snow layers determined?

 

• The Methods have been moved to the previous section. The rationale for the experiment should now be more specific.
• The text was rewritten to clarify how the densities of the snow layers was determined: “[o]bvious layers were identified in each of the snow cores extracted from the snow tube; within the core as each layer was removed, the remaining depth was estimated and remaining mass or SWE was measured. The density was estimated for each layer from SWE divided by depth by subtracting the measurements after removal of the layer from the measurements prior to removal of the layer.”

 

S4 Discussion. It isn’t always clear how the data described in S3 feed through to the discussion.

 

• The results presented in S3 are now referred to more in the Discussion. The Discussion has also been reorganized to follow the Results better.

 

(72) ‘Light distance and radar’ → ‘Light detection and ranging’ (or variants – but certainly not ‘radar’)

 

• “radar” changed “ranging”

 

 

(72) ‘optical portion’ More often in the near-infrared in fact

 

• Change to “short-wave”

 

 

(82) ‘vertical resolution… millimeters’ This is a bit optimistic, especially for spaceborne systems.

 

• This is the stated resolution. The text has been changed to: “The stated vertical resolution of the lidar is millimeters, but is typically in the range of 10-20 cm”

 

(90) ‘error’ → ‘errors’

 

• changed

 

(94) ‘now’ → ‘not’

 

• changed

 

(108) Explain that ‘ds’ means depth of snowpack.

 

• “ds” replaced with “snow depth”

 

(109) ‘does snow’ → ‘does not’

 

• The sentence has been changed

 

(114) ‘form’ → ‘from’

 

• changed

 

(115) ‘Density is computed as… depth.’ Strictly, this statement cannot be true since the ratio is dimensionless and actually represents the ratio of the mean snow density to that of liquid water.

 

• Changed to “Snowpack density, compared to the density of water,”

 

(128) ‘a second roughness was inserted’ - is there word missing here? These details in any case don’t belong in this section.

 

• a second roughness “board” – the word “board” was added

 

(153) ‘presents of snow accumulation’ - something has gone wrong with this sentence.

 

• The word “presents” has been deleted. The sentence now starts with “The third example of snow accumulation illustrates …”

 

(160) ‘It is called the semivariance...’ The given reason is incorrect and doesn’t make sense.

 

• The sentence has been removed, as it adds nothing

 

(177) ‘The snow depth and vertical position were estimated...’ Do you mean by lidar, or by snow depth probe? I don’t understand what you mean by ‘vertical position’ - is height above some datum?

 

• Changed to “The snow depth was estimated …”

 

(192) ‘Buffalo Pass’, ‘Spring Creek’ - where are these?

 

• Figure 1 is now cited

 

(193) ‘nine-six’ what does this mean?

 

• Changed to “96”

 

(204) ‘to horizontally measure’ - what does this mean?

 

• Changed to “to measure the horizontal distance”

 

(259) ‘Depth probe measurements are known...’ shouldn’t this statement be in section 2.2?

 

• This sentence is now incorporated into the next sentence: “Since the average difference is between -0.03 and 0.48 m (Figure 1a), over-probing, while known to over-measure (over-probe) snow depth in the order of 5 cm due to penetration into the soil below the snow [62; Sturm and Holmgren 2018], is likely not the cause of these differences.”

 

(278) ‘… but there was 10% difference...’ I am not sure I follow the implication of the word ‘but’ here.

 

• The sentence has been changed to: “The NWS measurements are taken only 10 meters apart, yet there was 10% difference in snow depth.”

 

(281) ‘...only 1% different.’ From what – one another?

 

• Yes, the words “from one another” have been added.

 

(319) ‘… causing the snow to compact...’ How do you know that no other factors were involved in snowpack metamorphosis? e.g. what were the air temperature and wind conditions?

 

• Yes, other factors are involved. The clause was changed to “which helps drive snowpack compaction and began also melt.”

 

(365) ‘measures depth’ → ‘measurements of depth’ (?)

 

• changed

 

(401) delete second ‘the’ for logical consistency

 

• removed

 

(403) ‘forested openings’ - surely, not forested, otherwise the tree canopy would interfere with the measurements? I suppose you mean forest clearings?

 

• Changed

 

(407) ‘shade differential’ - not sure I understand what you mean by this.

 

• “differential” have been changed to “differentially”

 

(408) ‘differences in snow pillow materials’. This is presumably part of the justification for performing this experiment, so it would be really helpful to tell the reader about these differences before, rather than after, presenting the results.

 

• A description of the three snow pillows was added to section 2.4 and here the sentence was revised.

 

(463) ‘these can now be derived from lidar...’ Can land cover really be derived from lidar? Please give a reference for this.

 

• This was asked for by another review. This has been rewritten (with reference) to state that canopy structure can be derived and that can be used determine evergreen forests versus other land cover type – in a broad snow hydrology context, those are the two main distinctions (with reference).