High Latitude Dust Transport Altitude Pattern Revealed from Deposition on Snow, Svalbard

Round 1

Reviewer 1 Report

The last few years in dust research have revealed that dust generated at high latitude environments such as Antarctica or Alaska can be lifted in significant quantities and observed over snow surfaces nearby or far away. There is a growing number of reports demonstrating that such dust activity in cold environments is more common than previously thought. This paper is a contribution to the general body of new knowledge on this subject. While not reporting something totally new, it does highlight a particular aspect of dust activity in the Arctic region that has not been widely noted. This paper reports observations of dust sampled in snow at different heights at the mountains of  Svalbard island in the N. Atlantic. Dust concentrations and composition are parsed as a function of height. A clear decrease of dust concentration in snow as a function of height up to a threshold relative height above which such correlation is broken. It is suggested that the dust deposited below the threshold height is from local sources whereas the one above is not. These results are consistent with previous studies in other high latitude glacier environments. Data analysis and readability is acceptable. I recommended to accept the paper for publication. Some minor comments are suggested. 

Abstract

 

Line 17 "replace with , "between the altitude of sample taken and its concentration

 

Line 20 replace implies with "suggests a "

 

Line 22 "drier"

 

Line 23 replace "allow" with "favor"

 

Line 134 what four samples are you talking about? not clear , this is the first instance that four samples were taken. Please clarify

 

Line 268  There is not reason to add this sentence, these measurements were made in  Antarctica, . If you choose so, add more context to this line as it does not make sense just leave as it is. 

 

292 This is a very different environment, not clear if the comparison is valid. Please show why this is valid. 

Author Response

Dear Reviewer,

first of all, thank you for your comments and suggestions, we appreciate the time you spent with reviewing the manuscript. We tried to incorporate all your suggestions and we listed our replies in the uploaded word file (replies are highlighted in yellow).

On behalf of the authors,

Jan Kavan

 

Author Response File: Author Response.docx

Reviewer 2 Report

The paper deals with the deposition of dust on snow in Svalbard at two different locations, one being more on the south tip of the archipelago and with larger ice and glaciers fields around, the second one being more central and drier in general, with more exposed ground to the atmosphere. The paper is in the scope of the journal, is concise and generally well written. The gathered data support the analyses and conclusions, but the discussion should be expended, particularly on the melting of snow. More details on the sampling and uncertainties should be given. Specific comments are below.

Major comments:

On uncertainties, there are basically none, while it exists for any kind of measurements. First, the snow sampling section 2.1 should be a bit more detailed, and the sub-sampling method described before the fact you merged them. Density is then estimated from the 4 sub samples, and uncertainty should be given, which impacts the estimate of dust fluxes deposited on the snow.

Same for the dust deposition estimation. Blank control filters on weight (dry) should be given, there is likely a variability in the weight of each individual filter. Same when they got wet and dried. Since the mass measured is not given, neither the weight of the filter, it is impossible to estimate what is the impact of that on the results. Uncertainties and scale type should be given.

The results show that there is more dust deposited at the bottom of the slopes, which seems to be coherent with the fact that it is more difficult to lift dust higher up, unless it is long range transport, as discussed in the paper. It goes also in the way (which is not discussed), that the higher parts of the slopes get snow first, and get lower dust amount, so a reduced deposition time. This is also in favor of a local source and not long range transport otherwise the entire slope will likely be covered by such a deposition event, i.e. long range.

However, there is no discussion on the snow cover, which could of course largely affect the amount of ground available to become a source. Maybe some more sentinel pictures could be obtained to discuss that. More important is that there is no discussion on snow melting. The lower part could have experienced melting, during winter maybe, but for sure in Spring, as shown on the temperature records. That will for sure concentrate the particles in the snow. Now the changes in terms of concentration from the bottom to the top is a factor of 10, and I doubt there is 10 times more snow on top. Nevertheless, melting is an important discussion point, especially when sampling snow in summer time.

Minor comments:

Introduction: it is not very clear where the study is located in the introduction as the literature presented covers many different places, from north to south poles. It is in fact only in section 1.1. that we discover Pyramiden as being a studied location, while I would have thought Ny-Ålesund and Hornsund from the introduction. Some rephrasing and clarification should be done.

Line 100: “and location of the meteorological site”. What do you really mean ? That the site is influenced by the seashore close by, or the location of the met station itself is influenced by it (so the data), more than the surrounding mountain ?

Fig 1: box C and D are mixed between text and figure, either swap in the fig or in the legend

Line 114: delete “The”

Line 115: “… eastern slope incline is in between”

Line 270: only Spring ? There was no snow in winter ? I guess it simply covers the snow season, which now the duration is hard to predict in the Arctic.

While the comparison with Antarctica work is fine in the introduction, I am not sure it stands much place in the discussion as Antarctica is 98% ice covered, so the local source amount versus the ice field is just ridiculous compared to Svalbard (Line 292-295). What is deposited in Antarctica is almost only long range transport.

Author Response

Dear Reviewer,

first of all, thank you for your comments and suggestions, we appreciate the time you spent with reviewing the manuscript. We tried to incorporate all your suggestions and we listed our replies and comments in the word file below (replies are highlighted in yellow).

On behalf of the authors,

Jan Kavan

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

This is the second version of the paper. The authors answered to the main points but there is still an issue which needs to be addressed regarding the sampling and the estimate of the dust load.

In the first version, the sampling was described as subsamples of 10 cm deep, and 10 cm in diameter. In the new version, V2, it is now 480 mm deep and 750 mm in diameter ? What is the correct value ? And what the word "depth" refers to ? Is it the depth of the sampling ? Or the thickness ? But if the thickness, is it from the surface ? Does that mean you do not take the entire snowpack ? That needs to be extremely clear as any sampling strategy for snow defines the interpretation of the results. Furthermore, the estimate of the volume of the sampling for the density (bulk from the 4 subsamples) refers to a volume 1.54 dm3, but how that number of 1.54 dm3 has been obtained ?? I could never reach the same value when calculating the volume of a cylinder using the V1 or V2 version of the given sampling volume. Such confusion between two papers versions is not very professional.

Thanks for having provided the estimate of the uncertainties for the weight measurements, but please add these value into the table (at least under, so the reader does not need to search into the text). What about uncertainty in measuring the snow density ? What about the natural snow variability ? It is often considered at least a 10% in density measurement, simply due to the variability, and 4 samples are not enough to overcome it. If you would have measured the density of each individual one, then you would have got an idea of it, but as you merged them, you can't. You have to at least add a 10% in density uncertainty, and include it in dust load uncertainty, on top of the filter weight one.

Some English editing would be welcome.

Conger, S. M., & McClung, D. M. (2009). Comparison of density cutters for snow profile observations. Journal of Glaciology, 55(189), 163–169.

Author Response

dear Reviewer,

thank you for your comments, here we provide our replies:

1/ first of all, thanks for noticing the mistake regarding the snow sampling volume! I realised that there were dimensions of another type of sampler in the first version of the manuscript during the final check before resubmitting and quickly corrected it. Unfortunately, as I was pressed by the deadline, I used, by mistake, the external dimensions of the sampler (tube with 75mm diameter and 480mm length). The inner diameter is 70mm and I took upper 10cm of snow. This gives the total volume of 1.54 dm³ for all the four samples together.

The dimensions were corrected and the text was adjusted to clarify this issue.

2/ I agree that the snow density estimates are subject of certain level of uncertainty, which I tried to overcome using a bulk of four subsamples. Unfortunately these subsamples were directly merged, which restricts the quantification of such uncertainty (as you mentioned) – the snow density estimation is now commented in the discussion using the reference you provided. However, the snow density uncertainty does not affect the estimate of dust deposition as this is completely the issue of material being deposited on the sampled AREA (g.m^-2).

 

thanks for your work, with regards,

Jan Kavan