|Macdonald et al revised|
This draft of the manuscript is significantly improved from the original submission. I am still far from convinced that the “effective deposition velocity” is a useful construct, but it is obvious that the authors feel that it is.
I also feel that the further attempt to explain the very high EVD for Ca^2+ is not convincing. It is troubling that the detection limit (reported in Supplemental Table 6) is so much higher (more than a factor of 7) than all other ions, suggesting very large variability in the blanks, or a serious analytical problem. (Note this is not true for the aerosol composition detection limits in Supplemental Table 8). The authors assert that qualitative agreement with ICPMS measurements of Ca gives confidence in both data sets, but only about half of all possible samples were above detection limits on both instruments. I find it hard to believe that so many samples were below ICPMS detection limits, raising concern about that data set as well. Given that the authors can not really explain the high EVD, they might be better served leaving Ca^2+ out of the story due to analytical issues causing large uncertainties.
One thing that the authors may want to expand on a bit more is the suggestion that the study period may have been unusual due to the active Icelandic volcano (mentioned in just one sentence near the end of discussion; pg 17 lines 2-5). Provision of the detailed aerosol data in Supplement Table 7 really makes it clear that the canonical “Arctic Haze” pattern for sulfate (and other pollution and dust derived constituents that usually covary with sulfate) with elevated concentrations in late winter into spring, generally peaking DJF, was not obvious. The really striking observation to me is that sulfate deposition to/with snow is so weak in the winter time , even in a year that may have had the usual haze peak in the atmosphere plus additional enhancement from the volcano. Cliff Davidson suggested very weak scavenging of sulfate by winter snow near Dye 3 way back (several papers before DGASP), based on snowpits that showed winter enhancements that were not proportional to increase in aerosol sulfate at Barrow, Alert, NyAesund during the haze season. DGASP showed that Haze really does not impact sites high on the Greenland ice sheet, so the inferred weak scavenging was not confirmed over Greenland. (Cliff actually claimed marked seasonality of sulfate scavenging, stronger in summer, since he did have aerosol measurements in summer that showed low sulfate compared to what he assumed it was in winter/spring.) Now it may be that you are confirming weak removal of sulfate during winter for Alert, and perhaps other parts of the Arctic basin, but this probably means untangling any impacts on this winter from the volcano.
As noted last time, the authors really need to look at Dibb et al., 2007 (Seasonal variations in the soluble ion content of snow at Summit, Greenland: Constraints from three years of daily surface snow samples, Atmos. Environ., 41, 5007-5019). As suggested by the title, this paper reports on 3 years, through the winters, of daily snow sampling at Summit, including ion concentrations and fluxes, and specifically separates freshly fallen from surface snow aged days to weeks. It probably needs to be considered in relation to the summary of previous work at top of page 2, and contradicts the statement made in lines 4-5 on page 3. Also, a main point of Dibb et al., 2007 was to compare seasonality of ions in snow falling at Summit to the seasonality inferred from snowpit studies, so it would seem an additional valid source for the comparisons made at the end of section 3.1 in this manuscript (page 9, lines 5-15).
P3/L30 is unavoidable and natural within natural snowpack systems. ---→is unavoidable within natural snowpacks and still expected to smaller extent on the snow table.
P3/L30-32 It is actually not necessary to know the depth of a sampled surface stratigraphic layer to calculate fluxes, in fact the depth can, and often does, vary horizontally. All that is needed is the area sampled and the mass of snow in that area (e.g., g snow/cm^2). Given concentrations in mass or mole per g of melted snow, conc x g snow/cm^2 give the amount of chemical in the layer per cm^2. The harder challenge is to only sample from single stratigraphic layer, but that can be done with care.
P4/L3-7 Possible underestimate of snowfall up to a factor of 10 seems too large to just pass over. Here in the text it is stated that large discrepancies between operator reports and a met station separated by 6 km justify ignoring the station date, referring to section 4.2 in the supplement for details. However, there are no details on the discrepancies in the supplement, just a rephrased version of the same statement.
P4/L30 Be more specific what you mean by underestimate of BC mass due to size cut-off. First point is that 0.02-50 fg is not directly a size (need to assume shape factor and density). Also not clear whether you are worried about truncation on the small end or the large end of the distribution (both could be possible). It also seems that the cited mass range is based on the SP2 detection principle, probably also need to think about any issues related to transmission efficiency of the nebulizer.
P7/L3-6 and Table 1 It is a little confusing that the list in the text does not overlap very well with the elements in the table (e.g. Ba, Ti, K, Cd, Ca, Cr, and Na are in text but not table while V, Se, Sb, and Tl are in the table but not listed in the text). Could either make the list in text agree with this table, or expand the list to include all of the elements in supplemental tables S3 and S4 and make it clear you are referring to those tables and not Table 1.
Footnotes to Table 1. MSA, ACE, PRP, and FOR are abbreviations, so OK to define, but C2O4^- and NH4+ are not. If you need to define these ion formulae, you probably need to define all of them. Better choice would seem to stop after the first 4.
Related comment, MSA usually signifies methylsulfonic acid, not the ion methylsulfonate that you correctly indicate was measured by IC. I think that MS^- would be better shorthand to indicate methylsulfonate (here and everywhere else in the MS).
Sections 3.2.2, 3.2.3 and section 4 Need to be consistent and show the valences for all of the ions everytime they appear in the text. Right now only a few are shown. This is especially important since you also discuss results of elemental analyses for Na, K, Mg and Ca. Throughout these 3 sections the references are all to the ions, not elements.
P12/L6 Not so clear that NH4^+ is always dominant compared to NH3. Maybe in the Arctic, and during cold season, but this reads like a global declaration.
P12/L8-9 As noted earlier, Ca^2+ is a striking exception to this statement, for no plausible reason. Do you really think this is real?
P13/L8 seems “in-cloud scavenging” should be “below-cloud”