|The authors have taken on board many of the most serious comments, and they have made important improvements to the paper based on this. I still find the interpretation too strong (too definite) in places, and feel that one more round of edits is needed before publication, to ensure that tentative conclusions do not become embedded in the literature as facts. However none of the changes are major ones and it should be possible for the editor to manage the remaining changes. Once the changes are made, this will be a useful paper.|
Abstract, line 19. I suggest “We propose that enhanced atmospheric NO3-…” to emphasise that this is a new suggestion that needs further confirmation.
Abstract lines 25-27. This statement about the melt should go up to line 16. Then the next sentence would start “Accordingly, we use records to investigate the emission sources and…”.
Page 8, line 21. I still emphasise that, if melt was important at the decadal timescale then this would not be likely to cause a correlation between melt and chemistry at the decadal timescale, because it is very complex to estimate how a pattern of melt will result in chemistry going in and out of particular layers. I propose “Although we do not expect a simple correlation between melt in a particular layer and net chemical flux, this result suggests that there is at least not a net loss of chemistry from layers of extreme melt, which together with the findings of ….supports our conclusion that melt has not been a major influence on the ion concentrations”
Page 8, line 24. In principle sulfate could be an indicator of marine biogenic. This would not be ruled out by the PC1 influence but would perhaps be expected to be seen in PC2 with MSA. So your conclusion may be OK, but needs a few extra words.
Page 9-10. I think the conclusion that the MSA change around 1900 is due to a major change in ice extent in the west Nordic Seas needs to be made extremely tentative, as it rests on a lot of very uncertain conjectures.
1. It assumes there are good sea ice reconstructions, but these are really tentative, with many of the proxies involved being generally considered proxies for temperature rather than ice extent. However, having said that, the calibration data that Macias-Fauria et al used (Vinje et al) does also show a sharp decrease in ice extent across 1900, and this should really be included as part of the argument.
2. You claim a good correlation between MSA and W Nordic sea ice over 800 years. However, this is probably dominated by the step change at 1900, and anyway since you are about to explain MSA through nitrate, you should not be expecting a strong correlation before 1900 with the ice extent (otherwise the nitrate argument is redundant).
3. While it is clear that MSA is affected by sea ice, the exact relationship you propose is handwaving and unclear (more ice does not necessarily mean more ice edge, and may mean the ice edge is further away so this could be argued either way. Probably, as in Antarctica, it is the seasonal contrast that matters as it is the retreating ice that leaves the ocean primed for productivity). It is certainly unclear why a 25% reduction in ice extent would give a 50% reduction in MSA, especially given that the nitrate increase would have been expected to lead to a significant increase in MSA.
So, in my view your discussion of Fig 6 essentially leads to the observation that there is a coincidence in timing of the MSA step change and the sea ice change, and this might tentatively be causal, but should be definitely labelled as tentative.
Page 12, line 23. The preindustrial ammonium values are not the same order of magnitude as at Lomo97 unless you take a very literal meaning of the phrase. They are a factor 2-3 different, which you already implied might be due to contamination of Lomo97. Please change this statement.
Page 13, line 1. I don’t understand on what basis you are asserting that increasing anthropogenic emissions are not important for ammonium until 1940. Fertilisers have a long history. On the last line of page 12, you say “As discussed above” but I couldn’t see where you made that case. The reasoning seems circular to me, and probably to be based mainly on the fact that you already made the same assertion for Belukha. Please make it more tentative again.
Page 14, line 4. As per comment on page 12, ammonium values do not “reasonably agree” with those of Lomo 1997. Please remove this statement.