General Comments

The Authors present a detailed foray into stable isotope source apportionment of NHx in a urban area to ascertain the relative importance in accounting for vehicle and industry NH3 emissions when setting air quality policy. They combine a robust observational dataset and isotope mixing model with archived NHx observations to unravel the sources contributing to ambient levels in Providence, RI. Overall, the manuscript presents a strong case for the prevalence and significant contribution of vehicular NH3 to urban air. Pending minor revisions below, the manuscript will be suitable for publication in Atmospheric Chemistry and Physics.

Minor Revision

1. The most substantial oversight in interpreting the long-range transport component of the NHx observations comes from the inferred agricultural regions of Canada that are upwind. While these areas are agriculturally intensive, the Authors seem to have overlooked that one of the 3 largest cities in North America, Toronto, is in this same fetch. It is a potent source of vehicular and industrial NH3 emissions. Similarly, Montreal and additional industry along the St. Lawrence River in Quebec are overlooked. The Authors should consider these and probably revise their stress on wildfire (biomass burning) emissions towards those from more urban and industrial sources.

Technical Comments

Page 6, Line 165-168: Given the sampling setup, it seems strange that you'd find NO2- in these extracts. Are the sources for this known? Is it an impurity in the reagents used for coating and sample preparation? A bit of elaboration on this could help reduce confusion on why this issue arises. Maybe give the breakdown of the criteria that were unsatisfied here (e.g. XX% of rejected samples were because of criterio 2)? This way the propensity of each issue is clear.

Page 9, Line 238: Should ‘fertilization’ be fertilizer?

Page 11, Line 294: Should % instead be ‰?

Page 11, Lines 308-310: A technical oversight here that there could also be increased physical losses driving the observations. The dry deposition loss of NH3 to the Earth’s surface is sustained during transport, and this would happen more readily in the winter. Suggest revising with the alternative process also used to provide context. There is plenty of literature on the subject.

Page 12, Line 332: ‘sewerage’ should be ‘sewage’ and appropriate references need to be added at the end of this sentence.

Page 12, Section 3.5: Implement consideration of the minor revision regarding urban and industrial areas in Ontario and Quebec throughout.

Page 12, Line 350: Deposition and revolatilization is not limited only to agricultural NH3, but any emitted NH3, which will then hop downwind and fractionate along the way. Suggest revising here.

Page 12, Line 354: Wildfires are not that common. Suggest retuning for urban emissions upwind.

Page 13, Lines 368-370: This is likely mixed in with volatilized NH3 from Canada. Revise the discussion in this section to be a bit more aware of sources in the region.

Page 13, Line 375: It isn't clear how SCR differs from vehicle sources of NH3? Here and below these are mixed and there doesn't seem to be a definition on why these are considered separate sources? I eventually found a description in the SI, but that example should be given here. From the SI it is not clear whether all power generation in the US is equipped with SCR by law? Couldn’t there be quite a lot that is not? If that is the case, should SCR be used so specifically in the main manuscript? Consider revising for clarity.

Page 13, Line 377: Indoor sources requires an appropriate reference(s).

Page 14, Lines 409-411: In summer, your source region is strongly influenced by the heavily urbanized Toronto and shoreline area.

Page 14, Line 424: This line of reasoning about wildfires doesn't have much factual justification. Is this speculative or can quantitative metrics be brought to bear on this? Suggest removing throughout if it is speculation as a major Canadian source. Biomass burning/wildfire events intruding into the US are largely confined to those originating in Alberta and British Columbia.

Page 30, Figure 7: Can an accurate representation of the great lakes be added to these figures? Since there's discussion regarding Ontario and Quebec linked to this, the major geographical features that make them identifiable should be properly plotted here.

Supporting Information

Page 4, Figure S3: Given the specific commentary on sources of NH3 from Canada, have the authors looked into acquiring CAPMoN or NADP datasets to include the source region more completely in this visualization?

Page 7: ‘may closer represent’ should be ‘may more closely represent’

Page 7: ‘…. Representative of all industry-related NH3 emissions’. Since steel industry uses coal in the coking process, this isn't too surprising to see a similar trend to the values found from fuel combustion. Values from natural gas or oil-fired facilities would provide a greater level of confidence that all fossil fuel fired power or industrial activities yield a similar NH3 signature.

Page 7: why does ‘14’ follow the per mil value?

This study presented very interesting results on the ammonia and ammonium concentrations and isotopic compositions at an urban site, highlighting the importance of urban ammonia emission to the regional atmospheric composition. The year-long observation provides valuable information for evaluating the ammonia emission inventory which is currently under strong debate. In addition, this paper presented detailed observation about the isotopic compositions of both NH3(g) and NH4(p), emphasizing that the isotopic fractionations between these species are significant and highly variable. In general, this manuscript is well written and easy to follow. However, some part of the discussion could use a little bit more clarification. I suggest a minor revision.

Lines 246-248: as you mentioned in the previous section, pNH4 is highly correlated with nitrate and sulfate aerosols in the atmosphere, meaning the partitioning between NH4 and NH3 should be mostly controlled by the atmospheric composition. In another word, the amount of pNH4 should be determined by multiple factors, such as the total emission of NHx, NOx and SO2 concentration in the atmosphere, and maybe water content. So, simply based on the observation that pNH4 concentration are similar among these sites, I would not argue “[pNH4+] may be more regional representative than [NH3] due to its extended atmospheric lifetime relative to NH3”.

Figure 1a: for such a small figure, perhaps it is better if you can only show the sites discussed in this figure and remove places like Boston and New Haven as they are not talked about in this paper.

Lines 286-299: I would recommend a more detailed discussion here as it presents interesting results of the isotopic fractionation between the two species, at least moving Figure S2 into the main text and discuss a little bit more about why the seasonal cycle occurred.

I would argue against the role of R4 in explaining the variable isotopic fractionation. NH3 has a very high pKa (9.26), and aerosol water is usually acidic (at least pH<8), so in aqueous aerosols, NH3(aq) is almost non-existent, almost 100% NH4+. So, only R2 and R3 should be important, and the equilibrium fractionation factor should be somewhere between 31 to 34 permil. The lower isotopic fractionation in the summer may be more related to 1) lower isotopic fractionation factor observed by Kawashima and Ono (2019), and 2) a more important role of kinetic isotopic fractionation.

Line 400: discussing how the relative contribution change by season without considering the change in concentration can be misleading. For example, the relative contribution of vehicle emission is lower in the summer and higher in the winter – however once you fold in the concentrations of NHx, you can see the contribution of vehicle emission (in ug/m3) may not change that much. I would recommend the authors revise Figure 8 to include the concentration information, so it is clear for the readers to see the variations of each source, as a sanity check.

In addition, using the source appointment results, the authors should discuss how it is different from NEI2014. It is clear that our observation showed strong discrepancy vs. emission inventory. For example, emission inventory suggests higher emission in the winter, considering lower mixing height winter, the NHx concentration should be significantly higher than summer, but we see the opposite trend. Isotopic evidence should at least tell us if the seasonal trends of the sources in the emission inventory agree with our observations.