General comments

This paper describes the implications of standard/state-of-the-art NH₃ measurements taken in Beijing between 2009 and 2020. NH₃ is an air pollutant that is principally emitted from human activities, affecting ecosystems, air quality and climate. The manuscript is a good fit for the ACP audience. I found the methods used to be generally sound relative to the conclusions drawn, and the paper is well written and clear. I would recommend this article for publication after the following comments are addressed (some major, some minor).

Specific comments

Major

• I feel that there should be better acknowledgement of the uncertainty inherent in relocating the measurement in 2017. While there have been references showing that the NH₃ vertical profile is relatively constant up to 300 m in Beijing, a measurement on the 14 story could be on the cusp of surpassing 300 m. There has also been recorded greater variability in the vertical profile in other locations. The manuscript raises that there is spatial variability in NH₃, so it’s not clear why this couldn’t also extend to measurements of the vertical profile (in other words, could the vertical profile of NH₃ differ in different parts of Beijing?). In my opinion, it would be better to acknowledge the uncertainty associated with this move, raise plausible hypotheses, and indicate that there is too much uncertainty to draw confident conclusions about the drop in 2017. Could emissions or changes in the urban topography have contributed? These aren’t discussed here but might also be relevant.
  • The authors may also consider splitting the dataset in 2017, but it is probably worth maintaining the dataset as continuous because the horizontal distance between the two locations is so small.
• I found it strange that there is an anticorrelation between NH₃ and temperature (on an average daily basis) and feel that the authors should include more interpretation of this finding. Are there any other papers showing an NH₃-T anticorrelation?
  • Figure 3 - To my eye there generally appears to be a positive correlation between air T and NH₃. The text starting on L241 also suggests there is a pretty even balance of days with positive or negative correlation between NH₃ and air T (my understanding is that this is correlating hours within a day). It seems to me that there could plausibly be a weak negative correlation on a daily average basis, but I don’t understand how there are strong negative correlations between air T and NH₃ for most seasons in L246. Is this correct? Otherwise, could the interpretation among these relationships between air T and NH₃ be expanded to clarify?

Minor

• Generally, I found the abstract to be a well written and succinct summary that draws out keen points of interest for this manuscript. My only suggestion is to strengthen the motivation for controlling NH₃ concentrations in the last sentence. The preceding sentence expresses that SIA concentrations are not very sensitive to NH₃, so it could be helpful to provide additional rationale (for example, that reducing NH₃ is cheaper? It’s more effective once NH₃ has been reduced to some extent?—both discussed in the paper).
• L38 – suggest clarifying that “particulate matter 2.5” is “particulate matter with a diameter less than 2.5 µm in size”
• Sentence starting on L36 (“However, long-term…”) – I think that this paragraph motivates the utility of long-term trends in ground-based atmospheric NH₃ well, but it would benefit from expanding on the urban aspect of this study. For example, are NNMDN sites generally in rural areas?
• L78 – Describe source of emission data in Figure S1 either in the main text or figure caption.
• L79 – I assume that the inlet is outside of the building—could you clarify?
• L80 – Later in the manuscript it is noted that there is not a strong vertical profile of NH₃ in Beijing (although this has been observed in other locations). It may be worth mentioning that here as well.
• L100 – Do you have any thoughts on why the slope is higher in this study than in Zhang et al. 2021?
• L112 – Could you please clarify why you switched the source of the met data?
• L122 – Could you please clarify where is the Chinese Academy of Meteorological Sciences relative to the other measurement locations of interest?
• Section 2.2.1 – It would be helpful to briefly state why this particular method is being used (to gap fill?).
• Section 2.2.2 – Could you please describe the model configuration? For example, are the PM₅ ion measurements used to run the model? It might be helpful to make that connection and remind the reader of that available data.
• Table S1 – It’s not totally clear what’s gained from comparing this study’s results to other studies located in different parts of the world. I am also unclear on what the numbers in the “This study” column represent – are those the averages from the same timeframe as the study that they are being compared with? Please expand/clarify.
• L227 – I don’t understand how the variation is being calculated. It is described as “average annual temperature” but then specified by season. Please clarify?
• Generally, I felt that the authors could be clearer about the specificity of their results for Beijing (as appropriate), in particular in the conclusions and when discussing the policy prioritization in the abstract (e.g. “measures to control NH₃ concentrations should be prioritized [in Beijing].”).
  • L367 – reiterate that this is for Beijing
  • L376 – “Therefore, reducing acidic gas emissions is still a primary focus for controlling fine particulate matter pollution [in Beijing].”
• L376 – Could you recall the evidence supporting “And in the future, more attention will be needed to focus on controlling NH₃” ?

Technical comments

• There is often a missing space between a word and the opening parenthesis for the citations.
• Figure S3 caption – It’s a little confusing that the subpart labels first follow and then precede the description of the content. It would be more intuitive to keep it consistent (e.g. “Monthly (a) and annual (b) variations and correlations between satellite observations… 116.5E) (c) and the average observations… ~118.5E) (d).”
• Consider referencing Figure 1 earlier, where the results are first mentioned (I believe in L172?)
• L374 I found this wording awkward: “the current reduction of SIA remains less significant in response to NH₃ than acid gases.” Consider: “SIA formation is more sensitive to acid gases than NH₃.”

This study presents 11-year NH3 data in an urban environment and explored the long-term trends of NH3, the influence meteorological variables played on NH3, and the role NH3 played on secondary inorganic aerosol formation. While the data presented here are useful, the analysis can be improved. More importantly, the discuss and presenting quality needs significant improvement. More specific comments are provided below.

Abstract:

Abstract needs a significant revision to better summarize major findings. Too many general statements, but lack of specific results.

Line 13, better replace “The total average” with “The 11-year average”.

Lines 14-15: Why not show the total percentage increase between 2009-2017 and percentage decrease between 2017-2020? This way can better reflect the two contrasting trends during the 11-year period.

Line 17: This sentence does not provide any useful information. Most pollutants would have seasonal and diurnal variations. You need to specify what kind of seasonal and diurnal patterns.  

Liens 18-19: the non-linear relation is well known in literature. You need to show some specific results. Same comment for several other sentences in this section.

Introduction

This section also needs some major rewriting. The topic of this research is on urban NH3 and its long-term trends. Thus, the Induction should cover these areas: (i) brief discussion on the important role NH3 played in various research areas (this is covered in the current Introduction, but could be simplified since such materials are rich in literature); (2) brief discussion on the major sources of NH3 in urban environments and major debate on this topic in literature, noting that existing studies have different opinion on the major sources (this is not mentioned at all in the current Introduction and should be added in the revision. Such materials may help explain the trends in Section 3.1); and (3) brief discussion on the studies of NH3 long-term trends, first worldwide (see Yao and Zhang. 2009, ACS Omega, 4, 22133-22142, as an example), then China, then Beijing. Then point out the knowledge gaps based on the summary of the knowledge presented above. Finally present the goals of this study.

Materials and methods

Line 79-80: need to specify the height above the ground of the two measurement sites (after the third floor and 14^th floor).

Line 90: change “subjected” to “subject”

Line 126: change “2.1 Methods” to “2.2 Data analysis methods”. You already used Methods for Section 2, here you need to use a more specific sub-section title.

Line 134-135: This statement is not accurate. EEMD has also been used in air-quality trend analysis studies in more recent years (for example: Yao and Zhang, 2016. ACP 16, 11465-11475.  Wang et al., 2022. Environment International, 159, 107031. Wang and Zhang, 2023, Environmental Pollution, 333, 122079). You need to cite more relevant studies instead of not-so-relevant studies.

3 Results and discussion

This section needs a better organization and more in-depth analysis.

A large portion of Section 3.1 is used to compare the NH3 data with literature, and the discussions on the trends and associated causes are limited. While comparing to literature data is needed, it should not be the main focus of the discussion. Besides, from Lines 190-191: if this statement is true, then it means that the uncertainties in the obtained trends (due to changing location and measurement height) are larger than the actual trends, making your discuss on trends meaningless.

I would recommend reorganizing Section 3.1 in this order: First present the trends from the monitored data using a quantitative statement, e.g., either using annual decreasing/increasing rate, or percentage decrease/increase during a period. Split the 11-year period into two periods since contrasting trends were observed during the whole measurement period. Use quantitative statements wherever possible and show the significance level of the trends. Then present the trends generated from the satellite data using the same rules as described above. Then discuss the similarities and differences between these two sets of trends, only at this stage you need to cite literature data to support your results and/or provide explanations on the causes of the differences between different data sets.

In general, long-term trends are mainly caused by emission changes and to a much less degree by meteorological factors. After discussing the trends in Section 3.1, you can then discuss driving factors of these trends in section 3.2, first focus on emission and then on meteorology. Emission inventory related discussion in Section 3.1 can be moved to section 3.2 to support the discussion.  See Lin et al. (2022 ACP, 22, 16073-16090) to get more ideas related to this comment.

Section 3.3 also has too many introductory materials in the first two paragraphs. Follow this rule, in the Results section, present your own results first and use literature data to support your discussion, instead of summarizing literature results separately (which really belong to Introduction section).

Conclusions

Line 353: “3” should be “4”. Avoid such simple typos.

Lines 354-356: Preneet the two different trends (in two periods) using a quantitative statement.

Line 356: have you tried to identify the actual causes of such discrepancies between NH3 concentration and NH3 emission?  See possible causes on the same topic in Yao and Zhang (2009, ACS Omega, 4, 22133-22142).