General comments

In this work, the authors established a comprehensive vehicular NH3 emission model and compiled a gridded on-road NH3 emission inventory with high spatial (3 km × 3 km), and temporal (monthly) resolutions for mainland China using published NH3 emission factors of motor vehicles and their relevant impact factors. With this high-resolution emission inventory, vehicular NH3 emissions during the period of 2000-2019 were estimated. The authors showed that vehicular NH3 emissions could exceed agricultural emissions in the densely populated areas, especially for the extreme populous megacities such as Beijing and Shanghai. Although this conclusion is not unexpected, the paper gives a quantifiable and reliable result, which is valuable for future study. The paper is overall well written. I still have some doubts about the uncertainties of the vehicular NH3 emission inventory and some minor questions as listed below. I recommend publication after these issues are addressed.

Specific comments

1. Line 100: “Bottom-up estimation of long-term vehicular NH3 emissions”. My major concern here is whether the authors have considered the additional impacts of the enhancement of driving conditions caused by traffic congestion in densely populated areas on vehicular NH3 emission factors? Or maybe the authors can discuss some uncertainties caused by this factor in the consequent sections of the text. After all, this paper focuses on the importance of NH3 emissions from motor vehicles in densely populated areas.

2. Line 190: I suggest that the temporal distributions of vehicular NH3 emissions can be moved to the main text, because this topic is one of the novelties of this study. It would be better if the authors could provide a set of temperature-depended NH3 emission factor correction factors for reference.

3. Line 196: If possible, I suggest that the authors could add more discussions on NH3 emissions from residential sources. According to Figure S7, Beijing and Shanghai also have a relatively high proportion of NH3 emissions from residential sources. Their emissions in each season are even higher than those from motor vehicles, and their emissions, if not unexpected, should also be mainly concentrated in densely populated areas.

4. Line 219: Due to the lack of measurements of vehicular NH3 emission factors, NH3 emission inventory still has large uncertainty on the whole. Especially for diesel vehicles, abnormal urea use and different SCR control strategies will affect its NH3 emission factor. Therefore, it is suggested that the authors could discuss more on the uncertainty of vehicular NH3 emission inventory.

During the past years, PM_2.5 pollution have been reduced substantially, while the occasionally occurred heavy PM_2.5 episodes and its driving forces still need to be explored. One of non-common sense hotspots is the role of NH₃. The relative importance of traffic sources to NH₃ emissions is still under debate.

Base on emission measurement data throughout different cities in China, this study developed high-quality traffic emission inventory of vehicular NH₃ emissions. This work can give a better insight into the absolute value and relative importance of vehicular NH₃ emissions in different regions, seasons and population densities in China. According to the results, they show that the significant role of on-road NH₃ emissions in populated areas have been underappreciated, which is quite important in terms of the atmospheric chemistry and air quality implications.

Overall, this manuscript is well organized and presented with some new insights on NH3 emissions and its contribution, thus, I think this paper is suitable for publishing in ACP after well addressing the following comments, questions and suggestions.

1. As shown in Fig. 2, NH₃ emissions from gasoline vehicles have already declined since 2010 while emissions from diesel vehicles grew significantly since 2014. What’s the most possible trend in total on-road NH₃ emissions in the near future under the join effects of vehicle growth and turnover?
2. I witnessed an overall higher level of EFs estimated from remote sensing than dynamometer measurements in Fig. 1. What’s the possible reason? Some discussion should be added.
3. Why the shares of vehicular NH₃ emissions in total anthropogenic NH₃ emissions show a large difference between the urban areas Shanghai and Beijing? More detailed discussion is required to quantitatively address the uncertainty (e.g., NH₃ emissions from residential or industrial sectors).
4. This paper mainly focused on emission inventory but didn’t reach to the real impacts on air quality. I understand that the scope of this paper may not be stretched further, but still wonder whether there’s any evidence for the air quality impacts of on-road NH3 emissions in urban areas?
5. The significance, shortage and implications of this study is suggested to be added in the Conclusions sections.

This study provides a comprehensive vehicular NH3 emission model with useful insight into spatial and temporal variations of vehicular NH3. The important role of NH3 emissions from vehicles in urban areas with higher population densities is highlighted, which could have important implications for PM2.5 and haze events. Overall the paper is well written and I recommend publication if the comments below can be addressed. 

• Section 2.1. Please clarify how the NH3 emission factors were obtained. For gasoline vehicles, was NH3 measured directly or predicted based on correlation with MCE? Further information on sample sizes and whether the data represents a wide range of driving conditions is needed. What are the uncertainties associated with the NH3 emission factors?
• Line 94 - 95 explains that NH3 emission factors of other diesel vehicles were calculated based on the relative fuel consumptions compared with HDDVs. It would be useful to highlight any limitations of this approach. It is also stated that the NH3 emissions varied significantly among tested HDDVs. How did you account for this?
• Many findings e.g. total vehicular NH3 (32.8 kt to 87.1 kt NH3 from 2000-2019), proportions of NH3 in different provinces (e.g. 8.91%) will be affected by the uncertainties in the NH3 emission factors. Provide estimates of uncertainty associated with these statistics.
• Does the compilation of gridded NH3 emission inventories account for any effects of different traffic conditions?
• Figure 1. The authors should refer to the SI, which explains how g/kg EFs have been converted to mg/km. It is useful to explain potential reasons for observed differences. For example, the derivations of mg/km emissions from remote sensing have not been adjusted to account for different driving conditions / fuel consumption, whilst dynamometer measurements may be lower than on-road emissions. Farren 2020 (ES&T) could be useful for mg/km NH3 EFs.
• Section 3.1. The literature suggests NH3 emissions from gasoline vehicles can increase as vehicles deteriorate / vehicle mileage increases. Do the trends consider this effect, which may be particularly important in the future if gasoline car ownership is increasing? It would also be useful to state the proportion of the proposed increase in NH3 from diesel vehicles that can be attributed to HDDVs and therefore how this may change with implementation of China VI.
• Conclusion. This study provides useful insight into vehicular NH3 emissions. It is recommended that the conclusions address the limitations of this study and how this could be improved in the future to better understand the air quality impacts of vehicular NH3.

Technical corrections:

• Use of informal language e.g. line 41 ‘What’s more’, line 154, line 176.
• Line 144: ‘The monthly variations compare well’
• Line 168: ‘might be probably controlled’ - be more specific
• Line 198: ‘among various population densities.’
• Line 207: should this be 20,000 person/km2?
• Line 236: ‘more severe’
• Line 244: ‘Euro 7/VII vehicles comply’