Unchanged PM2.5 levels over Europe during COVID-19 were buffered by ammonia

Evangeliou, Nikolaos; Tichy, Ondrej; Svendby Otervik, Marit; Eckhardt, Sabine; Balkanski, Yves; Hauglustaine, Didier

doi:https://doi.org/10.5194/ar-2024-22

Preprints

https://doi.org/10.5194/ar-2024-22

Preprints

02 Sep 2024

| 02 Sep 2024

Status: a revised version of this preprint is currently under review for the journal AR.

Unchanged PM2.5 levels over Europe during COVID-19 were buffered by ammonia

Nikolaos Evangeliou, Ondrej Tichy, Marit Svendby Otervik, Sabine Eckhardt, Yves Balkanski, and Didier Hauglustaine

Abstract. The coronavirus outbreak in 2020 had devastating impact on human life, albeit a positive effect for the environment reducing emissions of primary aerosols and trace gases and improving air quality. In this paper, we present inverse modelling estimates of ammonia emissions during the European lockdowns of 2020 based on satellite observations. Ammonia has a strong seasonal cycle and mainly originates from agriculture. We further show how changes in ammonia levels over Europe, in conjunction with decreases in traffic-related atmospheric constituents modulated PM2.5. The key result of this study is a -9.8 % decrease in emissions in the first half of 2020 compared to the same period in 2016–2019 attributed to restrictions related to the global pandemic. We further calculate the delay in the evolution of the emissions in 2020 before, during and after lockdowns, by an sophisticated comparison of the evolution of ammonia emissions during the same time periods for the reference years (2016–2019). Our analysis demonstrates a clear delay in the evolution of ammonia emissions of -77 kt, that was mainly observed in the countries that suffered the strictest travel, social and working measures. Despite the general drop in emissions during the first half of 2020 and the delay in the evolution of the emissions during the lockdown period, satellite and ground-based observations showed that European levels of ammonia increased. On one hand, this was due to the reduction of SO₂ and NO_x (precursors of the atmospheric acids with which ammonia reacts) that caused less binding and thus less chemical removal of ammonia (smaller loss – higher lifetime); on the other, the majority of the emissions persisted, because ammonia mainly originates from agriculture, a primary production sector that was not influenced by the lockdown restrictions, as practically agricultural activity never ceased. Despite the projected drop in various atmospheric aerosols and trace gases, PM2.5 levels stayed unchanged or even increased in Europe due to a number of reasons attributed to the complicated NH₃ - H₂SO₄ - HNO₃ system. Higher water vapour during the European lockdowns favoured more sulfate production from SO₂ and OH (gas phase) or O₃ (aqueous phase). Ammonia first neutralised sulfuric acid (due to higher atmospheric abundance) also producing sulfate. Then, the continuously accumulating free ammonia reacted with nitric acid shifting the equilibrium reaction towards particulate nitrate. In high free ammonia atmospheric conditions such as those in Europe during the 2020 lockdowns, a small reduction of NO_x levels drives faster oxidation toward nitrate and slower deposition of total inorganic nitrate causing high secondary PM2.5 levels.

Received: 26 Aug 2024 – Discussion started: 02 Sep 2024

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.

Download & links

Preprint (PDF, 2417 KB)

Supplement (2352 KB)

Download & links

Nikolaos Evangeliou, Ondrej Tichy, Marit Svendby Otervik, Sabine Eckhardt, Yves Balkanski, and Didier Hauglustaine

Status: final response (author comments only)

RC1:
'Comment on ar-2024-22', Anonymous Referee #1, 24 Sep 2024

Evangeliou and co-authors have modelled ammonia emissions and concentrations over Europe during the 2020 COVID-19 lockdown, and further assessed its effect on aerosol particle concentrations (in particular PM2.5). The topic is interesting and timely, and the paper is certainly worth publishing after some rewriting. I am an atmospheric chemist, but do not have expertise in either remote sensing or bayesian modelling. I am thus unfortunately unable to comment in detail on the technical aspects of the manuscript.
The main issue I have with the manuscript is that it is rather sloppily written - there is much repetition (e.g. the fact that agricultural activity continued during lockdown is repeated very many times) , unclear sentences and even completely incorrect statements (e.g. on line 572 in the conclusion, “ambient pollution levels did not INCREASE as expected” should presumable read “decrease”). As an extreme example of sloppiness, the caption of Figure 1 contains the phrase “Error! Reference source not found”. How did something like this pass by both the authors and the copy-editors? Overall the text gives the impression of a very promising but rather unfinished draft, not a submitted ready manuscript. While most of this sloppy writing is just irritating, rather than scientifically problematic, it also makes it hard to understand exactly what the authors are arguing. When they talk about various “decreases” (or increases), for example, it is often unclear if they mean absolute decreases/increases, decreases/increases compared to some trend or average, or something else.
The discussion on atmospheric chemistry (around for example lines 460-747 or 527 and onward) is also somewhat confusing or confused. For example, ammonium sulfate is certainly not formed directly in the gas phase (even the notation used, “(s)”, indicates SOLID ammonium sulfate, thus emphatically NOT gas phase). The net effect of NH3 + H2SO4 clustering may indeed be particles with the net composition of ammonium sulfate, so for the purposes of this study (focusing on PM2.5, not gas-phase chemistry), the distinction may be irrelevant, but to a gas-phase chemist such as myself it almost hurts to read this. Similarly, it is certainly not generally true that sulfuric acid is more abundant in the atmosphere than nitric acid, at least not if gas-phase H2SO4 and HNO3 are compared (total sulfate may well be larger than total nitrate though). Also, I would say that after e.g. the CLOUD experiments, the “effect” of NH3 on promoting H2SO4 nucleation IS quite well understood (though since the authors seem to base their understanding on references from 1999 and 2000, their confusion is quite understandable).
I recommend a thorough rewriting of the manuscript to avoid unnecessary repetition, and to make the key arguments and results clearer. After that, I have no objection to the manuscript being published.

Citation: https://doi.org/10.5194/ar-2024-22-RC1
- AC1: 'Reply on RC1', Nikolaos Evangeliou, 05 Dec 2024
  
  RC1: 'Comment on ar-2024-22', Anonymous Referee #1, 24 Sep 2024
  Evangeliou and co-authors have modelled ammonia emissions and concentrations over Europe during the 2020 COVID-19 lockdown, and further assessed its effect on aerosol particle concentrations (in particular PM2.5).
  
  The topic is interesting and timely, and the paper is certainly worth publishing after some rewriting. I am an atmospheric chemist, but do not have expertise in either remote sensing or bayesian modelling. I am thus unfortunately unable to comment in detail on the technical aspects of the manuscript.
  
  RESPONSE: We appreciate Reviewer’s positive attitude with respect to our manuscript. Below, we have tried to improve the manuscript taking into consideration all his comments. The corrections are highlighted with Track Changes and line numbers are also given for ease of the reviewing process.
  The main issue I have with the manuscript is that it is rather sloppily written - there is much repetition (e.g. the fact that agricultural activity continued during lockdown is repeated very many times) , unclear sentences and even completely incorrect statements (e.g. on line 572 in the conclusion, “ambient pollution levels did not INCREASE as expected” should presumable read “decrease”).
  
  RESPONSE: We have corrected this specific comment (L572) and the manuscript has been edited carefully to remove any repetition and unclear statements.
  As an extreme example of sloppiness, the caption of Figure 1 contains the phrase “Error! Reference source not found”. How did something like this pass by both the authors and the copy-editors?
  
  RESPONSE: We acknowledge Reviewer for pointing this out. This is a common problem in MS Word when using “cross references”. However, when shifting between different computers and office versions, broken links might appear, and this is what happened here. We have corrected this now (See legend in Figure 1).
  Overall the text gives the impression of a very promising but rather unfinished draft, not a submitted ready manuscript. While most of this sloppy writing is just irritating, rather than scientifically problematic, it also makes it hard to understand exactly what the authors are arguing. When they talk about various “decreases” (or increases), for example, it is often unclear if they mean absolute decreases/increases, decreases/increases compared to some trend or average, or something else.
  
  RESPONSE: We have tried to simplify and correct several expressions throughout the manuscript, so that it gives a better overview of what we argue. Please see the manuscript with Track Changes enabled.
  The discussion on atmospheric chemistry (around for example lines 460-747 or 527 and onward) is also somewhat confusing or confused. For example, ammonium sulfate is certainly not formed directly in the gas phase (even the notation used, “(s)”, indicates SOLID ammonium sulfate, thus emphatically NOT gas phase). The net effect of NH3 + H2SO4 clustering may indeed be particles with the net composition of ammonium sulfate, so for the purposes of this study (focusing on PM2.5, not gas-phase chemistry), the distinction may be irrelevant, but to a gas-phase chemist such as myself it almost hurts to read this.
  
  RESPONSE: The paragraph starting at line 460 has been corrected and more recent references have been added (as suggested by Reviewer 2). The respective changes are shown in page Page 15 (around L510 and below – See Track Changes). Section 4.2 has been carefully amended (see the manuscript with Track Changes enabled).
  Similarly, it is certainly not generally true that sulfuric acid is more abundant in the atmosphere than nitric acid, at least not if gas-phase H2SO4 and HNO3 are compared (total sulfate may well be larger than total nitrate though).
  
  RESPONSE: Yes, the reviewer has a fair point here and we acknowledge for this comment. This is not the correct argument to explain ammonia’s preference for sulfuric acid. We have now corrected this in Pages 15-16 (See Track Changes).
  Also, I would say that after e.g. the CLOUD experiments, the “effect” of NH3 on promoting H2SO4 nucleation IS quite well understood (though since the authors seem to base their understanding on references from 1999 and 2000, their confusion is quite understandable).
  
  RESPONSE: Indeed the CLOUD experiment has given valuable insights on ammonia’s atmospheric chemistry. We have now removed this unnecessary statement and added the highlights on NH3/H2SO4 nucleation from the CLOUD experiment (see Track Changes at Page 6).
  I recommend a thorough rewriting of the manuscript to avoid unnecessary repetition, and to make the key arguments and results clearer. After that, I have no objection to the manuscript being published.
  
  RESPONSE: We have followed all the comments and suggestions by both reviewers, restructured and added calculated uncertainties in our manuscript (see revised manuscript with Track Changes enabled).
  
  Citation: https://doi.org/10.5194/ar-2024-22-AC1
RC2:
'Comment on ar-2024-22', Anonymous Referee #2, 11 Oct 2024

Evangeliou et al. investigated how COVID-19 impacted ammonia (NH3) emissions in Europe. They used the CrIS satellite to constrain NH3 and conducted inverse modeling to calculate the emissions rates. They found that the NH3 emissions during COVID-19 were lower than prior years. However, NH3 concentration in the atmosphere was higher. They suggest it is due to lower SO2 and NOx emissions, leading to less acidic gases. The paper is of potential interest to Aerosol Research; however, there are many concerns the authors need to address prior to publication.
1) I agree with Reviewer #1 that the paper as presented is confusing and needs both reorganized as well as edited for understanding. One major reorganization is that the errors in the emission estimates needs to be presented much sooner (Section 3.1 ideally) as understanding these errors is necessary to understand if the differences in the results are statistically different or not (more on this later).
2) A major concern, similar to Reviewer #1, is that the authors do not appears to understand the thermodynamics that is controlling inorganic aerosol chemistry. First, it takes a large concentration of NH3 in the atmosphere to charge neutralize sulfate (Weber et al., 2016). Charge is emphasized as the term neutralization, which is used, is assumed to reference the pH of the system. Most environments rarely achieve pH neutralization (Pye et al., 2020; Weber et al., 2016). Further, the sulfate does not need to be charge neutralized by NH3 in order to start forming ammonium nitrate (NH4NO3). Instead, the formation of NH4NO3 is controlled by a combination of aerosol pH, temperature, and aerosol liquid water (Guo et al., 2016), and can happen when ammonium bisulfate is a dominant species in the aerosol phase. I highly recommend the authors review Pye et al. (2020) to improve the discussions about the thermodynamics controlling ammonium (NH4) in the aerosol phase as well as the discussions throughout the paper.
3) Further, it is not clear what thermodynamic model was used for the inverse modeling/chemical transport model. A description of this is important, as it entails what species are included, how well it performs at high or low relative humidity/temperature, and whether it was operated with constraints on all species (gas + particle) or just particle, which leads to higher uncertainty (Hennigan et al., 2015).
4) Important controlling knobs of NH3 were never discussed and compared throughout the manuscript. This includes temperature and relative humidity (both for emissions and thermodynamics), precipitation and other meteorology (wet/dry deposition), and comparisons of NOx, SO2, sulfate (SO4), and nitrate (NO3). Without these comparisons, it is not clear if the amount of emissions needed for the different years are being driven by the correct mechanisms or not. Clouds are important both for the lifetime of NH3 but also retrievals of NH3. How were observations with clouds dealt with, as optically thick clouds block any retrieval of NH3 (Shephard and Cady-Pereira, 2015)? Were there areas with optically thick clouds during any of these time periods that introduce more uncertainty?
5) Overall, the results are confusing as they are currently written. The authors argue that agriculture was not impacted by lockdowns, and thus why NH3 remained high; however, they then discuss how lockdowns reduced the NH3 in some different countries. The authors discuss how lockdowns reduced travel and human activity in urban areas; however, the results indicate that the urban areas may have had minimal changes with NH3 emissions. So, did the lockdowns impact agriculture, or was there other influences such as weather (delayed application due to rain/snow/temperature vs temperature/relative humidity reducing releases)? Lines 351 - 363, among others, make this very confusing, especially as the authors state that fertilizer application is tightly regulated but then give a wide range of months to apply.
6) Why did the authors only select a subset of stationary measurements to compare the NH3 from the new emission inventory? E.g., they did not select any stations that were near points of interest they discussed (e.g., northern Italy/Switzerland, the Netherlands, etc.) that also show hotspots that need to be validated before discussion. Also, one of the highest hot spots, Belgium, does not have any ground observations to verify the new emissions. How certain are the authors in the new emissions for this location, especially as the other emission inventories in the supplement do not show a similar hot spot? Should this be an area of higher uncertainty as there is nothing to constrain it, even though Fig. 5 shows relatively lower relative uncertainty?
7) Furthermore, the time series of these ground observations for the old vs new emissions do not validate that the new emission inventory is work. The root mean square error and mean absolute error do not look significantly improved between the old and new emission inventory and beyond the "peaks" the authors mentioned in the manuscript, there are many instances that the new inventory is phased-shifted (high when observations are low and vice versa) and/or shows high bias/baseline.
8) The authors argue there is no meteorological impacts as 2016 - 2019 NH3 emissions (average) were similar year to year. However, at the time of lockdown (March - April) for those years, the largest variation in the averages is seen. Thus, how much is meteorology (temperature, rain, etc.) is impacting the application and/or release of NH3. Also, as the values shown in Fig. 2e & f is the average, what is the standard deviation of those means? What is the median? Are the values really statistically different or not (more on this later). Finally, the authors skim over month of May where suddenly the NH3 emissions match prior years--what led to this jump?
9) Statistical analysis. As suggested in comment 1), the error analysis for the emissions needs to be presented at the beginning of the results. With it being there, it can be analyzed if the values in Fig. 2, 3, and 4 are statistically different or within the error of the emissions. If Fig. 5 is taken at face value, the combined uncertainty for many locations is 400 - 1000 mg m-2. The differences between lockdown and non-lockdown years and after the lockdown period are then all within this uncertainty, meaning that there was no statistically different changes in the NH3 emissions during 2020. If Fig. 5c is an error for scale, it is still confusing what is the error as it is stated as 11% in Line 438 and 48% in line 1118. If it's 48%, then that would mean there is no statistical difference in Fig. 2, 3, and 4. If it's 11%, then there may be statistical difference, but depends what the actual year-to-year variation (spread of observations about the mean, comment 8)).
10) Discussion. This goes with the discussions above about needing a better discussion about the thermodynamics controlling NH3 to make this section more easily understood. In addition, Fig. 6a does not provide any evidence that there should be more NH3 in the atmosphere, as the modeled predicted NH3 lifetime increased by ~0.02 days, which is well within the combined uncertainty of the emissions and chemistry as well as the spread of lifetimes for the region. Further, as discussed in Weber et al. (2006), there needs to be large reductions in SO2 in order to observe noticeable differences in NH3 and aerosol pH. Finally, the model was never shown for validation of aerosol composition. The model shows that most of the PM is primary (even during lockdown). However, as shown in Chen et al. (2022), most of the aerosol is secondary in nature. Thus, there is overall concern in the models ability to predict the aerosol composition and that it is getting PM2.5 correct for the incorrect reasons.
References
Weber et al. High aerosol acidity despite declining atmospheric sulfate concentrations over the past 15 years. Nature Geoscience. 2016
Pye et al. The acidity of atmospheric particles and clouds. Atmospheric Chemistry and Physics. 2020.
Hennigan et al. A critical evaluation of proxy methods used to estimate the acidity of atmospheric particles. Atmospheric Chemistry and Physics. 2015.
Shephard and Cady-Pereira. Cross-track Infrared Sounder (CrIS) satellite observations of tropospheric ammonia. Atmospheric Measurement Techniques. 2015.
Chen et al. European aerosol phenomenology - 8: Harmonised source apportionment of organic aerosol using 22 year-long ACSM/AMS datasets. Environment International. 2022.
Guo et al. Fine particle pH and the partitioning of nitric acid during winter in the northeastern United States. Journal of Geophysical Research-Atmosphere. 2016.

Citation: https://doi.org/10.5194/ar-2024-22-RC2
- AC2: 'Reply on RC2', Nikolaos Evangeliou, 05 Dec 2024
  
  RC2: 'Comment on ar-2024-22', Anonymous Referee #2, 11 Oct 2024
  Evangeliou et al. investigated how COVID-19 impacted ammonia (NH3) emissions in Europe. They used the CrIS satellite to constrain NH3 and conducted inverse modeling to calculate the emissions rates. They found that the NH3 emissions during COVID-19 were lower than prior years. However, NH3 concentration in the atmosphere was higher. They suggest it is due to lower SO2 and NOx emissions, leading to less acidic gases. The paper is of potential interest to Aerosol Research; however, there are many concerns the authors need to address prior to publication.
  
  RESPONSE: We acknowledge Reviewer for his valuable comments in an effort to improve the present manuscript. We have followed all his comments in the new version of the manuscript. The corrections are highlighted with Track Changes and line numbers are also given to ease the reviewing process.
  1) I agree with Reviewer #1 that the paper as presented is confusing and needs both reorganized as well as edited for understanding. One major reorganization is that the errors in the emission estimates needs to be presented much sooner (Section 3.1 ideally) as understanding these errors is necessary to understand if the differences in the results are statistically different or not (more on this later).
  
  RESPONSE: We have followed this comment and now present the calculated errors of the posterior emissions in section 3.2, right after we discuss the posterior emissions over Europe (section 3.1). We have also amended the whole manuscript to avoid repetition and confusing statements.
  2) A major concern, similar to Reviewer #1, is that the authors do not appears to understand the thermodynamics that is controlling inorganic aerosol chemistry. First, it takes a large concentration of NH3 in the atmosphere to charge neutralize sulfate (Weber et al., 2016). Charge is emphasized as the term neutralization, which is used, is assumed to reference the pH of the system. Most environments rarely achieve pH neutralization (Pye et al., 2020; Weber et al., 2016). Further, the sulfate does not need to be charge neutralized by NH3 in order to start forming ammonium nitrate (NH4NO3). Instead, the formation of NH4NO3 is controlled by a combination of aerosol pH, temperature, and aerosol liquid water (Guo et al., 2016), and can happen when ammonium bisulfate is a dominant species in the aerosol phase. I highly recommend the authors review Pye et al. (2020) to improve the discussions about the thermodynamics controlling ammonium (NH4) in the aerosol phase as well as the discussions throughout the paper.
  
  RESPONSE: Here, we think there’s an overall misunderstanding that is beyond comprehension of the inorganic aerosol chemistry. As in all models, we make several assumptions and simplifications in this global model too. The pH is not calculated “per se” as we do not have all the ions in the model. Instead, the ammonium-to-sulfate ratio is calculated as an indicator of the pH. See for instance equation (12) of Pye et al. (2020) explicitly saying “More recently, the total ammonium-to-sulfate ratio has
  
  been proposed as an indicator of pH (Murphy et al., 2017):”.
  
  Hence, in this model, we calculate the mass balance and not really the pH.
  
  Step 1: We calculate the total nitrate TN, total ammonia TA and total sulfate (TS). If TA < 2TS we are in ammonia poor conditions. In this case, there is insufficient NH3 to neutralize the available sulfate and the aerosol phase is assumed acidic (but the pH not calculated explicitly) . if TA > 2 TS then there is excess ammonia and the sulfate is assumed neutralized (but since the pH is not explicitly calculated it doesn’t mean the real aerosol is neutralized, just that the sulfates are assumed neutralized in the model). Maybe it is worth mentioning here that different species for different aerosols are represented in the model. So, excess ammonia (which did not react with sulfate) will be available to react with gas phase nitrate to produce nitrate particles.
  
  Step 2: Excess ammonia and nitric acid react to form ammonium nitrate. Depending on the RH and T, ammonium nitrate can be either as an aqueous solution or as a solid. In the model we assume it as ‘particles’ (as other particle types in the model, it has a solid part and an aqueous part). The formation of nitrates is an equilibrium reaction, and the dissociation constant is calculated depending on RH and T. Based on this equilibrium, if the system is supersaturated, a fraction of NH3 and HNO3 is transferred to the aqueous phase and the rest remains in the gas phase.
  
  We have removed the terms neutralization and pH from the manuscript since we do not calculate the pH in this model. We have further tried to clarify section 4.2.
  
  REFERENCES
  
  Pye, H. O. T., Nenes, A., Alexander, B., Ault, A. P., Barth, M. C., Clegg, S. L., Collett Jr., J. L., Fahey, K. M., Hennigan, C. J., Herrmann, H., Kanakidou, M., Kelly, J. T., Ku, I.-T., McNeill, V. F., Riemer, N., Schaefer, T., Shi, G., Tilgner, A., Walker, J. T., Wang, T., Weber, R., Xing, J., Zaveri, R. A., and Zuend, A.: The acidity of atmospheric particles and clouds, Atmos. Chem. Phys., 20, 4809–4888, https://doi.org/10.5194/acp-20-4809-2020, 2020.
  
  Murphy, J. G., Gregoire, P. K., Tevlin, A. G., Wentworth, G. R., Ellis, R. A., Markovic, M. Z., and VandenBoer, T. C.: Observational constraints on particle acidity using measurements and modelling of particles and gases, Faraday Discuss., 200, 379–395, https://doi.org/10.1039/c7fd00086c, 2017.
  3) Further, it is not clear what thermodynamic model was used for the inverse modeling/chemical transport model. A description of this is important, as it entails what species are included, how well it performs at high or low relative humidity/temperature, and whether it was operated with constraints on all species (gas + particle) or just particle, which leads to higher uncertainty (Hennigan et al., 2015).
  
  RESPONSE: At this point, we refer the reader and reviewer to Hauglustaine et al. (2014) for a description of thermodynamics in the model and the overall assumptions. We have avoided repeating all these details, as the LMDzORINCA model, that we used here, is a well-established and validated model. In the Supplementary Material of Hauglustaine et al. (2014), further details are given about the model performance. Furthermore, a comparison of the nitrate-aerosol module has been performed against results of the thermodynamic model ISORROPIA version 2.1(Nenes et al., 1998) using the same input parameters. The input parameters were then varied in order to compare the behavior of the two models across the range of variation in key inputs (i.e., total sulfate TS, total ammonia TA, total nitrate TN, relative humidity RH, and temperature T). Overall an excellent agreement was obtained between the INCA module and ISORROPIA. The sensitivity of the two models to key environmental parameters as encountered in the global atmosphere were very similar and the simulated concentrations in close agreement (see Supplementary Information in Hauglustaine et al., 2014).
  
  We are not sure that results from this paper need to be repeated here.
  REFERENCES
  
  Hauglustaine, D. A., Balkanski, Y., and Schulz, M.: A global model simulation of present and future nitrate aerosols and their direct radiative forcing of climate, Atmos. Chem. Phys., 14, 11031–11063, https://doi.org/10.5194/acp-14-11031-2014, 2014.
  
  Nenes, A., S. Pandis, and C. Pilinis, ISORROPIA: a new thermodynamic equilibrium model for multiphase multicomponent inorganic aerosols, Aquat. Geochem., 4, 123-152, 1998.
  4) Important controlling knobs of NH3 were never discussed and compared throughout the manuscript. This includes temperature and relative humidity (both for emissions and thermodynamics), precipitation and other meteorology (wet/dry deposition), and comparisons of NOx, SO2, sulfate (SO4), and nitrate (NO3). Without these comparisons, it is not clear if the amount of emissions needed for the different years are being driven by the correct mechanisms or not. Clouds are important both for the lifetime of NH3 but also retrievals of NH3. How were observations with clouds dealt with, as optically thick clouds block any retrieval of NH3 (Shephard and Cady-Pereira, 2015)? Were there areas with optically thick clouds during any of these time periods that introduce more uncertainty?
  
  RESPONSE: Details about the settings used are now described in Page 5 (see Track Changes). All the recommendations about the use of the different flags were followed in discussions with the developers of the retrieval algorithms of CrIS NH3. Thin clouds appear to be the main issue in the CrIS retrievals, but such data were removed from the dataset using Cloud Flag  1.
  
  A detailed comparison of modelled NO2, SO2 (and NH3) against ground observations from EMEP over Europe for the period 2013-2020 have been already performed and presented in Tichy et al. (2023) (see Supplementary Information Figures S8-S10). We plot average temperature, specific humidity and precipitation for the region of our study for all years 2016-2020 in Figure R2 from ERA5 (Hersbach et al., 2020). Temperature, humidity and precipitation were as any other of the previous 5 years in Europe during the same period and cannot justify more volatilisation of ammonia. This is additional evidence that the impact from meteorology did not drive ammonia or PM2.5 formation (see Figure R2)!
  REFERENCES
  
  Tichý, O., Eckhardt, S., Balkanski, Y., Hauglustaine, D., and Evangeliou, N.: Decreasing trends of ammonia emissions over Europe seen from remote sensing and inverse modelling, Atmos. Chem. Phys., 23, 15235–15252, https://doi.org/10.5194/acp-23-15235-2023, 2023.
  5) Overall, the results are confusing as they are currently written. The authors argue that agriculture was not impacted by lockdowns, and thus why NH3 remained high; however, they then discuss how lockdowns reduced the NH3 in some different countries. The authors discuss how lockdowns reduced travel and human activity in urban areas; however, the results indicate that the urban areas may have had minimal changes with NH3 emissions. So, did the lockdowns impact agriculture, or was there other influences such as weather (delayed application due to rain/snow/temperature vs temperature/relative humidity reducing releases)? Lines 351 - 363, among others, make this very confusing, especially as the authors state that fertilizer application is tightly regulated but then give a wide range of months to apply.
  
  RESPONSE: One of the key questions we try to answer within this study is the reason why NH3 remained high in Europe in 2020, since air quality improved. This is justified if agriculture was slightly affected by the COVID lockdowns. Spring concentrations are somewhat enhanced over Europe in all years (even when there were no lockdowns), but the exact start of the increases right after the lockdowns were applied triggered our attention. We have rejected the impact from weather influences, after performing the same analyses using same observations (CrIS), but for previous years (2016-2019); NH3 levels were found very close to those of 2020. The only difference appears to be a delay in the emissions in 2020 (Figure 4 in the revised version). Nevertheless, several observations show that NH3 in 2020 remained high in Europe (Lovarelli et al., 2021; Rennie et al., 2020; Viatte et al., 2021; Kuttippurath et al., 2023).
  
  About the “tightly regulated fertilisation of NH3 in Europe”, this is true! It is well known that fertilisation is generally allowed in late-February to summer in Europe; therefore the two peaks in the concentrations over Europe (among environmental parameters). However, we should acknowledge that each country has specific regulations. The range of months given in this paragraph is not “wide” and preserves very well the late winter to summer fertilisation. We have tried to clarify these points now in a more consistent way and we have shortened this confusing paragraph.
  REFERENCES
  
  Lovarelli, D., Fugazza, D., Costantini, M., Conti, C., Diolaiuti, G., and Guarino, M.: Comparison of ammonia air concentration before and during the spread of COVID-19 in Lombardy (Italy) using ground-based and satellite data, Atmos. Environ., 259, 118534, https://doi.org/10.1016/j.atmosenv.2021.118534, 2021.
  
  Rennie, S., Watkins, J., Ball, L., Brown, M., Fry, M., Henrys, P., Hollaway, M., Quinn, J., Sier, A., and Dick, J.: Shaping the development of the UKCEH UK-SCAPE Data Science Framework. Workshop report, 2020.
  
  Kuttippurath, J., Patel, V. K., Kashyap, R., Singh, A., and Clerbaux, C.: Anomalous increase in global atmospheric ammonia during COVID-19 lockdown: Need for policies to curb agricultural emissions, J. Clean. Prod., 434, 140424, https://doi.org/10.1016/j.jclepro.2023.140424, 2023.
  
  Viatte, C., Petit, J. E., Yamanouchi, S., Van Damme, M., Doucerain, C., Germain-Piaulenne, E., Gros, V., Favez, O., Clarisse, L., Coheur, P. F., Strong, K., and Clerbaux, C.: Ammonia and pm2.5 air pollution in paris during the 2020 covid lockdown, Atmosphere (Basel)., 12, 1–18, https://doi.org/10.3390/atmos12020160, 2021.
  6) Why did the authors only select a subset of stationary measurements to compare the NH3 from the new emission inventory? E.g., they did not select any stations that were near points of interest they discussed (e.g., northern Italy/Switzerland, the Netherlands, etc.) that also show hotspots that need to be validated before discussion. Also, one of the highest hot spots, Belgium, does not have any ground observations to verify the new emissions. How certain are the authors in the new emissions for this location, especially as the other emission inventories in the supplement do not show a similar hot spot? Should this be an area of higher uncertainty as there is nothing to constrain it, even though Fig. 5 shows relatively lower relative uncertainty?
  
  RESPONSE: This is not entirely true. The ground-based measurements that we used from EMEP consist of >50 European stations. We could not present >50 line plots, so we randomly selected 8 stations. The comparison of modelled NH3 with all observations from the EMEP are plotted as scatterplot in Figure 1c (see Figure 1c in the manuscript).
  
  Of cource, there is uncertainty in the calculated posterior emissions as the reviewer kindly states in this comment, and different prior emissions can affect the posterior. As we describe in section 3.2 of the revised version (Title: Uncertainty of the posterior emissions), this type of uncertainty has been included in the sensitivity, by performing the same inversions with different prior emission datasets for NH3 (4 in total).
  7) Furthermore, the time series of these ground observations for the old vs new emissions do not validate that the new emission inventory is work. The root mean square error and mean absolute error do not look significantly improved between the old and new emission inventory and beyond the "peaks" the authors mentioned in the manuscript, there are many instances that the new inventory is phased-shifted (high when observations are low and vice versa) and/or shows high bias/baseline.
  
  RESPONSE: We understand Reviewer’s concern here and we try to explain some principles. There are certain limitations when it comes to Bayesian linear inverse modelling by use of satellite observations. This is more pronounced when it comes to the CrIS observations that are performed in the log space.
  
  As explained in Section 2.3, Eq. 1 (in manuscript) gives the relationship of the satellite retrieved versus the true (or modelled) value of the measured NH3 as follows:
  
  ln(v^ret )=ln(v^a )+A(ln(v^mod)-ln(v^a ) )
  
  where v^ret is the retrieved profile concentration vector (in other words the model equivalent value for the retrieved concentration at each model level ), v^a is a priori profile concentration vector (provided by CrIS), v^mod=v^true is the true profile concentration vector, and A is the averaging kernel matrix in logarithmic space (for each 0.5°×0.5° resolution grid-cell).
  
  When we try to minimise the cost function of the inversion in the algorithm, we need to minimise the model-observation mismatches, namely the CrIS observation v^sat with the instrument operator v^ret (Eq.2 in manuscript). Note that in classic inversions with ground-based observations, v^mod (and not any retrieved vector like here) is directly compared with observations. From the log equation (Eq. 1), we see that (i) the optimisation of the modelled concentrations (v^mod) is tiny, due to the log space of the equation, and (ii) v^ret has a very large dependency of the prior column used in CrIS (v^a).
  
  The former suggests that even if v^ret and v^sat are very close, solving for v^mod to calculate the posterior emissions in a logarithmic equation such as Eq. 1 (manuscript) causes small improvements, due to the dependence from CrIS variables (such as the Averaging Kernel, A, and v^a). This is shown very clearly in Figure 2d from Tichy et al. (2023), where the posterior emissions (in blue) take the same tendency as the prior column used in CrIS (v^a in the figure) due to the logarithmic retrieval used.
  REFERENCE
  
  Tichý, O., Eckhardt, S., Balkanski, Y., Hauglustaine, D., and Evangeliou, N.: Decreasing trends of ammonia emissions over Europe seen from remote sensing and inverse modelling, Atmos. Chem. Phys., 23, 15235–15252, https://doi.org/10.5194/acp-23-15235-2023, 2023.
  8) The authors argue there is no meteorological impacts as 2016 - 2019 NH3 emissions (average) were similar year to year. However, at the time of lockdown (March - April) for those years, the largest variation in the averages is seen. Thus, how much is meteorology (temperature, rain, etc.) is impacting the application and/or release of NH3. Also, as the values shown in Fig. 2e & f is the average, what is the standard deviation of those means? What is the median? Are the values really statistically different or not (more on this later). Finally, the authors skim over month of May where suddenly the NH3 emissions match prior years--what led to this jump?
  
  RESPONSE: We partially agree with the Reviewer and we appreciate for this comment. In Fig.2, we plotted the minimun and maximum posterior emissions for the period 2016-2019 together with the mean. As pointed out nicely by the reviewer, to show whether changes are statistically significant the standard deviation of the 2016-2019 emissions is needed together with the mean value. We have now updated Fig.2 (see Track Changes around Page 34). As expected the standard deviation of the posterior of 2016-2019 is smaller than the min-max range that we gave in the previous version. In addition, we also plot the uncertainty of the posterior as calculated in section 3.2. In that sense, we can now have a nice overview whether changes in 2020 posterior are statistically significant are regards to posterior 2016-2019. And in fact, very likely they are, since the posterior uncertainty is outside the standard deviation of the 2016-2019 emissions for the later part of the lockdown. The same is also seen in the country-level statistics presented in the updated version of Figure 4 (see Track Changes Page 36). In addition to this, we also check if temperature over europe was somewhat different or exceptional in 2020, so that it would justify that these changes are due to meteorology. We plot average temperatures, specific humidity and precipitation in Europe for all years 2016-2020 in Figure R2 from ERA5 (Hersbach et al., 2020). Temperature, humidity and precipitation change insignificantly in March-May over the years and cannot justify and meteorological impact that would drive NH3 or PM2.5 (Figure R2)!
  
  As regards to the increase of the emissions after the lockdown ceased (rebound/relaxation period), it is likely a combination of two processes, namely (i) all activities returned back to normal, hence the associated emissions, (ii) it was already summer, hence higher soil temperatures caused volatilisation of ammonia as a result of agricultural practices that did not stop during the lockdowns. Besides, the same pattern for many other species has been already reported (e.g., Forster et al., 2020).
  REFERENCES
  
  Forster, P.M., Forster, H.I., Evans, M.J. et al. Current and future global climate impacts resulting from COVID-19. Nat. Clim. Chang. 10, 913–919 (2020). https://doi.org/10.1038/s41558-020-0883-0
  
  Hersbach, H., Bell, B., Berrisford, P., Hirahara, S., Horányi, A., Muñoz-Sabater, J., Nicolas, J., Peubey, C., Radu, R., Schepers, D., Simmons, A., Soci, C., Abdalla, S., Abellan, X., Balsamo, G., Bechtold, P., Biavati, G., Bidlot, J., Bonavita, M., De Chiara, G., Dahlgren, P., Dee, D., Diamantakis, M., Dragani, R., Flemming, J., Forbes, R., Fuentes, M., Geer, A., Haimberger, L., Healy, S., Hogan, R. J., Hólm, E., Janisková, M., Keeley, S., Laloyaux, P., Lopez, P., Lupu, C., Radnoti, G., de Rosnay, P., Rozum, I., Vamborg, F., Villaume, S., and Thépaut, J. N.: The ERA5 global reanalysis, Q. J. R. Meteorol. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803, 2020.
  
  9) Statistical analysis. As suggested in comment 1), the error analysis for the emissions needs to be presented at the beginning of the results. With it being there, it can be analyzed if the values in Fig. 2, 3, and 4 are statistically different or within the error of the emissions. If Fig. 5 is taken at face value, the combined uncertainty for many locations is 400 - 1000 mg m-2. The differences between lockdown and non-lockdown years and after the lockdown period are then all within this uncertainty, meaning that there was no statistically different changes in the NH3 emissions during 2020. If Fig. 5c is an error for scale, it is still confusing what is the error as it is stated as 11% in Line 438 and 48% in line 1118. If it's 48%, then that would mean there is no statistical difference in Fig. 2, 3, and 4. If it's 11%, then there may be statistical difference, but depends what the actual year-to-year variation (spread of observations about the mean, comment 8)).
  
  RESPONSE: We have now moved the section where the posterior uncertainty is calculated in section 3.2, right after the presentation of the posterior emissions (Section 3.1). We appreciate the reviewer for pointing us to the error in L1118. The average uncertainty for the domain is 11% (not 48%). The changes are highlighted in page 11 (L310). The Figures presented in the manuscript have changed and the posterior uncertainty if now plotted together with the posterior emissions, so that the standard deviation of the posterior of the previous years (2016-2019) instead of the min-max range. These changes show very well that there’s statistical difference in 2020.
  10) Discussion. This goes with the discussions above about needing a better discussion about the thermodynamics controlling NH3 to make this section more easily understood. In addition, Fig. 6a does not provide any evidence that there should be more NH3 in the atmosphere, as the modeled predicted NH3 lifetime increased by ~0.02 days, which is well within the combined uncertainty of the emissions and chemistry as well as the spread of lifetimes for the region. Further, as discussed in Weber et al. (2016), there needs to be large reductions in SO2 in order to observe noticeable differences in NH3 and aerosol pH. Finally, the model was never shown for validation of aerosol composition. The model shows that most of the PM is primary (even during lockdown). However, as shown in Chen et al. (2022), most of the aerosol is secondary in nature. Thus, there is overall concern in the models ability to predict the aerosol composition and that it is getting PM2.5 correct for the incorrect reasons.
  
  RESPONSE: As regards to the change in lifetime of NH3, the change in the lifetime is from 0.49 at the start of the lockdowns to 0.54 during the peak of the lockdowns, or from 11.7 to 13 hours . The 10-year average modelled lifetime of NH3 is 11.60.6 hours (see section 3.1 and Figure 1d in Evangeliou et al., 2021). Hence, the change in the lifetime of NH3 during lockdowns is much larger than the uncertainty range of the modelled lifetime.
  
  As explained in a previous comment, the pH was not explicitly calculated in the model, but an indication of the pH was calculated from the total ammonium-to-sulfate ratio (Pye et al., 2020). However, to account for SO2 and NOx reductions during COVID, we have used the data from Doumbia et al. (2021) who reported SO2 reductions over Europe of 12-15% and NOx reductions of 20-25% during the 2020 lockdowns (see Figure 8 in Doumbia et al., 2021). These changes should be sufficient to justify differences in NH3.
  
  As regards to validation of the aerosol composition in the model, it is important to note that we have different species for the different aerosols, but we do not have a total aerosol species representing all of them. Regarding the different aerosol species, such as different aerosol modes of mineral dust, Black Carbon etc…, the model often participates in intercomparison exercises organized by AeroCom (https://aerocom.met.no/publications, specifically Gliß et al., 2021; Sand et al., 2017; Koffi et al., 2016; Tsigaridis et al., 2014 and others).
  
  So, in conclusion, it is of low probability that this well and long validated model fails to capture aerosol composition.
  
  We calculate primary and secondary PM2.5 as in the GEOS Chem model (Gu et al., 2023): Primary=dust+bc+poa+ammonium+small part of seasalt
  
  Secondary=nitrate +sulfate + SOA. Chen et al. (2022) results focuses on the Organic Aerosol from stations that are likely urban, suburban, remote (non-urban) in PM1, quite different than what we do here. Also, their results are based on PMF using measurements (at urban, suburban and remote sites), whereas here we use several thousand satellite observations per day of high resolution everywhere in the European domain.
  REFERENCES
  
  Evangeliou, N., Balkanski, Y., Eckhardt, S., Cozic, A., Van Damme, M., Coheur, P.-F., Clarisse, L., Shephard, M. W., Cady-Pereira, K. E., and Hauglustaine, D.: 10-year satellite-constrained fluxes of ammonia improve performance of chemistry transport models, Atmos. Chem. Phys., 21, 4431–4451, https://doi.org/10.5194/acp-21-4431-2021, 2021.
  
  Pye, H. O. T., Nenes, A., Alexander, B., Ault, A. P., Barth, M. C., Clegg, S. L., Collett Jr., J. L., Fahey, K. M., Hennigan, C. J., Herrmann, H., Kanakidou, M., Kelly, J. T., Ku, I.-T., McNeill, V. F., Riemer, N., Schaefer, T., Shi, G., Tilgner, A., Walker, J. T., Wang, T., Weber, R., Xing, J., Zaveri, R. A., and Zuend, A.: The acidity of atmospheric particles and clouds, Atmos. Chem. Phys., 20, 4809–4888, https://doi.org/10.5194/acp-20-4809-2020, 2020.
  
  Doumbia, T., Granier, C., Elguindi, N., Bouarar, I., Darras, S., Brasseur, G., Gaubert, B., Liu, Y., Shi, X., Stavrakou, T., Tilmes, S., Lacey, F., Deroubaix, A., and Wang, T.: Changes in global air pollutant emissions during the COVID-19 pandemic: a dataset for atmospheric modeling, Earth Syst. Sci. Data, 13, 4191–4206, https://doi.org/10.5194/essd-13-4191-2021, 2021.
  
  Gliß, J., Mortier, A., Schulz, M., Andrews, E., Balkanski, Y., Bauer, S. E., Benedictow, A. M. K., Bian, H., Checa-Garcia, R., Chin, M., Ginoux, P., Griesfeller, J. J., Heckel, A., Kipling, Z., Kirkevåg, A., Kokkola, H., Laj, P., Le Sager, P., Lund, M. T., Lund Myhre, C., Matsui, H., Myhre, G., Neubauer, D., van Noije, T., North, P., Olivié, D. J. L., Rémy, S., Sogacheva, L., Takemura, T., Tsigaridis, K., and Tsyro, S. G.: AeroCom phase III multi-model evaluation of the aerosol life cycle and optical properties using ground- and space-based remote sensing as well as surface in situ observations, Atmos. Chem. Phys., 21, 87–128, https://doi.org/10.5194/acp-21-87-2021, 2021.
  
  Sand, M., Samset, B. H., Balkanski, Y., Bauer, S., Bellouin, N., Berntsen, T. K., Bian, H., Chin, M., Diehl, T., Easter, R., Ghan, S. J., Iversen, T., Kirkevåg, A., Lamarque, J.-F., Lin, G., Liu, X., Luo, G., Myhre, G., Noije, T. V., Penner, J. E., Schulz, M., Seland, Ø., Skeie, R. B., Stier, P., Takemura, T., Tsigaridis, K., Yu, F., Zhang, K., and Zhang, H.: Aerosols at the poles: an AeroCom Phase II multi-model evaluation, Atmos. Chem. Phys., 17, 12197–12218, https://doi.org/10.5194/acp-17-12197-2017, 2017.
  
  Kofﬁ, B., et al. (2016), Evaluation of the aerosol vertical distribution in global aerosol models through comparison against CALIOP measurements: AeroCom phase II results, J. Geophys. Res. Atmos., 121, 7254–7283, doi:10.1002/2015JD024639.
  
  Tsigaridis, K., Daskalakis, N., Kanakidou, M., Adams, P. J., Artaxo, P., Bahadur, R., Balkanski, Y., Bauer, S. E., Bellouin, N., Benedetti, A., Bergman, T., Berntsen, T. K., Beukes, J. P., Bian, H., Carslaw, K. S., Chin, M., Curci, G., Diehl, T., Easter, R. C., Ghan, S. J., Gong, S. L., Hodzic, A., Hoyle, C. R., Iversen, T., Jathar, S., Jimenez, J. L., Kaiser, J. W., Kirkevåg, A., Koch, D., Kokkola, H., Lee, Y. H., Lin, G., Liu, X., Luo, G., Ma, X., Mann, G. W., Mihalopoulos, N., Morcrette, J.-J., Müller, J.-F., Myhre, G., Myriokefalitakis, S., Ng, N. L., O'Donnell, D., Penner, J. E., Pozzoli, L., Pringle, K. J., Russell, L. M., Schulz, M., Sciare, J., Seland, Ø., Shindell, D. T., Sillman, S., Skeie, R. B., Spracklen, D., Stavrakou, T., Steenrod, S. D., Takemura, T., Tiitta, P., Tilmes, S., Tost, H., van Noije, T., van Zyl, P. G., von Salzen, K., Yu, F., Wang, Z., Wang, Z., Zaveri, R. A., Zhang, H., Zhang, K., Zhang, Q., and Zhang, X.: The AeroCom evaluation and intercomparison of organic aerosol in global models, Atmos. Chem. Phys., 14, 10845–10895, https://doi.org/10.5194/acp-14-10845-2014, 2014.
  
  Gu, Y., Henze, D. K., Nawaz, M. O., Cao, H., & Wagner, U. J. (2023). Sources of PM 2.5-associated health risks in Europe and corresponding emission-induced changes during 2005–2015. GeoHealth, 7, e2022GH000767. https://doi.org/10.1029/2022GH000767 .
  References by the Editors
  
  Weber et al. High aerosol acidity despite declining atmospheric sulfate concentrations over the past 15 years. Nature Geoscience. 2016
  
  Pye et al. The acidity of atmospheric particles and clouds. Atmospheric Chemistry and Physics. 2020.
  
  Hennigan et al. A critical evaluation of proxy methods used to estimate the acidity of atmospheric particles. Atmospheric Chemistry and Physics. 2015.
  
  Shephard and Cady-Pereira. Cross-track Infrared Sounder (CrIS) satellite observations of tropospheric ammonia. Atmospheric Measurement Techniques. 2015.
  
  Chen et al. European aerosol phenomenology - 8: Harmonised source apportionment of organic aerosol using 22 year-long ACSM/AMS datasets. Environment International. 2022.
  
  Guo et al. Fine particle pH and the partitioning of nitric acid during winter in the northeastern United States. Journal of Geophysical Research-Atmosphere. 2016.
  Figure R2: Average surface temperature, specific humidity and precipitation over Europe from January to June for the years 2016-2020 from ECMWF ERA5 (Hersbach et al., 2020). Temperature, humidity and precipitation are not significantly different than any of the previous years and cannot justify more volatilisation of ammonia. This is additional evidence that the impact from meteorology did not drive ammonia or PM2.5 formation.
  
  Citation: https://doi.org/10.5194/ar-2024-22-AC2

Nikolaos Evangeliou, Ondrej Tichy, Marit Svendby Otervik, Sabine Eckhardt, Yves Balkanski, and Didier Hauglustaine

Supplement

https://doi.org/10.5194/ar-2024-22-supplement

Data sets

Unchanged PM2.5 levels over Europe during COVID-19 were buffered by ammonia Nikolaos Evangeliou, Ondřej Tichý, Marit Svendby Otervik, Sabine Eckhardt, Yves Balkanski, and Didier Hauglustaine https://datadryad.org/stash/share/Wgbc9UiXwtMH44366myWh2bt7MQc92JKhJBz7UwQlgY

Nikolaos Evangeliou, Ondrej Tichy, Marit Svendby Otervik, Sabine Eckhardt, Yves Balkanski, and Didier Hauglustaine

Viewed

Total article views: 320 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
198	69	53	320	30	5	5

HTML: 198
PDF: 69
XML: 53
Total: 320
Supplement: 30
BibTeX: 5
EndNote: 5

Views and downloads (calculated since 02 Sep 2024)

Month	HTML	PDF	XML	Total
Sep 2024	95	26	27	148
Oct 2024	53	17	14	84
Nov 2024	28	8	1	37
Dec 2024	22	18	11	51

Cumulative views and downloads (calculated since 02 Sep 2024)

Month	HTML	PDF	XML	Total
Sep 2024	95	26	27	148
Oct 2024	53	17	14	84
Nov 2024	28	8	1	37
Dec 2024	22	18	11	51

Viewed (geographical distribution)

Total article views: 346 (including HTML, PDF, and XML) Thereof 346 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 13 Dec 2024

Download

Preprint (2417 KB)
Metadata XML

Short summary

The COVID-19 lockdown measures in 2020 reduced emissions of various substances improving air quality. However, PM2.5 stayed unchanged due to NH₃ and related chemical transformations. Higher humidity favoured more SO₄^2- production so as accumulated NH₃ did. Excess NH₃ neutralised HNO₃ towards NO₃^-. In high NH₃ conditions such as those in 2020, a small reduction of NO_x levels drives faster oxidation toward NO₃^- and slower deposition of total inorganic NO₃^- causing high secondary PM2.5.


Total:	0
HTML:	0
PDF:	0
XML:	0