the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 09 Jan 2024
Concentration and source changes of HONO during the COVID-19 lockdown in Beijing
Abstract. Nitrous acid (HONO) is an important precursor of OH radicals which affects not only the sinks of primary air pollutants but also the formation of secondary air pollutants, whereas its source closure in the atmosphere is still controversial due to a lack of experiment validation. In this study, the HONO budget in Beijing has been analyzed and validated through the coronavirus disease (COVID-19) lockdown event, which resulted in the largest changes in air pollutant emissions in the history of modern atmospheric chemistry. A home-made Water-based Long-Path Absorption Photometer (LOPAP) along with other instruments were used to measure the HONO and related pollutants from January 1, 2020 to March 6, 2020, which covered the Chinese New Year (CNY) and the COVID-19 lockdown. The average concentration of HONO decreased from 0.97 ± 0.74 ppb before CNY to 0.53 ± 0.44 ppb during the COVID-19 lockdown, accompanied by a sharp drop of NOx and the greatest drop of NO (around 87 %). HONO budget analysis suggests that vehicle emissions were the most important source of HONO during the nighttime (53 %) before CNY, well supported by the decline of their contribution to HONO during the COVID-19 lockdown. We found that the heterogeneous conversion of NO2 on ground surfaces was an important nighttime source of HONO (31 %), while that on aerosol surfaces was a minor source (2 %). Nitrate photolysis became the most important daytime source during the COVID-19 lockdown compared with that before CNY, resulting from the combined effect of the increase in nitrate and the decrease in NO. Our results indicate that reducing vehicle emissions should be an effective measure for alleviating HONO in Beijing.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(1962 KB)
-
Supplement
(984 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1962 KB) - Metadata XML
-
Supplement
(984 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-3139', Yang Zeng, 13 Feb 2024
The authors have made commendable efforts to conduct a comprehensive field measurement in urban Beijing, spanning the unique transition period from the pre-pandemic COVID-19 era to the subsequent lockdown period. While the controversies outlined in the introduction section (lines 95-122) remain unresolved, this study observes a notable decrease in HONO concentration from the pre-pandemic eve to the lockdown phase. Furthermore, the study offers a reasonable explanation regarding the contributions of HONO sources. I find this work particularly valuable as it underscores the significance of vehicle emissions and highlights the rapid temporal shifts in the roles of HONO sources, which could be of great interest to readers of ACP. I recommend publication of the manuscript following minor revisions. Below are my specific comments:
The abstract requires improvement. Please remove lines 36-38 “which resulted in … to measure the HONO and related pollutants”, and restructure the sentences for better clarity.
Lines 67 and 73: The second and third paragraphs contain overlapping information. I recommend integrating the introduction about HONO concentrations into the first paragraph. Subsequently, the second paragraph should focus solely on direct emissions, while the third paragraph should discuss secondary formation.
Line 77: Here, "harvest season" may not be the most accurate term. Biomass burning encompasses a range of activities, including wildfires, which can be particularly significant under certain circumstances.
Line 90: The authors omitted the acid displacement (VandenBoer et al. 2015, Liu et al. 2019)?
Line 176: Livestock HONO emission is negligible in urban Beijing. Similarly, the emission of HONO from soil, which may peak following fertilizer application, likely does not need to be considered.
Line 236: heterogeneous yield of HONO instead of NO2?
Section 3.1 Please present the result following the order of Figure 1 (a)～(i).
Line 348: How about the heterogenous conversion on ground?
Line 417: Is HONOcorr equivalent to CHONO,corr,t as indicated in Eq. (6)? Please maintain consistency in the use of abbreviations.
Line 449: In fact, Song et al. (2023) proposed two HONO production pathways distinct from the homogeneous reaction between NO and OH.
Line 472: “a change of 25% and 95% of HONO sources, respectively.” What does this signify? It appears that the selection of the photolysis rate constant could have a significant influence.
Lines 472-477: Move the results regarding soil emissions to the subsequent paragraph.
Line 504: The NO2 concentration decreased significantly from 26.9 ppb to 17.2 ppb (-36%) from P1 to P2, indicating a substantial reduction rather than a slight decrease.
Lines 513-514: It’s noteworthy that HONO exhibits minimal sensitivity to both the uptake coefficient (γ) and surface area concentration (As). However, the authors should provide an explanation for this phenomenon.
Line 526: These variations (-9% to +40%) are significant and warrant attention.
Lines 552-555: The method utilized to estimate the overall uncertainty of the parameterization should be presented in the Experimental section.
References
Liu, Y. H., K. D. Lu, X. Li, H. B. Dong, Z. F. Tan, H. C. Wang, Q. Zou, Y. S. Wu, L. M. Zeng, M. Hu, K. E. Min, S. Kecorius, A. Wiedensohler and Y. H. Zhang (2019). "A comprehensive model test of the HONO sources constrained to field measurements at rural North China Plain." Environmental Science & Technology 53(7): 3517-3525.
Song, M., X. Zhao, P. Liu, J. Mu, G. He, C. Zhang, S. Tong, C. Xue, X. Zhao, M. Ge and Y. Mu (2023). "Atmospheric NOx oxidation as major sources for nitrous acid (HONO)." npj Climate and Atmospheric Science 6(1): 30.
VandenBoer, T. C., C. J. Young, R. K. Talukdar, M. Z. Markovic, S. S. Brown, J. M. Roberts and J. G. Murphy (2015). "Nocturnal loss and daytime source of nitrous acid through reactive uptake and displacement." Nature Geoscience 8(1): 55-60.
Citation: https://doi.org/10.5194/egusphere-2023-3139-RC1 -
AC1: 'Reply to reviewer 1', Yongchun Liu, 06 Apr 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2023-3139/egusphere-2023-3139-AC1-supplement.pdf
-
AC1: 'Reply to reviewer 1', Yongchun Liu, 06 Apr 2024
-
RC2: 'Comment on egusphere-2023-3139', Anonymous Referee #2, 24 Feb 2024
This paper presents an analysis of measurements of gas-phase nitrous acid (HONO), and related air pollutants / atmospheric chemical species, performed in the boundary layer in Beijing prior to and during the Covid-19 lockdowns in early 2020. The variation in abundance of HONO, its precursors and related species are used to quantify different chemical source and sink terms for HONO, and from their diurnal variation and change across the start of lockdown, inferences drawn re their relative importance – in particular direct vehicle emission.
The abundance and sources of HONO are very much a current topic in urban atmospheric chemistry / air pollution, as besides being a pollutant in its own right HONO is usually the major precursor to (reservoir for) the key oxidant OH. Lockdown presents a unique near-step-change opportunity to explore these processes. This paper adds to the developing understanding of these sources, and in that respect is a valuable contribution to the literature.
The measurements appear to have been carefully performed and are clearly presented, and the analysis is interesting, with clear signal in the change in mean diurnal variations (which are used to develop qualitative arguments).
However, I have substantial reservations about several key assumptions / aspects of the quantitative analysis approach, which in my opinion do not allow the authors to draw the conclusions they reach. These are
1 Meteorology. Obviously, weather changes affect pollutant advection, import, dispersion, chemical processing etc. The paper presents "before" vs "after" comparisons of concentrations either side of the start of lockdown, without consideration (beyond a basic vertical dispersion parameterisation) of these effects. There is significant literature on the importance of correcting for meteorology – for example applying de-weathering approaches – several using the specific example of air pollutant abundance in Beijing across lockdown (e.g. Jiabo et al., 2021: https://doi.org/10.1016/j.aosl.2021.100060; Shi et al., 2021: https://doi.org/10.1126/sciadv.abd6696; Lv et al., 2022: https://doi.org/10.1016/j.apr.2022.101452). How much of the observed change (or lack of change) in each species is due to changes in the weather between the two time periods P1 and P2 ? It would be interesting to repeat the analysis using deweathered concentrations.
2 Chemical lifetimes. The paper in effect performs a steady state analysis on HONO, assuming any rate of change of concentration can be related to an imbalance in the in-situ source and sink terms. You can do this for short-lived species – such as OH – but only with care for species such as HONO, with lifetimes (these are midlatitude winter measurements) of tens of minutes. The measured HONO reflects the integrated chemical variability over the sampled airmass trajectory prior to its arrival at the measurement point. This will be heterogeneous – especially at ground level in the middle of a city (i.e. the OH and NO2 etc will have varied a lot over the period of time – given by the HONO lifetime – that the airmass has travelled to the measurement point within the city). See arguments developed by Lee et al, JGR 2013 (https://doi.org/10.1002/2013JD020341) and related papers.
3 Statistical significance / precision / uncertainty. The paper does not consider sufficiently the uncertainty in the (many) source and sink terms considered, and their propagation together (beyond the monte carlo result, which I cant believe includes the uncertainty in the individual inputs, e.g. OH concentrations). Is there any statistical power resulting from their combined uncertainties ? Are the uncertainties given in the paper 1 or 2 standard deviations ? Are differences statistically significant ? The +/- ranges – even assuming these are 1 sd – suggest not (e.g. HONO/NO2 ratio – changing from 0.38+/- 0.035 to 0.042 +/- 0.034 – this is not a meaningful (statistically significant) change). There are several examples.
I’ve noted some more points below but the authors need to address the points noted above vs the overall approach, to have confidence that the results of their analysis allow them to draw the conclusions presented in the paper.
Introduction – reviews different sources for HONO and their contribution, but this mixes together very different environments (i.e. the relative importance of different sources will vary for the measurement site vs a road tunnel vs a bare soil location vs the marine boundary layer vs livestock). Suggest to distill this to assess the key factors at the measurement location, ie city centre
L157 – NO2 measured by 42i – selectivity for NO2, the NO2 data will include other Noy species (including HONO).
L168 how was j(HONO) calculated
L175 heterogeneous reactions of NO2 have been accounted for – I didn’t follow this section, may need rewording.
L193 what is the conversion factor alpha
L203+ using NO2 and CO – NO2 concentrations will vary with a lifetime of a minute or so (with respect to the NOx-O3 PSS) and 6-12 hours our so (with respect to NOx removal). CO concentrations will vary with a lifetime of several weeks. Is it valid to use both in the same in situ emission analysis ?
L230 / table 1 – what is the estimated OH concentration & how does it compare with measurements – the values in L436 (presumably 24 hour mean, around 4-7e5) seem much lower than those observed in wintertime Beijing (2.7e6 as 24-hour mean; Slater et al., https://doi.org/10.5194/acp-20-14847-2020).
L241 Parameterisation of OH concentrations. We might expect that the main source of OH in bejing is HO2 + NO, the main sink for OH is OH + NO, and the main primary source of OH is HONO photolysis (one can then argue about if HONO is acting as a primary source or a reservoir). The parameterisation given was developed for rural sites (as the authors note), where NO levels would be much lower, and was developed prior to more recent understanding of HONO abundance (it is over 20 years old). Is it valid to use ?
L259 – considering thee above I do not agree that we can be optimistic about the estimated OH concentrations
L316 – PM2.5 components increased obviously in P2 vs P1 – from the plot it is not obvious to me that they do: is there a statistically significant change ? What about changes in the meteorology…
L351 the traffic index data are useful. Consider showing P1/P2 on these plots. Consider changing the box/whisker plot to linear (not log) – this will assist the reader to follow which differences are statistically significant
L367 concentrations of NO changed: You cannot conclude this without a statistical test, esp for NO2 and HONO.
L387 Figure 3 the shift in diurnal profiles is interesting, explore further (esp panel 5, HONO/NO2) ?
L418 contribution (not interference) of other HONO sources. This matters vs literature discussion of interferences in some HONO measurement approaches
L490 ifs there any info on changes in fleet composition ? Big change in total traffic, but the greatest change may be in discretionary journeys (private cars) while deliveries etc (potentially much greater per-vehicle emitters) may have continued ?
L539 the agreement is good to see but is there really confidence in the combined uncertainty of the terms entering the calculation – especially [OH] – to have confidence ? L546 same point
L565 plus – these changes must be considered in the context of (potential) changes in meteorology between the two periods – it can be very misleading to simply compare means calculated from two different, fairly short, date periods.
Citation: https://doi.org/10.5194/egusphere-2023-3139-RC2 -
AC2: 'Reply to reviewer 2', Yongchun Liu, 06 Apr 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2023-3139/egusphere-2023-3139-AC2-supplement.pdf
-
AC2: 'Reply to reviewer 2', Yongchun Liu, 06 Apr 2024
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-3139', Yang Zeng, 13 Feb 2024
The authors have made commendable efforts to conduct a comprehensive field measurement in urban Beijing, spanning the unique transition period from the pre-pandemic COVID-19 era to the subsequent lockdown period. While the controversies outlined in the introduction section (lines 95-122) remain unresolved, this study observes a notable decrease in HONO concentration from the pre-pandemic eve to the lockdown phase. Furthermore, the study offers a reasonable explanation regarding the contributions of HONO sources. I find this work particularly valuable as it underscores the significance of vehicle emissions and highlights the rapid temporal shifts in the roles of HONO sources, which could be of great interest to readers of ACP. I recommend publication of the manuscript following minor revisions. Below are my specific comments:
The abstract requires improvement. Please remove lines 36-38 “which resulted in … to measure the HONO and related pollutants”, and restructure the sentences for better clarity.
Lines 67 and 73: The second and third paragraphs contain overlapping information. I recommend integrating the introduction about HONO concentrations into the first paragraph. Subsequently, the second paragraph should focus solely on direct emissions, while the third paragraph should discuss secondary formation.
Line 77: Here, "harvest season" may not be the most accurate term. Biomass burning encompasses a range of activities, including wildfires, which can be particularly significant under certain circumstances.
Line 90: The authors omitted the acid displacement (VandenBoer et al. 2015, Liu et al. 2019)?
Line 176: Livestock HONO emission is negligible in urban Beijing. Similarly, the emission of HONO from soil, which may peak following fertilizer application, likely does not need to be considered.
Line 236: heterogeneous yield of HONO instead of NO2?
Section 3.1 Please present the result following the order of Figure 1 (a)～(i).
Line 348: How about the heterogenous conversion on ground?
Line 417: Is HONOcorr equivalent to CHONO,corr,t as indicated in Eq. (6)? Please maintain consistency in the use of abbreviations.
Line 449: In fact, Song et al. (2023) proposed two HONO production pathways distinct from the homogeneous reaction between NO and OH.
Line 472: “a change of 25% and 95% of HONO sources, respectively.” What does this signify? It appears that the selection of the photolysis rate constant could have a significant influence.
Lines 472-477: Move the results regarding soil emissions to the subsequent paragraph.
Line 504: The NO2 concentration decreased significantly from 26.9 ppb to 17.2 ppb (-36%) from P1 to P2, indicating a substantial reduction rather than a slight decrease.
Lines 513-514: It’s noteworthy that HONO exhibits minimal sensitivity to both the uptake coefficient (γ) and surface area concentration (As). However, the authors should provide an explanation for this phenomenon.
Line 526: These variations (-9% to +40%) are significant and warrant attention.
Lines 552-555: The method utilized to estimate the overall uncertainty of the parameterization should be presented in the Experimental section.
References
Liu, Y. H., K. D. Lu, X. Li, H. B. Dong, Z. F. Tan, H. C. Wang, Q. Zou, Y. S. Wu, L. M. Zeng, M. Hu, K. E. Min, S. Kecorius, A. Wiedensohler and Y. H. Zhang (2019). "A comprehensive model test of the HONO sources constrained to field measurements at rural North China Plain." Environmental Science & Technology 53(7): 3517-3525.
Song, M., X. Zhao, P. Liu, J. Mu, G. He, C. Zhang, S. Tong, C. Xue, X. Zhao, M. Ge and Y. Mu (2023). "Atmospheric NOx oxidation as major sources for nitrous acid (HONO)." npj Climate and Atmospheric Science 6(1): 30.
VandenBoer, T. C., C. J. Young, R. K. Talukdar, M. Z. Markovic, S. S. Brown, J. M. Roberts and J. G. Murphy (2015). "Nocturnal loss and daytime source of nitrous acid through reactive uptake and displacement." Nature Geoscience 8(1): 55-60.
Citation: https://doi.org/10.5194/egusphere-2023-3139-RC1 -
AC1: 'Reply to reviewer 1', Yongchun Liu, 06 Apr 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2023-3139/egusphere-2023-3139-AC1-supplement.pdf
-
AC1: 'Reply to reviewer 1', Yongchun Liu, 06 Apr 2024
-
RC2: 'Comment on egusphere-2023-3139', Anonymous Referee #2, 24 Feb 2024
This paper presents an analysis of measurements of gas-phase nitrous acid (HONO), and related air pollutants / atmospheric chemical species, performed in the boundary layer in Beijing prior to and during the Covid-19 lockdowns in early 2020. The variation in abundance of HONO, its precursors and related species are used to quantify different chemical source and sink terms for HONO, and from their diurnal variation and change across the start of lockdown, inferences drawn re their relative importance – in particular direct vehicle emission.
The abundance and sources of HONO are very much a current topic in urban atmospheric chemistry / air pollution, as besides being a pollutant in its own right HONO is usually the major precursor to (reservoir for) the key oxidant OH. Lockdown presents a unique near-step-change opportunity to explore these processes. This paper adds to the developing understanding of these sources, and in that respect is a valuable contribution to the literature.
The measurements appear to have been carefully performed and are clearly presented, and the analysis is interesting, with clear signal in the change in mean diurnal variations (which are used to develop qualitative arguments).
However, I have substantial reservations about several key assumptions / aspects of the quantitative analysis approach, which in my opinion do not allow the authors to draw the conclusions they reach. These are
1 Meteorology. Obviously, weather changes affect pollutant advection, import, dispersion, chemical processing etc. The paper presents "before" vs "after" comparisons of concentrations either side of the start of lockdown, without consideration (beyond a basic vertical dispersion parameterisation) of these effects. There is significant literature on the importance of correcting for meteorology – for example applying de-weathering approaches – several using the specific example of air pollutant abundance in Beijing across lockdown (e.g. Jiabo et al., 2021: https://doi.org/10.1016/j.aosl.2021.100060; Shi et al., 2021: https://doi.org/10.1126/sciadv.abd6696; Lv et al., 2022: https://doi.org/10.1016/j.apr.2022.101452). How much of the observed change (or lack of change) in each species is due to changes in the weather between the two time periods P1 and P2 ? It would be interesting to repeat the analysis using deweathered concentrations.
2 Chemical lifetimes. The paper in effect performs a steady state analysis on HONO, assuming any rate of change of concentration can be related to an imbalance in the in-situ source and sink terms. You can do this for short-lived species – such as OH – but only with care for species such as HONO, with lifetimes (these are midlatitude winter measurements) of tens of minutes. The measured HONO reflects the integrated chemical variability over the sampled airmass trajectory prior to its arrival at the measurement point. This will be heterogeneous – especially at ground level in the middle of a city (i.e. the OH and NO2 etc will have varied a lot over the period of time – given by the HONO lifetime – that the airmass has travelled to the measurement point within the city). See arguments developed by Lee et al, JGR 2013 (https://doi.org/10.1002/2013JD020341) and related papers.
3 Statistical significance / precision / uncertainty. The paper does not consider sufficiently the uncertainty in the (many) source and sink terms considered, and their propagation together (beyond the monte carlo result, which I cant believe includes the uncertainty in the individual inputs, e.g. OH concentrations). Is there any statistical power resulting from their combined uncertainties ? Are the uncertainties given in the paper 1 or 2 standard deviations ? Are differences statistically significant ? The +/- ranges – even assuming these are 1 sd – suggest not (e.g. HONO/NO2 ratio – changing from 0.38+/- 0.035 to 0.042 +/- 0.034 – this is not a meaningful (statistically significant) change). There are several examples.
I’ve noted some more points below but the authors need to address the points noted above vs the overall approach, to have confidence that the results of their analysis allow them to draw the conclusions presented in the paper.
Introduction – reviews different sources for HONO and their contribution, but this mixes together very different environments (i.e. the relative importance of different sources will vary for the measurement site vs a road tunnel vs a bare soil location vs the marine boundary layer vs livestock). Suggest to distill this to assess the key factors at the measurement location, ie city centre
L157 – NO2 measured by 42i – selectivity for NO2, the NO2 data will include other Noy species (including HONO).
L168 how was j(HONO) calculated
L175 heterogeneous reactions of NO2 have been accounted for – I didn’t follow this section, may need rewording.
L193 what is the conversion factor alpha
L203+ using NO2 and CO – NO2 concentrations will vary with a lifetime of a minute or so (with respect to the NOx-O3 PSS) and 6-12 hours our so (with respect to NOx removal). CO concentrations will vary with a lifetime of several weeks. Is it valid to use both in the same in situ emission analysis ?
L230 / table 1 – what is the estimated OH concentration & how does it compare with measurements – the values in L436 (presumably 24 hour mean, around 4-7e5) seem much lower than those observed in wintertime Beijing (2.7e6 as 24-hour mean; Slater et al., https://doi.org/10.5194/acp-20-14847-2020).
L241 Parameterisation of OH concentrations. We might expect that the main source of OH in bejing is HO2 + NO, the main sink for OH is OH + NO, and the main primary source of OH is HONO photolysis (one can then argue about if HONO is acting as a primary source or a reservoir). The parameterisation given was developed for rural sites (as the authors note), where NO levels would be much lower, and was developed prior to more recent understanding of HONO abundance (it is over 20 years old). Is it valid to use ?
L259 – considering thee above I do not agree that we can be optimistic about the estimated OH concentrations
L316 – PM2.5 components increased obviously in P2 vs P1 – from the plot it is not obvious to me that they do: is there a statistically significant change ? What about changes in the meteorology…
L351 the traffic index data are useful. Consider showing P1/P2 on these plots. Consider changing the box/whisker plot to linear (not log) – this will assist the reader to follow which differences are statistically significant
L367 concentrations of NO changed: You cannot conclude this without a statistical test, esp for NO2 and HONO.
L387 Figure 3 the shift in diurnal profiles is interesting, explore further (esp panel 5, HONO/NO2) ?
L418 contribution (not interference) of other HONO sources. This matters vs literature discussion of interferences in some HONO measurement approaches
L490 ifs there any info on changes in fleet composition ? Big change in total traffic, but the greatest change may be in discretionary journeys (private cars) while deliveries etc (potentially much greater per-vehicle emitters) may have continued ?
L539 the agreement is good to see but is there really confidence in the combined uncertainty of the terms entering the calculation – especially [OH] – to have confidence ? L546 same point
L565 plus – these changes must be considered in the context of (potential) changes in meteorology between the two periods – it can be very misleading to simply compare means calculated from two different, fairly short, date periods.
Citation: https://doi.org/10.5194/egusphere-2023-3139-RC2 -
AC2: 'Reply to reviewer 2', Yongchun Liu, 06 Apr 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2023-3139/egusphere-2023-3139-AC2-supplement.pdf
-
AC2: 'Reply to reviewer 2', Yongchun Liu, 06 Apr 2024
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 459 | 116 | 31 | 606 | 56 | 13 | 22 |
- HTML: 459
- PDF: 116
- XML: 31
- Total: 606
- Supplement: 56
- BibTeX: 13
- EndNote: 22
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 181 | 30 |
| China | 2 | 141 | 23 |
| undefined | 3 | 41 | 6 |
| Germany | 4 | 32 | 5 |
| United Kingdom | 5 | 25 | 4 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 181
Yusheng Zhang
Feixue Zheng
Zemin Feng
Chaofan Lian
Xiaolong Fan
Wei Ma
Zhuohui Lin
Chang Li
Gen Zhang
Ying Zhang
Veli-Matti Kerminen
Federico Bianch
Tuukka Petäjä
Juha Kangasluoma
Markku Kulmala
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1962 KB) - Metadata XML
-
Supplement
(984 KB) - BibTeX
- EndNote
- Final revised paper