the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 23 Sep 2022
Measurement Report: Long-range transport and fate of DMS-oxidation products in the free troposphere derived from observations at the high-altitude research station Chacaltaya (5240 m a.s.l.) in the Bolivian Andes
Abstract. Dimethyl sulfide (DMS) is the primary natural contributor to the atmospheric sulfur burden. Observations concerning the fate of DMS oxidation products after long-range transport in the remote free troposphere are, however, sparse. Here we present quantitative chemical ionization mass spectrometric measurements of DMS and its oxidation products H2SO4, MSA, DMSO, DMSO2, MSIA, MTF, CH3S(O)2OOH and CH3SOH in the gas-phase as well as measurements of the sulfate and methane- sulfonate aerosol mass fractions at the Global Atmosphere Watch (GAW) station Chacaltaya in the Bolivian Andes located at 5240 m above sea level (a.s.l.).
DMS and DMS oxidation products are brought to the Andean high-altitude station by Pacific air masses during the dry season after convective lifting over the remote Pacific ocean to 6000–8000 m a.s.l. and subsequent long-range transport in the free troposphere (FT). Most of the DMS reaching the station is already converted to the rather unreactive sulfur reservoirs dimethyl sulfone (DMSO2) in the gas phase and methanesulfonate (MS−) in the particle phase, which carried nearly equal amounts of sulfur to the station. The particulate sulfate at Chacaltaya is however dominated by regional volcanic emissions during the time of the measurement and not significantly affected by the marine air masses. In one of the FT events, even some DMS was observed next to reactive intermediates such as methyl thioformate, dimethyl sulfoxide, and methane sulfinic acid. Also for this event, backtrajectory calculations show, that the air masses came from above the ocean (distance >330 km) with no local sur- face contacts. This study demonstrates the potential impact of marine DMS emissions on the availability of sulfur-containing vapors in the remote free troposphere far away from the ocean.
-
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
(11556 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(11556 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
AC1: 'Comment on egusphere-2022-887', Wiebke Scholz, 28 Sep 2022
The reference to Hoffmann et al., 2020 in lines 491-492:
"Such high DMSO2 yields are in contrast to chemical models, where DMSO2 yields are smaller or DMSO2 formation is even neglected completely (Hoffmann et al., 2020),"
is not fully correct. In their model, Hoffmann et al. only neglect liquid phase DMSO2 formation and further oxidation of DMSO2, while gas-phase DMSO2 production via DMSO + NO3 and DMSO + BrO are included.
We will correct this in future iterations.
Wiebke Scholz
Citation: https://doi.org/10.5194/egusphere-2022-887-AC1 -
RC1: 'Referee comment on egusphere-2022-887', Anonymous Referee #1, 05 Oct 2022
The paper presents a very rare high altitude data set of state-of-the-art measurements of DMS and its oxidation products from CHC in the Bolivian Andes. A modified FLEXPART scheme was used to identify source regions of the chemical species measured and the degree of influence of the boundary on the measured gas and aerosol phase composition. Strong evidence is presented for long range transport of DMS and oxidation products in the free troposphere across the Pacific to CHC.
Line 110: Why was the figure of the measurement site put into Appendix B rather than the main text? It would be useful to have it in the same section as the station description for those not familiar with the geography surrounding CHC. Also, please label the location of Lake Titicaca and the other lagoons mentioned in line 121.
Table 1: The Q-ACSM species should be labelled as non-refractory in the table.
Line 142: It is stated that the nitrate-CIMS made measurements from April 19th to 25th but based on what is shown in Figure 1, it looks like those data were not analyzed. An explanation of the reason why not all data was analyzed (for all instruments) would be useful.
Line 200 says the FIGAERO-CIMS was operational April 10th to June 2nd but Figure 1 indicates it became operational in early May. Vertical lines corresponding to dates on the x-axis would help.
Section 3.1: The different panels within Figure 2 should be described in the order they are mentioned in the text. Currently, the order is 2A, 2E, 2C, etc.
Lines 274 – 275: It doesn’t seem necessary to say that they “appear anticorrelated”. Can’t the regression be performed to see what the degree of anticorrelation is?
Lines 297 – 302: The panels within Figure 3 are also mentioned out of order in the text.
Line 300: It is difficult to understand what is being said here since it is not clear what Figures 3 B1 and B2 are. Is this statement based instead on Figure 3F (“during the afternoon of the same day, shown in Fig. 3 B1 and B2, a still small, but larger fraction of the air travelled uphill close to the surface”)?
Two methods are used for designating the sampled air masses as primarily FT with little influence from the boundary layer or influenced by the local boundary layer - the FLEXPART analysis and the value of the identifiers listed in Table 2. Some discussion of why two methods were used and the unique information each provided would be useful.
Lines 312 – 315: It would be helpful to include a figure showing the FLEXPART domain when it is discussed in Lines 251 – 254.
Line 314: Should be “Intertropical” Convergence Zone.
Throughout: please use uniform date-time stamps in the text and figure captions.
Figure 4: Again, panels are mentioned out of order in the text.
Figure 5: How many data points are the boxplots constructed from? Would it be more appropriate to use an average and standard deviation?
Figure 7 is mentioned in the text before Figure 6.
Figure 7: Gridlines would help guide the eye to see correlations between peaks and valleys of the plotted parameters.
Figure 6: It would be helpful to add the monthly wind direction to the figure.
Figure 8 could be included earlier in the manuscript to show the FLEXPART domain and the location of CHC and Lake Titicaca.
Citation: https://doi.org/10.5194/egusphere-2022-887-RC1 -
AC2: 'reply to RC1 - discussing open questions', Wiebke Scholz, 10 Oct 2022
Dear Colleague!
Thank you very much for your early and helpful review comments! I like to use the opportunity to respond to your questions now in case any of my answers would trigger any further discussion.
1. Operational vs. Analyzed periods of the different instruments:
- Nitrate CIMS: The nitrate-CIMS was operational for an extended period, as shown in figure 1. However, due to technical difficulties and many power cuts during the wet season, the instrument was not always in a good state, so large fractions of the MSA data are very uncertain. The 19th-25th of April would be of good quality. We however decided not to go into further detail with these data, because – as mentioned a few times throughout the manuscript - we focused our analysis on periods of typical dry season conditions (clear sky, westerly winds) that were not met in that time period. Maybe the word „analyzed“ is also not perfect, as we have looked into all data, however, this paper only focuses on the marked periods. To clarify this, we will add a note in the caption of figure 1 and change the word „analyzed“ to „herein presented“ data.
- I⁻-Figaero: Yes, the inconsistency probably occurred due to a misunderstanding within the group, that was resolved shortly before submission and we forgot to update the figure. The dates given in the text are correct. The reason for not focusing on the data from April is again, that our aim was to focus on typical dry-season conditions for this manuscript.
2. Lines 274 – 275 (Anticorrelation):
„Anticorrelation“ might be the wrong word, as we have no close-to-linear relationship between these data, but the scatter plot (appended) shows that the processes leading to high MSA vs. high summed C8H10Ox, are counteractive.
3. Line 300:
Yes, that must‘ve slipped my eye. Previous figs. 3 B1 and B2 are now figs. 3 E and F and we will correct it in the next version.
4. Why we used two methods for free troposphere (FT) identification:
The thresholds of the identifiers are more precise than the Flexpart analysis to analyze whether the station is above the boundary layer or not:
The FT intrusions happen on a timescale of hours which is close to the limit of the model time resolution, while the various identifiers have a higher temporal resolution.
In addition, the identifier approach uses direct on-site measurements that are generally more reliable than the Flexpart analysis, which builds upon the WRF model that cannot perfectly represent all the complex mountain meteorology in the vicinity.
The model, when evaluated against short-range BL influences does perform very well on average (see Bianchi et al 2021, doi.org/10.1175/BAMS-D-20-0187.1 ). Also, fig 3 of this paper shows this great performance and the agreement between these methods gives us confidence in our FT identification. However, for single events, there are generally higher chances of the model missing the complex local meteorological features that drive the FT/BL interactions.
To summarize, for a temporally accurate subdivision into „above“ and „within“ the boundary layer, the indicators are more targeted.
On the other hand, identification of the air mass origin with Flexpart is an excellent tool for determining the most likely source regions of a compound (footprint analysis). The footprint analysis requires as much data as possible to get a good statistic (we used the whole time series from the PTR3). It is still somewhat uncertain because it does not account for processes other than passive transport.
However, one advantage of the footprint analysis is that it shows where the air masses transporting the compound of interest came from. An origin analysis in the horizontal direction is only possible with Flexpart and not with the parameters measured directly at the station. It also identifies the height at which the long-range transport occurred.
5. Figure 5:
Each boxplot is based upon 32 data points (pre-averaged to 30min intervals beforehand). The night- and daytime boundary layer conditions were chosen to be as close as possible (time-wise) to the FT periods with the same prevailing wind direction (horizontal air mass origin) for the best comparison. I think it makes sense to show the mean value as you suggested (see the appended figure as an example. The mean value is the green triangle). In a few cases (like DMS), a few very high values impact the mean. The data are also not gaussian distributed, so giving the standard deviation is maybe not so purposeful. An alternative to the boxplots would be violin plots if anything.
As mentioned, I reply to you now to spark a short discussion on these topics, if necessary. Therefore, this reply does not resolve all your more technical comments regarding the different figures. We will however address all your comments and adjust our figures accordingly in our next version of the manuscript after receiving further reviewer comments.
All the best, Wiebke Scholz
-
AC2: 'reply to RC1 - discussing open questions', Wiebke Scholz, 10 Oct 2022
-
RC2: 'Comment on egusphere-2022-887', Anonymous Referee #2, 29 Nov 2022
This manuscript summarizes a large dataset of gas and aerosol data collected at the GAW station in the Bolivian Andes. The authors deployed five state of the art mass spectrometers to measure DMS and its oxidation products. The data are further supported by the regular station measurements. The authors used chemical markers and FLEXPART to eliminate local contamination and to select periods of free troposphere flow. The relative mixing ratios of DMS and its oxidation products could be explained in these periods of free troposphere flow based on the compound lifetimes. It is a very interesting data set. The paper is well written and supported by five appendices. I recommend publication as is.
Citation: https://doi.org/10.5194/egusphere-2022-887-RC2
Interactive discussion
Status: closed
-
AC1: 'Comment on egusphere-2022-887', Wiebke Scholz, 28 Sep 2022
The reference to Hoffmann et al., 2020 in lines 491-492:
"Such high DMSO2 yields are in contrast to chemical models, where DMSO2 yields are smaller or DMSO2 formation is even neglected completely (Hoffmann et al., 2020),"
is not fully correct. In their model, Hoffmann et al. only neglect liquid phase DMSO2 formation and further oxidation of DMSO2, while gas-phase DMSO2 production via DMSO + NO3 and DMSO + BrO are included.
We will correct this in future iterations.
Wiebke Scholz
Citation: https://doi.org/10.5194/egusphere-2022-887-AC1 -
RC1: 'Referee comment on egusphere-2022-887', Anonymous Referee #1, 05 Oct 2022
The paper presents a very rare high altitude data set of state-of-the-art measurements of DMS and its oxidation products from CHC in the Bolivian Andes. A modified FLEXPART scheme was used to identify source regions of the chemical species measured and the degree of influence of the boundary on the measured gas and aerosol phase composition. Strong evidence is presented for long range transport of DMS and oxidation products in the free troposphere across the Pacific to CHC.
Line 110: Why was the figure of the measurement site put into Appendix B rather than the main text? It would be useful to have it in the same section as the station description for those not familiar with the geography surrounding CHC. Also, please label the location of Lake Titicaca and the other lagoons mentioned in line 121.
Table 1: The Q-ACSM species should be labelled as non-refractory in the table.
Line 142: It is stated that the nitrate-CIMS made measurements from April 19th to 25th but based on what is shown in Figure 1, it looks like those data were not analyzed. An explanation of the reason why not all data was analyzed (for all instruments) would be useful.
Line 200 says the FIGAERO-CIMS was operational April 10th to June 2nd but Figure 1 indicates it became operational in early May. Vertical lines corresponding to dates on the x-axis would help.
Section 3.1: The different panels within Figure 2 should be described in the order they are mentioned in the text. Currently, the order is 2A, 2E, 2C, etc.
Lines 274 – 275: It doesn’t seem necessary to say that they “appear anticorrelated”. Can’t the regression be performed to see what the degree of anticorrelation is?
Lines 297 – 302: The panels within Figure 3 are also mentioned out of order in the text.
Line 300: It is difficult to understand what is being said here since it is not clear what Figures 3 B1 and B2 are. Is this statement based instead on Figure 3F (“during the afternoon of the same day, shown in Fig. 3 B1 and B2, a still small, but larger fraction of the air travelled uphill close to the surface”)?
Two methods are used for designating the sampled air masses as primarily FT with little influence from the boundary layer or influenced by the local boundary layer - the FLEXPART analysis and the value of the identifiers listed in Table 2. Some discussion of why two methods were used and the unique information each provided would be useful.
Lines 312 – 315: It would be helpful to include a figure showing the FLEXPART domain when it is discussed in Lines 251 – 254.
Line 314: Should be “Intertropical” Convergence Zone.
Throughout: please use uniform date-time stamps in the text and figure captions.
Figure 4: Again, panels are mentioned out of order in the text.
Figure 5: How many data points are the boxplots constructed from? Would it be more appropriate to use an average and standard deviation?
Figure 7 is mentioned in the text before Figure 6.
Figure 7: Gridlines would help guide the eye to see correlations between peaks and valleys of the plotted parameters.
Figure 6: It would be helpful to add the monthly wind direction to the figure.
Figure 8 could be included earlier in the manuscript to show the FLEXPART domain and the location of CHC and Lake Titicaca.
Citation: https://doi.org/10.5194/egusphere-2022-887-RC1 -
AC2: 'reply to RC1 - discussing open questions', Wiebke Scholz, 10 Oct 2022
Dear Colleague!
Thank you very much for your early and helpful review comments! I like to use the opportunity to respond to your questions now in case any of my answers would trigger any further discussion.
1. Operational vs. Analyzed periods of the different instruments:
- Nitrate CIMS: The nitrate-CIMS was operational for an extended period, as shown in figure 1. However, due to technical difficulties and many power cuts during the wet season, the instrument was not always in a good state, so large fractions of the MSA data are very uncertain. The 19th-25th of April would be of good quality. We however decided not to go into further detail with these data, because – as mentioned a few times throughout the manuscript - we focused our analysis on periods of typical dry season conditions (clear sky, westerly winds) that were not met in that time period. Maybe the word „analyzed“ is also not perfect, as we have looked into all data, however, this paper only focuses on the marked periods. To clarify this, we will add a note in the caption of figure 1 and change the word „analyzed“ to „herein presented“ data.
- I⁻-Figaero: Yes, the inconsistency probably occurred due to a misunderstanding within the group, that was resolved shortly before submission and we forgot to update the figure. The dates given in the text are correct. The reason for not focusing on the data from April is again, that our aim was to focus on typical dry-season conditions for this manuscript.
2. Lines 274 – 275 (Anticorrelation):
„Anticorrelation“ might be the wrong word, as we have no close-to-linear relationship between these data, but the scatter plot (appended) shows that the processes leading to high MSA vs. high summed C8H10Ox, are counteractive.
3. Line 300:
Yes, that must‘ve slipped my eye. Previous figs. 3 B1 and B2 are now figs. 3 E and F and we will correct it in the next version.
4. Why we used two methods for free troposphere (FT) identification:
The thresholds of the identifiers are more precise than the Flexpart analysis to analyze whether the station is above the boundary layer or not:
The FT intrusions happen on a timescale of hours which is close to the limit of the model time resolution, while the various identifiers have a higher temporal resolution.
In addition, the identifier approach uses direct on-site measurements that are generally more reliable than the Flexpart analysis, which builds upon the WRF model that cannot perfectly represent all the complex mountain meteorology in the vicinity.
The model, when evaluated against short-range BL influences does perform very well on average (see Bianchi et al 2021, doi.org/10.1175/BAMS-D-20-0187.1 ). Also, fig 3 of this paper shows this great performance and the agreement between these methods gives us confidence in our FT identification. However, for single events, there are generally higher chances of the model missing the complex local meteorological features that drive the FT/BL interactions.
To summarize, for a temporally accurate subdivision into „above“ and „within“ the boundary layer, the indicators are more targeted.
On the other hand, identification of the air mass origin with Flexpart is an excellent tool for determining the most likely source regions of a compound (footprint analysis). The footprint analysis requires as much data as possible to get a good statistic (we used the whole time series from the PTR3). It is still somewhat uncertain because it does not account for processes other than passive transport.
However, one advantage of the footprint analysis is that it shows where the air masses transporting the compound of interest came from. An origin analysis in the horizontal direction is only possible with Flexpart and not with the parameters measured directly at the station. It also identifies the height at which the long-range transport occurred.
5. Figure 5:
Each boxplot is based upon 32 data points (pre-averaged to 30min intervals beforehand). The night- and daytime boundary layer conditions were chosen to be as close as possible (time-wise) to the FT periods with the same prevailing wind direction (horizontal air mass origin) for the best comparison. I think it makes sense to show the mean value as you suggested (see the appended figure as an example. The mean value is the green triangle). In a few cases (like DMS), a few very high values impact the mean. The data are also not gaussian distributed, so giving the standard deviation is maybe not so purposeful. An alternative to the boxplots would be violin plots if anything.
As mentioned, I reply to you now to spark a short discussion on these topics, if necessary. Therefore, this reply does not resolve all your more technical comments regarding the different figures. We will however address all your comments and adjust our figures accordingly in our next version of the manuscript after receiving further reviewer comments.
All the best, Wiebke Scholz
-
AC2: 'reply to RC1 - discussing open questions', Wiebke Scholz, 10 Oct 2022
-
RC2: 'Comment on egusphere-2022-887', Anonymous Referee #2, 29 Nov 2022
This manuscript summarizes a large dataset of gas and aerosol data collected at the GAW station in the Bolivian Andes. The authors deployed five state of the art mass spectrometers to measure DMS and its oxidation products. The data are further supported by the regular station measurements. The authors used chemical markers and FLEXPART to eliminate local contamination and to select periods of free troposphere flow. The relative mixing ratios of DMS and its oxidation products could be explained in these periods of free troposphere flow based on the compound lifetimes. It is a very interesting data set. The paper is well written and supported by five appendices. I recommend publication as is.
Citation: https://doi.org/10.5194/egusphere-2022-887-RC2
Peer review completion
Journal article(s) based on this preprint
Data sets
Data and Code for figures of "Long-range transport and fate of DMS-oxidation products in the free troposphere derived from observations at the high-altitude research station Chacaltaya (5240 m a.s.l.) in the Bolivian Andes" Wiebke Scholz, Jiali Shen, Diego Aliaga, Cheng Wu, Samara Carbone, Isabel Moreno, Qiaozhi Zha, Wei Huang, Liine Heikkinen, Jean-Luc Jaffrezo, Gaelle Uzu, Eva Partoll, Markus Leiminger, Fernando Velarde, Paolo Laj, Patrick Ginot, Paolo Artaxo, Alfred Wiedensohler, Markku Kulmala, Claudia Mohr, Marcos Andrade, Victoria Sinclair, Federico Bianchi, Armin Hansel https://doi.org/10.5281/zenodo.6866115
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 526 | 168 | 15 | 709 | 6 | 6 |
- HTML: 526
- PDF: 168
- XML: 15
- Total: 709
- BibTeX: 6
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Jiali Shen
Diego Aliaga
Cheng Wu
Samara Carbone
Isabel Moreno
Qiaozhi Zha
Wei Huang
Liine Heikkinen
Jean Luc Jaffrezo
Gaelle Uzu
Eva Partoll
Markus Leiminger
Fernando Velarde
Paolo Laj
Patrick Ginot
Paolo Artaxo
Alfred Wiedensohler
Markku Kulmala
Claudia Mohr
Marcos Andrade
Victoria Sinclair
Federico Bianchi
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(11556 KB) - Metadata XML