the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 04 Jul 2024
Quantifying the Impacts of Marine Aerosols over the Southeast Atlantic Ocean using a chemical transport model: Implications for aerosol-cloud interactions
Abstract. The southeast Atlantic region, characterized by persistent stratocumulus clouds, has one of the highest uncertainties in aerosol radiative forcing and significant variability across climate models. In this study, we analyze the seasonally varying role of marine aerosol sources and identify key uncertainties in aerosol composition at cloud-relevant altitudes over the southeast Atlantic using the GEOS-Chem chemical transport model. We evaluate simulated aerosol optical depth (AOD) and speciated aerosol concentrations against those collected from ground observations and aircraft campaigns such as LASIC, ORACLES, and CLARIFY, conducted during 2017. The model consistently underestimates AOD relative to AERONET, particularly at remote locations like Ascension Island. However, when compared with aerosol mass concentrations from aircraft campaigns during the biomass burning period, it performs adequately at cloud-relevant altitudes, with a normalized mean bias (NMB) between −3.5 % (CLARIFY) and −7.5 % (ORACLES). At these altitudes, organic aerosols (63 %) dominate during the biomass burning period, while sulfate (41 %) prevails during austral summer, when dimethylsulfide (DMS) emissions peak in the model. Our findings indicate that marine sulfate can account for up to 69 % of total sulfate during high DMS period. Sensitivity analyses indicate that refining DMS emissions and oxidation chemistry may increase sulfate aerosol produced from marine sources, highlighting their overall importance. Additionally, we find marine primary organic aerosol emissions may substantially increase total organic aerosol concentrations, particularly during austral summer. This study underscores the imperative need to refine marine emissions and their chemical transformations to better predict aerosol-cloud interactions and reduce uncertainties in aerosol radiative forcing over the southeast Atlantic.
- Preprint
(1744 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2024-1948', Anonymous Referee #1, 23 Jul 2024
Overall, the paper and analysis are solid. As noted in the Conclusions, “This study highlights the importance of constraining marine emissions and their chemical transformations by incorporating satellite-retrieved datasets and extending field campaign efforts during non-biomass burning periods.” This is a useful contribution to the literature on the aerosols that influence clouds in the SE Atlantic stratocumulus region – where most of the focus has been on the influence of biomass burning emissions. The inclusion of the section on Uncertainties, and quantifying how different emission inventories and marine organic emission affect the papers’ conclusions, is very good.
I have just a few more substantial comments, and then a set of smaller points. My recommendation is that the paper be published after minor revision, per the following comments:
More substantial points:
While there’s no harm in comparing the modeled and observed AODs, the stated focus of the paper is on the role of marine aerosols in aerosol-cloud interactions over the SE Atlantic. In this region, especially during the season when in-situ observations are available for testing the model, the majority of AOD is above-cloud. The real information informing model biases of boundary-layer / in-cloud aerosols comes from the vertically-resolved observations from CLARIFY and ORACLES. I think Section 3.1.1 could be significantly shortened, to focus on just giving the big picture of sources of aerosols over the region.
In contrast, given the focus of the paper it seems to me that Figure A3, showing the seasonal emissions sources for sulfur, and Table A3, giving the fractional contribution of marine sulfate at cloud height by month, should be shown in the main part of the paper. Also possibly Figure A4. I think the AOD plots could instead be moved to the Appendix if wanting to limit the number of figures in the main paper.
Regarding Figure A3: The Anthropogenic, Volcanic and Ship and Aircraft sources of sulfate appear to have no seasonality. Particularly for the Anthropogenic contribution – which is the largest of all emissions sources – this seems odd at best. Is this due to a lack of better information? An assumption? If this is based on not knowing the seasonal variability, versus there not being any actual seasonal variability, this needs to be at least acknowledged, as it for sure will affect conclusions about the relative contributions of each in different seasons.
Lines 200-206: Here, it is posited that several known biases in GEOS-Chem could be responsible for the 26.5% NMB in AOD during the biomass burning season. However, in Figure 5 it is pretty clear that there are significant low biases in GEOS-Chem in a) the amount of aerosol at higher altitudes, and significant low biases in the amount of NO3 – in addition to low biases in OA. These are likely bigger factors than inaccuracy in the aerosol optical properties of the aerosol. It would make more sense to discuss the source of model biases after discussing Figure 5, because then you can put known model issues in the context of the observed biases.
Line 225: a low bias of 20->55% is not “moderate agreement”… This is a significant low bias.
Figures 3 and Figure 7: Please add latitude and longitude to these maps. The longitude markers in Figure 3 need fixing as they are given as negative in both the E and W directions. Also, these maps don’t cover the 0-30S, 20E-20W box specified in the text & Fig 6 caption – so they either should, or that box should be indicated in the figure.
On lines 86-87, it says that in the model: “Organic aerosol follows the “simple” scheme which treats primary organic aerosol (POA) as non-volatile and includes irreversible direct yield of SOA from precursors.” i.e.: The model represents the contribution of both POA and SOA to total OA in the model. However lines 356-357 then lines 360-361, say: “Beyond marine sulfate and sea-salt aerosols, organic matter also makes a significant contribution to marine aerosol … However, the standard GEOS-Chem model does not account for these organic aerosol emissions. We analyzed the impact of marine POA on cloud-altitude aerosols over the SEA by incorporating POA emissions based on satellite-derived chlorophyll-a concentrations.” Can you please better explain what is not being accounted for in the model, that this analysis is looking at? Does the model account for land-based emissions that contribute to POA and SOA but not ocean-based? I was confused by this.
Lines 431-432: It’s asserted that “These underestimations are primarily due to limitations in representing natural aerosol emissions, transatlantic aerosol transport, particle mixing states, and the oxidation levels of organic aerosols.” However, these are really hypothesized sources of bias; the analysis in the paper doesn’t involve analyzing the contributions of each of these to biases in simulated AOD, so I don’t think it should be asserted here that this is the case, as no evidence given to support this. Again, I think the part of the analysis that focuses on AOD really doesn’t add much to addressing the main question of the paper, which is how marine sources contribute to sulfate and organics at cloud altitudes.
Smaller points:
Abstract, pg 1, lines 17-18: “At these altitudes, organic aerosols (63%) dominate during the biomass burning period, while sulfate (41%) prevails during austral summer, when dimethylsulfide (DMS) emissions peak in the model.” It’s not clear whether these numbers are also coming from the model. If all are from the model, reword to e.g.: “At these altitudes, in the model organic aerosols (63%) dominate during the biomass burning period, while sulfate (41%) prevails during austral summer, when dimethylsulfide (DMS) emissions peak.” If not from the model, where are these numbers from?
Abstract, pg 1, lines 19-21: “Sensitivity analyses indicate that refining DMS emissions and oxidation chemistry may increase sulfate aerosol produced from marine sources, highlighting their overall importance.” It’s not clear to me how demonstrating that changing the oxidation chemistry has an effect on sulfate production indicates that DMS itself is necessarily important (important in what regard?). I'd say instead that this analysis highlights that there remains large uncertainty in the role of DMS emissions in marine boundary layer sulfate, which is important given that DMS appears to make a significant contribution to sulfate concentration in the Sc clouds in the SE Atlantic region.
Abstract, pg 1, lines 22-24: “This study underscores the imperative need to refine marine emissions and their chemical transformations to better predict aerosol-cloud interactions and reduce uncertainties in aerosol radiative forcing over the southeast Atlantic.” Nothing given up to this point in the abstract (or in the paper) quantifies how biases in the contribution of marine sulfate actually affect aerosol-cloud interactions – only that marine aerosol account for a large fraction of the sulfate at cloud altitudes outside of the biomass burning season. So I don’t think it can be asserted that this is a significant source of bias or uncertainty in aerosol-cloud interactions over the SE Atlantic. It might indeed be; but the analysis here does not allow one to conclude that.
Pg 1, lines 30-31: “However, aerosol radiative forcing in the region exhibits highest uncertainty and one of the largest intermodel spread,” Across what set of models? And relative to the aerosol radiative forcing in other stratocumulus regions? Or in clouds anywhere? This needs to be contextualized.
Pg 2, lines 36-37: “and sources of uncertainty affecting aerosol composition" The paper focuses on this for the boundary layer/marine aerosol only – not for the biomass burning aerosol aloft, which is important since the elevated biomass burning aerosol layer can significantly contribute to aerosol-cloud interactions for the year.
Pg 2, lines 47-48: “leading to largest uncertainty of aerosol radiative forcing within climate models” As above, this needs context. Are you really saying that the formation of sulfate from DMS is the largest source of uncertainty in aerosol radiative forcing? Such a strong statement would need to be supported by a more recent set of publications than the 2013 reference cited, as climate models have evolved significantly in the past decade.
Lines 59-60 & 68-70: A very small point, but it would be helpful to the reader if the references given on lines 59-60 were mapped to the field campaigns given on lines 68-70. This could be done by giving (again) the citation for the overview paper of each campaign, e.g.: “(CLARIFY; Haywood et al., 2021)”
Line 81: “as 10 minutes” --> “of 10 minutes”
Lines 85 & 86: I am not sure what you mean by “follows”. In terms of emissions? In the optical / size / other properties of these constituents?
Lines 98-99: I assume that sea salt emissions are also wind-dependent!
Figure 3: There is clearly a peak in emissions / AOD in the vicinity of Johannesburg. The text around lines 164-168 comments on the other features of the AOD but not this one; it would be good to do so just briefly. (It’s referenced when discussing Fig 3b, but not the other panels – and it’s present in all three)
Lines 173-174: “This increase, combined with dust emissions from the Namib desert, contributes to an AOD hotspot as depicted in Fig. 3a on the southwestern coast.” Figure 7 shows the contribution from sulfate. Presumably you can therefore make a statement here about the relative contributions of sulfate vs dust to this ‘hot spot’. Correct?
Line 190: Not just a low correlation coefficient – but a negative correlation coefficient - !
Lines 199-200: There is an incomplete sentence here.
Lines 450-451: “The limited spatial and temporal coverage of the Lana dataset across our domain resulted in a 51% overestimate in emissions in July and a 38% underestimate in December relative to Galí.” Unless you know that the Galí dataset is “truth”, this would be better phrased as a difference between the two rather than a bias in the Lana dataset.
Citation: https://doi.org/10.5194/egusphere-2024-1948-RC1 - RC2: 'Comment on egusphere-2024-1948', Anonymous Referee #2, 12 Aug 2024
- AC1: 'Comment on egusphere-2024-1948', Mashiat Hossain, 26 Sep 2024
Status: closed
-
RC1: 'Comment on egusphere-2024-1948', Anonymous Referee #1, 23 Jul 2024
Overall, the paper and analysis are solid. As noted in the Conclusions, “This study highlights the importance of constraining marine emissions and their chemical transformations by incorporating satellite-retrieved datasets and extending field campaign efforts during non-biomass burning periods.” This is a useful contribution to the literature on the aerosols that influence clouds in the SE Atlantic stratocumulus region – where most of the focus has been on the influence of biomass burning emissions. The inclusion of the section on Uncertainties, and quantifying how different emission inventories and marine organic emission affect the papers’ conclusions, is very good.
I have just a few more substantial comments, and then a set of smaller points. My recommendation is that the paper be published after minor revision, per the following comments:
More substantial points:
While there’s no harm in comparing the modeled and observed AODs, the stated focus of the paper is on the role of marine aerosols in aerosol-cloud interactions over the SE Atlantic. In this region, especially during the season when in-situ observations are available for testing the model, the majority of AOD is above-cloud. The real information informing model biases of boundary-layer / in-cloud aerosols comes from the vertically-resolved observations from CLARIFY and ORACLES. I think Section 3.1.1 could be significantly shortened, to focus on just giving the big picture of sources of aerosols over the region.
In contrast, given the focus of the paper it seems to me that Figure A3, showing the seasonal emissions sources for sulfur, and Table A3, giving the fractional contribution of marine sulfate at cloud height by month, should be shown in the main part of the paper. Also possibly Figure A4. I think the AOD plots could instead be moved to the Appendix if wanting to limit the number of figures in the main paper.
Regarding Figure A3: The Anthropogenic, Volcanic and Ship and Aircraft sources of sulfate appear to have no seasonality. Particularly for the Anthropogenic contribution – which is the largest of all emissions sources – this seems odd at best. Is this due to a lack of better information? An assumption? If this is based on not knowing the seasonal variability, versus there not being any actual seasonal variability, this needs to be at least acknowledged, as it for sure will affect conclusions about the relative contributions of each in different seasons.
Lines 200-206: Here, it is posited that several known biases in GEOS-Chem could be responsible for the 26.5% NMB in AOD during the biomass burning season. However, in Figure 5 it is pretty clear that there are significant low biases in GEOS-Chem in a) the amount of aerosol at higher altitudes, and significant low biases in the amount of NO3 – in addition to low biases in OA. These are likely bigger factors than inaccuracy in the aerosol optical properties of the aerosol. It would make more sense to discuss the source of model biases after discussing Figure 5, because then you can put known model issues in the context of the observed biases.
Line 225: a low bias of 20->55% is not “moderate agreement”… This is a significant low bias.
Figures 3 and Figure 7: Please add latitude and longitude to these maps. The longitude markers in Figure 3 need fixing as they are given as negative in both the E and W directions. Also, these maps don’t cover the 0-30S, 20E-20W box specified in the text & Fig 6 caption – so they either should, or that box should be indicated in the figure.
On lines 86-87, it says that in the model: “Organic aerosol follows the “simple” scheme which treats primary organic aerosol (POA) as non-volatile and includes irreversible direct yield of SOA from precursors.” i.e.: The model represents the contribution of both POA and SOA to total OA in the model. However lines 356-357 then lines 360-361, say: “Beyond marine sulfate and sea-salt aerosols, organic matter also makes a significant contribution to marine aerosol … However, the standard GEOS-Chem model does not account for these organic aerosol emissions. We analyzed the impact of marine POA on cloud-altitude aerosols over the SEA by incorporating POA emissions based on satellite-derived chlorophyll-a concentrations.” Can you please better explain what is not being accounted for in the model, that this analysis is looking at? Does the model account for land-based emissions that contribute to POA and SOA but not ocean-based? I was confused by this.
Lines 431-432: It’s asserted that “These underestimations are primarily due to limitations in representing natural aerosol emissions, transatlantic aerosol transport, particle mixing states, and the oxidation levels of organic aerosols.” However, these are really hypothesized sources of bias; the analysis in the paper doesn’t involve analyzing the contributions of each of these to biases in simulated AOD, so I don’t think it should be asserted here that this is the case, as no evidence given to support this. Again, I think the part of the analysis that focuses on AOD really doesn’t add much to addressing the main question of the paper, which is how marine sources contribute to sulfate and organics at cloud altitudes.
Smaller points:
Abstract, pg 1, lines 17-18: “At these altitudes, organic aerosols (63%) dominate during the biomass burning period, while sulfate (41%) prevails during austral summer, when dimethylsulfide (DMS) emissions peak in the model.” It’s not clear whether these numbers are also coming from the model. If all are from the model, reword to e.g.: “At these altitudes, in the model organic aerosols (63%) dominate during the biomass burning period, while sulfate (41%) prevails during austral summer, when dimethylsulfide (DMS) emissions peak.” If not from the model, where are these numbers from?
Abstract, pg 1, lines 19-21: “Sensitivity analyses indicate that refining DMS emissions and oxidation chemistry may increase sulfate aerosol produced from marine sources, highlighting their overall importance.” It’s not clear to me how demonstrating that changing the oxidation chemistry has an effect on sulfate production indicates that DMS itself is necessarily important (important in what regard?). I'd say instead that this analysis highlights that there remains large uncertainty in the role of DMS emissions in marine boundary layer sulfate, which is important given that DMS appears to make a significant contribution to sulfate concentration in the Sc clouds in the SE Atlantic region.
Abstract, pg 1, lines 22-24: “This study underscores the imperative need to refine marine emissions and their chemical transformations to better predict aerosol-cloud interactions and reduce uncertainties in aerosol radiative forcing over the southeast Atlantic.” Nothing given up to this point in the abstract (or in the paper) quantifies how biases in the contribution of marine sulfate actually affect aerosol-cloud interactions – only that marine aerosol account for a large fraction of the sulfate at cloud altitudes outside of the biomass burning season. So I don’t think it can be asserted that this is a significant source of bias or uncertainty in aerosol-cloud interactions over the SE Atlantic. It might indeed be; but the analysis here does not allow one to conclude that.
Pg 1, lines 30-31: “However, aerosol radiative forcing in the region exhibits highest uncertainty and one of the largest intermodel spread,” Across what set of models? And relative to the aerosol radiative forcing in other stratocumulus regions? Or in clouds anywhere? This needs to be contextualized.
Pg 2, lines 36-37: “and sources of uncertainty affecting aerosol composition" The paper focuses on this for the boundary layer/marine aerosol only – not for the biomass burning aerosol aloft, which is important since the elevated biomass burning aerosol layer can significantly contribute to aerosol-cloud interactions for the year.
Pg 2, lines 47-48: “leading to largest uncertainty of aerosol radiative forcing within climate models” As above, this needs context. Are you really saying that the formation of sulfate from DMS is the largest source of uncertainty in aerosol radiative forcing? Such a strong statement would need to be supported by a more recent set of publications than the 2013 reference cited, as climate models have evolved significantly in the past decade.
Lines 59-60 & 68-70: A very small point, but it would be helpful to the reader if the references given on lines 59-60 were mapped to the field campaigns given on lines 68-70. This could be done by giving (again) the citation for the overview paper of each campaign, e.g.: “(CLARIFY; Haywood et al., 2021)”
Line 81: “as 10 minutes” --> “of 10 minutes”
Lines 85 & 86: I am not sure what you mean by “follows”. In terms of emissions? In the optical / size / other properties of these constituents?
Lines 98-99: I assume that sea salt emissions are also wind-dependent!
Figure 3: There is clearly a peak in emissions / AOD in the vicinity of Johannesburg. The text around lines 164-168 comments on the other features of the AOD but not this one; it would be good to do so just briefly. (It’s referenced when discussing Fig 3b, but not the other panels – and it’s present in all three)
Lines 173-174: “This increase, combined with dust emissions from the Namib desert, contributes to an AOD hotspot as depicted in Fig. 3a on the southwestern coast.” Figure 7 shows the contribution from sulfate. Presumably you can therefore make a statement here about the relative contributions of sulfate vs dust to this ‘hot spot’. Correct?
Line 190: Not just a low correlation coefficient – but a negative correlation coefficient - !
Lines 199-200: There is an incomplete sentence here.
Lines 450-451: “The limited spatial and temporal coverage of the Lana dataset across our domain resulted in a 51% overestimate in emissions in July and a 38% underestimate in December relative to Galí.” Unless you know that the Galí dataset is “truth”, this would be better phrased as a difference between the two rather than a bias in the Lana dataset.
Citation: https://doi.org/10.5194/egusphere-2024-1948-RC1 - RC2: 'Comment on egusphere-2024-1948', Anonymous Referee #2, 12 Aug 2024
- AC1: 'Comment on egusphere-2024-1948', Mashiat Hossain, 26 Sep 2024
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 296 | 89 | 81 | 466 | 9 | 11 |
- HTML: 296
- PDF: 89
- XML: 81
- Total: 466
- BibTeX: 9
- EndNote: 11
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1