the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Quantified effect of seawater biogeochemistry on the temperature dependence of sea spray aerosol fluxes
Karine Sellegri
Theresa Barthelmeß
Jonathan Trueblood
Antonia Cristi
Evelyn Freney
Clémence Rose
Neill Barr
Mike Harvey
Karl Safi
Stacy Deppeler
Karen Thompson
Wayne Dillon
Anja Engel
Cliff Law
Abstract. Future change in sea surface temperature may influence climate via various air-sea feedbacks and pathways. In this study, we investigate the influence of surface seawater biogeochemical composition on the temperature dependence of sea spray number emission fluxes. Dependence of sea spray fluxes was investigated in different water masses (i.e. subantarctic, subtropical and frontal bloom) with contrasting biogeochemical properties across a temperature range from ambient (13–18 °C) to 2 °C, using seawater circulating in a plunging jet sea spray generator. We observed sea spray total concentration to increase significantly at temperatures below 8 °C, with an average 4-fold increase at 2 °C relative to initial concentration at ambient temperatures. This temperature dependence was more pronounced for smaller size sea spray particles (i.e. nucleation and Aitken modes). Moreover, temperature dependence varied with water mass type and so biogeochemical properties. While the sea spray flux at moderate temperatures (8–11 °C) was highest in frontal bloom waters, the effect of low temperature on the sea spray flux was highest with subtropical seawaters. The temperature dependence of sea spray flux was also inversely proportional to the seawater cell abundance of the cyanobacterium Synechococcus, which facilitated parameterization of temperature dependence of sea spray emission fluxes as a function of Synechococcus for future implementation in modelling exercises.
Karine Sellegri et al.
Status: final response (author comments only)
-
RC1: 'Comment on acp-2022-790', Anonymous Referee #1, 16 Mar 2023
This paper presents measurements of sea spray aerosol particle number size distributions and calculated fluxes for particles produced from four different water masses during one sampling period. Surface tension measurements were done on water samples at different temperatures, and relationships between sea spray flux and water temperature are presented for the different water masses. This work is put in the context of previous work and relates the nanophytoplankton numbers to cloud condensation nuclei as well. Overall, this paper is interesting and relevant. Understanding sea spray production flux as a function of water temperature will be a good contribution to global modeling. There are some areas of the paper that can be improved, mostly through clarification of the methods and presentation of the results. Some general and specific comments are noted below.
General
The abstract states that future changes in seawater temperature will be important with climate change. In that case, why were such low temperatures chosen for the experiments? 2°C is outside of the range of the ambient temperatures measured (13-15°C). It would be helpful to have a comment on the global distribution of current sea surface temperatures and how they may change. For example, what fraction of the world’s ocean has a sea surface temperature at 2°C? How will that change in the future? Is it physically relevant to cool these specific water masses to 2°C? The statement made in the Conclusions is not fully supported. This study cooled ambient water, and many other factors will change with a warming ocean. It would have been more relevant to warm up the ambient water or present a larger range of temperatures.
The Methods should contain more details on the experiments and calculations, as mentioned below in the Specific Comments. There are some pieces in the Results and Discussion that should be moved into the Methods (i.e., the calculation of FCCN). Additionally, while references are included, there are some places that could use more explanation in the calculations. It also seems unnecessary to use “CCN” when it is simply defined as all particles over 100 nm. That could just be stated as its own variable.
Could you add some comments on how the temperature of the air would affect the particle number flux? The air temperature may influence the lifetime of the bubble at the surface. Would a large gradient in the temperature from the surface of the seawater to the air change the lifetime of the bubbles at the sea surface? What air temperature were these experiments done at? Was the air temperature in the headspace held constant, or was the whole system cooled?
More text needs to be included discussing role of different factors on the surface tensions measured. The temperature at the time of measurement affects the surface tension, as stated. Additionally, salt concentrations in the water can change the surface tension as well. Because these are different water masses, it is likely that their salinities also change. Reporting a total surface tension might not be relevant, unless it is in the context of these changes in temperature and salinity.
The discussion of the different biological factors influencing the particle flux at lower temperatures needs to be clarified. These are not necessarily species that would live at these temperatures. And it seems like any surfactants that would be in the water would have already been emitted at the ambient temperature. Is it possible that these species could die at these temperatures and then emit more surfactants or organics? Some more discussion of this would be useful. Towards the end of the paper, it is stated that the bubble films are more stable at colder temperatures due to the stabilization from surfactants. Why would there be more stabilization of surfactants at colder temperatures? This implies there are more or different surfactants when the temperature is changed.
The Results and Discussion section is somewhat short and could use more discussion of what these results mean. Some sections as mentioned below, could use more support for the statements. The Conclusions section starts as more of a discussion that could be moved up to the Results and Discussion. It would be better to have the key findings summarized in the Conclusions, and it should be clear what was measured in this study. The comparisons to the previous studies could be discussed in the previous section.
Specific Comments
Line 105: Nothing to change. Just noting that March 2020 must have been an extra stressful time to be on a research cruise, and it is great that you were still able to finish the cruise and complete the experiments.
Line 121: FCCN is used in this sentence but has not yet been defined. Because it is defined later, it might be better to remove it from this sentence, or move the definition up. It seems like “F” is for “flux”, but that should be stated explicitly here. And sub-scripting the “CCN” or not, should be consistent.
Line 122: Please expand this sentence. It is unclear what two fluxes are being compared. It is interesting and relevant that the wind speeds are related to the air entrainment in the plunging jet system. It would be helpful to have the wind speeds and corresponding flow rates that were used written out here.
Line 127: It would be helpful to have more details on the surface tension measurements. Were these all done on board the ship, with fresh seawater, directly after sampling? How did the ship movement contribute to any uncertainties in these measurements? What volume of sample was collected for the temperature gradient experiments? Were these mixed to ensure a constant temperature throughout the sample?
Line 132: How were the samples in the 10L carboys stored to prevent changes in the chemistry and biology? Please add more details on the aliquots and their storage prior to analyses. Were these analyzed on board the ship or later in the laboratory?
Line 166: The paragraph prior to this line could be moved into the Methods section. This is mainly describing the water masses that were sampled and their dates. Starting at Line 166, there are some results from the measurements.
Line 176: Add more explanation on how all of the particles greater than 100 nm can be considered CCN at 0.2% supersaturation.
Figure 1 should be improved. The y-axes labels all appear to be stretched. The panels should be merged together into one figure, with one common x-axis. The shading and labels are nice to have. It would be helpful if this figure also had sea surface temperature and salinity. Maybe b and c could be combined, and SST could be added to a on the right axis. In d, why are there more surface tension markers than flux markers? Because the flux was calculated from the size distributions which were measured continuously, it seems like those could be at higher resolution or the same resolution as the surface tension. Error bars should also be added to these markers, especially the flux and the surface tension, to see any overlap in variability. (Also, change “STT” to “SST” in caption.)
In Figure 2 it is not clear what are measurements from this study and those from previous studies. The text states that there are four other datasets, but I do not see markers for the Sea2Cloud data. “SO” needs to be defined as well. If that is the Sea2Cloud data, then I am not sure what the fifth dataset is that is referenced in the text. Overall, the caption could be a little more descriptive.
Line 190: I think this equation and all of the description of the FCCN calculations should be moved up into the methods. It would be helpful to have just an FCCN equation as well.
Line 202: This should be Figure 3.
Figures 3, 4, and 5 could be combined into a multi-panel figure, since they all have the same x-axis. Figures 3 and 5 should be combined, since they both contain sea spray flux as a function of temperature. It would be easier to compare the figures if they were together.
Line 266: Interesting result that there is a shift in the shape of the size distribution.
Figure 6 needs a legend to describe the different colors. Additionally, some of the marker colors do not match the line colors (i.e., black dashed line with orange x’s). It would be useful to have the same color scheme as Figure 7, to be consistent.
Line 283: Can you add a little more description on the calculations going into Figure 7. It seems like it would be interesting to compare the Dp values shown in Table 1 for both temperature ranges.
Line 294: This explanation is not consistent. Why would there be higher concentrations of surfactant in colder waters that would further stabilize the bubble film? Surfactant concentration is not the only thing contributing to the surface tension and thus bubble lifetime.
Line 348: More explanation is needed to make this claim. It is not entirely clear that the results of this study support the idea that with a warmer ocean, there will be less sea spray flux. The temperatures measured here were colder than ambient. In order to make this statement, it would have been better to warm the water instead. There are a lot of factors that will change in a warming climate, so this needs to be clarified.
Citation: https://doi.org/10.5194/acp-2022-790-RC1 -
AC1: 'Reply on RC1', Karine Sellegri, 05 Jun 2023
The comment was uploaded in the form of a supplement: https://acp.copernicus.org/preprints/acp-2022-790/acp-2022-790-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Karine Sellegri, 05 Jun 2023
-
RC2: 'Comment on acp-2022-790', Anonymous Referee #2, 21 Mar 2023
In this manuscript, the authors conducted a study during a research cruise in the Southern Ocean using a sea spray simulation chamber to generate nascent sea spray aerosol. The authors varied the seawater temperature in the chamber while steaming across different water masses to investigate the impact of seawater temperature and the biogeochemical state of the ocean on the sea spray aerosol flux.
As the authors allude to in their introduction, this research question has received significant attention in the scientific community. Although it is increasingly evident that seawater temperature impacts sea spray aerosol flux, there are still differences in both the magnitude and direction of the relationship between seawater temperature and sea spray aerosol flux, which may be due to variations in experimental approaches and differences in water chemistry and biology across different studies.
Although the study did not resolve the longstanding issue of the differences in the relationship between seawater temperature and sea spray aerosol flux found in the literature, my major criticism of the work is not the lack of scientific significance. Rather, the quality of the manuscript is compromised due to a general lack of attention to detail. The methods section omits critical details, rendering it impossible to evaluate the authors' findings. Furthermore, the work is poorly presented, making it difficult to understand the authors' intended message. In addition, the study overlooks numerous uncertainties, making it impossible to verify some of the authors' claims. In summary, the study feels rushed and fails to do justice to the authors' effort in collecting the dataset.
Therefore, I strongly recommend that the authors revise the manuscript substantially and focus on presenting the methods and results more clearly. As a result, I am afraid that I can only recommend rejecting the article in its current form. Below I outline in more detail the major issues I have identified with the manuscript.
Major issues
- The authors must provide a clear description of the sea spray simulation system they used in the methods section, as it is increasingly evident from the literature that the scale of laboratory systems used to generate nascent sea spray impacts the relationship between the number and size of aerosols generated as seawater temperature changes. In the current version of the manuscript, the authors have only referred to another paper (Sellegri et al. 2022) which is not included in the reference list. Given this, I assume that the system used is the same as that described in Schwier et al. (2015), which is relatively small compared to other systems used for simulating sea spray aerosol generation in the laboratory. The water depth in the system used by Schwier et al. (2015) was only 10 cm deep, which may have resulted in increased interaction between bubbles and the chamber walls. Moreover, the sea spray chamber used by Schwier et al. (2015) had multiple plunging jets, and there may be differences between this setup and those that use single plunging jets. However, since the details of the chamber used are not clear, it is currently impossible to determine this.
- Regarding the air entrainment rates described by the authors, how they were obtained is unclear. As the authors normalized their fluxes to this parameter, a clear description of the methodology used to determine the air entrainment rates is necessary in the manuscript.
- The authors need to provide a more detailed description of how they controlled the temperature of the sea spray chamber. While they mention the use of a 50 L temperature-controlled reservoir, it is unclear how this was connected to the sea spray chamber. Specifically, it is not clear whether the sea spray chamber was immersed in the reservoir or connected to it via some other means. A schematic of the experimental setup would be useful in clarifying this point. Further, how were the temperature experiments conducted? The authors mention that they applied temperature gradients ranging from 2°C to 15°C to the seawater over approximately 1 hour, but the exact form of these experiments is unclear. Given that these experiments are crucial to the study, it would be helpful for readers to see a typical experiment as a figure in either the main manuscript or the supplementary materials. This would provide more context and enable better understanding of the results. In addition, these are fast temperature ramps which leads to the question of how repeatable the measurements were. Were any experiments conducted over a longer time period at constant temperature to determine the impact of quickly ramping the temperatures on the fluxes versus holding the system at a steady temperature?
- There are some important details missing about the aerosol measurements. It is unclear how the instruments were connected to the sea spray chamber. Were all instruments connected through a single connection, and was the sampling conducted isokinetically? The type of differential mobility particle sizer used is also not specified. Was it purchased or custom-built in-house? Additionally, it is unclear whether an impactor was used to prevent particles larger than 500 nm or some other cutoff from entering the instrument. It is also unclear whether the aerosols were dried before sampling or whether they were measured under ambient conditions in the sea spray chamber, and if so, what was the relative humidity of the sample. Again, some of these details could be better explained with the inclusion of a schematic of the setup.
- In their study, the authors use the flux of particles larger than 100 nm as a proxy for cloud condensation nuclei (CCN). However, they should provide a detailed explanation of how they obtained this flux. Although they mention using the flush air flow and water surface of the tank, they fail to clearly explain the process. To help the reader understand, the authors should provide a mathematical explanation of the process and how they normalized it to wind speed. Moreover, the authors mention using air entrainment, but the details of how this was done are unclear, and it should be explained to the reader. To make it easier to understand, the authors should describe the process used mathematically.
- The authors conducted an experiment to measure the surface tension of seawater at different temperatures. They froze the water samples and then allowed them to warm up to 15°C over approximately one hour. It would be helpful to include a typical experiment's data plotted in the supplement or manuscript (this maybe figure 4 but it is not completely clear). The authors suggest that the short time period of the experiment reduced the impact of changing biogeochemistry on their measurements. However, they do not discuss the potential impact of freezing on the surface tension. For instance, freezing could rupture phytoplankton cells present in the sample, releasing organic matter into the water, which could impact the surface tension. Did the authors conduct an experiment where they measured the surface tension of a fresh sample at ambient temperature, froze it, and then returned it to the same temperature? If so, the measurements should be similar if freezing had limited impact, and this potential issue should be discussed. Furthermore, to rule out contamination from the sample tubes, were any measurements of pure water taken in the Falcon tubes? Including this information would enhance the clarity of the authors' findings.
- The authors have presented time series of different parameters in Figure 1, but there are some significant issues with the data presentation. Firstly, the linear axis appears to have the same distance between the values of 0.00E+00 and 1.00E+06 and the values of 1.00E+06 and 2.00E+06. Additionally, all the subscripts are missing from the axis labels and legend entries, which may seem minor, but it suggests carelessness on the part of the authors. Moreover, the authors argue for a trend in this data, which is impossible to determine considering there are no error bars on the data points. Without an idea of the uncertainty on each data point, it is difficult to find a trend in the data when it is so noisy. Furthermore, it is not clear what the authors have plotted in Figure 1. Have they plotted the integrated number flux of all particles following normalization, or is it the integrated number flux of all particles larger than 100 nm? The authors should clarify this point to help readers better understand their results.
- The authors initially mentioned fitting "single lognormal modes" to their measurements of aerosol size distribution. However, in the following sentence, they reported finding "four modes in the submicron range and two modes in the supermicron range". This description is unclear and adds to the manuscript's overall confusion. Upon examining Figure 6, it becomes evident that the authors actually fitted a series of lognormal modes, not "single lognormal modes". However, they did not provide any information about their fitting procedure or the quality of the fits. It is generally recommended to present data and its associated uncertainty (e.g., mean/median and standard deviation/error) before comparing contrasting datasets. If the authors wish to include fits, they could be presented in another panel or the supplement. Comparing the actual data across different temperatures is crucial to determining whether differences exist between the presented temperature regimes. Without the data, it is impossible to determine whether such differences exist.
Citation: https://doi.org/10.5194/acp-2022-790-RC2 -
AC2: 'Reply on RC2', Karine Sellegri, 05 Jun 2023
The comment was uploaded in the form of a supplement: https://acp.copernicus.org/preprints/acp-2022-790/acp-2022-790-AC2-supplement.pdf
Karine Sellegri et al.
Karine Sellegri et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 356 | 129 | 20 | 505 | 8 | 10 |
- HTML: 356
- PDF: 129
- XML: 20
- Total: 505
- BibTeX: 8
- EndNote: 10
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1