Atmospheric bromine deposition as a source of seasonal snowpack bromide reservoirs for Arctic bromine explosion events: evidence from linking chemistry-climate simulations with snow observations using a layered snow model
Abstract. Bromine explosion events (BEs) are an important source of reactive bromine in the Arctic boundary layer and contribute to episodic ozone depletion in spring. While bromide-rich sea ice and snow are generally regarded as the primary source, the extent to which atmospheric deposition of bromine can establish and replenish a snowpack bromide reservoir remains poorly constrained. We combine simulations with the ECHAM/MESSy Atmospheric Chemistry model (EMAC), field observations, and laboratory analyses of snow samples collected around Ny-Ålesund (Svalbard) in April 2024 to investigate the role of atmospheric bromine deposition for BEs. To connect atmospheric deposition fluxes with observed snow chemistry, we developed the Snow Model with Evolving Layers (SMELs), a simple offline snow model that tracks bromide accumulation within a dynamically evolving snowpack.
EMAC simulations indicate that bromine dry deposition from very short-lived substances (VSLS) and sea-salt aerosol (SSA) dehalogenation can establish snow bromide reservoirs sufficient to sustain typical Arctic springtime BrO abundances. Bromine species partitioning upon deposition suggests that a lack of HBr deposition may suppress heterogeneous Br2 production during fall. SMELs reproduces the observed vertical bromide enhancement and depletion patterns and demonstrates that finite snowpit sampling can introduce substantial biases. The strong agreement between modeled and observed normalized bromide profiles suggests that bromine deposition and post-depositional processing operate on regional scales around Ny-Ålesund. Our results identify atmospheric bromine deposition as an important seasonal bromide reservoir and highlight the role of bromine speciation and snowpack processes in controlling Arctic bromine activation.
Falk et al use a global model, a snow layer model, and observations of snowpack bromide in Ny-Alesund to examine the role of atmospheric bromine deposition for bromine explosion events. They conclude that atmospheric bromine deposition is an important source of bromine for bromine explosion events in the Arctic.
I found this paper very difficult and confusing to read for many reasons as described below. It’s hard to evaluate because I’m not sure what they are simulating in their models, but it seems that they are missing important processes and making assumptions that are not supported by the literature.
Introduction:
The introduction is overly long, and gets some of the chemistry wrong. For example, R1 should read:
HOBr + Br- + H+ à Br2 + H2O
This is a well-known acid-catalyzed reaction.
In addition to R2, the following reaction is also important:
BrNO3 + X- à BrX + NO3-
Where X = Br-, Cl-, or I-.
HOBr as a bromide reservoir. Is there any evidence that HOBr serves as a bromine reservoir in snow? This should be supported with a citation. It is thought that the reservoir is mainly bromide. Thus, Figure 1 seems wrong. It is bromide that is the important reactant in the aqueous phase, with HOBr reacting with aqueous-phase bromide according to the modified R1 shown above. The sentence below figure 1 says that “release of Br2 or Br is limited by the availability of Br-“. Much of the arguments in the paper later on rely on this assumption of HOBr serving as a reservoir, and I don’t know of any evidence of this.
Page 4: “smaller aerosol particles would gain bromide from the gas-phase during coagulation”. Coagulation refers to the aerosol phase only. What you are referring to here is gas-phase uptake onto aerosol.
Methods:
The models are not described well. For example, it doesn’t even mention that their model is a global model. It also doesn’t describe the halogen chemistry in the model, which is particularly relevant for this paper. What is T42L90MA? I’m really confused after reading the entire paper whether or not bromine release from snow is actually parameterized in the model, and if it’s not, how this is being estimated since this is mostly what the paper is about. Did you actually implement the Toyota mechanism? If so, briefly describe it and the assumptions made in your implementation of it.
Brorg is never explicitly defined. Nor is Bry. What species are being considered in these families? I’m surprised that deposition is a significant sink for species like CH3Br, which I think is what is being said here but difficult to tell since Brorg is never defined.
I’m also confused about the dynamic snow layer model. Figure 3 is very confusing. I have no idea what it is telling me. The variables are not defined. There is a variable in the figure caption (Li) that is not in the figure itself. What is the “active layer”? Is bromine lost from the snowpack via heterogeneous reactions or is deposition considered to be a permanent sink from the atmosphere?
It is also very unclear if and how the two models (global and snow) are coupled.
Figure 4 is also confusing. Height above what? Are you plotting snow depth here? Where is the top and where is the bottom? If this is snow depth than plot “depth” and depth = 0 should be the top layer in direct contact with the atmosphere. Also, how much time does this snow depth represent? I don’t know the snow accumulation rate. Is it one year of snow? More? Less? There are other figures with height in the snow which is confusing.
Results:
The fact that aerosol bromine is not deposited along with gas-phase bromine is a major model deficiency. The paper says later that it is 50% of total deposition. Where does this 50% come from? Since fine mode aerosol bromide tends to be enriched relative to sea salt in fine-mode aerosol while the coarse-mode is depleted, simulating fine and coarse mode aerosol deposition is needed to not only examine the total bromine deposition to snow but also the enrichment/depletion.
Figures 6 and 7 are also confusing. How can the Br- reservoir be negative?
HOBr can also react with chloride, so just because you run out of Br- from HBr, doesn’t mean that HOBr stays in the snow:
HOBr + Cl- + H+ à BrCl + H2O
Is the above reaction not considered in the model?
You mention the importance of speciation (BrNO3 and HBr) for the seasonality of bromine explosion events. What determines this seasonal change in the relative importance of these species in the model?
Figure 10. The Br- values show a depletion relative to sea salt. In Greenland there is an enrichment. What accounts for this difference? Since you have a global model perhaps you can get insight into this.
Figure 12. The text says that Figure 12 shows bromide from VLSL and SSA bromine deposition. But before you said you don’t have SSA bromine. Do you or do you not have the aerosol deposition source of bromine?
Section 3.4. I read this twice and have no idea what it is saying. The writing throughout the paper is unclear, but especially in this section.
Figure 13. How is depth different from height in the earlier figures? At this point I’m very confused about what the snow figures are showing.
You talk about different layers of snow (e.g., the second and third ice layers). It would be helpful to know how long ago this snow was deposited to the surface.
Figure 14. How can the x-axis possibly be negative according to Eq. 3? This would require the bromine concentration from BE, VSLS, or SSA to be negative, but how can a concentration be negative?
Conclusion and Discussion:
There are at least two other papers that have used a global model to simulate bromine release from snowpack and/or bromine explosion events (Swanson et al. 2022; Zhai et al. 2023), and the authors don’t try to compare their results with this previous work. Such a comparison could provide additional insight.
Zhai et al. (2023) found that atmospheric deposition is not sufficient enough to explain the observed springtime BrO at Summit, Greenland. You are saying the opposite, even though you don’t have aerosol bromine deposition. What accounts for this difference?
Page 28, end of second paragraph. The Zhai et al. (2023) and Swanson et al. (2022) papers do have these processes (explicit representation of SSA ageing, bromine release, transport, and incorporation of the snowpack). Which one of these does your model not have? It seems that explicit representation of SSA ageing is not represented, since you assume that 50% of SSA bromide is released into the gas phase. Do you not have bromine incorporation or release in the snowpack? Doesn’t your model include transport?
References:
Swanson, W. F., and Coauthors, 2022: Comparison of model and ground observations finds snowpack and blowing snow aerosols both contribute to Arctic tropospheric reactive bromine. Atmos. Chem. Phys., 22, 14467-14488.
Zhai, S., and Coauthors, 2023: Implications of Snowpack Reactive Bromine Production for Arctic Ice Core Bromine Preservation. J. Geophys. Res., 128, e2023JD039257.