Review of Sellitto et al. – The optical properties of stratospheric aerosol layer perturbation of the Hunga volcano eruption of January 15^th, 2022

Sellitto et al. built on published results of the Hunga aerosol size distribution to derive the optical properties characteristic of the volcanic plume. They present original results and compare them to historical volcanic events, El Chichon and Mount Pinatubo eruptions and their subsequent plume properties. The authors use a suite of instrument and publicly available datasets to constrain a Mie code allowing a closure on the optical, microphysical and radiative properties of these sulfate aerosols. The manuscript is generally well written and organized. Given its magnitude and particularities, the Hunga eruption has resulted with a strong interest from the Earth Science community, in particular atmospheric scientists. Numerous studies have already been published aiming at characterizing the Hunga aftermath, and this study is adding to this corpus with a new angle. On the downside though, I found a lack of references to acknowledge the work already achieved by peers. Although it presents some new results, a better referencing of previous studies may underline even more why this work is impactful. After consideration of minor revisions, this work should be suitable for publication in ACP.

Main comments:

Introduction in general: it is well written but could use more references to highlight what is already known and publish and where this study sits in that context. Li et al. (2024) (https://doi.org/10.1029/2024GL108522), for example, presents some similar computation on the extinction efficiency of the Hunga aerosol compared to Pinatubo.

In section 2.3, it is stated that Mie calculations are performed in two ways with two assumptions Me fixed or N0 fixed. These two quantities are linked together through Eq.2. The authors should clarify what is the benefit of doing the computations as such.

Figures 4 and 5, the linear scale renders the readability difficult as all the curves are close together. Enhancing the figures with some zooms or a logscale may help. Additionally, although the IR and FIR are crucial for radiative budget, a lot of the variability occurs in the SW UV/Vis, I would also suggest a finer resolution in this part of the spectrum. In particular, we see an increase in the extinction efficiency from the UV to Vis, is this expected? Even for the background state?

It looks like the authors already have the necessary parameters to plot the El Chichon and Pinatubo cases alongside the background and the different Hunga cases of Figure 4 and 5. It would be relevant to do so to underline precisely how different they were.

Figure 7, although the uncertainties on the observations overlap with the result of the model presented, the two values are ~20% different. Instead of comparing the AE which is stated to be highly dependent on the pair of wavelengths selected to compute it, would it be possible to plot OMPS or SAGE-III channels on top of the extinction plot you give in fig. 5 to see if the slope is generally reproduce by the model with the assumptions used.

Minor comments:

Abstract: The abstract reads well, some explicit values (e.g. Angstrom Exp., mean radius) would increase the precision of the results in the abstract. In particular, AE computed between which wavelengths?

Line 22: In two words, what was the main characteristic of Pinatubo and El Chichon?  Larger aerosol sizes?

Line 34-35: Please explicit the atypical conversion time to sulfates. Please, also add other references, maybe reflecting the range of results (e.g. Carn et al. 2022, Asher et al. 2023, …).

Line 43: Limb observations or solar occultation? Duchamp et al. 2023 mainly used SAGE-III from what I read.

Line 48: “A local maximum in number”. Also see Norgren et al., 2024 (https://onlinelibrary.wiley.com/doi/abs/10.1029/2024JD040992).

Table 2: Why not giving the AE for all the different wavelength pairs? One AE for Vis to MIR is trying to resume the complexity shown in figure 4 and 5 to a simple power law. It is shown to be wrong in this very study.

This paper compares the optical impacts of stratospheric sulfate aerosol (SA) for background conditions and following Hunga-Tonga, Pinatubo, and El Chichón. The results are useful but the authors due a spotty job in the presentation and in characterizing the aerosol size distributions from Pinatubo and El Chichón. They also needlessly confuse things by introducing quantities that are not useful. These issues need to be addressed.

Following the eruptions of El Chichón there were very few measurements available to draw from, but that is not the case following Pinatubo when there were many measurements available both remote and in situ. So perhaps following El Chichón using just one reference to characterize the aerosol is perhaps justified. Still the effective radius determined of 1.7 µm is rather large. Perhaps it should be mentioned that this is 1.5 months after the eruption while 7 months later it was down to 0.9 µm (Hoffman and Rosen, 1983).

But using just one measurement for Pinatubo is hardly justified, particularly when that reference Asano (1993) is misquoted. The distribution width from Asano is 1.05 not 1.5, which lowers the effective radius from 0.9 to 0.6  µm for Pinatubo, which is likely more characteristic of other observations. Asano’s estimate came within a month or so of the eruption. In any case the authors could do a much better job in finding an effective radius for Pinatubo. At least one that is more in line with most observations, or at least check that the observation they are using is consistent with other observations.

The other major point that the authors ignore in the rush to characterize the effectiveness of the Hunga eruption is that the effective radius, and hence the concomitant optical effects, continuously evolve as the aerosol decays. For the comparisons here the time frame, e.g. the time since eruption, should be stated clearly.

Here follow specific comments by line number split into major and minor. Quotes from the text are set off with ellipses.

Major:

31 … as high as 58 km … 58 km is not “up to the lower mesosphere”. What is this latter claim based on? It is not supported by Carr et al. 2020.

37-39 This sentence is superfluous and should be deleted. It is exactly what Zhu et al., 2022 showed.

39-41 This is not quite true. There were some CALIOPI observations of ash over one or two days as Sellitto et al. (2022a) state.

47 … direct measurements with sondes balloons … What kind of instrument is this?

73-74 It should be mentioned that these are chosen as that was what was used by Hummel et al. In fact this point is mentioned later so this sentence can be deleted.

Fig. 2 The SDs shown are differential size distributions, i.e. dn/dr, so the units should be cm-3 µm-1 on the ordinate label. What number densities are used for the Pinatubo and El Chichón distributions? Are these still No=1 /cm3?

107-110 What is meant by “effective SA mass” and “effective number density”? How are they defined and where are the equations from? Why are they used?

What is the difference between effective SA mass and SA mass? Ultimately Me (Eq’n 2) is equivalent to the SA mass if one does the algebra. That is Me=4/3 pi ρ Re^3 Ne =4/3 pi ρ Re^3 No exp(-3 ln^2(σ)). Then Re=Rm exp(5/2 ln^2(σ)). Cubing Re and including it above leads to Me = 4/3 pi ρ Rm^3 exp(9/2 ln^2(σ)) which is exactly the SA mass for a unimodal lognormal distribution. So why all of these definitions rather than just say that: The SA mass can be recast as Eq’n 2 using Ne = No exp(-3 ln^2(σ))? Then ultimately since Me is just the SA mass why invent a new symbol rather than just call it the SA mass and define it as M? This needless additional definition of a different mass is unnecessary.

Table 1 Why are the SA masses for El Chichón and Pinatubo left out of the table? Is there a reason why No is given for an SA mass of 1 µg/m3? If so it should be mentioned here.

115 Why are El Chichón and Pinatubo included in parentheses. Aren’t they going to be playing a central role in this paper as points of comparison? Suggest … The optical properties of background, Hunga, El Chichón, and Pinatubo perturbed stratospheric aerosol layers are …

Fig. 4 What are the parameters on the background distribution?

144 How is the number density fixed? In Table 1 it is fixed by setting the mass to a fixed value. Is that what is done here?

146 Isn’t this just the mass concentration?

153 roto-vibrational? Is this rotational-vibrational or something else?

Figure 6 Panels b) and c) are mixed up.

Fig. 7 The information content in this figure is so low it doesn’t justify a figure. The text is sufficient due to the small variation in the Angstrom exponent showing little difference between HT-HH and BG and both well within the uncertainty of the observations.

234-235 Confusing sentence. Suggest …estimate the radiative impact, through the perturbation of optical properties due to the impact of the Hunga eruption on the SD of stratospheric aerosol.

238 What is TIR? FIR?

250—253 Split this four line long sentence into several sentences for readability.

Fig. 8 Why make this figure difficult? Don’t make the reader flip between positive and negative values, both shown on the same ordinate. Use an ordinate that reflects the quantities as they are; SW < 0, LW > 0, … The LW is so small the axis would only have to run from -0.2 – +0.02. There is also a problem with the units given for the ordinate label.

Considering that effective radii for various eruptions are a bit uncertain, and that they are always changing, as eruption sizes vary and the volcanic aerosol decays, why limit the calculations to just 4 effective radii? Using all the assumptions of Eq’n 4 the authors could produce a much more useful figure. Obviously there is structure in delta F, and it is not well reflected with just 4 calculations.

257-260 This is not the first time this difference has been noted, e.g. (Li et al., 2024), as such it would be useful to compare these results with the previous work of others.

265-266 Probably much smaller.

Table 2. There is plenty of room to use 7 columns in this table making it much easier to read and use. Column 1 is the same but can be reduced by 4 rows if it is followed by 6 more columns: BKG (Ae, SSA, g) and HT (Ae,SSA, g).

294-295 Why is that? Is it because it is SA with not a large increase in effective radius, or?

298-300 Doesn’t this need to be qualified that the cooling rate is normalized by SA mass emitted? Then to get the total actual cooling one would have to multiply by the emitted mass.

Minor:

76 Hummel’s et al. or Hummel et al. The latter works fine.

88 Not S but σ?

87 … In “this” work or “that” work? If the authors are implying Duchamp et al., then it should be “that”. If they mean the paper here “this” is correct, but this latter doesn’t seem to be the meaning.

127 … has been calculated …

149 … in most of the …

192 … were present …

291 …eruption has … or eruptions have …

Li, C., Peng, Y., Asher, E., Baron, A. A., Todt, M., Thornberry, T. D., et al. (2024). Microphysical Simulation of the 2022 Hunga Volcano Eruption Using a Sectional Aerosol Model. Geophysical Research Letters, 51(11), e2024GL108522. https://doi.org/10.1029/2024GL108522