Review of "Radiative impacts of the Australian bushfires 2019-2020 - Part 2:

Large-scale and in-vortex radiative heating" by Sellitto et al.

General comments:

On the positive side, the overall organization of the manuscript is good, and the sentence level flow is also very good, making the paper a pleasant read. The intended purpose of the study is mostly clear and well defined.  The topic is an important one with a lot of recent literature, and the heating rate calculations will be of interest to readers of the journal.

On the negative side, however, the paper seems rushed with many holes where important details are left out, and the methodology is therefore often not clear at all.  For instance, there's some emphasis on clouds in the introduction but they are not mentioned again and there's no information about what underlying albedo was used to calculate the heating rates.  This may be a major flaw, if the clouds were really not taken into account despite the need for it being already emphasized in the authors' Part 1 paper.  Next, a "background" condition is mentioned in the methods but is not shown in figures or equations, so I can't tell whether it was really used or not. The section on use of the lidar data (which is new to this paper compared to Part 1) is quite superficial and leaves a lot of questions unanswered. Secondary conclusions about the heating rates being "consistent" with quantified plume rise rates are not supported at all.  Finally, the effort to compare with other similar research concerning radiative effects for the same smoke event, both in the introduction (to give perspective on this paper's unique contribution) or in the discussion (to compare/contrast results) is quite limited.  It's a very popular topic and various easy Google searches turn up many apparently relevant papers that are not referenced here.  Also, for the the lidar methodology section in particular, there are no lidar references at all to explain or support the authors' methods.

Specific comments:

(Abstract)

L20 Add "for February 2020", because this result is specific to that month, correct?

L21 The logical link between optical properties and aging is not established in this paper.

L24 The quantitative consistency between the heating rates and the lofting rates is not established in this paper.

(Introduction)

L46 I'm confused by the discussion of clouds.  In Part 1, some calculations were made indicating that there could be a very significant impact of surface albedo and it was indicated that clouds should certainly be taken into account in further work. However, this introductory summary of Part 1 is the only time that clouds are mentioned in the current work.  There's no information about what surface albedo was used in the current calculations.  Were clouds ignored again?

L64 "to cover both the radiative effects mentioned previously" isn't clear.  Is this referring to the two factors at line 46?  But it doesn't actually seem to cover the 2nd factor (the presence of clouds) at all.

L62 At least one other paper has heating rate calculations for the same event. This should probably also be included in the discussion of the context for the current paper.  Wu, D., X. Niu, Z. Chen, Y. Chen, Y. Xing, X. Cao, J. Liu, X. Wang, and W. Pu (2022), Causes and Effects of the Long-Range Dispersion of Carbonaceous Aerosols From the 2019–2020 Australian Wildfires, Geophys Res Lett, 49(18), e2022GL099840. doi: https://doi.org/10.1029/2022GL099840.  

There are many other papers addressing the same event. It would be helpful to readers to include a broader discussion of these and how the current manuscript fits in to this extensive literature, and to provide more insight on whether this manuscript is or isn't consistent with hypotheses and findings of other researchers.  For instance, at least one paper also uses CALIPSO data to calculate radiative effects: Papanikolaou, C.-A., P. Kokkalis, O. Soupiona, S. Solomos, A. Papayannis, M. Mylonaki, D. Anagnou, R. Foskinis, and M. Gidarakou (2022), Australian Bushfires (2019–2020): Aerosol Optical Properties and Radiative Forcing, Atmosphere, 13(6), 867. 

L64-65 Is the inclusion of longwave calculations new for this paper compared to Part 1? It seems like an important addition.  It would be good to mention it here.

L70-72 The description of the sections is cut-and-pasted from the Part 1 paper, and does not properly reflect this paper.

(Data and Methods, radiative transfer)

L83 In Figure 1 or in the discussion of it, it would be helpful to specify what ALL of the required inputs for the radiative transfer model are.  That is, spell out which "non-measured aerosol optical properties" are needed. Also, is the extinction an altitude-dependent profile of extinction, or a layer-integrated AOD? Also, are the layer height and depth needed? What about surface albedo? Solar geometry?

L120 This discussion of the "clean background" is confusing in multiple ways.  First, there would be some biomass burning in the the same months in any year, so this should not be called "clean background".  In fact, the authors discuss this in part 1, but that discussion isn't repeated in part 2, so the confusion occurs all over again.  Better to use a different phrase than "clean background".  In fact, the authors aren't attempting to calculate a forcing with respect to an aerosol free atmosphere, only an "anamolous forcing" due to this particular event, correct? 

L120 (continued) But it's also confusing that there's no further mention of the "background" conditions in section 4.  Why doesn't it appear in the figures or equations?  Is the background even relevant to this paper?

L129 The CALIPSO products lidar ratio, color ratio and depolarization ratio don't really give quantitative constraints on SSA and g.  I suggest rewording this to reflect the qualitative nature of the relationship, perhaps something like "Assumptions on SSA and g are aided by inferences of aerosol composition supported by CALIPSO products lidar ratio, colour ratio, and depolarisation ratio".

L139 somewhere in this section I would expect to see a discussion of what underlying albedo is used in the calculations

(Data and Methods, limb measurements)

L152 "spatial coverage".  Do the authors mean "spatial resolution"?  That is, I think SAGE III covers the same latitude bands (or else it would not be able to provide a useful Angstrom exponent) but the resolution is too coarse to be useful for the study of details.

(Data and Methods, CALIPSO)

L159-162 This description of the aerosol optical depth and lidar ratio from CALIPSO is very superficial and I'm left with many, many questions about the assumptions and uncertainties, whether it's done correctly, and even why an alternate AOD and lidar ratio calculation is done at all.  Adding lidar measurements is the primary new component of the methodology compared to Part 1, so it's strange that the section is so abbreviated. The authors should fill in this section with enough information about the methodology to allow for reproducibility.

CALIPSO standard products include layer AOD and extinction (see the CALIPSO Data Product Description https://www-calipso.larc.nasa.gov/resources/calipso_users_guide/data_desc/cal_lid_l2_layer_v4-51_desc.php), as well as type identification and lidar ratio.  The algorithms for producing AOD, extinction, and lidar ratio include a constrained transmittance retrieval similar to the one described here, for specific cases when the aerosol layers have the required aerosol-free air above and below them, and otherwise uses lidar ratios modeled for the inferred aerosol type (Young, S. A., and M. A. Vaughan (2009), The Retrieval of Profiles of Particulate Extinction from Cloud-Aerosol Lidar Infrared Pathfinder Satellite Observations (CALIPSO) Data: Algorithm Description, J Atmos Ocean Tech, 26(6), 1105-1119. doi: doi:10.1175/2008JTECHA1221.1.).  The manuscript does not say very clearly whether the CALIPSO standard products are used, but it seems not.  Why isn't the standard CALIPSO AOD product used? The motivation for discarding this and using an alternative algorithm should be made clear. 

Also, if the authors have made their own L2 retrieval for whatever reason, then this section should include the full methodology including equations or a prior published article that uses the same methodology.  Other papers that have used similar calculations should probably also be acknowledged (e.g. Prata, A. T., S. A. Young, S. T. Siems, and M. J. Manton (2017), Lidar ratios of stratospheric volcanic ash and sulfate aerosols retrieved from CALIOP measurements, Atmos. Chem. Phys., 17(13), 8599-8618. doi: 10.5194/acp-17-8599-2017). 

More specifically, are you using Platt's equation? (Platt, C. M. R. (1973), Lidar and Radiometric Observations of Cirrus Clouds, Journal of Atmospheric Sciences, 30(6), 1191-1204. doi: https://doi.org/10.1175/1520-0469(1973)030<1191:LAROOC>2.0.CO;2.)  How do you subtract the molecular backscattering, and what assumptions are involved in that?  Do you account for multiple scattering?

(The plume and the vortex)

L208, Consider writing in which direction the vortex is traveling.

L208, I don't understand "the frontal structure".  Can this be explained a bit more?

L218. The lidar results quoted for smoke are for smoke in the troposphere.  There is ample literature of lidar observations of smoke in the stratosphere to suggest that it is very commonly much more depolarizing.  Consider the following references:

Burton, S. P., et al. (2015), Observations of the spectral dependence of linear particle depolarization ratio of aerosols using NASA Langley airborne High Spectral Resolution Lidar, Atmos. Chem. Phys., 15(23), 13453-13473. doi: 10.5194/acp-15-13453-2015.

Haarig, M., A. Ansmann, H. Baars, C. Jimenez, I. Veselovskii, R. Engelmann, and D. Althausen (2018), Depolarization and lidar ratios at 355, 532, and 1064 nm and microphysical properties of aged tropospheric and stratospheric Canadian wildfire smoke, Atmos. Chem. Phys., 18(16), 11847-11861. doi: 10.5194/acp-18-11847-2018.

Hu, Q., et al. (2019), Long-range-transported Canadian smoke plumes in the lower stratosphere over northern France, Atmos. Chem. Phys., 19(2), 1173-1193. doi: 10.5194/acp-19-1173-2019. 

Sicard, M., Granados-Muñoz, M. J., Alados-Arboledas, L., Barragán, R., Bedoya-Velásquez, A. E., Benavent-Oltra, J. A., Bortoli, D., Comerón, A., Córdoba-Jabonero, C., Costa, M. J., del Águila, A., Fernández, A. J., Guerrero-Rascado, J. L., Jorba, O., Molero, F., Muñoz-Porcar, C., Ortiz-Amezcua, P., Papagiannopoulos, N., Potes, M., Pujadas, M., Rocadenbosch, F., Rodríguez-Gómez, A., Román, R., Salgado, R., Salgueiro, V., Sola, Y., and Yela, M.: Ground/space, passive/active remote sensing observations coupled with particle dispersion modelling to understand the inter-continental transport of wildfire smoke plumes, Remote Sens Environ, 232, 111294, https://doi.org/10.1016/j.rse.2019.111294, 2019. 

L220 Note that color ratio for attenuated aerosol backscatter is not the same quantity as the color ratio of (unattenuated) aerosol backscatter. In particular, if there is a strong spectral dependence of the absorption, 532 nm will be more strongly attenuated than 1064 nm and that will impact the color ratio. (Liu, Z., et al. (2019), Discriminating between clouds and aerosols in the CALIOP version 4.1 data products, Atmos. Meas. Tech., 12(1), 703-734. doi: 10.5194/amt-12-703-2019)

L221-223 This explanation for the large depolarization ratio is different from other published hypotheses of why stratospheric smoke aerosol tends to be depolarizing, see especially Haarig et al. 2018 and Sicard et al. 2019 given above, as well as Papanikolaou et al. 2022 . There should be additional discussion about the various hypotheses to explain why the authors prefer the ash hypothesis.  

L231 (Figure 3 caption) What altitude range is represented by the ECMWF-IFS images?

Table 1.  Define the colour ratio. I think you are using the opposite ratio to the CALIPSO product definition, which is fine, but it's not stated. Is it 532/1064 nm or 1064/532 nm?  Also specify, is it colour ratio of particulate backscatter or attenuated backscatter? 

L253-254.  This comment that it is not known how the HR depends on the optical properties is a bit odd with no followup discussion, considering the authors immediately go into a description of how they calculate the heating rate from the optical properties. What parts are uncertain? Are there aspects of the calculations used here that are uncertain?  What assumptions and uncertainties are associated with those calculations? 

L256. Earlier a "background" condition was mentioned.  How does this fit into the equation?  

L259. What value is used for the underlying albedo? The full CALIPSO curtain (which is shown here only clipped) shows  clouds at various altitudes below the plume.  In the Part 1 paper by the same authors, they demonstrate that bright clouds underlying the aerosol layer can have a very large impact on the radiative calculations, so how are the clouds accounted for?

L267. The quoted number is specific to SSA = 0.8, so I suggest replacing "for smaller SSA... absorbing aerosols" with "for the most absorbing aerosols with SSA = 0.8"

L269. Can the authors give any further insight relating to the simultaneous heating around 35 km?

L273. Clarify in the text what altitude range the quoted values refer to.

L275. Instead of averaging over SSA and g, it would be better to give the minimum and maximum results.  Since the different SSA and g inputs are not a statistical representation the aerosol but rather a grid of possibilities, an average doesn't have a lot of scientific meaning. A range of input was used because the values are unknown and a range better represents the state of knowledge.  The same should go for the output.

L276. "The radiative heating profiles are averaged in the altitude range 12-25 km".  This is from the Figure 5 caption text. Please put it in the main text also.

L278. Again I don't think it's a good idea to average over the different SSA and g values.  (To clarify, I have no concern about averaging over the altitude range or the weighted average used to combine the latitude bands.  It's only the average over the assumed SSA and g that is problematic, because the average removes important information about the range of results obtained and is too easily mistaken for a "best estimate").  

L280. Similarly, when the results are given quantitatively (here and in the conclusions and abstract), please give the min and max instead of the standard deviation.  The standard deviation is not a good representation of the distribution of possible results, since the 4 or 8 results are clearly not normally distributed, and the inputs were not any kind of statistical representation either.  The min and max will unambigously and correctly describe your findings.  Also, in the figures, please make the vertical bars represent the min and max.

L282. Since there's nothing in this paper specifically linking the SSA and g values to ageing, this statement about ageing probably should be supported by a reference to the appropriate literature.

L282. "see the error bars".  The error bar for the LW is so small it can't be seen.  Please include in the text a quantitative description of the spread in the LW, to distinguish it from zero.

L284. "variability... between 0.02 K/d (SSA=0.95)... and 0.12 K/d (SSA=0.8)".  Actually the variability is larger than that, and these values of heating rate are not the values for the specific SSAs given in parentheses.  The full range of variability should be used here, with the actual values for the specified SSAs, not the standard deviation, which has very little meaning in this case.

L285. "The net SW+LW HR is consistent with ... [a plume rise] of a few km in 4 months (see Fig 2a... of Yu et al. 2021).  I'm not following how this consistency is established.  I see at least the weaker conclusion that the authors have successfully found a combination resulting in a positive heating rate which they argue is required to explain any amount of plume rise.  The reference to the Yu et al. figure seems to be only to show that there is observational evidence that the plume did rise by that amount (which is shown in Figure 2 in this paper anyway), but I don't see that it shows a relationship between plume rise rate and heating rate that can be used to establish any kind of claim of quantitative consistency.  If I'm missing something, please explain it in more detail in the text. Otherwise, please soften the statement to remove any implication that a quantitative consistency has been demonstrated. (Here and in the conclusion and abstract).

L287. "consistent ... with Heingold et al. (2022) even if slightly larger".  Please explain this apparent contradictory statement in more detail.  What are the specific values from this work and from Heingold et al. that are being compared, and if there is a discrepancy, what's the likely explanation for it that makes them nevertheless consistent?

L287. Also how do the calculations in this study compare to heating rates given by Wu et al. (2022)?

Figure 4. If the background heating rate is used, it should be shown in Figure 4 also.

(Vortex calculations)

L309. I don't understand the motivation for having different input models in section 4.1 and 4.2.  That is, why does 4.2 use 3 prespecified models of SW and LW SSA and g, wherease 4.1 performed a full sensitivity study using all of the combinations?  Can this be clarified in the text, please?

L311. The definitions of the aerosol models should be included in the text, not just the figure caption, please

L311. What value of Angstrom exponent was used for the calculations in section 4.2?  What value for surface albedo? 

Figure 6.  Why is the LW result for the Large BC case so different from the other two cases, when the SSA and g inputs are very similar, and the earlier text suggests that the LW is not very sensitive to the particle SSA and g anyway?  This result is confusing, and since a significant aspect of the following argument depends on it, it's critical.  

L326 and L327.  Ranges are given for two of the inputs (SW g for black carbon and SW SSA for brown carbon) but that conflicts with there being only a single line for SW heating rate for each aerosol model. The specific values that are used in the test should be the ones given in the caption.

(Conclusions)

L351. See earlier comment about "consistency" between heating rates and observed plume rise rates.

Technical comments:

L17 suggest reversing the two features in the sentence, so you're isolating the vortex from the dominant plume.

L21 and L266 suggest replacing "and then" with "and therefore" or "and thereby".  (To avoid the ambiguity of "then" referring to a time-based relationship)

L45 suggest replacing "reconciliate" with "reconcile"

L56 suggest replacing "maintain" with "persistence"

L56 "attributes" should be replaced with "attributed"

L92 suggest deleting or replacing "at the basis", since it's a vague phrase that doesn't really convey anything (also in this particular case, it's ambiguously similar to "at the bases" which is not what the authors meant but would be valid grammatically, so it's especially confusing).  "diabatic heating of the compact anticyclonic vortices" seems sufficient.  Or if not, then I suggest replacing it with "due to".

L128 "discusses" should be "discussed"

L129 "constrains" should be "constraints"

L173 and L192 suggest replacing "double" with "dual"

L174 "compacts" should be "compact"

L193 suggest replacing "mutual" with "relative"

L226 suggest replacing "these latter" with "the vortices" for maximum clarity.

L231 I don't understand what the authors mean by "individuated" here (Figure caption).  Please replace with another word or phrase.

L233 Table 1 caption.  suggest replacing "individuated" with "specified".  

L237 Suggest spelling out "heating rates" in the section title, for the convenience of readers who may read the section headers before reading all the text in detail.

L244 Suggest "we represent cooling rates as negative HR" to replace the phrase beginning "we use the idea of..."

L281 suggest replacing "the evidence discussed" with what it refers to, perhaps "Reflecting the lack of variability in SSA in the LW and the lack of sensitivity to changes in g inferred from Mie studies as discussed above"

L290 Figure 4 caption, should "regional" be "zonal"?  That is, is each latitude band for the whole globe, and if not, what are the longitude limits?

Figure 4. The caption suggests that there are both dashed and dotted lines, but I can only see one line pattern.  

Figure 4 and Figure 5. Please include a visual key to the line colors and linestyles within each figure.  

Review of "Radiative impacts of the Austrialian bushfires 2019-2020 - Part 2: Large-scale and in-vortex radiative heating" by P. Sellitto et al.

General comments

In this new study, Sellitto et al. present OMPS and CALIOP aerosol satellite observations and associated short- and longwave radiative transfer calculations to estimate the hemispheric and in-vortex heating rates of the 2019-2020 southeastern Australian bushfires. The study finds a strong dependence of the radiative transfer calculations and heating rate estimates on aerosol type and microphysical properties (single scattering albedo and asymmetry parameter). Heating rates estimated for small-sized, strongly absorbing black carbon particles are found to be consistent with the observed self-lofting of the wildfire plume on global and local scale. The important role of longwave emissions on the heating rate estimates is particularly emphasized.

Overall, the paper is well-written, clear and concise. I found it interesting to read and think it fits well within the scope of the journal. The study is scientifically sound and the results seem plausible to me. I would recommend that the paper be considered for publication, subject to a few minor comments as listed below.

Specific comments

l62-68: In the introduction (and/or the conclusions), it would be good to discuss the broader implications and relevance of the study a bit more. Specific estimates of heating rates for the 2019-2020 Australian event are provided, but are they relevant overall? How do they compare with other events? Are the results relevant to chemistry transport or climate modeling?

l167-170: It might be good to add another 1-2 sentences regarding the ECMWF-IFS derived data sets for vortex tracking. Was this a specially generated data product/simulation or is it based on common IFS forecasts/operational analysis?

Figures 4 and 5: For these figures, I have trouble relating the individual curves to the individual aerosol properties listed in the caption. The caption refers to dotted lines, but I don't see any dotted lines in the plots? Also, I see only one sky blue line and three dark blue/black lines, but no medium blue line in my printout. The orange line in Fig. 5 is barely visible, and different types of orange dots are not found.

l334: The text refers to the impact of the plume on the UTLS region, but it seems that the radiative effects are mostly confined to the lower stratosphere (18-23 km altitude), but not to the upper troposphere?

Technical corrections

l40: about than ten times -> about ten times

l50-51: it seems the units (W/m^2) for the radiative forcing are missing?

l243: can in turns -> can in turn

l243: the vertical dynamics _of the_ lofting or sinking