the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Intensive aerosol properties of boreal and regional biomass burning aerosol at Mt. Bachelor Observatory: larger and black carbon (BC)-dominant particles transported from Siberian wildfires
Nathaniel W. May
Noah Bernays
Ryan Farley
Download
- Final revised paper (published on 28 Feb 2023)
- Supplement to the final revised paper
- Preprint (discussion started on 11 Apr 2022)
- Supplement to the preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on acp-2022-167', Anonymous Referee #2, 06 Jun 2022
This manuscript by May et al. presents field measurements of BB aerosols transported to the MBO site, including long-range transported smoke from Alaskan and Siberian boreal forest wildfires and emissions from regional wildfires.
The measurements characterized the physical and optical properties of BB aerosols for plume age from 10 h to 14+ days. The major findings include (1) This work supports the widespread influence of different wildfire emissions on aerosol properties in the western US; (2) The short and long-range transported plume aerosols present different physical and optical properties, which are related to sources and various chemical and physical processes during transport. The paper is well-written and is a valuable contribution to the BB studies. I have some comments detailed below.
Major comment:
1. Section 3.4, Page 12: the discussion of ΔPM1/ΔCO
1) Please check the unit of ΔPM1/ΔCO. It should be (μg m−3 ppbv−1)?
2) Line 285-287: Suggest adding more information, i.e. “The concurrent measurement of decreasing ΔOA/ΔCO with increased transport time in Farley et al. (2022) is still because that net OA loss through evaporation and deposition was great than the secondary processing. And OA is the majority component of PM1. Thus, the trend of ΔPM1/ΔCO follows the ΔOA/ΔCO, which can support the observation of PM1 loss in this study”.
2. Section 3.4, Page 13: the discussion of MAE and BC dominance
Line 318: I prefer not to use “BC enhancement” in this para, it is “the normalized enhancement ratios of BC (ΔrBC/ΔCO) in Siberian events were identified higher than other cases” (Farley et al., 2022). In Farley et al. (2022), the highest ΔrBC/ΔCO ratios were measured during both Siberia and Oregon events, suggesting that these events had more influence from flaming fires. It is still not explained why the Siberian events exhibited a higher average MAE (0.60 m2 g−1) than the SW OR events (0.30 m2 g−1). Suggest checking BC fraction in PM1.
The mean MAE in the 8/12 Siberian event (0.48 m2 g−1) is actually close to most other events, the mean MAE in the 8/17 Siberian event (0.72 m2 g−1) is much higher, which leads to the higher average MAE of Siberia events. What caused this difference between the 8/12 and 8/17 Siberia events?
The discussions indicate that Siberia events in this study had more influence from flaming fires, thus having a high ΔBC/ΔCO and MAE. I agree with this, more flaming fires do generate plumes with enhanced BC emissions (ΔBC/ΔCO). However, in this study, the ΔBC/ΔCO ratios are not only related to the source fire conditions but also the transport processes, i.e., Siberia events are suggested to experience wet deposition. It’s needed to mention that although the Siberia events experienced strong wet deposition, the ΔBC/ΔCO ratios are still higher than other events. A study ( https://doi.org/10.1029/2020GL088858) observed enhanced fraction of BC after vertical transport from the surface to the top of the boundary layer due to the lower removal efficiency of BC than the non-BC materials and the evaporation of other non-BC materials, which may be also related the transport processes.
3. Line 343: A recent laboratory study also found the imaginary part of BrC could be half decayed in a few hours, in line with the loss of its absorptivity after transport (https://doi.org/10.1021/acs.est.0c07569).
The laboratory work by Cappa et al., 2020 (hhttps://doi.org/10.5194/acp-20-8511-2020) and field observation from Wu et al., 2021 (https://doi.org/10.5194/acp-21-9417-2021) suggest that the evolution of AAE and BrC absorptivity with photochemical aging is dependent on the fire burn conditions and initial emission particle properties. There is an initial enhancement stage of AAE and BrC absorptivity followed by the decrease with longer aging times for more flaming fires, while more smouldering fires are suggested to experience a net decrease upon aging. The short-range transported (10-15 h) SW OR events with the highest AAE may experience the initial enhancement stage or decrease during this short-range transport period depending on the fire condition. Suggest adding more clarification here.
Line 393-395: “Notably, a portion of the 8/17 Siberian event exhibited a combination of elevated AAE and low SAE that is typically indicative of dust aerosols with enhanced absorption at short wavelengths.” Siberia events did exhibit lower SAE than other events due to larger size mode particles. However, the AAE values in Siberia events were not elevated compared to other events. What does the “elevated AAE” mean? The AAE values in Siberia events were close to another long-range transport event (Alaska event). I don’t think the AAE can be evidence of the dust aerosols mixing with plumes.
4. Page 22: For the aerosol size distribution in Siberia events, I agree that the observed <100 nm modes may be indicative of the influence of entrained background air and/or new particle formation. However, in Siberia events, the suggested wet deposition during transport would remove larger-size BB aerosols and would also result in a smaller size mode under 100 nm. Examples from Taylor et al., 2014 (https://doi.org/10.5194/acp-14-13755-2014) can support this. This is also related to the conclusion on Page 24.
Specific comments:
Line 143: need a full name for “PSAP”.
Line 171: Please check the correction of OPC PM1 measurements in the supplementary, not found.
Line 198: How do you get the BB event criteria of σscat > 20 Mm−1 and CO > 110 ppbv?
Line 214: In Figure1, need to add annotation (the blue, green and red lines in σabs and σscat plots represent blue, green and red channels respectively).
Line 220-221: Two BB events (7/5-7/7), rather than the (7/5-7/6) in the manuscript. The ΔWV of Alaska events is -1.52 in Table 1, not consistent with -1.53 here. Some values in Table 1 are not consistent with the manuscript, please check them.
Line 239: Which prior observations? If indicating previous work, please add the reference.
Page 11, Table 1: It would be good to add notes explaining why some of the data are missing in Table 1 (not measured or not good quality data?).
Line 313: “in the / during Siberia BB event” repeat word? Please re-phase
Line 314: the result "of" increased aqueous-phase cloud processing during transport
Line 330: ARCTAS-A aircraft measurements in Alaska "reported" a much larger BC/CO ratio.
Line 337: What is “PNW” BB event? BB event from where?
Line 369/370: Please explain what is “Dpm”?
-
RC2: 'Comment on acp-2022-167', Anonymous Referee #3, 05 Jul 2022
This paper summarizes the results of measurements made at Mt Bachelor over a roughly 2-month period. A variety of instruments were used to determine dry particle optical properties (absorption and scattering, each at 3 wavelengths), and particle number size distributions for particles in the diameter range of 30 to 600nm. Over this period, 13 biomass burning events were identified. The likely regions of wildfires producing the events were determined by back trajectory analysis, and properties of the events compared between sources (or ages of smoke). A possible dust event mixed with smoke is also identified. The paper is very well written, and care is taken in producing the data and evaluating uncertainties. However, sitting at a fixed site and observing periods of smoke blow by puts limitations on interpreting the smoke events. The authors spend considerable time trying to explain possible causes for the observed differences between events and those of other studies, which in most cases is largely speculation. I recommend the authors attempt a more robust analysis that would include more support (eg, statistical or theoretical calculations) for their many hypotheses to explain the differences. Overall, the methods nor the results are highly novel, but they do provide very useful data on smoke plumes of various approximate ages in the real atmosphere; possible causes for the variability discussed throughout the manuscript needs more attention.
Specific Comments:
Line 49-50, this definition of BrC (the aerosol overall AAE>2) is highly measurement specific, it reflects the fact that most methods, such as those deployed in this study, cannot detect low levels of BrC (ie, instruments that cannot isolate BrC, have very limited spectral resolution, and have lowest measurement wavelengths at rather high values where BrC contributions are not strong, such as in this study with lowest wavelength 450nm). The authors should clarify how this affects what they specify as BrC. For example, for a given wavelength, say 450 nm and 350 nm, for various AAEs, what fraction of the light absorption coefficient is due to BrC, and include an uncertainty in this ratio that considers the effect of the fit (ie, the assumption of a power law and the specific two wavelengths used, and the variability in pure BC AAE). Finally, does this method of determining contributions of BrC depend on the assumption that BrC falls on a continuum from weak to strongly absorbing BrC, see (Saleh, R. (2020), From Measurements to Models: Toward Accurate Representation of Brown Carbon in Climate Calculations, Current Poll. Rep., 6, 90-104). The point here is that there are subtleties and limitations in determining BrC based simply on AAEs – which should discuss.
Does the SAE and AAE calculated depend on the specific 2 wavelengths selected for the calculation. Eg, how much would they vary if the other two pairs were used in the calculation?
How much mass is missed by particles smaller than the OPC lower size limit? This could be determined by estimating the mass from the SMPS.
Line 178, at what wavelength was the single scattering albedo determined at?
Lines 275 to 295 is largely speculation since the NEMR’s at the sources are not known and NEMRs from only 2 other studies are used as a contrast. I suggest this analysis be changed by providing an in-depth discussion of various dPM1/dCO recorded close to wildfires from as many data sets as can be found, and then compare that and the variability to the data from this study. The authors could then more quantitatively assess if they are observing differences in processes or if it is hard to say bases on variability in emissions that have been recorded. I would point out that dOA/dCO could likely be used as a surrogate for dPM1/dCO since most of the mass is OA, as the authors state in the Introduction. This will likely significantly expand the published data that can be used given all the recent aircraft missions studying wildfires. Over-interpretation of the data is common throughout this paper. Broad statistical comparisons, such as shown in Fig 2 are more convincing than speculation. Can a Fig similar to Fig 2 be made for dPM/dCO?
Lines 329 to 333, the interpretation here is that differences in observed BC/CO can be used to infer BC/CO at the sources (ie, flaming vs smoldering). Again, this is speculation. It may be one possible reason but there could be others, such as differential loss of BC relative to CO during smoke transport (contrast the typical lifetimes of BC and CO if precipitation is encountered).
Ling 346-347. The authors might want to note a contrasting paper, Dasari, S., A. Andersson, S. Bikkina, H. Holmstrand, K. Budhavant, S. Satheesh, E. Asmi, J. Kesti, J. Backman, A. Salam, D. Bisht, S. Tiwari, and Z. H. O. Gustafsson (2019), Photochemical degradation affects the light absorption of water-soluble brown carbon in the South Asian outflow, Sci. Advances, 5, eaau8066.
Fig 4 and associated text. The correlations and slopes shown depend on two points out of 8 or so. Can one infer from this that there is a general relationship here? Provide statistical proof.
Line 459-460. This last line sounds like the authors are making a generalization based solely on two observed events. Is that reasonable? It is rather unfortunate that no filters were collected that could be used to measure dust (Ca2+) and smoke tracers (K+, along with the measured BC) in the same event. This would provide proof that the plumes were indeed mixtures of smoke and mineral dust, the analysis presented is only suggestive. (This was noted on line 490, but I would point this out earlier in this discussion.
Section 3.6 title and text within, define exactly what type of size distribution is being discussed, ie number distributions.
Lines 480 and on, would not a calculated volume distribution provide better evidence for a possible dust influence in the SMPS measurement size range. Ie, one could compare the shape of volume distributions for the non-dust and speculated dust events.
Lines 491 to the end is mainly speculation. Why not estimate the lifetime of an UF particle based on the measured number distributions to support this discussion.
In conclusions, bullet 2. What is meant by little BrC. (See earlier comment). Given the very insensitive method the authors used to determine BrC, this is rather a subjective statement. Same applies to the term little remaining…. The point is, using the method for determining BrC in this paper, exactly what fraction of the light absorption at a given wavelength (the authors may choose) is due to BC vs BrC and include an uncertainty. Maybe use this instead of the term little.
- AC1: 'Comment on acp-2022-167', D.A.J. Jaffe, 01 Feb 2023