This paper discusses observations collected over the winter to springtime biomass burning season (June-September 2017) at Ascension Island, describing the varying aerosol properties over time.  The analysis specifically focuses on measurements of ten BB plumes which were observed, divided into three different “Regimes” separated in time. 

The main major comment I have on this paper is regarding the Regime separation and its justification and significance. The authors state that the division is supported by differing burn conditions (fire density, surface RH, land use, and fuel type) over the season (Line ~228+, Table 3), but it is difficult to see how these differences are supported in the analysis (Figs 2; 3-5). 

The rBC:delCO distributions are described in the text (e.g. Line 16) as being used to determine the different fuel and burn conditions, but as these values are shown in Fig 2, they are not obviously subdivided according to a particular criterion.  Particularly looking at plume P2, it doesn’t seem to “suggest burning conditions remained mostly homogeneous over the six weeks” (Line 281).  I wonder if there’s any evidence to suggest that P2 is more in line with P9-10 in terms of origins? Figs 9, 10, and 12 also don’t clearly show 3 distinct sets of properties, so I would like to see a more concrete justification for that division, if you choose to keep it. (and actually, P2 is an obvious outlier in Fig 12 as well).

Particularly regarding the transition between Regime 1 and 2, Fig 3 vs Fig 4 show averages over two periods of 4-6 weeks each, with one day of separation between them. It’s not at all obvious that e.g. the spatial distribution of fire density is meaningfully different between these two periods (Table 4 also seems to indicate that many parameters may be statistically indistinguishable between the two of them); instead, these figures’ panels (a) seem to show that there is greater density in the latter period, but that the fires occur over largely the same spatial domain. A distinction of surface RH <50% or >50% (p. 10) also seems a bit arbitrary a cutoff, and e.g. it’s not actually clear that the locations of the fires in Fig 3 correspond to RH>50% (line 283).

In the same vein, the time intervals averaged in Figures 3-5, if I’m reading it correctly, show the continental conditions directly coincident with the observed conditions at ASI. Yet, obviously, airmasses don’t arrive at ASI instantaneously; according to Figs 13 and 19, the transport time to reach ASI is ~5 days minimum and may be even greater than 10 days, so do the continental conditions during the exact same time actually indicate changes in conditions? And if so, shouldn’t the conditions (and Regime definition) be lead/lagged relative to the ASI observations?  

This is also complicated by Table 2 vs Table 3; the Regimes are defined as these larger periods, and yet there are also “clean periods” within those periods. So do Figs 3-5 show the average including “clean” times? From Fig 1 it’s about half clean, half plume over a given “regime,” so even if transport time lag is taken into account, did the fires vary within these longer periods?

The “efficient”/”inefficient” burning condition distinction also was not clear: the low rBC:delCO mass ratios = inefficient combustion (Line 280, 806+); higher values in the second period = efficient combustion; and the third period is either “inefficient combustion” (Line 292) or a combination of the two (Table 3), despite being lower values than either of the other two which are supposedly distinct? If anything, it seems to me that Regime 1 with the one outlier P2 plume should be the mix of burning conditions.

This also gets a bit muddled with the discussion of oxidation and evaporation, transport, and other processes (Section 4, throughout), since it’s not clear to start how these plume events are similar to one another. If the transport pathways in Fig 13 vary from t~5 to 9 days in these examples (just from when it exited the continent, it’s difficult to follow how this can be used to state that the OA:rBC etc ratios vary at the time of emission, if they’re then oxidized at different rates over different times, if I’m following Sec 4. It’s not clear how “faster transport pathways (Line ~974+) are definitive in one regime over another.

All told, I think the authors’ argument could be strengthened with some additional trajectory analysis to better constrain the locations and timing of likely smoke origins for the different plume events. Trajectories are mentioned (Line ~187, 429) but it’s not clear how that informs how the figures and analysis are structured, so maybe that’s easy to clarify in a revision?  As it stands now it is not clear that the “regime” distinction is meaningful.  I’d recommend reworking the structure to either better support the regime designations being made, or simply do away with those distinctions and discuss seasonal evolution, perhaps addressing why P2 is such an outlier (more in line with the current Regime 3, if categories are being determined from rBC:delCO only). It doesn’t usually happen, but was there a precipitation event in its history, or some other potential explanation based on your trajectories?

These observations over a dedicated time period are interesting and worth publishing, but I think some of the justifications behind the differences need to be better supported, clarified, or restructured entirely before publication.

Minor comments:

• Figure 1: caption states “Pink boxes indicate selected plume events; blue boxes indicate selected clean periods” but I’m only seeing grey boxes edged in orange. It might also be nice to label the different plume events on this figure (P1-10).
• Check that subpanels for Figs 3-5 are labeled properly in the caption and in the text; the caption seems to describe a, c, b rather than a, b, c.
• A minor note: VIIRS is an instrument on the Suomi NPP satellite (e.g. Fig 3 caption)
• Line 51: suggest to report lat/lon in S, W coordinates, rather than negative N,E.
• Discussion ~Line 47-58: the context might benefit from discussing Eck et al 2013 (doi:10.1002/jgrd.50500), which saw an increase in SSA through the BB season at one site, i.e., changes in optical properties likely from similar geographical regions.
• Line 181: surely there aren’t many fires at 5.7W, 3.2 N? Typo?
• Fig 6: I’m curious what the bars are on this plot— is it some standard deviation rather than the percentile distribution over a given event? I ask because in contrast to many of the other figures, these ones seem to be uniform rather than varying from plume to plume, but I would imagine that the range in diameters would vary between plumes as well?
• Sentence starting on Line 205: sentence fragment or missing a verb, I think
• Line ~224: rBC units switch back and forth between micrograms and nanograms; I’d pick one. Also it might be good to show the rBC<20ng/m3 threshold in Fig 1, which has 200ng/m3 as the lowest tick mark, if I’m reading it correctly
• Table 3: I presume these values are means/standard deviations, but it would be good to confirm that in the caption, if you stick with the Regime construction.
• Table 4: last row might be missing a +/-
• Fig 8: it might be nice to show the plume events here, as in Fig1, or perhaps as the ratios between parameters as described in the text.
• Line 479: typo in “mass”
• Line 584: it’s a bit difficult to follow what this sentence is saying.
• Line 651/Figs 13-14: I’m curious why the 850hPa winds are shown (I think; would be nice to specify in the caption). Especially later in the season, the south African Easterly Jet for continental transport is 700 or 600hPa in Aug or Sep.
• Table 5: I’m a bit curious about the ATTO transport pathway that gets African biomass burning smoke to the Amazon sooner than to Ascension in June; is that just based on the season? (I might add that to the header, then). But this is a different value than stated in Fig 16.
• Line 863: what is the relevance of Arctic observations to the present study? I presume they were also of BB aerosol? But Figure 15 suggests that perhaps the present study and the past Arctic observations were not comparable. I presume the aging/transport time was much longer for Arctic aerosol?
• Line 902 and 946: seem to be conflicted as to whether wet deposition can happen in this region?
• Line 968: missing a word?
• Line 1079: wrong verb?
• Fig A1: what averaging is shown in this figure (I don’t think this is the 1Hz data)? Also, the caption and the legend seem to be saying two different things re: which line/color is the CAPS?