This study investigates the impact of anthropogenic land use change in the present day (2015) through perturbed emission of biogenic volatile organic compounds (BVOCs). They compare three simulations: with only natural vegetation, with present day deforestation and present-day deforestation only on crop land, while grass land is reforested. They find that present day deforestation leads to a decrease of 29% in biogenic secondary organic aerosol (SOA) compared to natural vegetation, and that their reforestation scenario (only crop land deforested) lead to an increase 26% compared to present day. In terms of the resulting radiative effect, they find a fairly small warming (0.06W/m2) as a result of deforestation and a similarly small cooling for potential reforestation (-0.038 W/m2).

The study is interesting, and mostly well written and it fits well within the scope of ACP. I do have a fair amount of comments, but most of them are about clarity in the text and the analysis/methods.

My main comments are:

• There is no estimation of the significance of the results in the results section. This should be fairly easy to fix. I suggest adding shadings to all figures where results are not significant, but it will be especially helpful for interpreting the cloud results I believe. There is a discussion of the significance of the results in the discussion section, but it is not clear what has been done and results are presented only for some properties. See more comments on this below.
• The study does not seem to take into account any changes in surface albedo from land use change and it is not entirely clear to me if BVOC emissions are allowed to affect oxidant chemistry and thus ozone and methane concentrations. I would assume not, since this is not discussed in the paper. The study would benefit, in my opinion, from highlighting early on and in the discussion and conclusion, which effects are and which are not included.
• The manuscript lacks information on SOA mass yields, new particle formation mechanisms and early particle growth. Both of these are important for interpreting the results in my view.

I have the following minor comments:

Abstract: The abstract is on the long side and in my view contains too much background. I would reduce the amount of detail and focus more on results.

L8: Is it correct to call it interactive vegetation if it doesn’t affect climate variables? Ref what is written in L136.

L9: “Given that …” seems a bit like the reader should know this already. I suggest to simplify this part of the abstract and only present what the simulations are meant to represent.

L29: There seems to be two versions of the same reference (L437-444) in the manuscript: Forester 2007a and b are the same?

L44: “… new particles, thus influencing SOA formation significantly”. I find this sentence confusing because SOA formation usually would indicate mass which is usually dominated by condensation. I would maybe rather say that the climate impact can be very different due to the impact on aerosol number and thus CCN.

L45: A 50 nm limit is too simplified both because it depends on a range of factors when a particle can act as a CCN and because the size at which it interacts efficiently with radiation directly is much higher. Please rephrase.

L47: Forester et al (2007) is the 4^th assessment report and the terminology has changed in the two last reports (5^th and 6^th). I suggest updating to aerosol radiation interactions (ARI) and aerosol cloud interactions (ACI) – especially because these terms are used later in the manuscript.

L51: “potential natural vegetation” should be explained here.

L52: Should it not say that this is a one way coupling in terms of everything except BVOCs? Or have I misunderstood the set up?

L59: What does it mean that the model represents atmospheric processes and their interactions with anthropogenic activities? Is there a two-way coupling there?

L64: I am not sure what details are in Jöckel et al. (2016) and which are also detailed below. Can you make this clear?

L69: “factors such as” can be deleted here, right? Because the list is exhaustive?

L84: “This allows the computation of OA particle capacity to serve as a cloud condensation nuclei” This seems to suggest that OA particles are externally mixed with the rest of the particle population, which should not be the case?

L86: Only in tropical regions? Are all the studies referenced below from tropical regions? What is the reference for the tropical region study? Consider revising these two sentences to make it clearer.

L94: Strike “however,” as there is no contradiction here.

Section 2.1: This section needs information here about new particle formation (NPF) and what role organics play there as well as SOA mass yields from BVOCs.

L130: Out of interest, what vegetation would this result in? Is there grass land and crops in the region?

L136: Which “fluxes” are referred to here?

L137: I think this point needs to be highlighted more (as mentioned above as well). Where does the albedo and the roughness come from? What model is used for these properties? Are the land properties which interact with the meteorology from present day land use or something else? One thing is that the climate effects via albedo, roughness etc. are not included, another is that there might be feedbacks also on the BVOC emissions. I think this deserves more space both here and in the discussion.

L174: I am not sure what “surface offline emissions” means. Is it aerosol and aerosol precursors? And is it then all except the BVOC emissions? Also, please give a reference to input for reproducibility.

L182: strictly speaking c and d do not show this.

L185: As far as I understand, grass is here used to represent both grass land and crops. Could you comment on how realistic this is in terms of BVOC emissions?

Fig. 2: “Area of vegetation per unit ground area”: is this the area covered by vegetation? Not e.g. leaf area index?

Fig.3: Are the units correct in e and f, or is this also supposed to be Tg(C) as in the above plots?

Fig. 4: 5^th-95^th percentile of what? The annual and spatial mean? The monthly and spatial mean?

Fig. 5: A bit the same question: the spatio-temporal mean is just one number, it doesn’t have a standard deviation? Maybe I misunderstood something? Also, the font size is a bit on the tiny end here.

L210: If you want, I would think you can remove the explanations of AOT and aerosol extinction since it should be text book material. But it’s fine either way.

L216: The sentence about which optical properties EMAC simulates seems out of place: consider deleting?

L215-220: An analysis of the significance of these differences would be appropriate here.

Fig. 6: the relative plots look a bit strange in the bottom here. Why is there no variability in f?

L225-231: Again, it would be good to get some idea of the significance of these results, in particular the tiny increase in CCN and CDNC in high latitudes.  

Fig. 7: Again, it would be useful to shade non-significant changes here.

L234: Check sentence structure.

L235: Figure 7 is not referenced here? Also, it feels like the blue blob over south-eastern Asia should be mentioned here already?

L267-268: Insignificance should be calculated. Also, I assume the percentage change is calculated from (Fclean – Fclean,clear-sky) in both simulations. This is the total cloud radiative effect for each simulation, not the total effect of aerosols on clouds, and in my opinion using relative change doesn’t make sense for this (just like it doesn’t make sense for temperature).

L279: Does the study confirm a decline in tree cover? It seems to me like the model is forced to have a reduction in tree cover, so this seems to be not a result, but an assumption in the study.

L284: It says “studies” but there is only one study referenced. Also, I assume you mean global BVOC emissions?

L294: Why is this particularly true because the study only looks at land use changes? I think it would be good to say something here about which studies include temperature feedbacks and which are purely forcing for example, and in which direction this would contribute (lower or higher change).

L294: MEGAN? It says MEAGN.

L325: I don’t quite follow the explanation here: Is it the growth of particles into the Greenfield gap and thus expanded life time that drives this, or the growth of particles into the size range where they have high light scattering efficiency? I also don’t particularly follow the argument that the absence of organics drives the growth. It would be helpful to know something about how NPF and early particle growth was treated here. It would also be good to see some evidence for these explanations – are you sure about them? In general, I suggest rewriting this paragraph to make the argument clearer.

I also wonder why this effect does not show up when you compare the DCL and DCGL runs.

L335: “Changes in cloud properties …” Is this sentence needed?

L336: Note that Scott et al (2014) look at surface level CCN and CDNC at cloud height, so it’s not really comparable to column burden. Consider revising.

L338: I guess one reason why the change in CCN is so much larger than the change in CDNC in your study is that you are looking at the change in column burden CCN, and a lot of these particles might be too high up to interact with clouds?

L241: Why do these results align well with Scott et al (2018)? Scott et al (2014) was the paper discussed previously in this paragraph, so should it still be Scott et al (2014)? If so, the question stands: I don’t think the results are quite comparable, so I am not sure you have reason to say either that it aligns or does not currently?

L345: This is in addition to roughness changes and albedo changes that can also influence cloud cover (see e.g. discussion in Cerasoli et al. (2021, PNAS)).

L350: Is Tilmes et al. (2019) the best reference for such a general statement? It’s a one model study. Maybe a review paper would be better.

L353-356: Consider revising. It’s a bit unclear what comes from which study and what are their assumptions.

L355: I would suggest sticking with ARI here instead of using direct and indirect effect.

L359: This would be a good place to discuss the impact of SOA yields (see e.g. discussion in Sporre et al. 2020, ACP) and potential oxidant chemistry (see e.g. Weber et al., 2022, Nat. Comms.).

L366-377: Why is this not part of the manuscript? These are results, not discussion. Also, it is not clear what has gone into the Student’ t test (Is a paired student t test with annual global means e.g.?).

Furthermore, statistical significance of a result is not a property of the effect (large, small etc.), but only a measure of if it can be detected or not (how likely is the result to be by accident). I write this because the text as it stands seems to suggest that bSOA either is or is not significantly affecting clouds, but this is highly dependent on model setup and if you run enough years, even a very small effect can be significant. Perhaps I am just misunderstanding the text, but please consider revising this paragraph.

L387: Is WASO changing its meaning? It’s defined as water-soluble organic compounds earlier.

Review of “Land use change influence on atmospheric organic gases, aerosols and radiative effects”

By Vella et al.

The aim of the manuscript is to detail changes in the impacts of biogenic emissions under different levels of deforestation. The global study changes the area of land currently defined as tree plant functional types towards more grass and cropland area as a future world requires more land for agriculture. The authors find that the reduction in biogenic emissions from deforestation lead to associated reductions in secondary organic aerosol production which leads to warming. In an alternative scenario, the impacts of an extreme reforestation scenario are explored where a cooling effect is produced from increased biogenic emissions.

I thank the authors for embedding the figures in the text, this greatly helps the reviewer!

The paper is well within the scope of ACP and I recommend publication after considering a few points.

1. Understanding the scenarios

I had real problems understanding the scenarios. I first thought PNV was the current vegetation. An explanation of what potential natural vegetation is would help when it is introduced on line 127 (i.e. the absence of humans).

In the methods section it says DCGL is sometimes referred to as ‘present day deforestation’ on line 163. Yet on line 241 it says ‘present-day land cover’. Perhaps this should be called ‘current land cover’? It took me a couple of reads to work this out.

I guess what I’m trying to get at is the study where we compare the current conditions the world finds itself in with either deforestation of afforestation, as it shows us what differences can be made if we collectively choose to adopt either scenario now. I’m finding the study using PNV confusing as it never existed – certainly not in the years 2000-2012.

2. Other comments

The responses of PNV and DCL look very similar to me – in figures 3 (e,f) and figure 8 (e,f). Is this because crops are not large bVOC emitters? And why do the isoprene responses in both figures peak in spring as opposed to summer months?

Line 166. The years being simulated should be stated at the start of the paragraph (rather than at line 170). And is the same meteorology driving all 3 of your scenarios, or is the model being driven by data from a climate run for an assumed future extreme afforestation run? Some comment on the associated impacts of warming on biogenic emissions and bSOA production is warranted.

Figure 1. The unit is ‘land transformation faction’. Does this figure already include land that is agricultural, or is this the land that is being changed in the model to crops and grazing land? I was surprised given the large land transformation fraction (what I understand as ‘deforestation’ from the figure caption) in the mid-USA that we don’t see more of an impact there in figures 5 onwards?

Figures 5 & 6. The line diagrams are missing a legend to distinguish between the colored lines.

Line 281. ‘by’