In this study, the authors analyse potential links between climate variability induced by the El Nino Southern Oscillation and BVOC emission fluxes in a coupled GCM framework with interactive vegetation. They focus on different aspects of ENSO and the associated impacts on the terrestrial biosphere and BVOC emissions. The study is well written and presents interesting results, and I appreciate that the model runs must have been a huge effort to set up. However, I have some concerns that need to be addressed before the manuscript can be accepted for publication and included some general and specific comments below.

General comments

• My major concern is the Results/Discussion section because the discussion is a bit thin. I might have miscounted, but there are only four references to contextualise the results! That’s not enough for a discussion and I would like to see a more critical view on the model set up and outcomes in the study. I wonder whether it would help to split the results and discussion into two separate parts of the paper (but this is up to the authors). Some things that could be discussed are:
  1. You used a coupled simulation stressing the importance of land-atmosphere interactions but you don’t really dig into describing processes that might influence the BVOC emissions (and how) except for the last sentence in the conclusions.
  2. Has your coupled GCM set-up been evaluated and demonstrated to capture BVOC responses sufficiently compared to observations (if observations are available)?
  3. How do other land surface schemes model BVOC emissions and would you expect different results using a different LSM or GCM? Would you expect that your model framework is more suitable to address your research question compared to other coupled models that enable BVOC simulations?
  4. Are there any caveats in the study itself or shortcomings in the model that could have inflated/ underestimated the results?

I want to stress that I don’t expect detailed answers to all the questions above, they are just some suggestions for potential discussion points.

• After reading the discussion and conclusions, it is not clear to me what the implications of the study are. By that I don’t mean it the study set-up and results are not sufficiently interesting, but in my view the authors could expand more on the significance of their study in the current climate and future. i.e. if in fact ENSO events do become more sustained and/ or more extreme in a changing climate, you expect increased BVOC emissions. But what does that mean for associated processes in the atmosphere?
• Your analysis relies on both isoprene and monoterpene emissions but the monoterpene emissions are largely neglected in the manuscript. Do the two types of emissions play different roles in the atmosphere or are they quite similar? You could pick this up in the discussion.

Specific comments

L1: It might be nicer to start with the umbrella term (BVOC emissions which is also in your title), and then divvy it up into isoprene and monoterpene emissions later on? But this is my personal preference and up to the authors.

L1: Can you give one example that explains the ‘significant role’ BVOC emissions play?

L5: ENSO is the most important mode of climate variability and you could state this in the abstract to motivate your study

L35: Typo (?) ‘The El Nino-Southern Oscillation (ENSO) is a periodic oscillation’

L58: Are there any assumptions that might explain the impact of ENSO on BVOC emissions in higher latitudes?

L83-84: The citations are a bit off - ‘aerosol-cloud interactions (e.g. Tost, 2017). In this study, version […] used in comprehensive model intercomparison studies (e.g. Joeckel et al., 2016)’

L87: Could define LPJ-GUESS in the first line of the section (L86, sorry for being pedantic)

L96: Why did you exclude land-use change? The use of PNV could also be a discussion point

L98-105: You very superficially describe the different components of the model, fair enough – but given this study is focussed on the BVOC it’d be nice to know if there is one core process or something that describes the BVOC module and how it links land surface and atmosphere in your model set up.

L107: Have you defined AMIPII somewhere?

L111: Are your thresholds defining weak, moderate and strong common practice? I.e. can you support this decision with a citation pointing to other research using the same thresholds?

L125: Not sure I understand the last sentence on the page. Are you saying you chose the seven regions because they are mostly in the tropics which are typically areas with high BVOC emissions (can you include a reference to support this statement)? You could further motivate the choice of regions by mentioning that they conveniently happen to be ENSO hotspots as well (except NE Australia)

L129: Have you defined the ‘base conditions’ somewhere?

L132: Does this mean that in the 31^st and 32^nd year you perturb the atmospheric circulation with your ENSO anomalies?

L142: Typeo - ‘Even though’

L151: Could you write out Jan and Dec to January and December please:)

L151: Capital Event?

L167: A bit convoluted.. Maybe something like ‘however, following the ENSO perturbation fluxes diverge’

L189: The definition of the aridity index belongs in the methods section

L190: Throughout your manuscript you’re not consistent with italic/ not italic ‘base’ scenarios/ conditions

L191: Are r-values the correlation coefficients?

Table 2: I like that you give both percentage and absolute changes for temperature in Table 2 to get a sense of magnitude. Can you also include the actual values for change in Radiation and AI in the table? It might make it easier to link the table to Figure 4.

Figure 3: I think the figure is very small and it’s quite hard to see anything on it. Maybe you could rearrange the panels. There is also a lot of white space at the top that maybe could be trimmed? But maybe I just can’t see the datapoints. Could you spell out the abbreviations in the caption too (AI, NPP, LAI)?

Section 3.2.1. is a description of the results – where is the discussion here? For example, are the anomalies shown in Figure 4 what you expect […]? As I said above, it might be easier to split results and discussion but that is up to you.

Figure 4: I appreciate the value of including anomalies over the ocean as the temperature and radiation plots show the typical ENSO anomalies over the ocean quite nicely. However I wonder, given this study is mostly focussed on land processes, whether you would consider to mask the ocean and include a supplementary figure of the SST anomalies to demonstrate that your experiment captures ENSO. Especially for the radiation anomalies, it is quite hard to see what’s happening for the majority of the land surface because the colorbar is maxed out to fit to the ocean anomalies. In this figure, I’m surprised that the bottom panels do not show a signal in Australia which pops up as one of the most impacted regions in Figure 5. Does the water limitation signal disappear because you use the aridity index here rather than direct precipitation anomalies? None of the other anomalies seem to able to explain the strong signal.

L219: You use the abbreviation SEASIA here but in Table 2 for example it is SEAsia. I’m not sure whether this happening in other places in the manuscript, but can you make sure you are consistent within the manuscript?

Figure 5: I’m a bit surprised about this figure but maybe I’m misreading it. The middle panels contrast vegetation anomalies in an extreme El Nino with that of an extreme La Nina right? The patterns almost look identical, especially for NPP, and I had to zoom in to see that they the top middle and right panel are not identical but show small differences in the hatching. I wonder whether there might have been a mistake in the plot. Typically for an extreme El Nino, you would expect a negative signal at least for parts of Australia due to increased water limitation while for a La Nina it would be positive as you show here. I also would have expected a negative signal in the tropical rainforests in South America and South East Asia. Figure 4 shows a somewhat contrasting signal in the Aridity Index and to some degree in the incoming SW radiation (but it's hard to tell due to the colorbar, see comment above). Can you confirm that Figure 5 indeed shows the ‘correct’ distribution of anomalies, and if so can you explain the signal given it is quite counter-intuitive?

L.234-235: Can you rephrase this? Nearly instantaneously and rather quickly sound like quite similar timescales to me. Are you showing the lag in vegetation response somewhere? If so can you point the reader to that information? If this is meant to be a more general discussion point, could you include a reference to support this statement?

L263: I probably just missed it, but where did you mention before that the emission changes may be exaggerated? I think this could be a good discussion point, can you unpack this more?

L279: What does this mean – high NPP and LAI = high isoprene emissions?

L271-284: Following this section, you define ‘strong correlations’ as values greater than 0.4? Often correlation coefficients are split into weak/moderate/strong classes, and values around 0.5 would typically considered moderate. I think you should be more careful with your phrasing here and/or define somewhere where your differentiation is coming from (based on significance?). You do need to be consistent with the actual values of the correlation coefficients though; the ones in written in the text do not always match the ones in the figure (small differences only).

L289-295: You found relationships based on a Pearson correlation, but you don’t explain why temperature anomalies drive isoprene fluxes in Africa, and LAI in the southern USA, north east South America, South Africa, Central Asia and Australia. Is this a surprising result? Is it what you expected? Do you know why this is emerging from the model?

L312: Your current data availability statement is not sufficient for Earth System Dynamics. You should at least make your analysis code publicly available. For the time being I’m sure a github link (or similar) will be enough but for publication you will be asked to publish a zenodo link anyway so you might as well get started on that now!

Reviews for “BVOC emission flux response to the El Nino-Southern Oscillation”

Isoprene and monoterpene emissions from the terrestrial biosphere play a significant role in major atmospheric processes. Biogenic volatile organic compound (BVOC) emissions are sensitive to climatic influences. This manuscript attempts to understand the relationship between BVOC emission and ENSO events using a global atmospheric chemistry-climate model with enabled interactive vegetation. Overall, the results are reasonable, and I recommend a major revision before acceptance.

Major comments:

(1) In Section 2.2 EMAC-LPJ-GUESS configuration, I prefer that you can list some key equations for the parameterizations of BVOC emissions in this study. So we can easily understand why you choose temperature, radiation, AI, NPP, and LAI to investigate their impacts on BVOC emission anomalies.

(2) Lots of sentences in the main text should appear in the figure captions. Please revise them through the whole text. For example, Page5 Line117-118, “The base year (ie. The 30-year average SST …. Blue (La Nina).” should be placed in the Figure 1 caption. Page 9 Line 191-192 “The r value for each grid is shown and correlations with p < .01 are marked with a start sign” should be placed in the Figure 3 caption.

(3) Page 8 Line 176-177 “During El Nino and the subsequent two years, SWUSA experiences a rise of 15.6% and 14.3%, respectively, while a decline of 24.4% is found in SWUSA during the two years following La Nina”. The responses seem to be asymmetrical for El Nino and La Nina. So why the response of BVOC to La Nina has the lowest decline in the following two years?

(4) English writing need to be improved further.

Some minors:

(1) Page 1 Line 1: “major atmospheric processes”, could you show one or two specific examples.

(2) Page 2 Line 45-46: “Several studies explored the sensitivity of the terrestrial biosphere to different ENSO phases (e.g. Ahlstrom et al., 2015; Chang et al., 2017; Bastos et al., 2018; Teckentrup et al., 2021)”, here is another paper well suitable here. See “Wang, J., Zeng, N., Wang, M., Jiang, F., Chen, J., Friedlingstein, P., Jain, A. K., Jiang, Z., Ju, W., Lienert, S., Nabel, J., Sitch, S., Viovy, N., Wang, H., and Wiltshire, A. J.: Contrasting interannual atmospheric CO2 variabilities and their terrestrial mechanisms for two types of El Ninos, Atmos. Chem. Phys., 18, 10333-10345, 2018.”

(3) Page 6 Line 142: “give realistic insights on changes” => give insights into changes. I think simulated results are not necessarily “realistic”.

(4) Page 10 Line 210: “… anomalies from very strong El Nino and La Nina scenarios”, the results in Figure 4 is composite results?

(5) Page 13 Line 240-244: Two sentences are duplicate.

(6) Page 13: “TeBe” => “TeBE”

(7) Page 14 Line 262-263: “statistically significant anomalies only occur in the very strong El Niño scenario with and increase from 34.13 Tg yr^-1 to 38.13 Tg yr^-1 (+11.72%) from base scenarios to very strong El Nino” => statistically significant anomalies only occur in the very strong El Niño scenario with the increase from 34.13 Tg yr^-1 during the base scenarios to 38.13 Tg yr^-1 (+11.72%) during the very strong El Nino.

(8) Figure 8 figure caption: The Person’s correlation => The Pearson’s correlation