The authors provide a thorough overview of the implementation of the MEGAN module into the SURFEX land model, and compare the model results in terms of isoprene emissions against reported inventories from the literature. The authors obtain mostly similar emission totals and distributions, compared to other MEGAN-based emission estimates, and clearly discuss the impact of model assumptions and input datasets which  drive the parameterization. The impact of changes in driving meteorology is relatively modest, and mostly explained by the PPFD. Changes in emission potential can have a larger impact, while the introduction of a sensitivity to drought, through the actual soil moisture relative to the wilting point, were shown to lead to very large differences in modeled emissions. Based on this information the authors rightly address this issue that further studies are necessary to reduce the uncertainties to soil moisture assumptions.

This raises the question whether the authors have any thoughts on how they intend to validate the assumed BVOC emissions. It would be useful if they could spend some words on this. It also relates to a concern I have whether there is evidence if some of the emission estimates (esp. the MEGAN-MACC) can be disregarded from this study, as they may appear beyond the range of reasonable values.

In general, this manuscript is very well written, and I recommend publication after addressing a few minor (mostly technical) comments:

minor comments:

line 8: “more precise” : more precise than what?

line 42”emission of ozone” -> “production of ozone” ?

Figure 1 / line 95 / line 96 / line 97: Please align the definition of different tiles, esp choose between ‘Town’ / Urban area / city , but not these three different names.

line 183 “its corresponding defined” please check wording

line 195 “is similar for most vegetation types”: Actually, apart from shrubs I also see relatively large discrepancies for grassland and needle-leaf trea - but maybe it ’s a colorscale issue (an/or not so relevant in eventual emissions.) - cant the authors comment?

line 200: “LAI value of the past 10 days” change to “LAI value of 10 days in the past” ?

Figure 3 shows that also the T24 and PPFD24 (previous day mean temperature and PPFD) are required as input to the ISBA-MEGAN processor. Out of curiosity, how are these parameters computed? Simply keeping track of temporal variations in PPFD and Temperature over all (24) hours for the last day, or is there a more intricate procedure for this?

line 217 “simulation of isoprene emissions”

line 224: The authors write: “As there are no available inputs for surface incident diffuse shortwave radiation .. a value of 0Wm-2 is assigned”. Maybe I do not understand this well, but wouldn’t it make more sense to assign a higher default value than 0 Wm-2 for this quantity?

line 275: “PAR calculated from incoming shortwave radiation” from which product is this PAR here derived? From ERA5 / ERA-Interim? please clarify

Figure 6: It would help if the authors re-organize the grouping such that one panel is given for each region, with bars for each emission product, to more easily intercompare the various estimates per region and month.

Oumami and co-workers present the coupling of the widely used biogenic emission model MEGANv2.1 and the surface model SURFEX. SURFEX aims to provide improved accuracy of vegetation-type specific parameterizations in MEGANv2.1 model and a framework for assessing the biogenic emissions in future climate scenarios. The SURFEX-MEGAN global simulation driven by ERA5 meteorological fields yields an estimated annual emission similar with previous studies, both in terms of magnitude and spatial and temporal distribution. Sensitivity to the input meteorological fields was found to be

very weak, of the order of 5%, with tropical regions being the most affected. The use of emission potential values per plant functional type, instead of an emission potential map, led to an overall decrease by 12%. In agreement with previous studies, the emissions are found to be sensitive to the soil moisture stress activity factor, with a calculated decrease of 38% with respect to the reference simulation which ignores this stress factor. The manuscript adds to the large body of MEGANv2.1 model studies in the literature. The results look reasonable and are well described. However, the innovative aspects of this work are small and concern mainly the coupling between SURFEX and MEGANv2.1. The manuscript should also 

be carefully checked throughout for spelling errors (e.g. 'for there ability', line 28; 'car' instead of care', line 30; 'with another models', line 53), missing references (e.g. line 43), use of undefined acronyms (e.g. ISBA, MERRA, PFT in the abstract), syntax errors (e.g. 'to be able to be forced', line 23), inconsistencies (e.g. 12% decrease reported in line 401 and 14% in line 475), and obvious mistakes (e.g. emission of ozone, line 42). A considerable

effort is therefore necessary to improve the presentation. In addition, I describe below a list of concerns to be addressed in a future version. I recommend major revision at the present stage.  

Comments:

- As often mentioned in the manuscript, the goal of this work is to force the the coupled model using climate change scenarios and assess the impact of climate change on the biogenic emissions. However, only a single year (2019) is presented in this paper. Unless you include long SURFEX-MEGAN simulation results in a revised version, it should be avoided to create more expectations than the manuscript can address.  

- You mention that the vegetation-type specific treatment in SURFEX improves the accuracy of isoprene estimates. Please provide evidence to prove this. Data from isoprene flux measurement campaigns (Seco et al. 2022, Emmerson et al., 2020; and many others) should be used to assess the SURFEX-MEGAN emissions. In a revised version, a detailed evaluation of the model results against flux measurements should be included.    

- The differences between the reference run and the sensitivity run using MERRA reanalysis are not adequately explained.  To ease the discussion, could you include comparisons of the monthly temperature from ERA5 and MERRA for the regions of Figure 8? Could you quantify the differences between the two datasets for temperature and solar radiation, which are the main driving factors of the isoprene emissions? Where are those differences more significant and what is the induced uncertainty due to the meteorology? 

- It is not clear what parameterization is used for the activity factor accounting for the isoprene inhibition due to enhanced CO₂ levels (Equ. 2). The impact of CO₂ levels on isoprene fluxes is highly uncertain, as past work reported contradictory findings (Sun et al., 2013; Tai et al., 2013, Bauwens et al., 2018). I urge the authors to discussion this aspect and quantify the impact of the CO₂ inhibition on the total annual emission. 

- This study does not account for recent efforts to improve the soil moisture activity factor parameterization in MEGAN (Jiang et al., 2018; Wang et al., 2022; Opacka et al. 2022). A discussion is needed and an estimation of the isoprene fluxes using (at least) one different parameterization should be added in the manuscript. 

 

- About Figures: (i) The readbility of Fig 2 is quite low. Could you change the scale? (ii) Not clear where Fig 8 is used. The figure seems to be at lower resolution. Can you improve the resolution?  

- l 74-75, "the coupling between SURFEX and MEGAN can create a feedback loop that takes into account both the impact on climate on vegetation and the impact of vegetation on climate'. Poorly worded sentence. Furthermore, these impacts are not presented in the manuscript. Could you include a tentative quantification of these impacts? 

- l 291: 'In low isoprene emission regions, such as North America'. This is not correct. Southeast US is among the strongest isoprene emitting regions in the world (Wiedinmyer et al., 2005; Kaiser et al. 2017 and others). 

- l 314: 'by a very low activity', replace by 'a very low flux'  

- l 323: 'These regions are active', replace by 'These regions are emitting'

- l 330: 'As stated in (Guenther et al., 2012), replace by 'As reported in Guenther et al. (2012)'. Here and throughout the manuscript, check your citations. 

- Table 4: can you add a column with the global total emission for each of the setups? Similarly in Figure 4: add the global emission estimate inside the subplots. In Table 4, the description 'use of isoprene emission potential' is vague. Be more specific. 

- l 346: correct problems with spaces

- Fig 13: Are the red dots over islands due to very low or zero e_map values over those regions? I recommand to add a global map of isoprene flux based on the e_PFT (with the global emission provided inset). 

- l 385: 'has been observed', replace by 'has been calculated' or 'has been found'

- l 392: 'was done to cover', replace by 'accounts for'

- l 425: correct the exponents

- l 421: '273Tg'. Add a space '273 Tg' here and throughout the manuscript. 

- code availability: provide links for the datasets presented in the paper

References

Bauwens et al. (2018) https://doi.org/10.5194/bg-15-3673-2018

Emmerson et al. (2019) https://doi.org/10.1016/j.atmosenv.2019.04.038

Jiang et al. (2018) https://doi.org/10.1016/j.atmosenv.2018.01.026

Kaiser et al. (2018) https://doi.org/10.5194/acp-2017-1137

  

Opacka et al. (2022) https://doi.org/10.3390/rs14092021

Seco et al. (2022) https://doi.org/10.1073/pnas.2118014119

Sun et al. (2013) https://doi.org/10.1093/jxb/ert318

Tai et al. (2013)  https://doi.org/10.1002/grl.50650

  

Wang et al. (2022) https://doi.org/10.1029/2022MS003174

Wiendinmyer et al. (2005)  https://doi.org/10.1029/2005JD005800