The manuscript “Enhanced dataset of global marine isoprene emission from biogenic and photochemical processes for the period 2001-2020” by Cui et al. is about satellite and ERA5 based global calculations of biogenic emissions and photochemical emissions of isoprene in the surface ocean. The authors use and update existing calculation methods for biogenic (BIO) and SML emissions of isoprene on hourly basis, compare emissions and surface concentrations with observations and correlate their calculated emissions with other parameters (e.g. SST, light, wind speed). Additionally, they investigate potential periodic changes of SML and BIO emissions in the twenty year period using MVEOF.

General comments

Understanding the mechanisms of isoprene production and consumption in the world oceans is crucial to finally investigate the influence of marine isoprene on climate. This work summarizes the current knowledge of production and loss processes in the surface ocean as well as the photochemical production in the SML and uses satellite and ERA5 data in order to calculate isoprene emissions with a high temporal and spatial resolution, which is of absolute importance in order to evaluate the impact of global marine isoprene emissions.

However, the authors like to stress the high temporal resolution, but they do not show any hourly or daily data. Results are shown mainly as global annual mean maps. Moreover, chla is used as a proxy for both, the BIO as well as for the SML emission calculations, but input chla data is on monthly basis. Apart from hourly ERA5 data which definitely is a key to describe diurnal cycles of isoprene emissions (which are also not shown) using monthly mean chla data, is somewhat contrary to “a high temporal resolution”. In a revised version of the manuscript I suggest to stress the “high temporal resolution” a bit less and concentrate on a proper discussion of the findings (in comparison to other models and/or observations). As a validation of the model, especially a proper comparison with observations is of fundamental importance.

The manuscript is very long in the method part, which is not a problem in general, but could be shortened, as some parts of the results and discussion in the method section should be moved. Comparisons to other model results or to observations are partly mentioned but more like in small bits here and there which makes it difficult to follow the flow. The authors should think about a few main scientific questions in which context they want to present, but also discuss (important!) their findings. This will help the reader to follow the “red line”.

On the other hand, in some parts information is missing which is needed to understand the reasoning of the authors (see specific comments).

The manuscript additionally needs an english language revision by a native speaker.

Therefore, I suggest to revise this manuscript related to content and structure before publication.

Specific comments

Abstract: The abstract contains a lot of unnecessary information which should be mentioned in the method section and not in the abstract. Please revise and focus on the main findings and implications.

Introduction: Please streamline the introduction dependent on the main questions which will be discussed in the manuscript. The first paragraph about BVOCs is very general and could potentially be used to concentrate on isoprene (which is in similar context discussed in ll.80-88).

ll.72: “ Previous estimates…” Please be more specific and/or give references.

ll.95: This paragraph is unnecessary.

ll.101: “…downwelling radiative flux diffuse attenuation coefficient…” Please be more specific. Is it 490nm? Somewhat later 490nm is mentioned but should be shifted to this section.

ll.110: It is not clear to me how the use of the water leaving radiance at 410nm from the period 2012-2020 is matched with the general period starting in 2001. I assume that the authors used average monthly values from the period 2012-2020 in order to determine the prevalent phytoplankton types, which would be a monthly climatology. Please revise. Additionally, if somehow monthly averages over time period 2012-2020 are used, this issue should be discussed later in the manuscript as all other input parameters are not averaged over 20 years.

l.115: ”…biological costs…”. This wording sounds wrong to me and appears a few times in the manuscript.

BIO emission module: It is not clear to me how the authors use equation 2 (from eq. 3 in Simó et al., (2022)) by phrasing that the resulting value α is dimensionless. In the correlation from Simó et al. (2022) the chl-a dependent loss term is a rate constant in 1/day. Please elaborate on this. Also, how is the maximum value α=0.373 (when chl-a conc. is higher than 5,77 mg/m3) calculated?

ll.156: “The mean annual…” This is a result and should be moved to the result section.

ll.176: I absolutely agree to fill grid cells with numbers if there are missing values in order to avoid underestimation of isoprene emissions. However, the authors should justify why a coefficient of 0.028 is used in areas with chl-a conc lower 0.04 mg/m3 or areas with missing values.

ll.181: “…a combination of 50% diatoms and 50% haptophytes in the grids…”. In the adjacent Table 2 it says “50% other types + 50% diatoms”. Please change accordingly.

ll.191 until end of paragraph: This section belongs to the discussion section.

ll.213: The citation of Flab is incorrect. It reads that the authors calculated a mean Flab value from the published values by Ciuraru et al. (2015a) and Conte et al. (2020). However, the authors use an average Flab value which already was calculated by Conte et al. (2020) and is dependent on the data from Brüggemann et al. (2017), Ciurau et al. (2015a) and Ciuraru et al. (2015b).

l.215: μ(photo) is not the radiation intensity in mW/m2. According to Brüggemann et al. (2018) it is the photochemical emission potential. Please revise wording and units.

Table 4 and Figure 3: Both, table and figure, present isoprene emission values. However, those values are hardly comparable as they are noted in three different units. Please just use consistent units throughout the manuscript.

l.301: The citation for the Schmidt number is incorrect. The correct reference is Palmer and Shaw (2005).

Section 3.2: The first paragraph shows results which is not appropriate in the method section. The rest of the section is mainly discussion (as the title also suggests) and should be moved accordingly.

ll.340: The authors highlight their use of hourly data “…which probably provides a more accurate representation of emission dynamics.” This sounds very vague. On the other hand, the authors could actually proof this using their dataset and compare to results using a lower temporal or spatial resolution and explicitly show the improvements (also in relation to one of the general comments).

ll.376 and Figure 5: The authors provide results of contribution (in percent) of different areas to the total isoprene emissions. Please provide information if these numbers are area normalized or if those emissions are absolute numbers. In the ladder, it is not surprising that the Southern Hemisphere contributes more to the total isoprene flux than the Northern Hemisphere does, just because of a larger oceanic surface area.

ll.394: “The emission rates in coastal areas…larger…by several orders of magnitude.” Please provide numbers, as this are results from your work. Also, I do not see this difference in the described Figure 6.

ll.398: Description of SML emission results is missing.

ll.408: The Arctic Ocean shows an increasing trend of isoprene emissions which is different to the Pacific or Indian Ocean. Perhaps the authors could discussion this within the context of sea ice retreat? Could that be a reason?

Figure 7: The Atlantic trends are shown in Figure 7c. However, this subfigure is not discussed in the text.

ll.424: correlation plots. What temporal resolution of datasets were used to perform the correlation calculations per each grid cell? Hourly, daily, monthly data? This information is missing in the text.

ll.437: “These two physical factors…show contrasting correlations”. The is an issue which should be discussed in the manuscript. How do these factors (SST and wind) influence the different emission modules (BIO and SML)?

ll.440: It is not clear to me what the authors mean with “large-scale air-sea system” within this context. Perhaps they can be a more specific.

The paragraph “Conclusions and Perspective” should be changed to “Summary” if the content stays as is.

References

Brüggemann, M., Hayeck, N. & George, C. Interfacial photochemistry at the ocean surface is a global source of organic vapors and aerosols. Nat Commun 9, 2101 (2018). https://doi.org/10.1038/s41467-018-04528-7

Ciuraru, R., Fine, L., van Pinxteren, M., D’Anna, B., Herrmann, H., & George, C. (2015a). Photosensitized production of functionalized and unsaturated organic compounds at the air‐sea interface. Scientific Reports,5(1), 12741. https://doi.org/10.1038/srep12741

Ciuraru, R., Fine, L., van Pinxteren, M., D’Anna, B., Herrmann, H., & George, C. (2015b). Unravelling new processes at interfaces: Photochemical isoprene production at the sea surface. Environmental Science & Technology,49(22), 13,199–13,205. https://doi.org/10.1021/acs.est.5b02388

Conte, L., Szopa, S., Aumont, O., Gros,V., & Bopp, L. (2020). Sources and sinks of isoprene in the global open ocean: Simulated patterns and emissions to the atmosphere. Journal of Geophysical Research: Oceans,125, e2019JC015946.https://doi.org/10.1029/2019JC015946

Palmer, P. I., and Shaw, S. L. (2005), Quantifying global marine isoprene fluxes using MODIS chlorophyll observations, Geophys. Res. Lett., 32, L09805, doi:10.1029/2005GL022592.

Simó, R., Cortés-Greus, P., Rodríguez-Ros, P. et al. Substantial loss of isoprene in the surface ocean due to chemical and biological consumption. Commun Earth Environ 3, 20 (2022). https://doi.org/10.1038/s43247-022-00352-6

I have included my comments in the attached PDF.