Review of the paper “Modeling Mediterranean ocean biogeochemistry of the Last Glacial Maximum”.

 General comments

 I have read the manuscript “Modeling Mediterranean ocean biogeochemistry of the Last Glacial Maximum” by Six et al. 2024. The manuscript falls in the scope of the journal and it provides the a modelling effort of the biogeochemical dynamics of the Med Sea during the last Glacial Maximum. I found the work very interesting, very well written with only minor issues to be addressed before the final acceptance. I think that the manuscript should undergo minor revisions before its eventual acceptance.

 Comments

 Line 23: I would say “was characterized”…then “is characterized”

Line 29: “Fresher” then “Less saline”

Line 32:  Probably I missed the point in the sentence and thus I’m asking for some additional clarification. As far as I remember the formation of LIW is driven by the salinity. The Authors states that in response to the reduction of the inflow at Gibraltar there is an increase in the salinity in the east Med.. thus I would expect an increase in the LIW formation not a reduction..Could you please explain better this point?

Line 79: I would say “coupled” than “combined”

Line 93 onwards: How many biogeochemical tracers considers HAMOCC? How do you represent the phytoplankton in your model? As functional group? Please explain better and provide more information about

Line 128: How do you consider the period of 1000 year as suitable for the spin up? Did you check the kinetic energy? What about drifts in the tracers?

Line 133: what about the zenith and albedo?

Line 145 onwards: The Authors use an open boundary at Gibraltar Strait but they do not impose any speed. Could you please why? Is you E-P+R balance over the Mediterranean sea equal to zero or did you apply any correction to preserve the total volume of the domain? Please explain

Line 185: “precipitation”

Figure 2: As far as I remember Medar medatlas also provides the uncertainties for each variable and for each depth. I would include them in the figure as error bars to see if they overlap the values simulated by the model. Did you have an explanation for the difference in the oxygen vertical profiles with depth? Respiration processes?

 Line 203 onwards: There are some recent results related to the Med ZOC simulated and discussed in Reale et al., 2022. I would include them here. Reale, M., Cossarini, G., Lazzari, P., Lovato, T., Bolzon, G., Masina, S., Solidoro, C., and Salon, S.: Acidification, deoxygenation, and nutrient and biomass declines in a warming Mediterranean Sea, Biogeosciences, 19, 4035–4065, https://doi.org/10.5194/bg-19-4035-2022, 2022.

Line 204: All the listed areas are characterized by deep water formation processes. The Rhodes gyre is the area for the intermediate waters. Please correct

 Line 209 onwards:  I would provide more quantitative information here about the differences between simulated and observed values. The Authors could use for example for the NPP the table 4 in Reale et al., 2020 Reale, M., Giorgi, F., Solidoro, C., Di Biagio, V., Di Sante, F., & Mariotti, L., et al. (2020). The regional Earth system Model RegCM-ES: Evaluation of the Mediterranean climate and marine biogeochemistry. 

Line 220 onwards:  10-20 % is not slightly higher. Please correct

Figure 7 and Paragraph 4 The authors talk about anomalies and biases. As far as I see they are differences. I would call them like that. I would test if these differences are statistically significant and I would mark the map with a dot where this happens.

Line 262-264: Could you explain better this point since it is not very clear

Paragraph 4.1 and 4.2 as far as I understand from the results the SST gets colder and thus I would expect a decrease in the vertical stability and increase in deep water formation processes. Could you please explain better why happens the opposite?

The authors reported original results using a regional physical-biogeochemical ocean model to study the Mediterranean Sea circulation and biogeochemical cycle during the LGM. The regional model was forced by the data of Earth System model taking into account topography and bathymetry changes, which is highly appreciated. Several interesting findings were shown, in particular the sensitivity test attesting the importance of the sill depth in the Straits of Gibraltar and Sicily that explained a large part of sluggish Mediterranean circulation at the LGM. The authors succeeded in reconciling apparent contradiction of proxy reconstruction showing reduced primary production in the surface and increased organic deposition on seafloor during the LGM by the interplay of water column stratification, zonal advection and slower respiration in glacial cold waters.  Also, the potential influence of seasonality changes during the LGM on paleo-temperature reconstruction was discussed.

I enjoyed reading this manuscript. The scientific theme fits perfectly with the topics treated by Climate of the Past and this is the first attempt of simulation of biogeochemical cycle in the Mediterranean Sea under glacial conditions. I wish to see this work published on Climate of the Past.

The manuscript is well structured with clear figures and tables. I am not a modeller and I have only limited comments / suggestions.

My first point is scarce comparison between proxy reconstruction and simulation results. Only SST reconstruction based on planktonic foraminiferal assemblages (Hayes et al., 2005) was used for comparison (Figs 8 and 9). The authors mentioned possible bias of reconstruction due to seasonality changes at the LGM taking alkenone as an example but the comparison between model results and UK’37-SST is not presented except for Fig. 9. I would suggest to add a map presenting the comparison using the data provided by Cacho et al. (2002) and Essallami et al. (2007) by updating the MARGO database used in the present manuscript. It will be interesting to examine whether the sites with large SST offset between UK’37-SST and delta T32 correspond to the areas of marked delta peakiness (Fig. 12).

The second point is about remineralization rate of organic matter and dissolved oxygen concentration. Slower remineralization rate in glacial cold water is proposed to increase transfer efficiency of detritus (Teff). Together with cold sea surface that increase oxygen solubility, generally high dissolved oxygen is expected but some cancelation due to sluggish ventilation is possible. Therefore, the distribution of dissolved oxygen concentration is highly interesting but only vertical profiles from the selected areas are shown in Fig. 2.  I would like to see transect of dissolved oxygen under different simulation scenarios at least in supplementary information.

I would suggest to accept this manuscript after minor revision.

Minor/specific comments

Line 28 and throughout the text. “Gibraltar” would be replaced by “the Strait of Gibraltar”.

Line 28. “ZOC”. Do you mean “zonal overturning circulation”?

Line 33. “Proxy data from foraminiferal shells” should be replaced by “Water mass proxy data recorded in foraminiferal authigenic fraction”.

Lines 155-156, about the nutrient supply from shelf areas during glacial sea level low stands. Was the similar effect considered inside of the Med Sea?

Line 160. About the use of the present-day nutrient concentration for the LGM simulation. I understand the reasoning of the authors but some sensitivity tests by modifying nutrient concentration will be interesting to examine the robustness of the results.

Lines 185-186, “PEM = preciptation + river runoff - evaporation”. The first letter of the words does not correspond to the abbreviation.

Line 187 and throughout the text, “1000 m3 s-1”. The authors may use always “Sv” or “mSv” because small fluxes like +0.06-0.076 Sv and +0.04-0.11Sv (line 194) are shown with Sv unit.

Lines 206-207, “slightly weaker...other modelling studies”. Add the range of flux obtained by other studies to be more precise.

Line 212, “Nile damming”. Do you mean “the Aswan High Dam in 1964”? Please revise.

Lines 233-234, “Hamann et al., 2008” is missing in the list of reference. The authors may use the map provided by Venkatarathnam and Ryan (1971) that presents a detailed distribution of calcium carbonate in the eastern basin sediments.

Line 247, “Kuhlemann et al., 2008”. The data provided by this reference is not used for the data model-comparison. Please revise.

Line 253. The authors may cite Fig. 8 in addition to Fig. 9.

Line 260, “Fig. 7d”. Isn’t it “Fig. 7b”?

Lines 304-305. Fig. A1e does not exist. Please revise.

lines 316-317. Why is phosphate concentration higher in the Alboran Sea during the LGM? Is this because of enhanced river discharge?

Lines 395-396, about the agreement between high organic matter accumulation and increased abundance of benthic foraminifera sensitive to food supply during the LGM. A map presenting simulated organic matter content in sediment (Fig. 6) and changes benthic foraminifera abundance will be interesting to show the trend.

Line 451, “can not” should be replaced by “cannot”.

Fig. 1. It will be helpful to add latitudes and longitudes scale since the white line indicates the zonal transect is used Fig. 13 that is presented with longitudinal scale.

Fig. 4. I am curious to see the result of zonal stream function of PI-Straits at least in supplementary information.

References

Cacho, I., Grimalt, J. O., and Canals, M.: Response of the Western Mediterranean Sea to rapid climatic variability during the last 50,000 years: a molecular biomarker approach, J. MARINE SYST., 33–34, 253-272, 2002.

Essallami, L., Sicre, M. A., Kallel, N., Labeyrie, L., and Siani, G.: Hydrological changes in the Mediterranean Sea over the last 30,000 years, Geochem. Geophys. Geosyst., 8, Q07002, 2007.

Venkatarathnam, K. and Ryan, W. B. F.: Dispersal patterns of clay minerals in the sediments of the eastern Mediterranean Sea, Marine Geology, 11, 261-282, 1971.