The authors investigated the dynamics of dissolved inorganic carbon in the deep convection area of the North-West Mediterranean Sea. The study was based on a good coupling between observations from mooring sites and cruises and 3 D coupled physical-biogeochemical model. 

The main findings were that the area:

-was a moderate sink of CO2 (0.47 mol C m^-2 yr^-1) with an increase during the spring phytoplankton bloom, the air sea flux represented only 12% of net community production in the upper lever of the water column;

- both biological processes and physical transport (vertical and horizontal) played a dominant role in the annual DIC budge;

- winter ventilation had a reducing effect of on the atmospheric CO₂ uptake;

- the region acted as a source of DIC for surface and intermediate waters.

Overall the approach is innovative and the results are relevant for a better understanding of the CO₂ system dynamics in the NW Mediterranean Sea.

I think that the discussion could be improved by a deeper comparison with one of the few other areas of the Mediterranean Sea where deep convection occurs as the Southern Adriatic Sea.  The Adriatic Dense Water formation plays an important role for the sequestration and storage of the anthropogenic carbon, as the anthropogenic CO₂ is transferred in the deep waters of the Eastern Mediterranean (Krasakopoulou et al., Deep Sea Res., 2011; Cantoni et al. Mar. Geol. 2016; Ingrosso et al. Deep Sea Res., 2017)

In the Chapter 5.5 “Contribution of north-western deep convection region to the carbon budget of the Mediterranean Sea” the discussion could be improved by taking into account not only the modelling studies but also the experimental studies showing that the Adriatic continental platform acts as a sink for atmospheric CO₂ (e. g.:  Turk et al., Jour. Geophys. Res., 2010; Cantoni et al., Est. Coast Shelf Sci.,2012; Catalano et al., Jour. Geophys. Res., 2014; Urbini et. al., Front. Mar. Sci., 2020).

In the sensitivity tests including the carbonate production the authors used a PIC/POC ratio of 0.5 but according to the results reported in the cited paper of Miquel et al. (2011) the ratio is subject to wide interannual variations ranging from 0.31 to 0.78.  It would be important to know how these natural variations would affect the sensitivity tests.

In the conclusion the authors state that the air -sea flux represents only 13% of the upper column Net Community Production (NCP) whereas in the chapter 5.4 that state that the flux represent 12% of NCP. The discrepancy should be solved.

In the conclusion the authors states that the physical fluxes in the upper layer is of 3.3 mol C m^-2 yr^-1 but in the figure 12 the difference between the lateral DIC transport and the vertical DIC transfer amounts to 3.4 mol C m^-2 yr^-1. The data should be checked.

The authors in the conclusion more clearly the in the discussion (L.547-552) state that calcification processes could lead to an underestimation by 23-58% of the annual uptake but the authors should take into account that the calcification processes although reducing the TCO₂ will increase the pCO₂ in seawater therefore counteracting the CO₂ intake from the atmosphere.

The authors use the terms “biogeochemical flow “and “physical flow”  which are not very appropriate terms as both are related to a mass flow of carbon generated by biological processes or by physical processes (advection, mixing, particle settling). I suggest to find a more appropriate alternative term e.g.:“physical transport”.

Specific comments

L. 49-50, “is one of the region where deep convection occurs” a specific reference to the Southern Adriatic SAD should be added.

L.  370-37, “upward flux of DIC into the upper layer of 41.40 mol C m^-2…” The units of a mass flux should be used. They should be expressed as the mass of carbon that passes through a defined cross-sectional area over a period of time.

L. 390- 395, L. 469-470; L. 529. same as above.

L. 452, “the DIC drawdown due to biological processes decreases and net DIC production events took place”: could the authors specify which are the processes driving the DIC production events.

L. 465 “an annual consumption of 3.7 mol C m^-2 of DIC”: the unit of time is lacking.

L.549-552. This sentence is not clear and the CO₂ production during calcification should be taken into account.

L. 578 physical flow? Do the authors mean physical transport?

L. 589 DIC exchange flows? Do you mean DIC flows?

Fig. 7. The units for fluxes are expressed as an inventory: mol C m^-2. The mass fluxes should be expressed as the mass of carbon that passes through a defined cross-sectional area over a period of time e.g. mol C m⁻² y⁻¹.

Fig. 12. The data represented are inventories or fluxes in the latter case they should be expressed as the mass of carbon that passes through a defined cross-sectional area over a period of time e.g. mol C m⁻² y⁻¹.

Please see the attched document for my comments.