Dear Dr. José,

First, I’d like to thank you for submitting your work to Biogeosciences and I’d like to apologize for the delay to come back to you after the submission of your responses.

Based on my reading of your responses to the referees’ comments, you give me the impression that you may be able to address most of these comments. In particular, I assume you will now include a proper discussion, and a detailled analysis and discussion of the advective fluxes of nitrate and nitrite in both eddies.

Incidently, it seems that you did submit a revised version of your manuscript as an addtional comment. I would encourage you to submit this revised version following the proper procedure as explained. This revised version will probably undergo further reviews with the same referees.

With best regards,
Laurent Bopp
Associate Editor

Dear Yonss,
First, I’d like to thank you for the thorough revision of your manuscript. In particular, I appreciate the fact that you now include some detailed analysis of advective fluxes at the eddies boundaries, and some additional particle-release experiments, as suggested by the reviewers. I also noted that you re-run simulations with adjusted parameters.
While one of the original referees is now fully satisfied, the second one still sees a few shortcomings in the revised version. He/She has some concerns about the separation of advective and local processes in explaining the nutrient distribution in the eddies, as well as some questions about the use of the Lagrangian experiments.
Based on my own reading, I can say that your manuscript has considerably improved and I would be ready to accept it for publication if you can respond to the above-mentioned criticisms in a rebuttal. I would also suggest that you take the opportunity of this last round to respond to the other minor comments raised by the reviewer (see below).
I hope you’ll be able to take these comments on board. Again, thanks for submitting your best work to Biogeosciences,
Sincerely yours,
Laurent Bopp


--- Specific comments by the reviewer ---
Specific Comments from the Reviewer

P4 L3 “However, the magnitude of simulated nitrite is lower compared to the observations.”
Is this sentence for the inshore nutrients?

P5 L7 “Despite reproducing the subsurface nitrite maxima, the simulated concentrations are an order of magnitude lower compared to the concentration within the observed eddy”
Which observation is this? Is it in-situ observation?

P6 L6 “and and”
Remove one of two “and”.

P6 L29 “These results show that the low nitrate (high nitrite) within the eddy Asim is not primarily controlled by local biogeochemical dynamics within the eddy Asim”
What missing in the presented analysis in Figure 7 and 9 is that there is no information about the time evolution of the eddy volume. In addition, as the eddy travels, some portions of the water can be transported as the eddy flux depending on the nonlinearity i.e., u/c, where u is the current velocity of an eddy and c traveling speed of an eddy, and some water is not. This is problematic. If the volume of water, which the authors tracked for the analysis in Figure 7 and 9 is all trapped within the eddy, all variations in tracers must be controlled by the local processes. In contrast, if the eddy is a completely linear wave, and no water is fluxed by the eddy, the biogeochemical processes within the tracked eddy (wave) is all nonlocal processes. Instead, it is just looking at different water column as the wave propagates. In reality, as authors pointed, some water is held inside the eddy and some not. In summary, it is very difficult to say if it is caused by the local processes or not, if the volume of the water is varied as the eddy evolves.

Figure 7 Caption “Bsim” should be “Asim”.

P7 L13 “In the eastern tropical ocean, eddies have been observed to carry low-oxygen waters out of the core region of the OMZ”
Is this applicable for only anticyclonic eddies or both cyclonic and anticyclonic eddies?

Figure 10 and 12, and related text
The unit of the flux is not for flux. It is flux divergence.
Similar to the comment for Figure 7 and 9, time variation of the volume tracked is needed.

Figure 11 and 13 and related text
These results are rather descriptive hydrography in the numerical model, and sounds more like discussion as the authors speculate the importance of the advection just by the T-S diagrams. If water properties are horizontally (along isopycnally) uniform, T-S properties are unchanged by the lateral (along isopycnall) advection. These water exchanges could be addressed more quantitatively by the Lagrangian particle tracking.


P8 L20 “(Fig. 11-e-f).n)”
There is no panel n.

P9 L7 “Subsurface anticyclonic eddies…”
I did not realized a fact that this study focused on the subsurface anticyclonic eddies. If so, these previous studies for mode water eddies or under current eddies should be in the introduction. If the study is focused on the subsurface eddies, the detection method needs to be modified for them.

P9L27 “similar the observed” should be “similar to the observed”.

P8 3.3 Eddy stirring and nutrient entrainment
It is good that the authors conducted a Lagrangian simulations. The results would show importance of the water exchanges, but are not clear. This is because these entrained particles do not necessarily represent the main body of the water inside the eddy. It is better if the particles are back tracked from the inside eddy with different depths. This is more direct to know where the water comes from.

Figure 14b includes the number of particles inside Asim and shows that it is not largely changing compared to that of other eddies. This is not consistent with the claim that there is strong advective inputs of nitrate into Asim.

P9 L27-33 “In the low nitrate (high nitrite) Asim eddy (Fig. 4-c), the on-going nitrate reduction by denitrification is lower than the nitrate production by nitrification. This fails to explain the low subsurface nitrate within the eddy, which is in contrast to the interpretation of the observations by Altabet et al. (2012) and Stramma et al. (2013). Further, we find in our simulation that the advective fluxes across the edge of the eddy Asim shows a strong nitrate supply into the eddy, in the first days prior to the eddy formation. This nitrate supply, which is predominately horizontal, is a consequence of the exchange of water masses with the surrounding environment during the eddy propagation.”

I’m confused by these sentences. The results indicate that local biogeochemical processes, which imply increase in nitrate, and the advective flux divergence also results in a net increase in nitrate. So both estimates cannot explain the low nitrate inside the eddy?

P10 L27- “The model results show a decoupling between local nitrate reduction (nitrite production) via biogeochemical processes and total changes in nitrate (nitrite) within the eddy”
As mentioned in my comments above, it is difficult to separate local and advective processes, because how much water is trapped affects the locality with respect to the moving eddy.

As it is clear in the conclusion, the results of this study could not successfully close the budget. The both local and advective processes do not explain the nitrate concentration within the eddy quantitatively.

P11 L1 “shaping the nutrient patterns within cyclonic eddies”
Where are the results for cyclonic eddies?