Review of "A Novel Approach to Estimate Carbon and Nitrogen Flux from In Situ Optics: Application to Cyclonic Eddies off the Cape Verde Islands" by Moradi and co-authors.

The authors describe field observations of sinking particle fluxes and its modeling from particle imagery using data from the Cape Verde Islands both inside and outside of cyclonic eddies. They introduce two methods for modeling sinking particle fluxes from in situ particle imagery to model POC and PON fluxes and their characteristics. They show a two-fold increase in flux inside cyclonic eddy compared with sites outside.  

Several issues preclude the publication of this paper at this time. First, there seems to be a significant error in the formulation of the model used in their second method for assessing sinking POC/PON fluxes. This issue must be reconciled before this paper can be considered for publication. I also have a couple suggestions about the first method as well. Second, the paper ignores uncertainty in the modeling performed, making it impossible to tell how robust any of the methods are. Third, I was unclear of the paper's intent. Is it a methods paper or is it about the roles of eddies on sinking particle fluxes? This needed to made clear from the onset (title, abstract and intro) and this was not accomplished. Last, the paper needed a good deal of editorial help before it was submitted. There were many writing and presentation issues that made this paper very hard to review. I do wish I could be more positive in this recommendation as I think there are several interesting analyses presented.

First, there were several methodological issues and omissions that need to be reconciled and missing analyses need to be included. Importantly, there appears to be a mistake in the formulation of the second method for assessing sinking POC/PON fluxes. Equations 4 & 5 (and the text before that) all state that the carbon content at depth should decreases linearly in time due to remineralization processes, implying that at some time the CarbonMass would be negative. This is incorrect. Typically, the loss of mass due to a first order decay process is modeled as...

            CarbonMass = InitialCarbonMass * exp(-DecayRate * Time)

This is fundamental and the analysis presented needs to be entirely reworked. After realizing that there is an important error, it seemed pointless to spend more time addressing the results from this method.  

Further, the comparison of the UVP and gel trap spectra could be very interesting. However, all what is presented are the one-to-one coefficients from the fits (Figure 7), and it is unclear what this represents. More analysis is needed here and it would be easy. For example, it would be good to know if the concentration of sinking particles seen by the gel traps is always less than the suspended seen by the UVP (taking into account uncertainties in the method too!). Also, what is the fraction of sinking to suspended particles, what its dependence on size would be, etc. These obvious results are missing and should be included in a revised manuscript.

Second, the paper ignores uncertainty in the modeling performed, making it impossible to tell how robust any of the methods are. The model parameters are developed using least squares fits and it is straight forward to estimate their confidence levels. All through the paper, the model fit parameters are reported to the 5th or some 6th digit, which is way more precision than Any revision of this paper needs uncertainties properly calculated and reported AND their propagation to final calculations.

Last, the paper was very difficult to read and review. Part of it was overall construction. The paper does not lead the reviewer through the story effectively. Not sure what the point was. The introduction should lay that path and that was a mess. Some of that was in the discussion. I found myself guessing what the point of the paper was. At the next level down, paragraphs need to focus on a single topic with a useful topic sentence that describe the goal of the paragraph. What was the hardest was the lack of polish on the paper. Many things were hard to figure out. For example, the acronym "SDPF" is used multiple times starting on page 7 before it is defined on page 11. There are many other examples that made this tough to review. Again, I wish I could be more positive in this recommendation, but the paper needs some serious attention to details that I know the authors are capable of.

Title: "A Novel Approach to Estimate Carbon and Nitrogen Flux from In Situ Optics: Application to Cyclonic Eddies off the Cape Verde Islands"

Authors: Moradi et al.

This is an interesting study of cyclonic eddies off the coast of the Cape Verde Islands, where the authors show an increase in particulate flux within the core of the eddy compared with outside.  This has implications for the flux of particles (and by inference Carbon and Nitrogen) within the global ocean where hotspots of particle sinking may be remaining unobserved (certainly) or unaccounted for (possibly) in our estimations of key elements within the Earth System.  These type of "funnel" features have been observed before, on a seasonal basis (e.g. Dall'Olmo and Mork, 2014).

The use of this type of approach allows the flux of particles to be captured in real-time (i.e. as it is happening) rather than being an integral over several days / weeks / months / years as is the case for sediment traps.

The methods themselves involve several steps and assumptions.  This includes assumptions about the sinking velocity of particles (as a function of size) which, as the authors show, in Figure 5 only explain 11% of the variance.  This isn't necessarily unusual in this field, as is shown by the intercomparison with other works (Cael et al.).  The other assumptions regard the conversion of particle diameter to carbon content (which they start off using a power law relationship, which then gets modulated through the water column by physics and chemistry (temperature and oxygen)).  The authors use an optimisation step for Acar and Bcar.  The PON content is then related to the POC content through another factor gamma.

It is here that the authors need to be clear.  They are estimating the POC and PON flux based on particle size spectra without any recourse to measuring POC or PON.  Anywhere that it is stated that these are observations, I think requires re-writing.  Also, there needs to be a thorough analysis done to determine uncertainties at every step.  This should be relatively straightforward - given they draw on the work of Guidi et al. (2008), and build on it, the use of the standard deviations (errors) should be invoked here to cascade down through the calculation chain.

Overall some of the reasoning was difficult to follow, and using non-standard nomenclature for things like number (#) should be changed.  At times the authors used in-paragraph equations and other times they were separated out of the text.