The paper presents and evaluates an updated version of the marine biogeochemical model ECOSMO II, that now also includes a parameterisation for a variable C:Chl ratio of phytoplankton. The biogeochemical is coupled to the HYCOM ocean model, configured for the Arctic Ocean. Model evaluation of a default setup is carried out against observed nutrients and Chl a from in situ measurements and remote sensing, and complemented by comparison to literature values of primary production. Model sensitivity to biogeochemical parameters is evaluated in two further experiments. In general, all model configurations perform quite well with regard to the observations, and improvement in surface chlorophyll when parameters are adapted to the hydrodynamics of the Arctic.

In general, the paper is well written and I very much appreciate the thorough comparison to observed tracers. One drawback is the - to my eyes - somewhat incomplete description of phytoplankton (see (1) below), and the slightly confusing description of how observations were used for comparison (2). I further encourage the authors to improve on the structure of the paper (3) and to extend a bit on the comparison of the three different experiments (see my comment no 4 below).

(1) For better understanding of how phytoplankton responds to nutrient and light limitation I suggest to present the equation(s) for phi (in Eqns 1 and 2). Without knowledge about this functional form it is difficult for the reader to understand its dynamics, in particular as DS2013 might not be accessible to everyone. Further, does the C:Chl ratio (or the amount of chlorophyll in phytoplankton) affect the light attenuation and P-I curve, i.e. the of phytoplankton growth rate? 

(2) The description of type of observations, and how these were used for model evaluation (section 4.1) is somewhat confusing (see also below, specific comments).

(3) Some sentences of the model setup (section 3) already anticipate the results (e.g., lines 183-186: "Using lower ... column stabilizes." and 189-191: During the continuous ... will replace the parameterization set in EXP2."). I would suggest to reorganise the paper a bit (see also comment (4)), and draw conclusions only from the model evaluation.

(4) I enjoyed reading the analysis in section 4.3. I think this analysis could be complemented by constrasting it with the results obtained with EXP2 and EXP3. For example, Table 2 shows that EXP1 always performs best with regard to the correlation coefficient, in 7 (8) cases out of 12 with regard to the normalised StdDev (Bias%), and in half of the cases with regard to RMSE. Likewise, it performs always best (regardless of metric) for phosphate and nitrate, and in half of the cases with regard to chlorophyll. Even if some differences to EXP2 and EXP3 are only small, I think this could be discussed a bit more (so far, there seems only be a sentence in line 451), and contrasted with the outcome presented in Tables 4 and 5, which indicate that EXP3 performs best with regard to surface Chl a. What is the reason for these differences? Is it because Table 4 and 5 only refer to surface values?  

Specific comments:

Line 36: Here I would prefer a more specific link to a model application at copernicus (I had to search around a bit).

Line 49: "introduces"

Line 60f: "(...) therefore there is already some zooplankton biomass present at the time of the start of spring." - Only if they ascend early enough, correct? perhaps add: “... and ascend to the surface at the onset of spring”?

Line 65: "Glacialis" should be lower case

Line 71: "description"

Line 89: "use the time stepping" - What time step lengths were applied?

Line 103: "matters" - Perhaps better: "components"?

Line 105: "discrete nutrients" - What are "discrete nutrients"? Skip "discrete"?

Line 105: " the a" - skip "the"

Eqn 1: Number seems to be missing.

Eqn 2: please show how phi is calculated

Line 172: Evaluation after just two years of spinup seems to be quite soon - do you have any indication if the model drift (with regard to the BGC components) has decreased to some specific value?

Line 188: better: "parameter sets" (?)

Lines 197 and 199: "GOTM": Please explain and reference the abbreviation before it is used first.

Line 201: ""from Station-M is" - "are"

Line 201-202: "The dynamics shown in Appendix A1 is expected to be valid for each model point." - How is this expectation justified?

Line 211: "for model assessment is" - "are"

Line 214 "the model's use as a tool" - perhaps better: "the model as a tool"?

Line 220-221: "Nitrate, silicate, phosphate and chlorophyll a in situ data from Institute of Marine Research (2018) were used for the statistical evaluation of the model results." Where were these data used and what is the difference to the comparison against WOA2013 data?

Line 226: "were used for evaluation" - of simulated Chl a?

Line 229 "Processed model chlorophyll" - What does "processed" mean in this context?

Line 230: "OC-CCI chlorophyll a" - I can guess what this is, but provide the abbreviation above, when citing Sathyendrant et al.

Figure 3 caption and elsewhere: "WOA18" - above you refer to WOA2013

Figure 3 caption: "EXP1: solid ..." - But in the figure the different experiments are indicated by different colours.

Line 278: "0.6-0.72" - Table 2 gives 0.79-0.83 for EXP1: confusion with  EXP3? 

Line 279ff: For the biases Table 2 gives the relative bias, but in the text you refer to absolute biases, which is somewhat confusing. 

Line 280: "2.47-3.34" - this seems to be the bias of EXP3.

Line 285: skip one "Fig."

Line 290: "0.74-0.78" - again, for EXP3?

Line 292: "modeled upper 100 m maximum" - please rephrase

Line 296: "The model code" - this sounds as if the model does it all by itself, I would rephrase this e.g. to "in the model uptake of silicate is limited to concentrations above"

Line 300: "0.81-0.89" - again EXP3?

Line 311: "0.97-1.2"  - EXP3?

Line 315: perhaps better: "chlorophyll a is always below 8mg ..."

Line 383-383: "silicate concentrations ... in the climatological data." - are discharge rates into the Arctic very different for different types of nutrients? (I.e., outside the assumed stoichometry)? Perhaps a note on this would be interesting ...

Line 425: "EXP1 is the least representative of the in situ data" - of Chl a, I assume? Because Table suggests that EXP1 represents in situ data of nutrients best.

Lines 443-446: It is not really clear to me what you want to say with this sentence - could it be rephrased?

Line 447: "overestimate" - perhaps better: "... curve leads to an overestimate of the bloom"?  

Line 451: "EXPeriment statistics for inorganic nutrients are very similar in all experiments (Table 2)." - Overall, to me it seems as if EXP1 performs better than EXP2 and EXP3 (see above my comment no. (4))

Lines 478-479: "using realistic ... spin-up period" - what are the realistic constant values? and: constant with respect to what - over time? over depth?

The paper presents a description and evaluation of a new configuration of the biogeochemical model ECOSMO II (ECOSMO II(CHL)) in which they have included chl in the three functional phytoplankton types as prognostic variables. Furthermore, the paper presents a comparison of model experiments using three different parameter sets. In the appendix, the authors have also included a comparison between ECOSMO II and ECOSMO II(CHL) in a configuration where the models are coupled to the 1d physical model GOTM.

The paper is well written and the methodology ambitious. However, I would like to have seen a larger focus on the comparison between ECOSMO II and ECOSMO II(CHL). Furthermore, apart from two sentences (451-453) I find no discussion on why EXP1 seems to perform better for nitrate and phosphate (Fig  3 and Table 2) than EXP 2 and 3. I find that the paper could be published in GMD after some minor revisions.

Specific comments

Lines 123-126: F (Eq (4)) is not explained. 

Lines 139-144. What limited the concentrations from becoming too small before you added this? Also, you state that: “The minimum concentration at which the loss terms are calculated are…”. Should it say: “ The minimum concentration at which the loss terms are switched off”?

Fig.3: It is not obvious to me that EXP2 and 3 perform better than EXP1. It looks the opposite from this figure.

Line 320: Is it not the net primary production your extracting? You don’t explicitly model respiration so what you model is gross primary production (photosynthesis) minus respiration i.e. the net?

Line 350: Please insert full names of functional types and their abbreviations in the caption.

Lines 388-389: What is the reason for the difference in N/P ratios in the Barents sea between WOA and model, I guess mostly as a result of the difference in nitrate in this area?

Fig. A1: It would be interesting to see a plot comparing ECOSMOII and ECOSMOII(CHL) to observations so as to get a clearer view of the benefits of variable C:Chl ratios.

Technical comments:

Line 105: “...the a single sediment layer” remove “a”.

Line 112: “...addition on” replace “on” with “to”

Line 122, Eq. (4): Specify index of summation.

Lines 241-242, Eqs (8) and (9). Add index of summation.

Line 265: Fig3: Change text to explain line colors. It now states: solid, dashed, dotted.

Line 258: Remove one “Fig.” from reference to figure 4

Narrative

The manuscript describes and evaluates the coupled HYCOM-biogeochemistry ECOSMO II model, configured for the Arctic Ocean. Its evaluation includes some suggestions for fine-tuning of the parameterization according to the geographical characteristics of the expanded domain.  The technical description of the ECOSMO II model is well written and constitutes a good baseline for the modeling community. The evaluation of the model results, against observational values, is also of value for a quantitative validation of the model and it is well complemented by the commentary and the figures therein.

The paper is very well written, but it would benefit, in my opinion, by a better and more organized description of what observations were used in what experiments. The selection of the areas (and experiments) to show and compare seems ad-hoc. What is the justification for these selections? Why not compare all of the areas in all comparisons and statistics? Seems that it wouldn’t add much more content and would be of better use in a comprehensive descriptive assessment of the model for this region at large. Perhaps a way to scope the comparative analysis would be to concentrate on one or two of the subregions of the model domain. The performance comparison for the model would be well prescribed in a table of EXPT vs. AREA with some objectively-derived performance metric value (i.e. rmse/skill score). This would allow for a good summary of the many combinations of experiments and areas covered. The many regions, very comprehensive, somewhat dilutes the detailed validation of the model itself, only providing a general descriptive validation with some statistical quantification for some of the areas.

While technically, this is a paper of excellent value, scientifically, it could have been further elaborated to elucidate what the model brings in terms of better understanding of the hydrodynamics and biogeochemistry of the Arctic region. It would also benefit from a more detailed description, via equations, diagrams, and text, of the plankton group dynamics and again, the scientific value of those results.

In summary, the paper is very well written and the technical topic is well covered, but the scientific value of the paper could be improved with more details on the model parameterization, inherent errors therein, and the “take home” message of what is learned scientifically.

Line-item comments

• There is still some “proof reading” needed to correct typos, inconsistencies, etc. throughout the paper, figures, equations. Some equations are not full described (e.g. Eq. 4) and some figures could use further clarity in terms of colors and legends. Following are mostly such items:
• Line 11: “online” -- shouldn’t that be “inline” -- it is a more common way to describe coupling “inline or offline”.
• Line 38 (entire paragraph could use rewording -- better “the transfer of the ECOSMO II model to a different…)
• Line 45: Comma after which, no comma after that ( consistency throughout the paper for easier reading and clarity).
• Line 71: “description”
• An introductory atlas-like figure (indicate where Station-M is, etc.) of the domain, with a temperature or density climatology background or similar, will benefit the introduction -- perhaps in combination with Figure 2 or complementary to Figure 2.
• Line 229: Not clear - reword.
• Line 271: “The consistent late spring…” needs rewording or a comma (before and) or separate into 2 sentences.
• Figure 3 and Figure 4 seems redundant for the fields that ovelap. Figure 3 may benefit from a ~surface-100m integrated quantity and will cut the redundancy and give a better assessment of the water column constituency.
• Line 285: Fig.Fig repeated
• Line 293: …maxima are…
• Line 313:… concentration ranges…
• Line 315: …the lower nstd is underestimating…
• Line 340: … “. The model…”
• Line 341: “primary production shows”
• Line 348: … “. The southern part…”