Sensitivity Analysis of a Regional Nutrient Budget Model for Two Regions with Intensive Livestock Farming in Korea
Round 1
Reviewer 1 Report
A brief summary
This paper contains relevant information of nutrient budgets in two Korean regions. It includes both a comparison of the budgets of the two regions as well as a sensitivity analyses (actually an uncertainty analysis). These aspects clearly fit within the scope of Sustainability. However, this is not very well presented and structured. This paper is still in a draft stage and sometimes hard to understand. Furthermore, the used methodology is poorly described. Therefore, I recommend the editor that a major revision is required.
General comments
This paper contains relevant information of nutrient budgets in two Korean regions. Unfortunately I found the paper difficult to read and to understand, which I will explain below.
- The research is presenting nutrient budgets for two regions, which is fine. But the reasoning for including those two regions is lacking. I am missing crucial background on the regions, which should preferably two contrasting regions. However, I have the impression that this is not the case.
- For performing a comparison is needed to present results also in kg/ha. Only than a comparison of between the two regions, and results from literature make sense.
Background on the model are published in Kim et al (2017), but difficult to access for the international community, therefor it is difficult and sometimes not clear what the background is of used formulations and data. The equations 2 in Table 2 are not clear and the used units cannot be correct. In N/P 1 resulted in % ton/ha (ton/ha × %) whereas N6 in g (ha × g/ha) etc. This becomes even more complicated by used rather uncommon descriptions of model parameters, e.g. “units of swine excreta” (excretion rates?) and “nitrogen coefficients of swine excreta” (nitrogen contents in excreta?). I suggest to look in international literature on how to address various terms of the nutrient balance.
- As far as I understood, this research did not cover a sensitivity analyse (all parameters are varied equally), but un uncertainty analysis (parameters are varied within their uncertainty range), usually a sensitivity analysis is preceding an uncertainty analysis in other to select the most relevant parameters for the uncertainty analyses. see eg. https://link.springer.com/chapter/10.1007/978-94-011-0962-8_27. Furthermore, there is a lot of literature available on uncertainty in nutrient balances at various spatial scales, but none of these are mentioned. See https://www.sciencedirect.com/science/article/pii/S0264837718313401#fig0025
and much others.
- The used method for the sensitivity analyses (in fact uncertainty analysis) is not well described. Did you performed a one-at-the time analysis or a Monte Carlo? How is the sensitivity calculated, as: (Xmax – Xmin)/Xdef?
- This is also true for presentation of the results, see also international literature on how to improve this (e.g. the above mentioned paper, but there are much others).
- My most important concern, however, is that background on the used parameter ranges is not given. Since the selection of parameters and the quantification of the parameter uncertainty is the most important step in an uncertainty analysis, this requires much more attention.
- The discussion should include a more in depth comparison with international literature.
- Finally, the paper is not very well organised. Parts of the M&M (on how the SA is performed) are presented in the results section (see specific comments).
In summary I recommend the following:
- The description of the model and used data must be improved.
- The used method of uncertainty analyses and details on quantifying parameter uncertainty must be clarified.
- Results must be presented in kg/ha and more emphasis should be put on the details of the two regions.
- The discussion should be made more sound by comparing the results with other studies.
Specific comments
l 2: Title: skip NIER-NB, mention that the study is focussing on regions, e.g. Uncertainty in nutrient budgets of two (contrasting?) Korean regions.
l 12: The addition ‘(hNS; excluding atmospheric nitrogen emissions)’ is confusing. What is meant is GNS-hNS. Please rephrase.
l 13: Not clear what ‘units of swine excreta” are. Be concise and use more commonly used (in international literature) descriptions.
l 14: Not clear what “nitrogen coefficients of swine ...” are. N contents in excretion? Be concise.
l 18: The addition “with sensitivities exceeding ±2 %” is confusing and not needed.
l 64: In Europe and in the CAPRI model animal manure is also recycled. Please reconsider.
l 70: What is LES?
l 75-l86: This paragraph belongs to M&M
l 96-99: The need for performing this study is not very well supported by this paragraph. That parameters are variable is rather trivial and that model results are sensitive to variability and uncertainty is also trivial. This must be made clear.
l 115-116: Be concise with using terms like ‘fertilizers’ and ‘manure’. They are used differently throughout the manuscript. The same is true for “composted solid manure” and “liquid manure”.
l 115: What is done with grazing manure?
l 118: Rephrase figure caption. Diagram is not only on nitrogen and excretion.
l 120-123: Provide more background on the include regions. It seems that the two regions are rather comparable. So what is the reason for including those two? I suggest to include two contrasting regions in order to invest the behaviour of the model for different input ranges.
l 136-186: The description of the model in combination with Table 2 is concise and difficult to understand. This should be completely rewritten.
l 187-122: “Estimation of nutrient budget model coefficients” is not an appropriate title. The section is (should be) about the quantification of the uncertainty model inputs and parameters. However, this section in combination Table 3 needs some improvement. It is not clear what the minimum and maximum values are and what their origin is. This should be included in de MS or if this is too much as supplementary material.
l 229: Reconsider this table. Check consistency of equations, units etc.
l 223: This section should be extended and combined with a revised and extended version of section 3.2.
l 230: Reconsider this table and provide the minimum and maximum values and make is consistent with Table 2 and text.
l 230: Why you are using a mixture of P2O5 and P? This is confusing, please reconsider.
l 230: Why didn’t you use P deposition as input?
l 239-252: Budgets should be presented in kg/ha as well
l 250: I presume is should be 9,963.
l 254: Figure should be improved and presented in kg/ha. I suggest to use more sound descriptions on X axes, e.g. Fertilizer (rather than N1), Liquid manure (rather than N2-1), etc.
l 269: Same as above.
l 270-285: Swap to M&M, rewrite and extend.
l 273-274: This equation is not clear. Please reconsider and provide more details.
Did you calculate: Fluctuation rate = (Xmax – Xsmin)/Xdef ?
And did you calculate this for each parameter separately (one-at-the-time) or did you varied all parameters at the same time (by performing Monte Carlo)?
I suggest to combine this with section 3.3 and swap to M&M.
l 317: Provide a rational on the ”sensitivities exceeding ±2 % criterion”, and introduce this in the M&M section.
l 333: This equation is not clear. Please reconsider and provide more details.
Did you calculate: Fluctuation rate = (PSmax – Psmin)/PSdef ?
And did you calculate this for each parameter separately (one-at-the-time) or did you varied all parameters at the same time (by performing Monte Carlo)?
I suggest to combine this with section 3.2 and swap to M&M.
l 360-etc. Include a more thoroughly comparison with international literature.
l 360-368: This per ha based comparison should be include in the results section.
l 446: This seems to be an interesting table, but is hard to understand. Please reconsider this. E.g. I do not understand the role of Max and Min in this context?
Author Response
We thank Reviewer #1 for his/her thorough comments on our manuscript. We are glad to learn that Reviewer #1 considers that “This paper contains relevant information of nutrient budgets in two Korean regions.” The comments by Reviewer #1 have helped improve the manuscript, we are very grateful for them.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Dear Authors,
I revised the manuscript “Sensitivity analysis of a nutrient budget model, NIER-NB” submitted to the Sustainability Journal. The paper is interesting. However, I have some concerns, which need to be addressed before considering for final publication.
General remarks:
1. Use units with a superscript "-1", e.g. kg ha-1.
2. Graphs - use square brackets on the vertical axis.
3. Use the equation numbering as recommended in the "sustainability-template.doc" file.
4. Modify the appearance of the tables to meet the requirements of the "sustainability-template.doc" file.
Introduction.
Write in a few sentences how this paper is related to sustainability and sustainable agriculture. Use this information in your conclusions.
Conclusions.
Point out the directions of development of agriculture in Korea in the aspect of sustainability.
Line 403-425. There are no references to the content of the text. Add them.
Line 478. Missing “Author Contributions”.
Author Response
We thank Reviewer #2 for his/her thorough comments on our manuscript. We are glad to learn that Reviewer #2 considers that “This paper contains relevant information of nutrient budgets in two Korean regions.” The comments by Reviewer #1 have helped improve the manuscript, we are very grateful for them.
Author Response File: Author Response.pdf
Reviewer 3 Report
General comments
Sensitivity analyses of nutrient budget (model NIER-NB) is a valuable model and it could be considered for publication in Sustainability as it related to local and regional nutrient security. However, there are several observations that need to be addressed before it could be published.
Specific comments
Line 10. How does the word “survey” fit here, and what is it? It is a weird sentencing
Line 12. Atmospheric N emissions are not part of the hydrospheric or lithospheric N surplus anyway
Line 9-23. It looks like the authors assume that all the readers might have reviewed all of their previous publications when they not even referred to those. That is why the abstract is built upon the earlier publications without even briefly explaining the contexts of intro and conclusions. For example, what is this nutrient budget for? Agricultural soil or what? Lots of assumptions are used.
Line 28. What are processed livestock products – authors could say “processed meat products”
Line 30. Replace fertilizer was with “fertilizers were”
Line 33. In combination with does not seem to fit here, please reword
Line 35. one-half?
Line 36. Drained externally?
Intro para # 2 need some rearrangement, for example, the 2nd last sentence could fit better in the earlier part of this para
Line 50. Do you mean excessive use of nutrients? It is a bit unclear sentence
Line 51-52. Somewhat this sentence would be required in the abstract to understand what authors did the research for
Line 54. What authors mean organically link – do they mean carbon link? Clarify
Line 62. Reword/correction
Line 85-86. This is not right, don’t agree, rework
2nd last paragraph of intro could have been much more inspiring as to why this study was carried out; not much motivation is expressed here. Last intro para should include specific objectives (with clear objectivity) and hypothesis
Line 116. I expected here how hNS and hPS were estimated and determined, must be explained here or provide with references
Line 128. I struggle to imagine how authors obtained animal headcount data from WEMS and why? Also, were all the metadata collected belong to the same region and were collected under similar environmental conditions (e.g., season/temperature, precipitation, humidity) and time scales, if yes, how was that possible to collect that enormous amount of extensive data, readers need to understand how was that possible? If not, how the data spatial and temporal scales were upscaled or downscaled and what were the methodologies adapted with references?
I notice that authors mixed up the concepts of estimation, determination and calculation. At several spots, they used word “estimation” instead of “calculation” for example, line 155 (it is calculation, not estimation). These unclear concepts are unacceptable and must be fixed through the ms thoroughly. Please report. Estimation is - when a value is determined by adopting a scientific methodology, for example, soil phosphorus is estimated by using the Olsens method (US Salinity Lab Staff 1958). Remember please – summing is a calculation not estimation
Line 169-170. Authors calculated (summed) for aNS, and how the industries measured the N losses (did they also summed from somewhere?) give some idea, an and example how they measured N2O using various methods (authors date).
Figures x-axes are overlapping and smudged with blue bars, should be legible
Line 450-479. The conclusions are rather too long. Much of the results and discussion are squeezed into conclusions. Conclusions are only take-home message; therefore, all results and discussion must be removed and may be added as the last paragraph in discussion if indispensable (better avoid). For example, line # 450-456 are pure results not conclusions, please remove.
One of my serious concerns is that authors never used any statistical analysis on the data used in order to see if the data were normal or what and then I wonder how they constructed a model based on values not statistically tested before model construction. These questions need to be answered before any publication could be warranted.
Author Response
We thank Reviewer #3 for his/her thorough comments on our manuscript. We are glad to learn that Reviewer #3 considers that “This paper contains relevant information of nutrient budgets in two Korean regions.” The comments by Reviewer #1 have helped improve the manuscript, we are very grateful for them.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The authors significantly improved their manuscript and concisely responded to all of my comments. Therefore, I consider this revised manuscript acceptable for publication in Sustainability.
