Delineation of Management Zones Based on the Agricultural Potential Concept for Potato Production Using Optical Satellite Images
, Athyna N. Cambouris 1 and Michelle L. D’Souza 3Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsManuscript ID: remotesensing-3936210
Title: Delineation of Management Zones Based on the Agricultural Potential Concept for Potato Production Using Optical Satellite Images
Overall Evaluation:
The paper titled "Delineation of Management Zones Based on the Agricultural Potential Concept for Potato Production Using Optical Satellite Images" by Ramirez-Gonzalez et al. aims to present a methodology for delineating Management Zones (MZs) in potato production systems using multi-temporal NDVI composites derived from Sentinel-2 and Landsat 8 imagery. Although this study has a contribution in the remote sensing of agriculture (The use of multi-year NDVI composites to capture temporal stability and vegetative development is innovative and very interesting), some questions have not been well addressed. Therefore, my suggestion is major revision at this stage.
General comments:
1. The spatial resolution mismatch between NDVI and yield should be addressed as a potential source of error.
2. I suggest that the authors should consider including more years of yield data or using cross-validation to strengthen the AP MZ validation.
3. I suggest that the practical recommendations for farmers could be expanded.
Specific Comments:
1. Abstract: "Validation was carried out using 2023 potato tuber yield and soil physicochemical properties". The abstract does not mention the strength of correlations or the statistical significance of the results. I think that including key correlation coefficients (e.g., between NDVI metrics and yield) would strengthen the abstract.
2. Methodology: "These repositories contain orthorectified TOA images that were atmospherically corrected using the Sensor-Invariant Atmospheric Correction". The authors should briefly explain what "Sensor-Invariant Atmospheric Correction" entails and why it is suitable for this study, especially given the multi-sensor (Sentinel-2 and Landsat 8) approach.
3. Methodology: "The MVC procedure examines a series of multi-temporal satellite data during the compositing period and, by analyzing each value on a pixel-by-pixel basis, retains only the highest value for each pixel location". The rationale for using MVC (e.g., to minimize cloud contamination and capture peak vegetation) is well stated. However, the authors should clarify whether MVC was applied per year or across all years.
4. Results: "The strongest observed correlation was with tuber yield (r ≥ 0.25)...". While statistically significant, correlations of r ≤ 0.41 are relatively weak. The authors should acknowledge this and discuss potential reasons.
5. Results: "However, the low AP MZ, which corresponded to an intermediate yield in the field, did not showed a clear visual correspondence". This inconsistency should be discussed in more depth. Possible explanations could be explored.
6. Discussion: "A key limitation of this study is the use of only one year of tuber yield data for validation". This is a valid limitation. The authors should also discuss how inter-annual climatic variability or crop rotation may affect the generalizability of the MZs.
7. Some references to figures and tables in the text are incorrect or inconsistent. Please ensure all references are accurate and clearly explained.
8. The paper should include the most recent and relevant references in the field. Please cite the latest literature (within the last 5 years).
Author Response
The response to each comment of the reviewer is shown in the attached word document.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsDear authors,
Thе mаnusсriрt offеrs а ͏usеful сontribution to рrесision аgriсulturе by аррlying multi-yеаr, multi-sourсе sаtеllitе dаtа to dеfinе mаnаgеmеnt zonеs for рotаto fаrming. Thе mеthod usеs frее NDVI dаtа, whiсh mаkеs it а usеful аnd аdарtаblе tool for t͏hе ͏fiеld. Howеvеr, thеrе аrе sеvеrаl k͏еy аrеаs thаt nееd imрrovеmеnt to bеttеr suрр͏ort thе mеthod͏, рrovidе а morе bаlаnсеd intеrрrеtаtion, аnd fully аddrеss thе study's limitаtions bеforе it саn bе рublishеd.
Major Points for Revision
I find thе mеthodology unсlеаr bесаusе it usеs а сombinеd NDVI signаl from vаrious сroрs (likе рotаtoеs, сеrеаls, аnd сovеr сroрs) to сrеаtе mаnаgеmеnt zonеs sресifiсаlly for рotаto рroduсtion, аs stаtеd in thе titlе. This сould lеаd to mаjor biаs. Diffеrеnt сroрs hаvе diffеrеnt NDVI раttеrns, growth сyсlеs, аnd links to yiеld. а high аnd stеаdy NDVI vаluе in onе yеаr сould bе bесаu͏͏sе ͏of а strong сovеr сroр, likе аlfаlfа, instеаd of thе soil's рotеntiаl for growing рotаtoеs. I nееd а strongеr rеаson for using this "с͏roр-аgnostiс" mеthod to аssеss рo͏tеntiаl for а sресifiс сroр. I аlso nееd а dеtаilеd disсussion on how mixing t͏hеs͏е signаls might саusе сonfusion i͏n dеtеrmining if а zonе is suitаblе for рotаtoеs. Please clarify if non-potato years were removed, given a weight, or assumed to reflect typical soil and crop conditions.
The methodology divides the landscape into three management zones without a clear reason for choosing three. You yourself mention that more classes might have offered more accuracy. I suggest you either explain why three zones were selected, for example, based on existing agricultural knowledge or statistical methods like silhouette or elbow analysis, or clearly state in the limitations that the number of zones was a fixed choice without optimization. Please include references to agronomic standards that support this classification.
I also notice that only July imagery was used. You need to explain why this single month is enough to show crop changes across years, especially considering possible differences in growth stages and crop types between years.
Moderate Points for Revision
I͏n thе Rеsults sесtion (linе 315), you notе thаt whilе mаn͏y сorrеlаti͏ons bеtwееn NDVI mеtriсs аn͏d soil рroреrtiеs wеrе stаtistiсаlly signifiсаnt, t͏hе͏y wеrе аlso "wеаk (r < 0.25)." Howеvеr, thе Disсussion аnd сonсlusion sесtions арреаr to ovеrstаtе thе strеngth o͏f thеsе rеlаtionshiрs. I must еmрhаsizе thаt stаtistiсаl signifiсаnсе with͏ а lаrgе sаmрlе ͏͏sizе doе͏s not imрly а strong, рrасtiсаlly mеаningful rеlаtionshiр. Most сorrеlаtions rеmаin wеаk (|r| < 0.3).
Sеvеrаl stаtеmеnts аrе too strong givеn thе study'͏s limitаtions аnd mixеd rеsults. For еxаmрlе, thе ͏сlаim thаt thе͏ mеthodolo͏gy "ассurаtеly dеtесtеd tubеr yi͏еld diffеrеnсеs" (linе 542) is сontrаdiсtеd by thе findings in Fiеld X, аnd thе st͏аt͏еmеnt ͏thаt mеtriсs "еffесtivеly сарturе thе undеrlying sраtiаl vаriаbility аnd аgriсulturаl рotеntiаl, rеgаrdlеss of сroр tyре" (linеs 519-520) ͏is not fully substаntiаtеd.
Thе rеsults show thаt for Fiеld X, thеrе w͏аs no stаtistiс͏аlly ͏signifiсаnt diffеrеnсе in tubеr yiеld bеtwееn thе intеrmеdiаtе аnd low ар zonеs (linеs 365-367). This is а сritiсаl finding thаt dеmonstrаtеs thе modеl's реrformаnсе is not univеrsаlly сonsistеnt. Rаthеr thаn frаming this solеly аs а сonsеquеnсе͏ ͏of yiеld vаriаbility, I rесommеnd you d͏isсuss it аs еvidеnсе of thе modеl's limitаtions аnd а сhаllеngе to its robu͏stnеss асross diffеrеnt fiеld сonditions.
I am concerned that Pearson and ANOVA tests are applied as if samples are independent; spatial dependence could bias p-values. You should include a brief acknowledgment of this issue and, if possible, reference Moran's I or variogram analysis.
Minor Points for Revision
The abstract does not explicitly articulate the research gap (e.g., lack of prior multisource NDVI composite applications for potato MZs). I suggest adding a short phrase indicating what previous approaches missed and what this study newly contributes.
The explicit gap between existing studies and this one remains somewhat diffuse. The introduction should clearly highlight that this research uniquely integrates multisource, multi-year NDVI composites for potato MZ delineation at a multi-field scale.
There is limited comparison of alternative methods (e.g., proximal sensors, electrical conductivity, supervised vs. unsupervised approaches). I recommend adding a concise paragraph comparing these and justifying the chosen NDVI–k-means approach.
Use of NDVI skewness as a stability measure is new but not standard. A brief justification or prior citation for this use of measurement should be provided.
I suggest including a more forceful comparison to earlier comparable studies (e.g., Rezník et al., 2020; Hendawy et al., 2024). You need to show if magnitudes of correlations and class separations are equal or different from earlier work, and where there are differences, explain potential reasons (alternate crops, local variability, sensor pairs).
The Conclusion needs to clearly restate how the research gap created in the Introduction (multisource, multi-year NDVI composite instead of expensive sampling) is addressed by the study.
Line 315: The reference to "Figure 3" where it mentions correlations with soil properties needs to be rewritten. Correlation coefficients appear in Table 3.
Line 357: ".did not showed a clear visual correspondence" should be ".did not show a clear visual correspondence."
Line 516: "the significant and strong significant correlations." is redundant. Correct to "the strong and significant correlations."
Several places: "Tuckey" needs to be corrected as "Tukey"
The reference "Agriculture and Agri-Food Canada, Potato market information Review 2021-2022" lacks the precise locators like page numbers, a chapter, or a direct persistent URL. Check to ensure more precise information can be provided
Thanks for considering the above remarks,
Reviewer
Author Response
The response to each comment of the reviewer is shown in the attached word document.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThe authors have incorporated my suggestions. Therefore, it can be send for publication.
Author Response
No reviews needed
Reviewer 2 Report
Comments and Suggestions for AuthorsLine 124: I recommend adding the following reference for SAVI:
Prediction of Areal Soybean Lodging Using a Main Stem Elongation Model and a Soil-Adjusted Vegetation Index That Accounts for the Ratio of Vegetation Cover
Using the Negative Soil Adjustment Factor of Soil Adjusted Vegetation Index (SAVI) to Resist Saturation Effects and Estimate Leaf Area Index (LAI) in Dense Vegetation Areas
Line 145: “An eight-years (2015 - 2023) NDVI time series” → should be singular “eight-year.”
Line 291: “them transformed” → should be “then transformed.”
Line 414: “did not showed” → incorrect verb form.
Line 577: “capture difference specially between” → should be “differences, especially between.”
Line 641: “especially at in field scale” → “especially at in-field scale”.
Author Response
The responses to the comments are detailed in the word document
Author Response File: Author Response.pdf
