Comments to the author:

Article entitled “High-resolution global map of closed-canopy coconut”, attempt is to produce the first global coconut map at high spatial resolution (20 meters) and estimate the global coconut area using satellite remote sensing. The study results could be a scientific base for the high-resolution, quantitative, and precise data necessary for assessing the relationships between coconut production and the synergies and trade-offs between various sustainable development goal indicators. Therefore, the topic fits the scope of the ‘Earth System Science Data’ journal.

While reading this paper, limited literature review, lacks logical justifications of contents doesn't lead the reader to the next step. Overall the organization, analysis, methodology is not as developed as I would expect to see such a reputed journal articles. Therefore, the papers truly needs a revision for data and context updation, robust literature review, necessary explanation of each employed methodology as well as results and findings.

However, I have several major concerns about the article, which I listed below:

Abstract

1. Abstract started with vegetable oil crops and their distributional statistics, but paper mainly focuses on coconut distribution. The writing should focus on coconut plantation only.

1. Why suddenly deep learning for semantic segmentation, e.g., ‘U-Net model’ used in this mapping not justified.

1. The study why produced ‘20-meter global coconut layer’ from sentinel 1 and sentinel 2, this is also confusing from 10m sentinel product. It may create confusion to the readers.

1. Again in the 3^rd sentence discussing on ‘mapping palm species’, reader may lose interest to read for such a confusing mixing between ‘coconut’ and ‘palm species’, not clear what authors want to focus.

1. Second last sentence should be focus on coconut production instead of ‘vegetable oil production’.

Introduction

1. The literature review section little poor and required robust literature review.

1. Writing should be focused on coconut production and distribution instead focusing on starting with very non-contextual sentence with ‘vegetable oil crops’ (1^st sentence), even discusses ‘1.7% of the global volume of vegetable oils’ (2^nd sentence).

1. 3^rd sentence “Coconut is generally overlooked ……. see this palm as a threat to biodiversity” is not also confusing, why suddenly palm came in discussion, it bit confusing. Need a few lines of introduction before mixing such terminologies.

1. Reference started with “Meijaard et al., 2020b” in the first sentence instead of ‘Meijaard et al., 2020a’, no reference sequence maintained.
2. The sentence “A high resolution global map of the distribution of coconut would serve as a basis for understanding their impacts, as well as help to shape environmental and biodiversity policy”,

• This is not rightly written, is it really possible to “understand their impacts from map”, how it can shape ‘environmental and biodiversity policy’ is also very confusing? This sentence has no value as written with superficial context.

1. “Research is therefore needed to map coconut on a global scale”, are authors map coconut, or ‘coconut tree’ or ‘distribution of coconut tree’ or distribution of any species, everything very confusing without any contextual writing and justification from previous sentences.

1. Last paragraph of introduction “the first global coconut map”, is it rightly said ‘Coconut map’ or something else, please see previous comment to write it appropriately.

1. “where climate was favourable for coconut growth” is climate only the factor, as native we know soil is priority concern for coconut growth and their distribution.

1. Very few literatures addressed in the research gap section, please try to add few more literature in context.

Methods:

2.1 Bioclimatic analysis

1. SPAM model suddenly introduced as acronym, without having any discussion.
2. Title ‘bioclimatic analysis’ what does it mean? For what purpose this term is used missing here.
3. Repetition of sentences, example ‘a literature search’ and ‘our literature search’.
4. What does it mean about this sentence - “The points were collected in coconut-producing regions, based on our literature search, and coconut regions according to the SPAM model”?
5. About previous sentence mentioned - from secondary data to collect primary location is not a good research, for species distribution model authors need to collect actual location to model it based on bioclimatic variables, soil, or/and other factors.
6. “we identified coconut” this is not right, coconut, tree or species, please specify as mentioned earlier.
7. After collecting bioclimatic data, need to know collinearity but here authors failed to understand collinearity before moving to understand bioclimatic responses.
8. Why only bioclimatic, what about soil, geology and geomorphology, which is most important factor of coconut distribution.
9. For verification “visualized images from Google 70 Street Maps to verify the presence of coconut”, would it be good idea to verify coconut areas? There will be lot of wrong areas which will be looks like other palm or arecanut and others, are not possible to use them to generate coconut distribution area.
10. “a pixel in the WorldClim dataset was considered 75 suitable for coconut growth if at least 18 of the 19 bioclimatic variables fell within the bioclimatic range”, why “18 of the 19” is very confusing, no justification added by authors.

Section 2.2, 2.3, and 2.4, very well written

2.5 Validation

1. “a cost-effective alternative by visually reviewing the sub-meter resolution images from Google Earth because coconut is easily identified using such data”, would it be possible to differentiate from other palm or arecanut and others, following figure ‘2 coconuts sparse’ left down figure high density tree is not coconut, coconut open canopy looks like arecanut, how authors validating - this is not convincing method.

Result:

1. “The annual temperature ranged between 22.4 and 28.8 °C”, is it valid? I know the region very well, the temperature of this region more than 40C now, is it justifiable for valued output.

1. “Indian state of Karnataka” [black soil] is a decan trap region, coconut production is very less in Karnataka, some of the southern section producing few coconuts, soil associated with geology and geomorphology is most important factor rather distance from and elevation.

1. “the coldest month was 11.5 °C, indicating that coconut cannot tolerate cold temperatures”, the regions coldest month temperature is less than 11.5C. I do not think, coconut responsive with climate rather geology and/or soil.

1. “Interestingly, we found that coconut grows in a variety of rainfall regimes. Coconut was found in humid tropical regions as well as relatively arid regions in India and East Africa, regions with no precipitation during the driest quarter and where annual precipitation was slightly above 100 mm.” This results totally invalid – there is no arid regions with coconut trees, Kerala called ‘land of coconut’ –average rainfall more than 3000mm. Such huge in-discrepancy results authors will be able to justify, if not include soil and geological condition. In some context local geomorphology also justify some of coconut production like, in the state of Andhra Pradesh.

1. “The separability analysis revealed a low separability between coconut and oil palm in the VV and VH bands”, as paper focusing coconut, suggest to mention coconut only.

1. Paragraph 4 in results section, discuss again tree species focusing non-palm pnattations only on cinnamon (Cinnamomum spp. Schaeff.) and mango. (Mangifera spp. L.), which is very rare species in the states with coconut tree available in India. These section can be avoided, as never mentioned earlier any species classification earlier throughout the manuscript.

1. “the global coconut area was 12.31 ± 3.83 x 106 ha”, if we see map 4 –estimated area will be higher, although in map now showing some prominent coconut areas in Thailand in North Pattaya region, higher in southern west Bengal, where very few coconut trees available. Not a single plantation available in west Bengal in reality. Lot of miss-classified species included in classification.

Discussion: this section very well written

Conclusion:

1. “Our global coconut layer is of considerable interest to researchers…. of vegetable oil crops (especially oil palm) to meet future demands for food, feed, biofuel, surfactants, and other oil uses” – focusing on coconut would be better in conclusion again avoiding ‘vegetable oil crops’.

1. Last sentence “coconut presents a spatial overlap with high levels of threatened species, species endemism, and species richness, in tropical islands and, thus, the global coconut layer might benefit studies that evaluate the associated environmental impacts of 430 coconuts in such biodiversity hotspots”- should be avoided, this is non-contextual from the present analysis.

Appendix:

1. Analyzed variables e for coconut distribution is not sufficient, need to include many indicators like geology, soil, geomorphology, else Indian coconut distribution will not be justified.
2. Figure A1, ‘no of bioclimatic variable’ and maps not meaning anything what authors want to mean.

This research developed the first 20-meter global coconut maps using U-Net and Sentinel imagery. According to the results shown in the paper, the dataset can provide essential and detailed spatial distribution for global coconuts, and the accuracy is relatively high. However, it is recommended to address the mentioned points to improve the clarity, comprehensiveness, and applicability of the research.

1. Please summarize the previous efforts and methods developed for coconut tree monitoring in the introduction section. This will help readers understand the novelty and significance of the current research.
2. Provide details of the training samples used in this study. Please present an example of one image (Sentinel-1 and -2 composite) and one corresponding per-pixel label image to illustrate the data used for training.
3. The feature differences between different crop trees (such as coconut, oil palm, mango as mentioned in the paper) should be given for better comprehension of the classification procedure.
4. Is it possible to investigate the correlation between the mapping results and FAO statistics on a country level (besides the three countries shown in Table 2)? This analysis will strengthen the validity and reliability of the mapping results.
5. Compare different semantic segmentation models or explain why U-Net was chosen as the final model for mapping. Additionally, discuss the potential use of other deep learning models, and elaborate on why semantic segmentation models were preferred.
6. Incorporate the coconut-producing regions obtained based on the literature review and SPAM. It might be useful to overlay these regions with the coconut locations shown in Figure 1a to demonstrate their spatial distribution.
7. As this study took the semantic segmentation model as the mapping network, it may be helpful to provide pixel-level validation results in addition to the accuracy obtained from validation points, as it will provide a more comprehensive evaluation of the model's performance and its ability to accurately delineate coconut tree boundaries.
8. Is it possible to classify the coconut to Sparse, Dense open-canopy and Closed-canopy? It will provide more detailed mapping results, and give the readers opportunity to choose the product they need, just like the previous global oil palm product the authors developed, which provide the smallholder oil palms and industrial oil palm plantations.
9. Include zoomed-in regions that contain both coconut trees and oil palms (or other plantations) to illustrate the model's capability to accurately extract coconuts from other tree types. This may help to address the statement made by the authors: "The separability analysis revealed a low separability between coconut and oil palm in the VV and VH bands."
10. Provide the uncertainty map of the classification results to help users better understand and utilize the dataset shared in this research.