Spatial Stratification Method for the Sampling Design of LULC Classification Accuracy Assessment: A Case Study in Beijing, China
Round 1
Reviewer 1 Report
Honestly, I remember to have revised this paper, but I cannot access to my specific report. So I am judging the revision based on the specific answer to the reviewers. Based on this, the only comment I have is regarding a few details on figure 3. For comparison, it would be better to place original and reclassified images side by side (as suggested in the previous revision), and avoid using the same letters (in capital vs no capital) for differentiating between both images. You can simply use letters A to E, and place a single legend for each type of classification. This will allow you to standardize the size of each map and make it more easy to compare.
Author Response
Please see the attachment. Thank you.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
I wish to recall the authors' attention to the article "Stratified even sampling method for accuracy assessment of land use/land cover classification: a case study of Beijing, China" published on Taylor & Francis (SN 0143-1161 doi: https://doi.org/10.1080/01431161.2020.1739349), by the very same authors., in the year 2020.
I might be mistaken, but the two manuscripts seem to overlap in many parts. For instance, chapter 2.3.2 of T&F looks almost like a duplicate of section 3.2.
I am sorry. It is on my advice beyond the acceptable limit of 25-30% of reused material from other publications.
Author Response
Please see the attachment. Thank you.
Author Response File: Author Response.pdf
Reviewer 3 Report
Interesting manuscript. There were many land cover products, but gridded data were usually quite limited because of the different spatial resolutions, classification types, and temporal gaps. Therefore, this manuscript had merits. And yes, I also agree the method proposed has a big potential.
However, there were too many examples figures, so it was quite hard to follow up on every detail. I think the quality of the maps was quite low, and the figures and tables need to be re-arranged to represent core results more clearly.
Figure 3 needs to be adjusted. Actually, They were the maps that just aggregate land cover classes following Table 2. I am not sure that they were that important to cover up many pages. In addition, Figure 7 and 8 were able to present at the same time.
Figure 6 looks the key results, and it had each 2 strata. I expected to compare and made more strata considering the Figure 2, but the case study did not. the case study also had own the other methodological steps. Thus similar methodological figures and tables made a lot of confusions. With this perspectives, I am wonder that why the author mentioned three land cover products, and aggregate three data proudct at first. Accoding to the Figure 5, aggregation of land cover classes of 10m GLC data is not necessary and it means, the information about 10m GLC in Table 1 and Table 2 was not required. Figure 3(C), (c) was also.
The question also goes to the accuracy assessment. The author used the 10m GLC data as the reference. If it was assumed as the correct date, why didn't we just use the 10m GLC data to know actual land covers? Maybe we did not need to concern the misclassification as the author mentioned in the introduction.
Furthermore, 10m classification also can be limited in some ways similar to the other land covers, so maybe the real field data and a cadastral map would be mentioned in some ways.
Therefore, with the above point of view, I recommend adding some more discussions with references. Maybe the author need to directly describe the case study first, and the theoratical expalantion about spatial stratification method can support the case study in the discussion. Anyway, the mansucrpit needs some edits for readers.
Author Response
Please see the attachment. Thank you.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Dear authors
Many thanks for your kind cover letter and your attention to my comments. I feel more comfortable with this new outfit of the manusctipt. The duplication issues with its antecedent (reference number 14) look been fixed.
I have no further comments or changes to improve the manuscript's quality before publishing. The idea is original, and the work is valuable.
That's all
Author Response
Please see the attachment. Thank you.
Author Response File: Author Response.pdf
Reviewer 3 Report
All comments were well revised. I noticed the similar but different paper (I think the difference would be the methodological approach of the stratification) in the International Journal of Remote Sensing titled "Stratified even sampling method for accuracy assessment of land use/land cover classification: a case study of Beijing, China", so I hope the author should more clarify the difference between these papers and avoid some duplication.
Author Response
Please see the attachment. Thank you.
Author Response File: Author Response.pdf
