Extracting Khmer Rouge Irrigation Networks from Pre-Landsat 4 Satellite Imagery Using Vegetation Indices
Round 1
Reviewer 1 Report
The manuscript is an interesting study, presenting the possibilities of extracting irrigation structure networks from satellite imagery using vegetation indices.
The main purpose of the work is to examine and evaluate irrigation systems created during the Khmer Rouge period.
A very good idea is to use old satellite images for this research. Thanks to these photos it was possible to reconstruct the land use in the studied area in the past. In a situation where it is necessary to compare the current situation with the past, the possibility of using archival materials is very important. Such materials are not always available or their quality and form cannot be used. It is necessary to adapt them properly. In this work, the authors proposed a methodology for using satellite images from Landsats 1-3. They showed differences in imagery from Landsats 1-3 and currently available satellite materials and pointed out the steps that had to be taken so that these materials could be compared with each other. The proposed research procedure allowed for analysis of old satellite images and comparison of both data sets. Thanks to this, it was possible to identify elements of the irrigation network built during the Khmer Rouge period and assess its current condition.
I think, that the research idea is very good and the results may be used in the future. This subject is interesting from a natural, social, historical and technical point of view.
I have no objections to the methodology of work. I think, that the objectives of the study were formulated properly. The article has sufficiently discussed the technical details of the research (such as data sets, photogrammetric processing, data analysis). The paper is clearly structured and well written. The introduction presents a sufficient literature review of analyzed issue. However, I found many editorial errors and inaccurate descriptions of some figures that need to be corrected. In its current form, the manuscript is not fully understood.
First of all, in my opinion most of the figures are too small. The authors discuss details that are difficult to find in the figures. There are a lot of interesting details on the images, which, at this size, may not be readable enough. I think, that the figures should be enlarged and even greater enlargement of some of their fragments should be applied (especially for figures 1, 5, 7, 8, 9, 10).
Figure 1In this figure, information about the scale is duplicated (white scale on the images, black below the images). In figure 1b the description "1 km" is not needed - I do not know what it concerns. Consider matching font sizes (the coordinates in the description of the images are much smaller than the descriptions "Figure 1 a", "Figure 1 b"). You can delete the rectangle with the scale and the north arrow under the images (the figure has a north orientation, the arrow is unnecessary, and the scale is in the image).
Figure 2 is missing !!!Section 2.3 (line 183, point 4)
Was the digitization of the water features carried out after removing the ancient channels that you write about in section 2.2? A figure showing this digitized network would be interesting.
Section 2.3.3I suggest to expand the description of vegetation indicators. A brief description about NDVI can be found in the text but CTVI and TTVI are described only by the formula. Interpretation of indicators would certainly be easier if their short characteristics and differences between them appeared in the manuscript.
Figure 6Not much can be seen in this figure. I think, it should be changed - instead of one large figure, I suggest creating a collage of four images (enlarged fragments of the area with intersections a and b). I think, it would be good to compare the enlarged fragments with the drawn and non-drawn intersections. Readers could then see how the two intersections are visible in the Landsat imagery. Now the white lines drawn cover these intersections.
Chapter 3.2 requires the most attention. There are a lot of editorial mistakes and a fragment of the text that is not entirely clear.
Figure 7It seems to me, that the images on the right side of the figure (site b false infrared and site b CTVI) are the same graphic. CTVI image is probably wrong.
Do figures 7 and 8 present the same area?Lines 324-332 (description of figure 8)
This figure is definitely too small. Details described in the text are not visible in the figure. In the description, the authors talk about the gray line running SE to NW - I do not see this line in the figure. The channels marked in the figure are perpendicular to the visible road, not parallel, as mentioned in the description (unless the description says about a different road…). The arrows in the figure not clearly show the course of the channels. Perhaps, the difficulties in interpreting the figure are due to its size. I suggest, that authors should verify this fragment of the manuscript and look critically at the figure. Something wrong happened here.
Line 364-365The example describe here could be illustrated, if it is possible (reservoir during high and low water stages).
Mistakes in numbering and reference to tables - both tables are described as “Table 1” (line 376 and line 441).
Line 389 – “Table ???”, this should be corrected. I think, it is Table 2. Line 442, table - in the column "Year Built", there are green triangles, probably left after exporting data from an Excel file. Line 427 – shortcut “KR” - everywhere in the text the full name is used (shortcut has not been explained anywhere). Line 436 – is “The Jan 6th/ Traing Krasaing” correct name of the system? Line 449 – is it correct figure citation?Figure 9 and 10 - fonts of different sizes in the figures.
I understand, that parts c and f are fragments of the picture shown in pictures a, b, d, e? Maybe it would be useful to mark this area with a rectangle on pictures a, b, d and e. I think, the scale of the enlarged images should be the same. Then the size of the water storage can be compared.
Line 466 – this is figure 10, not figure 1.Author Response
Hello,
I have made most of the changes in the figures that you suggested, hopefully they are more clear and straightforward now. I was unable to get imagery for high/low water levels at the Sla Reservoir dam for the dates needed, so I have not added that figure. I have updated the vegetation index descriptions and completely re-wrote much of section 3 to walk the reader through the new, updated figures.
In regards to your question on the January 6th/ Traing Krasaing name, yes that is the name of the location, as many locations in Cambodia have multiple names depending on which government document you read. The importance of the January 6th dam is that it was one of the first built by the CPK, with much propaganda surrounding it, so historically it has been known (and was called by the CPK) as the January 6th dam. However in most current water resource documents the names alternate. We have found that many locations have an original official name, a common name used by locals, and/or a new name after being rehabilitated, so naming irrigation schemes can be complicated.
I have added figures illustrating the results of the digital canal and reservoir mapping. I also split the problematic figures in section 3 into two separate, enlarged figures for each location. I have also corrected issues regarding table names and column appearance. Please let me know if there is anything that you think I have missed.
Thank you for your time.
Reviewer 2 Report
I thought this was a really interesting paper that would be a great contribution to the literature. Technically, it is well-written, describes the research thoroughly, and avoids common pitfalls like the use of jargon. The fact that this research can help inform an historical debate means that it is likely to have more significance than many similar studies that I have seen.
I only have a few comments for clarity in the writing, and on that alone, it may not even be necessary to send out for another round of review. That said, unfortunately, the only download that I can get does not contain any of the figures in the paper. I tried on several different computers and operating systems, so I'm not sure that the error is on my end. I would like to at least take a look at the figures before I consider my peer review complete. So I'm putting this as "major revision" for that reason alone, but I expect that it shouldn't require much more editing.
Here are some minor suggestions on the manuscript:
Page 1, line 36: "agriculture health" seems like an awkward phrase.
Page 2, line 77: typo "been be accomplished"
Section 1.1 Historical Background: I would encourage the authors to highlight this information a little more; it's really interesting to learn about the Khmer Rouge's emphasis on irrigation. Adding a line to the abstract could help capture attention for the paper.
Page 4, line 124: typo: "agricultural"
Pages 6-7, lines 191-213 or so: When discussing the software packages used, I encourage a little less hedging or justification. As now, it reads like it was written by someone who is still sort of new to the game or something. Sentences like "Although this step could be accomplished in..." There are dozens of different ways to do anything in the GIS world, and it's often just a matter of personal preference. To me, just describing what you used for each step is entirely sufficient.
Page 8, Equation 2: The written "times" in the middle of the equation can probably be removed.
Author Response
Hello! Thank you for your effort in reviewing my manuscript. I have made the spelling/typo changes that you pointed out in the article.
Although I have not added anything to the historical background section, I did rearrange the abstract and add a sentence about irrigation planning to it.
I tried to limit my discussions of the software to descriptions of actual methods, when possible. My original "hedging" was meant to encourage the use of open-source software, but I can see where that point was lost in my previous descriptions. I did keep some of the information (like the ways in which QGIS implements SAGA and the advantage those tools provide).
The word "times" in the equation was an export artifact from LaTex formatting, so that was fixed in this version.
I do apologize that you received a copy of my manuscript without figures, I'm not sure whether you got the LaTex version or the Word document, but hopefully this new Word document will download for you with images. Several figures have been added, and the rest edited for clarity.
Thank you again for your time!
Reviewer 3 Report
The paper is really interesting and I think that will have significant merit. The key idea which encompass the use of Landsat imagery for studying the history and social system is really amazing. A clear and detailed study is presented and interesting conclusions are shown highlighting the importance of these types of remote sensing products in the area of agronomist design. My recommendation is minor revision, in case that I've missed something on this excellent work
Author Response
Thank you for your very positive response to my manuscript. I have incorporated some of the figure changes suggested by other reviewers in this new version of the paper, as well as done my best to catch any typos or formatting issues. I have added a few figures to show mapping results and to break up overly-busy figures into more focused individual highlights.Please let me know if you feel that anything needs more explanation or still needs corrected.
Thank you again for your time!
Round 2
Reviewer 2 Report
This is a great paper, it was a pleasure to read. All of the figures are well-labeled and appropriate, as expected. I still could not see Figures 8 and 9, but I don't think that should hold up publication.
The only typo I saw was: Page 8, lines 190-1: See figures and for results.
Otherwise, no concerns on this end.
