Unveiling Sri Lanka’s Wilderness: GIS-Based Modelling of Wilderness Attributes

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

It is interesting to see this effort of making a wilderness map of Sri Lanka. The procedure seems straight forward and the result is analysed with respect to especially protected areas and need for future protection. This part of the article is clear and as far as I can judge, valid.

I have, however some comments I find important:

1. The authors state that the wilderness term is understood differently by different authors and in different parts of the world. This, however, is not picked up in the discussion. It seems that the index used favours a human-centred approach of what is felt or traditionally understood as wilderness. I guess it is therefore path-distance and viewshed analysis are used in the two first elements of the index. Another approach could be wilderness for nature itself, biodiversity with natural processes and biodiversity in its different aspects. I feel a deeper discussion of this would be of great interest, especially compared with other efforts of making indices of human interventions in nature found in international literature. How do this index work for different approaches? Could it be improved. I can see it is a comparison with the human footprint index, but to widen this discussion would be a great improvement.
2. It is stated that the wilderness analysis partly becomes somewhat subjective since the term wilderness varies in meaning. But a large part of the index is a straightforward calculation of mapped human interventions. These parts of the index and calculation can be made rather technical and objective before the term wilderness is used. A discussion around this would also be interesting.
3. What does wilderness quality mean? The map splits in wilderness and not wilderness. I cannot see that wilderness quality is discussed.
4. Generally, a bit more information about the calculation of the three different elements of the index could be useful. I would guess that the two first will be rather similar and the third very much dependent of the coarse scale it represents. Nevertheless, some comments of the classes available and geographical distributions of these together with expected variations within the classes would be of interest. What could be alternative information input?
5. Why is buffers and viewshed preferred used instead of a straightforward neighbourhood analysis. This could also shortly be discussed.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The Authors present a very interesting work where the wilderness of Sri Lanka is mapped. I enjoyed reading the manuscript. There are some parts of the manuscript in need of improvement, namely the methodology. Also, the figures can be reorganised to save space and avoid repetitions. In addition, figure captions should be more auto-explicative.

1. Remoteness from public roads: This section is not very clear. Why not calculate simply the distance to the nearest road? It is necessary to explain how the model was calculated. Only the variables are indicated, but not the procedure. Please, indicate the origin of the variables.

2. What were land cover data classified? Which land uses were considered as wild and which ones are not wild?

3. Which is the spatial resolution of the quality index analysis?

4. Figure 12: Please, do not show again the wilderness quality map. You can show this human footprint in Figure 6. Figure 1 can be improved by putting a hill-shading model below the digital elevation model. This figure should show some human structures, at least roads and main settlements. This will give a first impression to the reader about how humans impact Sri Lanka.

5. In fact, figures 3, 4, and 5 can be represented together. Figures 6, 7, 8, and 9 can be represented together, plus the human footprint. You can use two rows if five maps in a row are too much. Figures 10 and 11 can be represented together.

6. L 195: Please, define MCE.

7. L 206: What do you consider as public roads? What types did you consider? Did you include unpaved roads?

8. L 206: Can you explain better what are human interventions? Are they artificial structures? Impervious surfaces? Are agriculture fields considered human interventions? Did you consider mining zones and other extraction areas? What about artificial water bodies?

9. Figure 2: What is Viewshed Explorer? Please, provide information about this tool. What land cover data were used? Did you reclassify the land cover classes? Did you use the same classes as the ones used for the naturalness?

10. L 273: Please, indicate the six classes of land use.

11. L 288: When you refer to combining the attributes, do you mean that the values were added?

12. Please, provide more details on Carver et al methodology.

13. L 291-294: Why classifications the index in six classes, and then in four classes? Why not go directly to four classes or two classes? Even if you tried several classifications, for simplicity, I would indicate only the last one.

14. L 296: I do not understand what you did here. If the index was calculated as a raster, how did you obtain the segments? Are these segments vectors? I conclude from Figure 7 that the final a-class raster was vectorised.

15. L 404-408: This is methodology and should not be included in the results. Please, clearly indicate that the WDPA data is spatial. It is not clear how the overlap was performed.

 

16. L 430-434: This is also methodology.

Author Response

Please see the attachment. 

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Although this article does not use particularly complex research methods, it is a very practical study. Scientific research is not about being too complex, I think solving practical problems is very meaningful. A good study.

Author Response

Thank you very much for taking the time to review this manuscript. The changes are highlighted in the revised manuscript.