On How Crowdsourced Data and Landscape Organisation Metrics Can Facilitate the Mapping of Cultural Ecosystem Services: An Estonian Case Study
Round 1
Reviewer 1 Report
In this well written manuscript the authors nicely advance our use of social media data in a landscape pattern context. I only have a few suggested edits below that I feel would add clarity and likelihood of citations.
Methods
Could you expand on how to interpret landscape coherence. Is a high or low value “better”. What are conclusions drawn in past work (perhaps this latter needs to be in the introduction)
Results –
Section 3.2 – order the results in the same order as presented in the methods
Discussion
Line 196 – how was transport accessibility included in this project? I may have missed how this was estimated.
Results and Discussion
Could you discuss the value of the LDA algorithm? For example, how many times did you have to transfer from one bin to the other (like with the pets). What are other shifts? Were the three topics based on automated sorting by machine learning or a priori?
Author Response
First of all, thank you for your useful comments, increasing the overall clarity and value of manuscript. We attempted to address all of them.
Point 1. Methods. Could you expand on how to interpret landscape coherence? Is a high or low value “better”. What are conclusions drawn in past work (perhaps this latter needs to be in the introduction)
Response 1: accepted, the interpretation of landscape coherence values has been added (lines 173-174, 186-195). Introduction has been extended to explain the value of landscape coherence and colour harmony (lines 76-83).
Point 2. Results. Section 3.2 – order the results in the same order as presented in the methods
Response 2: accepted, in methods and results the indicators are discussed in the same order.
Point 3. Discussion. Line 196 – how was transport accessibility included in this project? I may have missed how this was estimated.
Response 3: accepted, this claim was initially supported with the visual map examination only (linear patterns of the photographs follow the road network). Now we have added GIS-based analysis of transport accessibility (lines 213-217), which support the idea of a crucial role of transport accessibility for CES use in Estonia.
Point 4. Results and Discussion. Could you discuss the value of the LDA algorithm? For example, how many times did you have to transfer from one bin to the other (like with the pets). What are other shifts? Were the three topics based on automated sorting by machine learning or a priori?
Response 4: accepted, the extensive discussion of the LDA application, its value and shifts between the CES categories has been added (lines 303-323).
Reviewer 2 Report
The very valuable, simple, and well-produced paper presenting a new set of tools for measuring cultural ecosystem services using crowdsourced data. The paper is short and very cohesive so needs some loosening to be easier to read and more suitable to use it. It can be used as a role model for other scientists, managers, and consultants.
The abstract is clear and reflects the content of the paper. The introduction is also properly designed, with all the references needed. Maybe it could be a bit longer in par about cultural ecosystem services. Please consider adding one more reference – Common International Classification of Ecosystem Services (CICES) or even CICES-Be (Turkelboom, Raquez, Dufrêne, 2013). Maybe it is too “local”, but maybe you will find it useful. As You wish.
Section of Material and Methods is properly and comprehensively presented. I only feel a shortage of “manually interpretation” explanation (lines 98-100). I mean not what did You find on the pictures but how have you been looking for it. That interesting method needs to be shown in the nice scheme, please try to provide it.
The scope of the work is properly and comprehensively presented in the part of the results. Only figure 2 is too small and looks illegible. Discussion is really interesting and dividing it into sections makes the argument easy to understand.
The conclusion also seems to be well done, answering points signalised in the Introduction, but lacking more information about possibilities of possibilities for wider use the scheme of research. It will be sound too local without it.
There is a large appendix after the main text, showing all the research data. For me, it could be compressed and partially used it the main text to make it lighter and easier for reading that way.
Author Response
First of all, thank you for your useful comments, increasing the overall clarity and value of manuscript. We attempted to address all of them.
Point 1. The paper is short and very cohesive so needs some loosening to be easier to read and more suitable to use it.
Response 1: accepted, the text has been expanded to include wider description of all the manuscript components.
Point 2. Maybe it could be a bit longer in par about cultural ecosystem services. Please consider adding one more reference – Common International Classification of Ecosystem Services (CICES) or even CICES-Be (Turkelboom, Raquez, Dufrêne, 2013). Maybe it is too “local”, but maybe you will find it useful. As You wish.
Response 2: accepted, the reference to CICES has been added along with explanation of how it is relevant to the study (lines 45-53). CICES-Be classes are less relevant than CICES V. 5.1 due to the Belgian specifics, so we have limited ourselves with CICES only.
Point 3. Section of Material and Methods is properly and comprehensively presented. I only feel a shortage of “manually interpretation” explanation (lines 98-100). I mean not what did You find on the pictures but how have you been looking for it. That interesting method needs to be shown in the nice scheme, please try to provide it.
Response 3: accepted, the explanation of the used methods has been expanded and additionally discussed in Discussion (lines 111-115, 123-130, 186-198, 303-323); the research workflow scheme has been added as well (Figure 2).
Point 4. The scope of the work is properly and comprehensively presented in the part of the results. Only figure 2 is too small and looks illegible.
Response 4: accepted, Figure 2 has been split to the two figures to increase the overall legibility of results (Figures 5-6).
Point 5. The conclusion also seems to be well done, answering points signalised in the Introduction, but lacking more information about possibilities of possibilities for wider use the scheme of research. It will be sound too local without it.
Response 5: accepted, the possibilities for the wider use of research have been described in a more comprehensive manner (lines 368-379).
Point 6. There is a large appendix after the main text, showing all the research data. For me, it could be compressed and partially used it the main text to make it lighter and easier for reading that way.
Response 6: accepted, the materials from the appendix have been compressed and partially used in the sections 3.1 and 3.2 (Figures 1 and 7). Only tables and plots with the research data, supporting the plots in the section 3.2, have been left in the appendix.
Reviewer 3 Report
The study adopts an interesting approach in studying cultural ecosystem services based on pictures uploaded in the social networks in combination with some basic landscape indices to investigate human preferences towards some landscape attributes. Despite the significant degree of bias inherent in data collected from social networks (most of them are acknowledged by the authors), they constitute a significant source of information regarding human’s perceptions on various aspects of natural environment. The manuscript is well written and reads nicely and the few minor mistakes I am sure they will be corrected in a revised version of the manuscript.
The introduction provides all the background information, necessary for the reader to understand the topic of the manuscript, it has a clear problem statement and the objectives are clearly presented in the last paragraph of the section.
In the materials and methods section, I would like to see a slightly more detailed description of the LCI calculations. Although the references are helpful, I believe a more detailed description will help the readers to follow the manuscript easier.
The results are presented with a good combination of figures and text while the discussion is in accordance with the results with no unnecessary speculations. In the discussion the limitations of the approach are also presented.
With all the above being said I believe this is an interesting manuscript worth to be published after a minor revision.
Author Response
First of all, thank you for your useful comments, increasing the overall clarity and value of manuscript.
Point 1. In the materials and methods section, I would like to see a slightly more detailed description of the LCI calculations. Although the references are helpful, I believe a more detailed description will help the readers to follow the manuscript easier.
Response 1: accepted, the description of the LCI calculations has been extended with more formulas and explanation (lines 177-196).
Reviewer 4 Report
Thank you for the chance to contribute to your paper.
Below are some contributions to be made to improve the paper.
Abstract.
It presents a precise, well-organized abstract. However, it is expected that the results obtained at the end of the study are highlighted at the end of the abstract.
Introduction.
The introduction has some spelling errors. The authors need to review the text.
For an academic text, it is necessary to put the main contribution of science made by the paper.
The organization paragraph of the paper is missing in the last paragraph of the introduction.
Materials and Methods
Place equations with unique symbols, not with callouts.
Results.
The settings of the model or features need to be presented.
Figures and results need to be better described.
Conclusion.
In conclusion, the positive and negative aspects of the approach need to be presented.
Future works, presenting the extensions of the paper, need to be presented.
Author Response
First of all, thank you for your useful comments, increasing the overall clarity and value of manuscript. We attempted to address all of them.
Point 1. It presents a precise, well-organized abstract. However, it is expected that the results obtained at the end of the study are highlighted at the end of the abstract.
Response 1: accepted, the results description in the abstract follows the order in which they appear in the text: CES mapping, landscape organisation indices mapping, relation to the land cover (lines 24-27).
Point 2. The introduction has some spelling errors. The authors need to review the text.
Response 2: accepted partially; the manuscript had undergone proofreading with Elsevier English editing service prior to the first submission. However, to ensure the English quality, additional proofreading was provided by a native speaker Ms. Edith Chenault, who is an expert in science writing and editing.
Point 3. For an academic text, it is necessary to put the main contribution of science made by the paper.
Response 3: accepted partially; the scientific contribution is directly presented in the discussion part (line 287 and further); the introduction discusses rather the problems, which are further addressed and therefore, indicate the potential contribution (lines 41, 64-65, 81-83).
Point 4. The organization paragraph of the paper is missing in the last paragraph of the introduction.
Response 4: accepted, the landscape organisation metrics are discussed in the introduction (lines 76-83).
Point 5. Materials and Methods. Place equations with unique symbols, not with callouts.
Response 5: this comment is not clear. Could you, please, explain your idea?
Point 6. Results. The settings of the model or features need to be presented.
Response 6: accepted, more information about the Wilcoxon test is presented on Figure 6 and its caption. Appendix contains all the other information about the statistical analysis.
Point 7. Figures and results need to be better described.
Response 7: accepted, more figures with wider description (lines 214-218, 235-237, 277-280) have been added to ensure the clarity of results. Analysis presented on the Figures 5 and 6 is designed now for the areas with median NDVI values lower than 0,1 and higher than 0,1 to see the differences for rather vegetated and non-vegetated areas.
Point 8. In conclusion, the positive and negative aspects of the approach need to be presented.
Response 8: accepted, in addition to the positive aspects (lines 366-372), negative aspects are also described (lines 376-378).
Point 9. Future works, presenting the extensions of the paper, need to be presented.
Response 9: accepted, further directions and possibilities for further research have been expanded (lines 368-376).
Round 2
Reviewer 4 Report
Dear authors. Thanks for improving the paper. Some terms have been met, and others remain to be done. I will try to be more evident in this review to avoid rework.
The paper organization paragraph is missing. In this case, it would be the following example.
Section II of the paper will present the concepts that guide the study. Section III presents the model. You can view existing papers in the journal with examples of how to build this paragraph.
The second confusing point was about the presented equations. Generally, an equation in the precise pattern has only variables, not texts. An example is equation 1 that presents: ???? ??????????y. This name to represent a context is not ideal. In equations, a symbol is used to represent this value. For example, you can use Greek letters like beta, sigma, alpha to represent this concept of GLCM.
The other items were attended completely.
Author Response
Dear reviewer, thank you for additional clarifications.
Point 1. The paper organization paragraph is missing. In this case, it would be the following example.
Section II of the paper will present the concepts that guide the study. Section III presents the model. You can view existing papers in the journal with examples of how to build this paragraph.
Response 1: accepted, the organisation paragraph has been added, following the example of some published papers (lines 86-92).
Point 2. The second confusing point was about the presented equations. Generally, an equation in the precise pattern has only variables, not texts. An example is equation 1 that presents: ???? ??????????y. This name to represent a context is not ideal. In equations, a symbol is used to represent this value. For example, you can use Greek letters like beta, sigma, alpha to represent this concept of GLCM.
Response 2: accepted, the GLCM Homogeneity in equation 1 is presented as abbreviation GLCMH, adopted in the remote sensing papers, for example https://doi.org/10.1007/s10661-014-3681-1 and https://doi.org/10.1016/j.ufug.2018.01.021 In text we use GLCM Homogeneity in full to increase the readability of the paper.
