Simulation and Spatio-Temporal Analysis of Soil Erosion in the Source Region of the Yellow River Using Machine Learning Method
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsI have reviewed the manuscript and found that the subject is interesting but that manuscript requires some interventions given below to ease its reading. It does not appear as a scientific paper, but rather as a study report. Therefore, there is need to insert more of a ‘scientific’ flavor by improving style of presentation. Authors should avoid ‘we’ and other personalization, i.e. to write in the ‘third person’ style’, and to correct all typos and some grammar errors. Even though I am not native speaking English, proofread is highly recommended.
Below are other comments and suggestions:
1. Title: Maybe slightly modified title: ‘Simulation and Spatio-Temporal Analysis of Soil Erosion in the Yellow River Source Region Using Machine Learning Methods’
2. Abstract: I suggest avoiding detailed information in ‘key findings’ part where many numbers appear. Rather keep them for introduction part and in abstract be more descriptive. Avoid here and everywhere in manuscript personalized wording such as ‘we’, ‘our’, etc. Also add as the last sentence what next should be done from the research point of view.
3. Introduction: It is well written with relevant references. To make it easier to read, I suggest dividing this section in several paragraphs with ‘topics’ elaborated in related references. Longest central paragraph ends with just two sentences about the random forest (RF) model which is proposed in the manuscript. Put them in separate paragraph and slightly extend it by giving details on what are the characteristic of this model and justification of proposing it. Some details given in abstract (key findings) could be given in separate closing paragraph starting with sentence ‘Objectives of this study were …’.
4. Materials and Methods (Study Area). Fig. 1 is clear, however, there is no information where it is in China. I suggest adding small picture of China with such indication. Mathematical part is mostly correctly presented; however long paragraphs still exist where meaning of variables in formulas is explained. I suggest rewriting for ease reading. Table 3 belongs to section Results.
5. Results and Discussion: These parts can be reduced by removing some figures into appendix.
6. Conclusions: It is unusual that in initial part of conclusions there are details ‘presented by numbers’. They belong to Results and Discussion. Here only descriptive conclusions should take place, followed by agenda for future research.
7. I suggest careful reading the manuscript after corrections given above, to improve style and grammar.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsDear authors:
It's my honor to review your manuscript. The manuscript “Simulation and Spatio-Temporal Analysis of Soil Erosion in the Source Region of the Yellow River Based on the Machine Learning Method” addresses an interesting topic, which adhere to Land journal policies. I think the study presents interesting findings suitable for this journal.
This is a very important study that aims to connect established approaches for modeling changes in soil erosion, from the aspect of spatio-temporal analysis of soil erosion in the source region of the Yellow River based on the machine learning method.
The results of the research are clearly presented, and the conclusions are supported by the results. Research methods are scientifically relevant and adequately described. References cited in this manuscript are appropriate and relevant to this research
I suggest that the manuscript be published after minor revision.
1) I suggest you make a clearer map of the study area.
Show the geographic location of your study area on a national scale. Where is your study area located within China?
2) I have several proposals for discussion.
Missing from the discussion is any validation or uncertainty analysis of model outcomes with measured values.
Are there measurements of suspended sediment on the rivers in the study area? Moreover, maybe there are empirical studies done on soil erosion or river suspended sediment in the region.
How do the modeled erosion in your research relate to the measured ones?
It would be desirable to add some additional discussion where you cross-reference these modeled outcomes with some empirical evidence. Without validation, it is impossible to know how representative your model outcomes are for real-world trends. Some qualitative cross-references with empirical observations or other studies can already increase the trust of the model outcomes.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
