Review Reports
- Selline Mutiso1,
- Keisuke Nakayama1,* and
- Katsuaki Komai2
Reviewer 1: Muhammad Nda Reviewer 2: Anonymous Reviewer 3: Anonymous
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe manuscript concept is quite intresting having been able to bring to fore suspended sediment morphology tracing using the XRF and XRD and then using the data deduced as input for a simple neural network analysis. But there are some drawbacks to be addressed presented below;
1. Line 79, "Bedded sediment" should read bedload sediment which is the proper phrase used for larger sediments close to the river bed.
2. Line 86, Collins et al. should be rewritten as Collins & Walling [21] and also check Line 92, 95, and 352 where the citation numbers were missing after the authors name and presented at the end of the sentence. For example, Ishida et al written alone should be Ishida et al [22] then continue.
Similarly, proofread the manuscript once more to get reed of some miss-match for instance in Line 56 “well-being of living population” seems out of context as used.
Lastly, using the NNA in the manuscript is a novel contribution in water resource application. The manuscript did not elaborate on the total number of data sets used in NNA algorithm and how they are subdivided into training, testing, and evaluation only a mention of this process in Line 359 – 361. Hence, the manuscript will be improved if the following inclusions are considered in the methodology and results section.
1. Total number of input data as well as the percentages used for training, validation, and testing in the methodology section.
2. NNA Performance indicators to show the accuracy and error level of NNA i.e., R2, MSE, MAE or any other acceptable and commonly used indicators in NNA related research, in the methodology section (equations and their description) and the arising results presented and discussed in the results section.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe manuscript "Estimation of sediment transport by a simple neural network" is a good contribution by the authors. I have the following queries:
1) Some of the references mentioned are not relevant as the manuscript refers to suspended sediments, other types of sediments and their impacts are not relevant to this study.
2) Title indicates "Estimation of sediment transport". However, abstract and manuscript work on suspended sediments. I found there is a conflict.
3) Authors use simple NNA tool. But nothing is mentioned about the data set, how the data is split into training and testing. Also the results of neural network are not included.
4) It looks more of chemical analysis of sediments rather than the measurement of suspended sediments.
5) Totally the manuscript leads to confusion about whether the reader understands the measurement of suspended sediments, the impact of land use on the generation of sediments, or the application of neural networks.
6) Hence, I suggest for the revision of manuscript to mainly highlight on suspended sediments and application NNA for estimation of the same.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsEstimation of sediment transport by a simple neural network
Review
Oct 16th, 2024
The objective of the manuscript is to use data for sampled sediment to infer the prominent sediment sources in a small basin in Japan. A neural network has been used for the data analysis.
While some misunderstanding may have arisen from my limited experience in machine learning, still I think that the content of the manuscript is not very clear. The description of the methods needs improvement; it is unclear which the target output of the analysis is, since sometimes from the text it seems to be a rate of suspended transport, while in the results are focused on contribution of different sources on what is found at the outlet (and, in this sense, a title like “estimation of sediment transport” would not be so appropriate).
My recommendation is to return the manuscript to the authors for a major revision.
18: it is not clear what the “output” should be. Please clarify.
31: why and how does the movement of suspended sediment create waves and currents? The statement is too generic.
42: in case there is another dam downstream of the flushed one.
80: to the outlet of the river basin?
101: a method for analysis cannot have an effect on sediment transport. Rather, it can account for nonlinear effects in the processes. Please rephrase.
130: what is the average time of suspended sediment concentration? Please clarify.
Fig 1: the red/yellow spot (colours are not defined) and the black/grey one do not seem to have the same shape. Please check.
142: not sure that a sampling procedure can be totally skipped, even if it is present in another work.
145: this proximity to the road comes for the second time (it was mentioned also in the abstract), but it is not clear why it is important. Maybe it will be understood later.
164: please define this “production”. Determining the particle size distribution does not determine the yield. Probably I misunderstood something, please clarify.
178-183: a description of the NNA does not fit the section title. Needs to be moved to somewhere else.
181: again (same as line 18), which is the output? Sediment transport rate, mentioned at line 240?
Fig 4 and 5: here, some results are obtained without the NNA, but just from the sediment composition. Basically, similarity of sediment to that at the outlet is used to infer a significant contribution of a certain origin. Would this need to be given more emphasis?
291: it is unclear how the NNA provides the percentage contributions.
In general, my highest uncertainty is in which is the target quantity to be estimated. Sometimes it seems that the authors intend to estimate a sediment transport rate (240, 349-358), but the results are never presented in terms of that quantity (rather, the weights, or contributions, are used). I am not following the line of thought.
In the conclusions, NNA just comes in a generic statement in the last two lines, while the other material seems related to the surface and sediment type.
Comments on the Quality of English LanguageNeeds improvement.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsAll the comments are addressed. I accept the manuscript
Reviewer 3 Report
Comments and Suggestions for AuthorsI am generally satisfied with the authors' revision; therefore, I am pleased to recommend acceptance.