Short-Term Hydrological Forecast Using Artificial Neural Network Models with Different Combinations and Spatial Representations of Hydrometeorological Inputs

Round 1

Reviewer 1 Report

Thank you for allowing me to review this article.

The manuscript entitled "Short-term hydrological forecast using Artificial Neural Network models with different combinations and spatial representation of hydrometeorological inputs" is an original contribution, and the topic is interesting for readers of the WATER journal.

The presentation is fine. The manuscript can be published as it is.

Author Response

We would like to thank Reviewer 1 for agreeing to read and review our manuscript.

We thank the reviewer for his approval to publish the article as it is.

Reviewer 2 Report

The paper presents an Artificial Neural Network (ANN) model framework for short-term hydrological forecast studying different combinations and spatial representation of inputs. It is a suitable topic for Water MDPI journal. The manuscript is professionally written, clear, and easy to read. The results are relevant, well presented and discussed.

I recommend a minor revision of the manuscript following my comments below.

• Line 79: Wouldn’t it be “three” instead of “two”?
• Lines 153-161: This paragraph should be removed.
• Line 193: It would rather be “7” instead of “9”.
• Line 205: this “virtual hydrological environment” could be better explained.
• Table 2: The simulation options should be at least briefly explained in the text.
• Figure 3: This figure should be more detailed in the text.
• Line 250: Double “for”.
• Line 321: The authors have not mentioned the “cross-validation approach before”, which makes this sentence being confusing.
• Figure 5: This figure should be more detailed in the text.
• Figures 7 and 8: A graphic legend could be added to these figures, to make it easier to the reader to identify the representation of each line.
• Figures 9, 10, 11, 12, and 13 are of low resolution.
• Line 818: Revise this sentence.

Author Response

Please see the attachment.

Author Response File: Author Response.doc

Reviewer 3 Report

Dear Editor, dear Authors

The authors evaluate what the Artificial Neural Network (ANN) brings for modeling and forecasting stream flows at the outlet of two watersheds used as case studies. It is not strictly speaking a research article but more a methodological approach aimed at evaluating the limits of techniques that have been well known for a few decades.

The watersheds studied have a very large surface area, several tens of thousands square kilometers for one of them, with a very heterogeneous land use, which makes the project very daring. The second basin is smaller with a more homogeneous land use, which means that the two case studies complement each other.

Proven distributed models are used to estimate the spatialized state variables that account for the heterogeneity of the basins from a hydrographic point of view, the water content at the surface and at depth.

The learning and cross-validation techniques are carried out over 16 consecutive years. The results are reviewed step by step using different inputs, soil water content at various depths, and past stream flow with different combinations, and with different degrees of spatial resolution according to the forecasting horizon.

The evaluation of the techniques is exhaustive, avenues are given to improve the quality of the different inputs, the references up to date. I propose that the article be published as it is with some minor revisions intended to improve the readability of the manuscript.

Minor revisions:

Remove lines 153-161

Figure 3, all the parameters should be explained.

Table 4 needs more explanations.

Author Response

Please see the attachment.

Author Response File: Author Response.doc

Reviewer 4 Report

The article still needs to be modified and explained more detailed.

 

Recommendations for addition:

1. In the introduction, it is necessary to remove the content unrelated to the study; please see the paragraph between the lines 153-161.
2. Main section 2. Materials and Methods, Line 163: Study Area, recommendation to present a subsection with numbering. Line 183-Figure 1: add some labels because the study area's presentation is not clear.
3. Line 198: Figure 2: Graphical presentation of The climatic and hydrological regimes is not readable, choose a better form of presentation, what does presents the axis x and axis y, add comment also.
4. Line 279: ANN model, present such as a subsection; see the comment from point No. 2.
5. Line 319: Methodology, the same recommendation as in comment before, please accept it in all parts of document.
6. Figure 8, Figure 9 are illegible.
7. Despite its scale, the discussion has insufficient links to works of similar importance. Compare the achieved results with a greater emphasis on scientific studies of given importance.
8. Do not authors consider supplementing datasets from results via GitHub? It would significantly increase the interest of readers.
9. The study needs to be minor revised.

Author Response

Please see the attachment.

Author Response File: Author Response.doc