SMODERP2D—Sheet and Rill Runoff Routine Validation at Three Scale Levels

Round 1

Reviewer 1 Report

This is a study of the calibration and validation of SMODERP2D on the Bykovicky catchment. SMODERP2D is a sheet and rill runoff model developed by the Department of Landscape Water Conservation at the Czech Technical University in Prague. The authors validated the model using experimental and observation data from three scales, which provided useful information on the model's applicability. The three scales included 16 m^2 (simulated rainfall), 50 m^2 (natural rainfall events), and 36.6 ha (natural rainfall events). However, the writing of the manuscript is not clear, and some of the essential details might have been lost in the process of drafting the manuscript. Therefore, I recommend that the manuscript be substantially revised to clarify the study's critical issues.

First, it appears that the manuscript was not thoroughly proofread, as there were numerous misspelled words. For example, "safety site" (side) in line 173, "tipping bucked" (bucket) in lines 236 and 258, and "DMT quality" (DTM) in lines 448 and 450 are just a few of them (more examples at the end). Furthermore, the MDPI template sentences are still included in the manuscript and have not been removed (lines 408-410). The numbering of the figures is likewise incorrect. Figure 1 on page 6 should have been Figure 3, and the subsequent figures should be renumbered Figures 4 to 10. All of these caused readers to lose faith in the manuscript's quality.

Second, one major mistake that stood out is the misuse of "calibration" or "calibrated" to represent "validation" in lines 352-353. The two terms refer to different stages of modeling work, and I am sure the authors were aware of the differences. Unfortunately, they did not write properly and then proofread the document to avoid such errors.

Third, the most critical issue deserving the authors' attention is the missing model calibration and validation detail, which is the manuscript's primary focus. According to the authors (lines 243-244), only four measurements were made for bare and vegetated soil separately. How did the authors calibrate three parameters and obtain statistically satisfying results with only four measurements? What was the purpose of the two-step calibration? How did the authors derive the three parameters in Eq. 1 (infiltration process, surface retention, and effective precipitation) and then Eq. 2 (X, Y, and b) based on the measurements? Please provide more explanations. Furthermore, how did the authors calculate the standard deviations? According to Table 2, all standard deviations were zero except for one. What was the reason for this? Then, how did the authors apply these numbers to three different types of validation (mean, mean + std, and mean - std)? Also, why were the values of Y and b the same for both bare and vegetated soil? "The surface flow equation parameters did not change because those values are obtained from soil texture, which is the same at both plots" (lines 330-331); what exactly did the authors mean?

Fourth, in addition to the issues raised above, the authors' figures may be improved to make them more readable. Many of the legends are too small to be legible as they stand now, and Figure 3 on page 6, Figure 9 on page 13, Figure 10 on page 14, and Figure 11 on page 16 are examples of this. The authors should also replace the correlation coefficient (r) in Table 3 with the Nash-Sutcliffe efficiency (NSE).

Finally, as for the sensitivity analysis, what values were assigned to the parameters in the 250 simulation runs?

Overall, I believe the manuscript's significance is obscured by its writing, and I hope to read more specifics in the revised version of the manuscript.

Here are a few additional comments:

- Table 3: Correlation coefficient is not a good metric for this. Why not use Nash-Sutcliffe efficiency (NSE)?

- Table 4: Misspelled word in the caption (fir). Inconsistent date format in Table 3 and Table 4.

- Figure 1 on page 6: This figure should be renumbered Figure 3. The legend of the figure is too small to be legible. What is DMR?

- Figures 5 to 10: These figures should be renumbered Figures 4 to 9.

- Figure 8b: extra letter n

- Figure 9: The legends are too small to be legible.

- Figure 10: The legends are too small to be legible.

- Figure 13: This figure should be renumbered Figure 10.

- Line 116: are shown

- Line 149: filled

- Line 159: sheet

- Line 173: side

- Line 195: decreases

- Line 224: Figure 3b

- Lines 281-282: The meaning of the sentence is unclear.

- Line 285: Figure 3b

- Line 302: underestimates

- Line 303: overestimates

- Line 355: There was no figure corresponding to 7/6/2010.

- Line 360: Do you mean 8/7/2012, as shown in Table 3?

- Line 367: prevalent?

- Lines 408-410: Delete!

- Line 438: redundant word (case)

- Line 448: DMT?

Author Response

Thank you for your comments on our article, they will certainly lead to a better understanding of the final text. Responses to your comments are provided in attached word document here and are reflected in the second version of the manuscript.

Author Response File: Author Response.pdf

Reviewer 2 Report

Dear authors,

thanks for sharing you research. In your paper you focus on the prediction of rills by the SMODERP2D model especially the location where rills occur after heavy rainfall events. I think that farmers are well aware where in their fields this may occur. I think the added value of your work lies be more on the contribution of rill and ephemeral gully erosion on the total soil loss as this is the weak point of empirical models like RUSLE (don't use USLE in your paper is this is an old version) as you also stated in your previous work published on this SMODERP2D model. 

I think the three way steps in which you use measurements at different scales for the calibration and validation of the model is well done. In the text  the terms calibration and validation are not always presented in the correct order.  First calibration should take place and then apply the calibrated model to validate it for measured events in the field. 

Some sections need to be corrected by a native speaking English editor.

More detailed comments can be found in the attached document.

Comments for author File: Comments.pdf

Author Response

Thank you for your comments on our article, they will certainly lead to a better understanding of the final text. Responses to your comments are in attached pdf document in links to your comments. 

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

This is the second review of a study of SMODERP2D on the Bykovicky catchment, which described the procedures for calibrating and testing the SMODERP2D rainfall-runoff erosion model on an experimental field plot with three types of data at varying scales. When I reviewed the manuscript for the first time, I found that it was missing many important details and was difficult to comprehend. The revised manuscript, on the other hand, is a pleasant surprise and a significant improvement over the previous version. In my opinion, the authors did an excellent job editing the paper in clear English and offering numerous important explanations to make the study more thorough and significant. As a result, I recommend minor revision prior to publication acceptance.

My recommendation is based on a couple of minor flaws that remain in the revised manuscript. First, a new author was added to the manuscript, but his role was not specified in the "author contributions" section.

Second, I'm still perplexed as to why Table 2's STD of X is 0.94 for bare soil and 0.00 for vegetation. Is it possible for the authors to explain it again?

Third, the issue with the inconsistency of date formats was not resolved. The date on line 381 is 7/8/2010, but the date on line 453 (as well as Figures 6-7 and Table 3) is 8/7/2010.

Finally, the rests are minor flaws and grammatical errors:

- Lines 23 and 224: Býkovický or Býkovice catchment?

- Lines 212, 214, and 216: sheet (in the equations)

- Lines 252-255: Exactly how many plots? Two bare soil plots and two additional vegetated plots?

- Line 332: SR (not LS) and NR (not SR)

- Table 1: underestimates (not inderestimates)

- Line 362: STD (not SDT)

- Line 391: SMODERP2D

- Line 408: Figure 7 (not Figure 10)

- There was no reference of Figure 9 in the text.

- Line 488: Where is Fig. 13b?

Author Response

Thank you for your second round of comments. Responses to your comments (italic) are in bold here and are reflected in (and linked to) the second version of the manuscript. We hope that lead to a better understanding of the text.

My recommendation is based on a couple of minor flaws that remain in the revised manuscript. First, a new author was added to the manuscript, but his role was not specified in the "author contributions" section.

We included the new author in the “author contributions”.

Second, I'm still perplexed as to why Table 2's STD of X is 0.94 for bare soil and 0.00 for vegetation. Is it possible for the authors to explain it again?

We have attempted to better explain it in the text on lines 352 to 359 (the version without revisions tracking). In the case of the parameters of the runoff equation, the attempt was not to use these parameters for calibration. Their derivation is based on setup measurements on a laboratory rainfall simulator from simulations on several soil textures and several slopes (was not published in a international journal). The X parameter of equation 3 explains, besides the texture, also partly the surface microtopography.

Third, the issue with the inconsistency of date formats was not resolved. The date on line 381 is 7/8/2010, but the date on line 453 (as well as Figures 6-7 and Table 3) is 8/7/2010.

We fixed the date in the text at line 381.

Finally, the rests are minor flaws and grammatical errors:

- Lines 23 and 224: Býkovický or Býkovice catchment?

We changes in the abstract Býkovický to Býkovice.

- Lines 212, 214, and 216: sheet (in the equations)

Corrected.

- Lines 252-255: Exactly how many plots? Two bare soil plots and two additional vegetated plots?

We hopefully clarified the sentence at the L255.

- Line 332: SR (not LS) and NR (not SR)

Fixed.

- Table 1: underestimates (not inderestimates)

Corrected.

- Line 362: STD (not SDT)

Corrected.

- Line 391: SMODERP2D

Changed.

- Line 408: Figure 7 (not Figure 10)

Corrected.

- There was no reference of Figure 9 in the text.

We add the reference in the line 488 and 489

- Line 488: Where is Fig. 13b?

No, it supposed to be Figure 9. (see the previous comment)

Reviewer 2 Report

Thanks for the resubmitted manuscript. The manuscrips has been improved well.  Two minor things on page 6 (see attached file) that need to be corrected.

Comments for author File: Comments.pdf

Author Response

Thank you for your second round of comments. The corrections were done and the reaction on the reviewer’s comments in in the PDF file.

Author Response File: Author Response.pdf