Regional soil erosion assessment based on a sample  survey and geostatistics

Yin, Shuiqing; Zhu, Zhengyuan; Wang, Li; Liu, Baoyuan; Xie, Yun; Wang, Guannan; Li, Yishan

doi:https://doi.org/10.5194/hess-22-1695-2018

Articles | Volume 22, issue 3

https://doi.org/10.5194/hess-22-1695-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/hess-22-1695-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 22, issue 3

Research article

|

08 Mar 2018

Research article |

| 08 Mar 2018

Regional soil erosion assessment based on a sample survey and geostatistics

Shuiqing Yin, Zhengyuan Zhu, Li Wang, Baoyuan Liu, Yun Xie, Guannan Wang, and Yishan Li

Download

Final revised paper (published on 08 Mar 2018)
Preprint (discussion started on 23 Aug 2016)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'General comments', Anonymous Referee #1, 29 Sep 2016
SC1: 'Response to general comments by Referee #1', Shuiqing Yin, 03 Oct 2016
RC2: 'Regional erosion assessment in China', Anonymous Referee #2, 10 Oct 2016
- SC2: 'Response to main comments by Referee #2', Shuiqing Yin, 17 Oct 2016
AC1: 'Final response to the comments from two referees', Baoyuan Liu, 15 Nov 2016
AC2: 'An inquiry concerning the progress of hess-2016-394', Baoyuan Liu, 17 Nov 2017

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

ED: Reconsider after major revisions (further review by Editor and Referees) (01 Dec 2016) by Axel Bronstert

AR by Baoyuan Liu on behalf of the Authors (25 Jan 2017)

ED: Referee Nomination & Report Request started (15 Feb 2017) by Axel Bronstert

RR by Anonymous Referee #2 (06 Mar 2017)

RR by Anonymous Referee #1 (24 Mar 2017)

Suggestions for revision or reasons for rejection

Re-Review of the manuscript “Regional soil erosion assessment based on sample survey and geostatistics” by Shuiqing Yin et al.

General comments:

As already state in my first review the paper addresses an interesting topic, which would fit into the scope of HESS. Regional soil erosion assessment is still challenging due to the often-missing input data needed for such assessment.

The authors provided an extensive reply to all my general comments and overall the paper substantially improved. However, I am still not convinced that the paper can be published as it is and suggest major revisions. This has the following reasons:

1. The structure of the paper is unbalance. The author provide an extensive introduction (approx. 5 pages; whereas it is not clear why all the different approaches to estimate regional/national erosion need to be presented here in so much detail), while the entire results and discussion is similar in length. A lot of information from the literature would be better placed/discussed in the discussion.
2. The authors present an interesting interpolation scheme but the validation of their approach is week (a few lines in chapter 3.1). In general, I would expect two major parts of the paper: (a) methods and (b) results and validation of results. I would, for example, expect different goodness-of-fit parameters as well as a more extensive discussion where the model performs appropriate and where major errors can be expected. Errors in interpolation and PSU data! The comparison with the data of Guo et al. (2015) (in discussion) shows that the results are in a similar rage but this is not a validation.
3. Any kind of evaluation how sensitive the results are regarding quality of input data is missing? Which are the most important co-variables. At least some sensitivity analysis would be very helpful to underline the quality of the method.
4. The authors argued that the available slope data (30 m Aster or 90/30 m SRTM) are not good enough to be included as co-variables in their interpolation. I agree that these data are far from perfect, but compared to an interpolated R factor, soil information derived from a relatively coarse map (K factor), I assume that the slope data show less uncertainty. As slope is one of the dominant small scale modulator of soil erosion (compared to all other data used) I disagree to omit slope as co-variable from the interpolation.
5. I strongly suggest to remove model I and II from paper. This has two reasons: (a) It is obvious from the results (e.g. line 6-8 and line 17-19 on page 15 in tracked changed document) that the interpolation without taking land use into account leads to an underestimation of erosion on farmland and an overestimation in forested areas. This is obvious and not worth to be published. Comparing models III to V with models I to II (e.g. Fig. 9) is misleading. (b) Land use produces discrete borders resulting in specific non-continuous changes in soil erosion. An interpolation without taking the ‘steps’ in erosion into account will always produce artificial results (and e.g. in geostatistics would violate general assumptions of the method).

Hide

ED: Publish subject to revisions (further review by Editor and Referees) (02 Apr 2017) by Axel Bronstert

AR by Baoyuan Liu on behalf of the Authors (26 Jun 2017)

ED: Referee Nomination & Report Request started (04 Jul 2017) by Axel Bronstert

RR by Anonymous Referee #2 (25 Jul 2017)

Suggestions for revision or reasons for rejection

I am surprised that authors get back after long time. Authors have taken into account my comments but there are still some issues to resolve. I would propose a major revision as I have seen the comments from the other reviewer.

Major issues:
- Authors did not convince both reviewers that it is better to use their proposal instead of applying an USLE with the available resolution of input data.
- The erosion rates in forest lands are very extreme and not justified. This can be connected to the erroneous outputs from NDVI calculations (I am very concerned about the NDVI. This is also confirmed by the difference with the study of Guo.
- The whole issue with LS factor and how authors present it. Previous studies in European LS-factor and the study of Borrelli in soil erosion in Italy demonstrated that the finer the resolution, the best are the LS-factor estimates and soil erosion model outputs. I do not agree with the what authors state in P12L13-14. The same confusion is created in P17L20-23. In the past studies in Europe it was shown that the finer the resolution, the best representation of the topography and then the more accurate the estimate of LS-factor and erosion rates.
- Table 1: quite confusion. What are the numbers in the row below the land uses? Moreover model numbers are missing in this table
- The low importance given in the model approach for R-factor is preoccupying.
- General comment about the methodology: Interpolating the soil erosion modelled point estimates is less appropriate than the application of model using input layers. Authors do not make clear statements of this and readers may be confused in making interpolation of erosion points even if they have the input layers. This point should be better addressed.
Some other issues:
- P2L15-15: The 2 sentences on water & wind erosion are repetitive
- P2L20. I don’t agree with this sentence. It maybe that land use is improved and the soil erosion risk decreased.
- P3L3. Experts have always expertise
- P3L29: “seamless” is not appropriate.
- P5L3-5: The sentence is confusing
- P6L25L : for first time you present the factors R, K, L, S….B, E: and the reader does not know what those factors are. In the next page you describe them…but there should be a logical sequence.
- Eq (1): What is uk? Describe below..
- P7L27: soil families are not a good term
- Equation 2: What is q?
- What are the measurement units of factors in equation 2?
- L10P17: What is f(,)
- Tabl3e 5: What not provide the total soil loss and discuss on the comparison of soil loss with sediment loss?

Referee Report: PDF

Hide

RR by Anonymous Referee #3 (06 Dec 2017)

ED: Publish subject to minor revisions (review by editor) (07 Dec 2017) by Axel Bronstert

AR by Baoyuan Liu on behalf of the Authors (16 Dec 2017) Author's response Manuscript

ED: Publish as is (17 Jan 2018) by Axel Bronstert

AR by Baoyuan Liu on behalf of the Authors (21 Jan 2018)