Estimating spatially distributed soil water content at small watershed scales based on decomposition of temporal anomaly and time stability analysis

Hu, W.; Si, B. C.

doi:https://doi.org/10.5194/hess-20-571-2016

Articles | Volume 20, issue 1

https://doi.org/10.5194/hess-20-571-2016

© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/hess-20-571-2016

© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 20, issue 1

Research article

|

02 Feb 2016

Research article |

| 02 Feb 2016

Estimating spatially distributed soil water content at small watershed scales based on decomposition of temporal anomaly and time stability analysis

W. Hu and B. C. Si

Download

Final revised paper (published on 02 Feb 2016)
Preprint (discussion started on 03 Jul 2015)

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

RC C2964: 'review', Anonymous Referee #1, 04 Aug 2015
RC C3002: 'review', Anonymous Referee #2, 05 Aug 2015
AC C3485: 'Response to the comments of referee #1 and referee #2', wei hu, 07 Sep 2015

Peer-review completion

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

ED: Reconsider after major revisions (17 Sep 2015) by Theresa Blume

AR by Wei Hu on behalf of the Authors (20 Oct 2015) Author's response Manuscript

ED: Referee Nomination & Report Request started (23 Oct 2015) by Theresa Blume

RR by Anonymous Referee #2 (17 Nov 2015)

RR by Anonymous Referee #1 (22 Nov 2015)

Suggestions for revision or reasons for rejection

Major comment

The authors have made a thorough job during the revision of their manuscript according to the reviewers comments. I however do not agree with the authors comply to the second reviewer´s request of including additional datasets into their study while at the same time maintaining the overall structure of the manuscript. As it is now, the manuscript targets both, a benchmarking of the new model against two additional datasets and a correlation analysis between model components and site factors at the Canadian site.
The problem is that the correlations are only discussed for the Canadian site whereas the two other sites are only very superficially introduced and discussed. I think introducing and discussing also the two other sites in similar detail like the Canadian site would make the manuscript too lengthy and complicated.

The manuscript would gain considerably if the authors would focus on either one of the two aims.

I personally recommend focusing on the introduction of the TA model at the Canadian site, including the comparison to the SA model and the discussion of correlations between site factors and TA model components.

An alternative would be introducing all three sites properly in the material and methods section and focus on a model benchmarking of TA and SA models. At the same time a discussion of correlations between site factors and model components at the Canadian site should be skipped.

I find that the manuscript is otherwise publishable, apart from some minor technical details. I therefore recommend another round of minor revisions.

Minor comments:

L71: equals or corresponds to ?

L122: Materials and methods: If the authors wish to keep the application of the models to the two other sites they should also be presented in similar detail as the Canadian site in the materials and methods section (see major comment).

L217: maybe better: "the spatial variance"

L243: I think "contribution" is not the correct term here (see comment below)

L256: The A_tn looks strange..

L321: of the original SWC

L363: I do not agree that something can contribute with more than 100% to another thing. sigma^2(M_tn) merely takes on such large values because it is countered by a negative covariance. I therefore suggest to avoid using the term "contributes" in this context (throughout the entire manuscript). I apologize for not having pointed this out in the first review. Since the percentages of the variance are not boundeed by 100% the authors should rather contrast it with the contribution of the other positive component in equation 8. Otherwise it is unclear whether the contribution was relly high or not. It occurred to me at this point that including a shrinkage regularization in the model fitting may be interesting in future studies.

L421 maybe better: "..validation methods. This is also supported by the increasing AICc.."

L542: just to make it easier for the reader to follow the manuscript: stable in time or in space?

Hide

ED: Publish subject to minor revisions (Editor review) (07 Dec 2015) by Theresa Blume

AR by Anne Brekerbohm on behalf of the Authors (18 Dec 2015) Author's response Manuscript

ED: Publish subject to technical corrections (17 Jan 2016) by Theresa Blume

AR by Wei Hu on behalf of the Authors (20 Jan 2016) Author's response Manuscript

Short summary

Spatiotemporal SWC was decomposed into into three terms (spatial forcing, temporal forcing, and interactions between spatial and temporal forcing) for near surface and root zone; Empirical orthogonal function indicated that underlying patterns exist in the interaction term at small watershed scales; Estimation of spatially distributed SWC benefits from decomposition of the interaction term; The suggested decomposition of SWC with time stability analysis has potential in SWC downscaling.