Improving operational flood ensemble prediction by the  assimilation of satellite soil moisture: comparison between  lumped  and semi-distributed schemes

Alvarez-Garreton, C.; Ryu, D.; Western, A. W.; Su, C.-H.; Crow, W. T.; Robertson, D. E.; Leahy, C.

doi:https://doi.org/10.5194/hess-19-1659-2015

Articles | Volume 19, issue 4

https://doi.org/10.5194/hess-19-1659-2015

© Author(s) 2015. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/hess-19-1659-2015

© Author(s) 2015. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 19, issue 4

Research article

|

09 Apr 2015

Research article |

| 09 Apr 2015

Improving operational flood ensemble prediction by the assimilation of satellite soil moisture: comparison between lumped and semi-distributed schemes

C. Alvarez-Garreton, D. Ryu, A. W. Western, C.-H. Su, W. T. Crow, D. E. Robertson, and C. Leahy

Download

Final revised paper (published on 09 Apr 2015)
Preprint (discussion started on 23 Sep 2014)

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

RC C4232: 'review', Anonymous Referee #1, 02 Oct 2014
RC C5356: 'Review', Uwe Ehret, 28 Nov 2014
AC C5810: 'Authors final response to referees', Camila Alvarez-Garreton, 26 Dec 2014

Peer-review completion

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

ED: Reconsider after major revisions (07 Jan 2015) by Erwin Zehe

Dear Mrs. Alvarez-Garreton,

I had a close look at your manuscript, the reviewer comments and your responses to the reviews. I agree with reviewer 1 that your study addresses an important issue in a thorough and innovative manner. I also agree with both reviewers, that there are several major points that need to be addressed within major revisions, before the paper is acceptable:

- As pointed out by reviewer 1 the adequacy of the model structure for semi-arid reasons needs to be reflected. As it stands it is not clear how soil moisture in the model controls the splitting of rainfall into direct runoff and subsurface runoff and whether this is appropriate for the governing runoff processes in the study area (which are by the way not discussed). I think the paper will greatly benefit from a more detailed explanation of the soil moisture accounting and runoff generation (including equations).
- The true added value of SSM assimilation can only be evaluated, when using a model that is calibrated based on the full information on stream flow. I think this could be easily achieved for the semi distributed model version.
- As already pointed out by reviewer I the method section needs to be better explained, variables and indices should be selected with greater care. E.g. Equation 1 the summation depends on i and the state is also i. This is not wrong, but misleading. It is also puzzling to switch among variable names.
- In the present form you use many abbreviations, this is partly confusing.

There are better ways to keep things concise. Last not least the paper would greatly benefit from a more precise language and a thorough proof reading e.g.:
- Page 10637 What is meant with water states in a model?
- What is meant “with unique runoff processes in semiarid catchments” do you mean distinct runoff processes?
- Page 10638 “ …for improving stream flow within” – maybe the word predictions is missing…

Best regards and best wishes for 2015,

Erwin Zehe

Hide

AR by Camila Alvarez-Garreton on behalf of the Authors (10 Feb 2015) Author's response Manuscript

ED: Referee Nomination & Report Request started (13 Feb 2015) by Erwin Zehe

RR by Anonymous Referee #1 (14 Feb 2015)

Suggestions for revision or reasons for rejection

I have carefully read the revised manuscript and the reply to reviewers and I believe the authors have successfully addressed the comments raised by the reviewers. Specifically, the major issue of the biases open loop simulation was solved and currently the paper shows a correct and clear analysis. As I already mentioned in the previous review, I believe that studies on soil moisture data assimilation into rainfall-runoff are strongly required and this study performs a good analysis and also introduces new techniques. Therefore, the paper deserves to be published.

The only major comment is related to the results after the assimilation of satellite soil moisture data that are unchanged with respect to the first submission. After the correction of the bias in the open loop simulation, I believed that also the results of the “updated” simulation were different. Am I wrong? Do the authors considered the bias corrected streamflow ensemble in the assimilation run? Can the authors add an explanation for that?

In the replies, it reads that different rainfall-runoff models were considered and that, actually, the PDM performed the best. I suggest adding in the paper as most hydrologists do not expect this finding. PDM was developed in a humid climate, and hydrologists will expect that it can’t be used in a semiarid climate. Therefore, this point should be mentioned in the paper. Otherwise, readers could think that the results shown in the paper are strongly influenced by the selection of a possibly wrong rainfall-runoff model (this was my thinking at the beginning), but this is not the case!

I would also revise the abstract. From its reading, only the positive impact of soil moisture assimilation is highlighted and not the issues encountered in the overall procedure: need of bias correction of streamflow, large errors in the simulation at ungauged sections, little or no improvement in terms of Nash efficiency. In my opinion, these points should be underlined in the abstract.

Finally, some small errors are present in the new text added in the paper. For instance, at lines 176-177, something is missing. At line 518, the reference to equation 12 is not clear. At line 898, it reads two times theta^ams. I suggest to check the full text.

Hide

RR by Uwe Ehret (24 Feb 2015)

ED: Publish subject to minor revisions (Editor review) (02 Mar 2015) by Erwin Zehe

AR by Camila Alvarez-Garreton on behalf of the Authors (06 Mar 2015) Author's response Manuscript

ED: Publish as is (09 Mar 2015) by Erwin Zehe

AR by Camila Alvarez-Garreton on behalf of the Authors (10 Mar 2015)

Short summary

We assimilate satellite soil moisture into a rainfall-runoff model for improving flood prediction within a data-scarce region. We argue that the spatially distributed satellite data can alleviate the model prediction limitations. We show that satellite soil moisture DA reduces the uncertainty of the streamflow ensembles. We propose new techniques for the DA scheme, including seasonal error characterisation, bias correction of the satellite retrievals, and model error representation.