Accounting for three sources of uncertainty in ensemble hydrological forecasting

Thiboult, Antoine; Anctil, François; Boucher, Marie-Amélie

doi:https://doi.org/10.5194/hess-20-1809-2016

Articles | Volume 20, issue 5

https://doi.org/10.5194/hess-20-1809-2016

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

https://doi.org/10.5194/hess-20-1809-2016

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

Articles | Volume 20, issue 5

Research article

|

10 May 2016

Research article |

| 10 May 2016

Accounting for three sources of uncertainty in ensemble hydrological forecasting

Antoine Thiboult, François Anctil, and Marie-Amélie Boucher

Download

Final revised paper (published on 10 May 2016)
Supplement to the final revised paper
Preprint (discussion started on 30 Jul 2015)
Supplement to the preprint

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

RC C4097: 'Review on 'Accounting for three sources of uncertainty in ensemble hydrological forecasting' by Thiboult et al.', Anonymous Referee #1, 06 Oct 2015
- AC C4545: 'Answer to the Anonymous Referee #1', Antoine Thiboult, 31 Oct 2015
RC C4246: 'Referee comment', Anonymous Referee #2, 14 Oct 2015
- AC C4569: 'Answer to the Anonymous Referee #2', Antoine Thiboult, 31 Oct 2015

Peer-review completion

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

ED: Reconsider after major revisions (19 Nov 2015) by Thomas Kjeldsen

AR by Antoine Thiboult on behalf of the Authors (27 Jan 2016) Author's response Manuscript

ED: Referee Nomination & Report Request started (14 Feb 2016) by Thomas Kjeldsen

RR by Anonymous Referee #1 (19 Mar 2016)

RR by Anonymous Referee #2 (21 Mar 2016)

Suggestions for revision or reasons for rejection

General comments:
In general, the comments I had to the first version of the paper have been adequately addressed in the revised version. I find the paper suitable for publication in HESS when the minor comments given below have been addressed.

Specific comments:
Line 26. ‘hydrometeorological chain’ instead of ‘meteorological chain’.
Table 1. Define ‘P’ and ‘Q’ abbreviations used.
Line 151. I suggest to add ‘…daily time step starting at 6 EST (12 UTC)’.
Line 203-205. I suggest to reformulate: ‘For all hydrological models the same snow accumulation and melting module and evapotranspiration formulation have been used’.
Line 224-225. Reformulate: ‘… model spin up to provide better estimates of initial catchment conditions’.
Line 256. ‘output error and prior state estimate, respectively’ instead of ‘output error respect to …’.
Line 261-263. Not clear what ‘intrinsic variability’ and ‘epistemic uncertainty’ refer to here.
Line 290-294. Not clear. I don’t see the argument for not using a combined state-parameter update here. You use a multimodel approach to account for model structural uncertainty and do not explicitly account for parameter uncertainty.
Line 365. Reformulate: ‘… and would increase computational costs considerably’.
Line 446. What does ‘even if the global tendency is respected’ refer to?
Line 515. Would be good to add a sentence that summarises the seasonal analysis results.
Line 576-579. You may refer to a recent paper that addresses this issue: Zhang, D., Madsen, H., Ridler, M., Refsgaard, J.C., Jensen, K.H., 2015, Impact of uncertainty description on assimilating hydraulic head in the MIKE SHE distributed hydrological model, Advances in Water Resources, 86, 400-413.
The paper includes numerous linguistic errors. A careful proof reading of the paper is needed.

Hide

ED: Publish subject to minor revisions (Editor review) (30 Mar 2016) by Thomas Kjeldsen

Dear Authors,

Both reviewers were satisfied that their comments on the early version had been adequately addressed. One reviewer had some minor corrections which I consider reasonable (see below). I would therefore ask you to implement the suggested changes (or argue why they are not applicable) and upload a final manuscript (which will only be subject to review of the editor).

Regards,
Thomas

Specific comments:
Line 26. ‘hydrometeorological chain’ instead of ‘meteorological chain’.
Table 1. Define ‘P’ and ‘Q’ abbreviations used.
Line 151. I suggest to add ‘…daily time step starting at 6 EST (12 UTC)’.
Line 203-205. I suggest to reformulate: ‘For all hydrological models the same snow accumulation and melting module and evapotranspiration formulation have been used’.
Line 224-225. Reformulate: ‘… model spin up to provide better estimates of initial catchment conditions’.
Line 256. ‘output error and prior state estimate, respectively’ instead of ‘output error respect to …’.
Line 261-263. Not clear what ‘intrinsic variability’ and ‘epistemic uncertainty’ refer to here.
Line 290-294. Not clear. I don’t see the argument for not using a combined state-parameter update here. You use a multimodel approach to account for model structural uncertainty and do not explicitly account for parameter uncertainty.
Line 365. Reformulate: ‘… and would increase computational costs considerably’.
Line 446. What does ‘even if the global tendency is respected’ refer to?
Line 515. Would be good to add a sentence that summarises the seasonal analysis results.
Line 576-579. You may refer to a recent paper that addresses this issue: Zhang, D., Madsen, H., Ridler, M., Refsgaard, J.C., Jensen, K.H., 2015, Impact of uncertainty description on assimilating hydraulic head in the MIKE SHE distributed hydrological model, Advances in Water Resources, 86, 400-413.
The paper includes numerous linguistic errors. A careful proof reading of the paper is needed.

Hide

AR by Antoine Thiboult on behalf of the Authors (10 Apr 2016) Author's response Manuscript

ED: Publish as is (12 Apr 2016) by Thomas Kjeldsen

Short summary

Issuing a good hydrological forecast is challenging because of the numerous sources of uncertainty that lay in the description of the hydrometeorological processes. Several modeling techniques are investigated in this paper to assess how they contribute to the forecast quality. It is shown that the best modeling approach uses several dissimilar techniques that each tackle one source of uncertainty.