Identifying rainfall-runoff events in discharge time series: a data-driven method based on information theory

Thiesen, Stephanie; Darscheid, Paul; Ehret, Uwe

doi:https://doi.org/10.5194/hess-23-1015-2019

Articles | Volume 23, issue 2

https://doi.org/10.5194/hess-23-1015-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/hess-23-1015-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 23, issue 2

Research article

|

19 Feb 2019

Research article |

| 19 Feb 2019

Identifying rainfall-runoff events in discharge time series: a data-driven method based on information theory

Stephanie Thiesen, Paul Darscheid, and Uwe Ehret

Download

Final revised paper (published on 19 Feb 2019)
Preprint (discussion started on 28 Aug 2018)

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

RC1: 'Interactive comment on “Identifying rainfall-runoff events in discharge time series: A data-driven method based on Information Theory” by Stephanie Thiesen et al.', Yiwen Mei, 13 Sep 2018
- AC1: 'Responses to comments posted by Dr. Yiwen Mei', Stephanie Thiesen, 01 Oct 2018
RC2: 'Review of hess-2018-427', Anonymous Referee #2, 27 Oct 2018
- AC2: 'Responses to comments posted by Referee #2', Stephanie Thiesen, 19 Nov 2018
EC1: 'Editor comment', Bettina Schaefli, 22 Nov 2018

Peer-review completion

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

ED: Reconsider after major revisions (further review by editor and referees) (22 Nov 2018) by Bettina Schaefli

AR by Stephanie Thiesen on behalf of the Authors (07 Dec 2018) Author's response Manuscript

ED: Referee Nomination & Report Request started (13 Dec 2018) by Bettina Schaefli

RR by Yiwen Mei (21 Dec 2018)

Suggestions for revision or reasons for rejection

Summary
The authors have addressed all my previous comments. In my opinion, the manuscript should be good for publication after some minor changes. Therefore, I suggest a minor revision.

Comments
• Section 3.2.1: In my opinion, there is no need to have more higher-level hierarchy in the section. I suggest to merge all those sub-sections under section 3.2.1. Same applied for section 3.2.3.
• Title of Section 3.2 and 3.3: I would suggest better titles like for example “Information theory method” and “Characteristic point method”, respectively, for section 3.2 and 3.3.
• Section 4: I would suggest to keep only 2 hierarchy here. The current section 4.1 seems redundant.
• Table 6: “Model-based group with 𝑄-based predictors” is not accurate. Perhaps call it “Model-based group with sorely 𝑄-based predictors”.
• P15 line 4: insert “.” After “repetitions”.
• Figure 7: I am not sure what criteria the “expert” used in manually identification of events. If I were the expert, I would group the last four events as two bi-model events. I feel like both ITM and CPM are doing better than the “expert” in these cases.
• P18 line 34: I would like to point out that the reason that CPM require precipitation data is because we need to study runoff coefficient and time lag in our original application. Without precipitation data, one can still use CPM to produce flow event period. This was implemented by Shen et al. (2017, Comprehensive Database of Flood Events in the Contiguous United States from 2002 to 2013) which developed a three-level flood event database. To produce the level 1 events, only the streamflow records were used and only the timing (begin, peak(s), end) of flood events were provided; level 3 events have all the event properties (e.g. runoff coefficient, event time lag, etc.,) that require the use of precipitation to compute. As the developer of the CPM program, I would say that the toughest aspect of implementing CPM could be the proper selections of model parameters. Also, I would like to take this chance to thank the authors for their idea of using a “synthetic true event database” to calibrate the model parameters. Though in my philosophy visual inspection by human is never a synthetic true, but it is a very clever attempt.

Hide

RR by Anonymous Referee #2 (10 Jan 2019)

ED: Publish subject to technical corrections (28 Jan 2019) by Bettina Schaefli

AR by Stephanie Thiesen on behalf of the Authors (05 Feb 2019) Author's response Manuscript

Short summary

We present a data-driven approach created to explore the full information of data sets, avoiding parametric assumptions. The evaluations are based on Information Theory concepts, introducing an objective measure of information and uncertainty. The approach was applied to automatically identify rainfall-runoff events in discharge time series, however it is generic enough to be adapted to other practical applications.