Development of a monthly to seasonal forecast framework tailored to inland waterway transport in central Europe

Meißner, Dennis; Klein, Bastian; Ionita, Monica

doi:https://doi.org/10.5194/hess-21-6401-2017

Articles | Volume 21, issue 12

https://doi.org/10.5194/hess-21-6401-2017

© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 3.0 License.

Special issue:

Sub-seasonal to seasonal hydrological forecasting

https://doi.org/10.5194/hess-21-6401-2017

© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 21, issue 12

Research article

|

15 Dec 2017

Research article |

| 15 Dec 2017

Development of a monthly to seasonal forecast framework tailored to inland waterway transport in central Europe

Dennis Meißner, Bastian Klein, and Monica Ionita

Download

Final revised paper (published on 15 Dec 2017)
Preprint (discussion started on 29 May 2017)

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

SC1: 'Short comment on the literature review', Mehmet C. Demirel, 01 Jun 2017
- AC1: 'Answer to SC1: Short comment on the literature review', Dennis Meißner, 02 Jun 2017
RC1: 'Very useful analysis, lacking sound discussion against related work', Massimiliano Zappa, 15 Jun 2017
- AC2: 'Response to Reviewer 1', Dennis Meißner, 14 Jul 2017
RC2: 'Revier of hess-2017-293', Anonymous Referee #2, 08 Aug 2017
- AC3: 'Response to the comments of the anonymous referee #2.', Dennis Meißner, 22 Sep 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) (18 Aug 2017) by Fredrik Wetterhall

AR by Dennis Meißner on behalf of the Authors (22 Sep 2017) Author's response Manuscript

ED: Referee Nomination & Report Request started (25 Sep 2017) by Fredrik Wetterhall

RR by Christophe Lavaysse (12 Oct 2017)

Suggestions for revision or reasons for rejection

Review of the paper “Development of a monthly to seasonal forecast framework tailored to
inland waterway transport in Central Europe” by Meißner et al..
First, I would like to thank the authors for their responses and the modifications done that, in my opinion, improve and clarify the document. I still have only one minor comment related to a previous comment.

About the score of each individual member:
Q : Some forecast assessments are not well calculated. For instance P13L6, the score of the ensemble is calculated by averaging the score of each individual member. The purpose of an ensemble forecast is not to generate good forecasts of each member.
Reply : We decided to calculate the deterministic metrics MAE and MSE for each ensemble member first and calculate the mean afterwards. The idea behind this procedure is that besides assessing the accuracy we want to consider the spread / sharpness of the ensemble, too. The method we used rewards accuracy, but punishes spread. Of course, you’re right with your statement and we agree that an ensemble forecast should be verified primarily probabilistically. But on the other hand each ensemble member represents an equally likely scenario of future conditions. As this way of generating the MAE / MSE is even stricter than treating the ensemble as one single trace (the ensemble mean), we would like to keep the corresponding results for the evaluation within this study.

Comment : I think this statement is questionable. The purpose of an ensemble model is not to produce an ensemble of perfect members. If it was the case, the interest of probabilistic forecasts would be useless. In a practical point of view, and according to the skill scores of deterministic forecast at medium range lead time, the errors and the chaotic nature of the atmosphere requested probabilistic forecast to take into account the lag of predictability. This spread is needed to compensate the uncertainties of the model. It has been shown in studies, that is important to have this ensemble spread, especially for extreme events. I am still very perturbed by this methodology of assessing each individual member. According to Fig. 12, for example, the ensemble mean of S4 is not necessary good all the months but according to the ensemble spread, it is maybe possible to extract a robust information. So ‘punish’ spread is not necessary correct and there is, so far, no clear link demonstrated in between large/low spread and errors of the model in that specific application. Maybe the authors should modified their methodology to assess the uncertainties and to discuss that point.

Hide

RR by Massimiliano Zappa (17 Oct 2017)

ED: Publish as is (02 Nov 2017) by Fredrik Wetterhall

Short summary

Inland waterway transport is a commercial sector relying on hydrological forecasts on different timescales. This paper describes the set-up of a monthly to seasonal forecasting system for the German waterways. Multiple approaches are tested, compared and combined. Despite the predictive limitations on longer lead times, this study reveals the existence of a valuable predictability on monthly up to seasonal timescales. Forecast quality depends on forecast location, lead time and season.