OpenForecast: An Assessment of the Operational Run in 2020–2021
Round 1
Reviewer 1 Report
I would like to see one year comparison of 7 day simulated and observed runoff at three gauges a) one with high NSE b) one with medium NSE c) one with low NSE .Select gauges at large basin without reservoirs .
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This study investigates the efficiency of a runoff forecasting system called OpenForecast in Russia. The authors claim that their system is able to provide 7-day ahead forecasts for numerous gauges across Russia. They try to address several questions in this paper in order to support their system and provide better insight. The manuscript includes some points that should be revised or improved to be considered as a scientific paper. Therefore, as a reviewer, I suggest that the manuscript can be accepted after minor revision. Comments and suggestions for improvement are presented below:
- The questions are provided at introduction part which provide a blueprint for the paper including consideration of social aspects (Question 5) of this system which is highly valuable. However, the same part is repeated at Result and Discussion (lines 153-160) as well as Conclusions (lines 337-344) sections which are redundancy and must be deleted.
- It is better to explain more about the hydrological models (HBV and GR4J) used in OpenForecast. Although both models are well-referenced, they need more explanation at least in the term of free parameters. The authors opened an important discussion about the free parameters; thus, more explanation would shed more light on this matter.
- At lines 137-138, authors claim that 30% of gauges could be considered representative. Especially after declaring that the gauges were selected based on data consistency (Line 135). Two questions popped up in mind in this regard.
First, what is the representativeness criteria? Based on the pervious studies, what percentage can be considered as representative (it needs a reference).
Second, how selected gauges based on a feature (data consistency) can be generalized for all other gauges? Is there any method to confirm this kind of sampling can be representative?
This is the only major problem in this paper which requires further explanation and references.
- The figure 5 needs additional clarification about the positive and negative amounts of differences between performances of calibration and hindcast.
- Violin plots are making the results very understandable; however, it would be great to use equal vertical axis for them as well (Figure 6 and A2). I know it is not applicable sometimes but it’s worth to try.
Comments for author File: Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Dear Authors of the paper ‘OpenForecast: an assessment of the operational run in 2020–2021’,
You did a very extensive assessment of the runoff forecasting system at a national scale. The assessment is versatile and provides a comprehensive characterization of the system. The paper is carefully organized, and the problem and methods are clearly explained. The style and graphics are correct.
However, I would recommend some minor changes by explanation of the issues given below so that the Reader has a better insight into the quality of the system:
- Please provide more information about the stations where NSE was a) less than 0 b) less than 0.5 in Fig 4 (Hindcast), namely what is the contribution of such stations to the total number of stations and (if possible) if there is a difference between them and the others as regards for example the catchment area, orography etc. Are there the same stations where both the GRJ4 and HBV models have NSE<0?
This is a valuable information for future users, eg. for local authorities, hydrological service, etc.
- Is it known how do NSE and KGE depend on catchment area or other characteristics?
- The caption to Fig 8: please add the names of the variables correlation of which is depicted.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
