Quantifying Uncertainty in Runoff Simulation According to Multiple Evaluation Metrics and Varying Calibration Data Length

Round 1

Reviewer 1 Report

This paper delves into the intricacies of runoff simulations, employing the Soil and Water Assessment Tool (SWAT) model in the context of the Yeongsan River basin. The study investigates the uncertainty in simulations by exploring the impact of varying calibration data lengths (ranging from one to 11 years) and employing multiple evaluation metrics. The topic is suitable for publication in Water, as it contributes to understanding uncertainty quantification in runoff simulations. To enhance the clarity and structure of the paper, I suggest emphasizing generalizable knowledge, providing detailed hydrological regime descriptions, offering comprehensive descriptions and comparisons of outflow regimes, and enhancing the methodology section to include crucial details about the calibration algorithm.

Conceptual Recommendations:

Generalizable Knowledge: The authors should explicitly highlight the generalizable knowledge that emerges from their work. For instance, can the optimal length of the calibration period for another river be determined based on the presented findings?

Data Selection: Address the question of data selection, especially when dealing with longer data series. How can one determine the most suitable part of the data for calibration? Isn't it more advantageous to use the entire available dataset, unless it is impractical due to the overwhelming volume?

Hydrological Regime Description: Provide a more detailed description of the hydrological regime of the investigated river. Is the regime relatively stable, or does it exhibit significant annual flow variations? Analyzing annual outflows, daily/monthly maxima, and minima can help determine the required calibration period.

Comparison of Outflow Regime: Conduct a detailed comparison of the outflow regime during the calibration and test periods. Utilize plots to illustrate variations across calibration periods with 1, 2, 3, … 11 years.

Annual Hydrograph: Include at least the annual hydrograph of the Yeongsan River to better understand the flow patterns and variations over time.

Methodology:

Calibration Algorithm: Provide an in-depth description of the calibration algorithm used in the study. Clarify whether calibrations are performed uniformly for all calculations. Include information on the number of iterations, convergence status, calibration time, and any relevant details affecting the calibration process.

Calibration Reproducibility: Address whether the calibration results are sensitive to chance. Have the authors attempted running calibration calculations more than once to assess the consistency of the results?

Conclusion:

By addressing these conceptual recommendations, the paper can enhance its clarity, structure, and overall contribution to understanding uncertainty in runoff simulations, ensuring its relevance to the broader scientific community.

no comments

Author Response

We attached reply file.

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper quantified the uncertainty of runoff simulations based on the calibration data length and selection of the evaluation metrics.The theme is relatively novel and the significance is high.Here are some questions and advices.

1.Please pay attention to the correspondence between the charts and text in the article（e.g. Table S4 in the line 273 unable to find）

2.there were no observational data for 2011 because of large-scale civil works(line 144).It is necessary to prove whether the underlying surface conditions of the watershed are consistent before and after 2011，otherwise comparison is not valid.

3.The catchment area is 3,371.4km², yet there are only 6 weather stations available for data collection(with only 1 situated within the catchment itself). Is it possible to obtain data from additional stations?

4.In the discussion, it is declare that  the optimal calibration data length is five to seven years,and  the analysis was performed using a calibration data length of P20.It's just a description of the experimental results. Is it possible to conduct a more detailed analysis based on hydrological physical principles and meteorological forcing attribution?

Author Response

We attached our reply file.

Author Response File: Author Response.pdf

Reviewer 3 Report

The manuscript "Quantifying uncertainty in runoff simulation according to multiple evaluation metrics and varying calibration data length" (water-2852150), It is an important work for the SWAT community, the results being significant and bringing a great impact to the work that, if published, will be a reference regarding the uncertainties for the SWAT community. However, before recommending the manuscript for publication, the authors must improve several aspects of the present study. Therefore, I am recommending this work for major revisions.

As small observations, which must be attended to, I highlight:

1 – I noticed some grammatical errors in writing, therefore, I suggest the revision of English by a native speaker.

2 – The authors need to reformulate the abstract, note that the first sentence (objectives) is practically identical to the last sentence (main conclusion), it is an obvious agreement, it is essential that the authors focus on a central conclusion or central result. I highlight that authors must bring methodological points and pay attention to Water's word limit (200 words, the maximum limit adopted by the journal).

3 – Authors must review the title page of the manuscript, as a call letter is not established in the authors' name or affiliation.

4 – The title of Figures 1 and 2 should be reformulated, it is too small and does not fully describe what the figures represent.

5 – In Figure 2, identify Figures 2a and 2b, and in the smaller figure, change the name “South Korea” to “South Korea”, the country should not have its name in lowercase.

6 – Why didn't the authors test regression models? Treating the data, as I believe the authors did, the next step would be to test regression models, which would provide greater representativeness and reinforce the analyzes already carried out.

7 – Rearrange Figure 9, the results of the boxplots are messy, it is difficult for the reader to interpret the results as they are. Align each boxplot and establish letters from (a) to (e).

As a minor and main note, I highlight:

1 – Use the Mendeley Reference Manager for references as well as citations, as both Water standards are not standardized in the body of every manuscript.

Moderate editing of English language required.

Author Response

We attached reply file.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

I accept revisions regarding my previous review made by authors

Reviewer 2 Report

Reviewer 3 Report

Based on the corrections made, I am accepting the current study for publication.