Using the Debiased Brier Skill Score to Evaluate S2S Tropical Cyclone Forecasting

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This is a quite focused research on the tropical cyclone forecast since it evaluates the influence of the forecast time window, tropical cyclone radius or the evaluation region. Four models from Europe, United States, Korea and United Kingdom were considered. The studied region is East Asia. Some tests are presented together with their results. Due to the cyclone impact on the human lives, this paper could be considered for publication after the introduction of the following minor changes.

The authors should justify the analysis robustness, i.e., they should indicate if the used data are representative enough.

The number of references to discuss the presented results should be increased. These references could establish the contrast between the current research and previous analyses.

The relevant results and conclusions should be highlighted to increase the scope of this study, i.e. to make it interesting for a broad group of readers.

References should be avoided in the conclusion. Perhaps its last paragraph could be placed in the Results section.

Finally, 21 citations correspond to F. Vitart over 35 references. Although the paper is quite focused, the authors should consider if these references are necessary.

Minor remarks.

Some labels should be introduced in Figures 4 and 6.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The paper performs a sensitivity analysis on the factors that influence the calculation of Daily Tropical Cyclone Probability (DTCP)  and Debiased Brier Skill Score (DBSS) of Subseasonal to Seasonal (S2S) TC predictions. The introduction provides adequate information to introduce the reader to the topic. Materials and methods are clearly presented. The text flows easily and the authors answer progressively to the questions posed in every subchapter. Small errors in the use of language should be taken care of. One certain thing that should be changed is the Figure’s captions; The captions should be more explanatory. For example, in Figure 5 the reader does not know which dataset provided the results. Reaching line 304 of the text the reader receives the information that the dataset used in Figure 5 came from ECMWF but this information should be present in the caption too. The same is true for almost all Figures. After addressing those minor comments, the paper should be published (even without a second revision round).

Comments on the Quality of English Language

Small corrections like

Line 17 : “…its forecast…” should be replaced by “…their forecast…”

Line 21: ”…the using of DTCP…” should be “...the use of DTCP…”

and others like that.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

When addressing the issue of validation or evaluation forecasts, specific results of specific products are expected. But the main goal of this study was declared as the examination of tropical cyclones forecast data from four models in the seasonal to subseasonal real-time database, testing multiple factors influencing model forecast skill evaluation. That is, not for a comparison and generalization of errors between different models, but for an assessment of validation procedures. So, it is expected to get some recommendations to use them in practice. We have only one: “It is recommended to use the theoretical value of Brier Score for reference climate forecast instead of its estimation”. Other results are acknowledgement of the influence of various factors, such as, forecast time window, cyclone influence radius, etc, which are naturally inherent for any methodology. It is desirable to formulate in conclusions more recommendations, at least short, for use by other evaluators.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf