PCLM-D-25-00140

Unsupervised Concept Discovery for Deep Weather Forecast Models with High-Resolution Radar Data

PLOS Climate

Dear Dr. Choi,

Thank you for submitting your manuscript to PLOS Climate. After careful consideration, we feel that it has merit but does not fully meet PLOS Climate’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Beyond the concerns raised by the reviewers (see below), I would like to ask you to publish with your manuscript all the software/code and data developed and used to produce your work. This include the code implementing the AI algorithms, all the data used for training, and all the output data produced. In this case, this is necessary to assure the replicability and reproducibility of your work. You should deposit the software and code in a permanent repository suitable for scientifc publication, such as Zenodo, FigShare or PANGAEA, and include in your manuscript the links and DOIs for the repositories containing such software and data. This is according the policy of our journal: 

https://journals.plos.org/climate/s/materials-software-and-code-sharing#loc-sharing-software

Please submit your revised manuscript by Aug 13 2025 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at climate@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pclm/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

We look forward to receiving your revised manuscript.

Kind regards,

Juan A. Añel

Academic Editor

PLOS Climate

Journal Requirements:

1. Please note that PLOS Climate has specific guidelines on code sharing for submissions in which author-generated code underpins the findings in the manuscript. In these cases, we expect all author-generated code to be made available without restrictions upon publication of the work. Please review our guidelines at https://journals.plos.org/climate/s/materials-and-software-sharing#loc-sharing-code and ensure that your code is shared in a way that follows best practice and facilitates reproducibility and reuse.

If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise. 

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Additional Editor Comments (if provided):

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Does this manuscript meet PLOS Climate’s publication criteria?>

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?-->?>

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available (please refer to the Data Availability Statement at the start of the manuscript PDF file)??>

The PLOS Data policy

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English??>

Reviewer #1: Yes

Reviewer #2: Yes

**********

Reviewer #1: I. Summary and Overall Impression:

The paper describes a novel technique to understand / explain the concepts learned by deep neural network based weather prediction models such as the one studied in the paper which predicts precipitation from radar data. The framework described aims to translate complex decisions made by the model into understandable and sometimes co-occuring meteorological concepts giving each a probability score.

The authors propose a framework that integrates example-based explanations with self-supervised concept vector analysis. It utilizes a multi-label self-supervised deep clustering algorithm to derive meaningful representations from an embedding space. The overall framework has 4 distinct components. The results are evaluated quantitatively comparing against a technique called Automatic Concept Extraction (ACE) using the Silhoutte Coefficient metric and showing that the refined space achieves a high score compared to baseline. They also show qualitative results from human evaluators that demonstrates high skill of this technique.

Overall, the authors solve a niche and challenging problem with a new framework that seems quite promising to this emerging field of AI based weather/climate forecasting. This kind of an explainable AI technique applied to this problem is certainly novel and has scope for very high impact especially as models may be forecasting never seen before weather/climate events. The paper is mostly well written, though sometimes, a more clear explanation of the technical concepts used with the main paper would be helpful.

II. Issues:

For the quantitative evaluation against the ACE method, it is not exactly clear what the validation dataset is. One concern here would be that the validation dataset is not large enough or possibly the training data may have some bias that allows better explanation of concepts by the framework proposed. The authors can ideally add more on how the training and validation dataset is defined.

It would be good to see more case studies than just the Polar Low vs Typhoon one to make sure the system is robust to different concept explanations.

The concepts derived are specific to precipitation and required some posthoc analysis from human labeled datasets, if I understand correctly, and therefore it is to be seen how scalable this would be to other variables.

III. Minor Issues:

For better understanding, the Figure labels could be made a little more detailed, especially Figure 4.

Typos, grammatical errors, and formatting issues:

Line 69-71, the example was similarities and differences in the samples seems to be reversed. Line 152: Space before ‘Each’

Line 175: Heading seems to be re-written

Line 183: The superscript of 2 next to the reference number seems erroneous

Line 336: “encapsulathe te” typo

IV. Concluding Remarks:

The proposed technique is novel, seems to be supported by reasonable evaluation methodology and has potential for high impact in other areas of AI based weather and climate models.

Reviewer #2: This manuscript introduces a novel and well-executed approach to enhancing the interpretability of AI-based weather forecasting models. The integration of watershed segmentation, masked autoencoders, and multi-label clustering is methodologically sound and addresses the critical challenge of explainability in AI-driven meteorological models. The validation through expert surveys and the ability to distinguish between complex precipitation mechanisms further demonstrate the practical value of the approach.

To enhance the clarity of the manuscript, I suggest the following minor revisions:

1. Visualization of All Concept Clusters: Please consider including the complete set of 24 extracted concept clusters in the main manuscript. Expanding Figure 5 to display all clusters—either directly or through an additional panel—would significantly improve reader comprehension and transparency regarding the diversity of discovered precipitation mechanisms.

2. Discussion on Broader Applicability: The current study focuses on radar data and meteorological patterns specific to the Korean Peninsula. I recommend adding a brief paragraph to the discussion section that addresses the applicability of the proposed method to other regions and datasets. This paragraph could address potential challenges and opportunities for adaptation.

3. Minor errors: Line 175. Duplicated.

**********

what does this mean? ). If published, this will include your full peer review and any attached files.

Do you want your identity to be public for this peer review? If you choose “no”, your identity will remain anonymous but your review may still be made public.

For information about this choice, including consent withdrawal, please see our Privacy Policy

Reviewer #1: No

Reviewer #2: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PCLM-D-25-00140R1

Unsupervised Concept Discovery for Deep Weather Forecast Models with High-Resolution Radar Data

PLOS Climate

Dear Dr. Choi,

Thanks for submitting the reviewed version of your manuscript. Unfortunately, when checking the DOIs for the code and data in the Data Availability Statement of your manuscript, we have found that they do not work. These DOIs are not found. Therefore, please, double-check them and provide a reviewed version with correct DOIs for your data and code.

Please submit your revised manuscript by Aug 28 2025 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at climate@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pclm/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

We look forward to receiving your revised manuscript.

Kind regards,

Juan A. Añel

Academic Editor

PLOS Climate

Journal Requirements:

If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise. 

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Unsupervised Concept Discovery for Deep Weather Forecast Models with High-Resolution Radar Data

PCLM-D-25-00140R2

Dear Dr. Choi,

We're pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you'll receive an e-mail detailing the required amendments. When these have been addressed, you'll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at https://www.editorialmanager.com/pclm/ click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. For billing related questions, please contact billing support at https://plos.my.site.com/s/.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they'll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact climate@plos.org.

Kind regards,

Juan A. Añel

Academic Editor

PLOS Climate

Additional Editor Comments (optional):

Reviewers' comments: