MESMER-RCM: a probabilistic climate emulator for regional warming projections

Pan, Hao; Gudmundsson, Lukas; Hauser, Mathias; Schwaab, Jonas; Quilcaille, Yann; Seneviratne, Sonia I.

doi:10.5194/npg-33-73-2026

Articles | Volume 33, issue 1

https://doi.org/10.5194/npg-33-73-2026

© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.

Collection:

A collection on artificial intelligence and the complexity...

https://doi.org/10.5194/npg-33-73-2026

© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 33, issue 1

Research article

|

12 Feb 2026

Research article |

| 12 Feb 2026

MESMER-RCM: a probabilistic climate emulator for regional warming projections

Hao Pan, Lukas Gudmundsson, Mathias Hauser, Jonas Schwaab, Yann Quilcaille, and Sonia I. Seneviratne

Download

Final revised paper (published on 12 Feb 2026)
Supplement to the final revised paper
Preprint (discussion started on 02 Jul 2025)
Supplement to the preprint

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2025-3052', Anonymous Referee #1, 10 Aug 2025

Please see the attached file.

Citation: https://doi.org/10.5194/egusphere-2025-3052-RC1
- AC1: 'Reply on RC1', Hao Pan, 30 Oct 2025
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-3052/egusphere-2025-3052-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2025-3052-AC1
RC2:
'Comment on egusphere-2025-3052', Anonymous Referee #2, 20 Aug 2025

In this study, the authors presented a generative AI model for the region climate simulation. This model generates the large ensembles and can well capture the intrinsic climate variability. The topic is very interesting. But there still are several questions that need to be addressed.
1.How was the MESMER-RCM model trained? Please provide more details about the training of the model. For example, how to divide the training and testing sets? How to set the model parameters?
2.There are some parameters in the model. Are the results sensitive to the choices of the parameters? The detailed tests should be done.
3.The figure 1 shows the 2-m temperature in a region. Why there are some blank areas? Additionally, could you add the latitude and longitude in the figure? Because the readers may be not familiar with that region.
4.In this study, only a simple example was displayed. To show the advantage of the model, more examples in different areas should be presented. Whether can this model be extended to other regions?

Citation: https://doi.org/10.5194/egusphere-2025-3052-RC2
- AC2: 'Reply on RC2', Hao Pan, 30 Oct 2025
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-3052/egusphere-2025-3052-AC2-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2025-3052-AC2
EC1:
'Comment on egusphere-2025-3052', Jie Feng, 24 Aug 2025

Based on the comments from two Reviewers, both of them gave positive encouragements on the value of the manuscript. I also think this study is interesting and worth publishing in this journal. However, the authors need to address all the comments, especially both Reviewers requested strengthening of the description of scheme and training procedures of MESMER-RCM model. One of them also mentioned that the results from another benchmark model should be provided for a more complete comparison, at least for subset of the metrics. The manuscript needs a major revision and I look forward to receiving a further improved version of this study.

Citation: https://doi.org/10.5194/egusphere-2025-3052-EC1
- AC3: 'Reply on EC1', Hao Pan, 30 Oct 2025
  
  We sincerely thank the Editor for the encouraging feedback and constructive suggestions.
  Following the reviewers’ comments, we have strengthened the description of the MESMER-RCM model scheme and the training procedures in the manuscript. In addition, we have included the comparison results from benchmark models, as suggested by Reviewer #1.
  We believe these revisions have further improved the clarity and completeness of the manuscript.
  
  Citation: https://doi.org/10.5194/egusphere-2025-3052-AC3

Peer review completion

AR – Author's response | RR – Referee report | ED – Editor decision | EF – Editorial file upload

AR by Hao Pan on behalf of the Authors (30 Oct 2025) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (06 Nov 2025) by Jie Feng

RR by Anonymous Referee #2 (19 Nov 2025)

RR by Anonymous Referee #1 (24 Nov 2025)

ED: Publish as is (02 Dec 2025) by Jie Feng

AR by Hao Pan on behalf of the Authors (04 Dec 2025)

Short summary

Existing regional climate model (RCM) emulators mainly provide deterministic emulations, while internal RCM variability is typically not represented. We develop MESMER-RCM, a probabilistic RCM emulator for annual 2-m temperature, using a simple and physically interpretable approach. We demonstrate its ability to emulate both RCM trends and internal variability in a high-dimensional spatial setting, where existing approaches typically struggle.

MESMER-RCM: a probabilistic climate emulator for regional warming projections

Download

Interactive discussion

Peer review completion

Suggestions for revision or reasons for rejection

Suggestions for revision or reasons for rejection