Lessons for multi-model ensemble design drawn from emulator experiments: application to a large ensemble for 2100 sea level contributions of the Greenland ice sheet

Rohmer, Jeremy; Goelzer, Heiko; Edwards, Tamsin L.; Le Cozannet, Goneri; Durand, Gael

doi:10.5194/tc-19-6421-2025

Articles | Volume 19, issue 12

https://doi.org/10.5194/tc-19-6421-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

Special issue:

Improving the contribution of the land cryosphere to sea level...

https://doi.org/10.5194/tc-19-6421-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 19, issue 12

Research article

|

02 Dec 2025

Research article |

| 02 Dec 2025

Lessons for multi-model ensemble design drawn from emulator experiments: application to a large ensemble for 2100 sea level contributions of the Greenland ice sheet

Jeremy Rohmer, Heiko Goelzer, Tamsin L. Edwards, Goneri Le Cozannet, and Gael Durand

Supplement

https://doi.org/10.5194/tc-19-6421-2025-supplement

Data sets

Codes supporting the manuscript ``Lessons for multi-model ensemble design drawn from emulator experiments: application to a large ensemble for 2100 sea level contributions of the Greenland ice sheet'' (zenodo\_release) Jeremy Rohmer https://doi.org/10.5281/zenodo.17761003

Model code and software

R Scripts fro analysing the emulator's experiments for MME design Jeremy Rohmer https://github.com/rohmerj/MMEdesign

Short summary

Developing robust protocols to design multi-model ensembles is of primary importance for the uncertainty quantification of sea level projections. Here, we set up a series of computer experiments to reflect design decisions for the prediction of future sea level contribution of the Greenland ice sheet in 2100. We show the importance of including the most extreme climate scenario and the implications of using a single type of numerical model for ice sheets or regional climate.