DIY the Integrated Climate Model and its computational performance
Pengfei Wang

TL;DR
This paper presents the development, optimization, and benchmarking of a DIY integrated climate model combining ECHAM5 and NEMO2.3, demonstrating high computational performance and operational capabilities on supercomputers.
Contribution
It introduces a flexible framework for building a coupled climate model with improved coding, restart features, and performance benchmarking on supercomputers.
Findings
ICM can simulate over 100 model years per day.
Successfully integrated ECHAM5 and NEMO2.3 with improved communication.
Achieved high-performance climate simulations on supercomputers.
Abstract
This article describes the software engineering framework and computation performance of a global climate system model which helps the user to understand the step-by-step technical to DIY(do it yourself) a climate model by your own. The model integrates ECHAM5 and NEMO2.3 using OASIS3 as the coupler. The program skill of the Integrated global Climate Model (ICM) is demonstrated here, including the porting of NEMO into the COSMOS framework, the organization of variable exchange, and component model communication synchronization. We describe how we successfully fixed some bugs in the component models and detail the new code and scripts that were added to ICM. In particular, an improvement of ICM's coding is that we enabled it to perform perfect restart runs, which is an important feature that was not implemented in the original version of NEMO2.3. ICM is designed as a model hybrid for…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsMeteorological Phenomena and Simulations · Climate variability and models · Distributed and Parallel Computing Systems
