DOE OSTI · 2543152
Recommendations for Comprehensive and Independent Evaluation of Machine Learning‐Based Earth System Models
Abstract
Abstract Machine learning (ML) is a revolutionary technology with demonstrable applications across multiple disciplines. Within the Earth science community, ML has been most visible for weather forecasting, producing forecasts that rival modern physics‐based models. Given the importance of deepening our understanding and improving predictions of the Earth system on all time scales, efforts are now underway to develop Earth‐system models (ESMs) capable of representing all components of the coupled Earth system (or their aggregated behavior) and their response to external changes over long timescales. Building trust in ESMs is a much more difficult problem than for weather forecast models, not least because the model must represent the alternate (e.g., future or paleoclimatic) coupled states of the system for which there are no direct observations. Given that the physical principles that enable predictions about the response of the Earth system are often not explicitly coded in these ML‐based models, demonstrating the credibility of ML‐based ESMs thus requires us to build evidence of their consistency with the physical system. To this end, this paper puts forward five recommendations to enhance comprehensive, standardized, and independent evaluation of ML‐based ESMs to strengthen their credibility and promote their wider use.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Ullrich, P. A. [Division of Physical and Life Sciences Lawrence Livermore National Laboratory Livermore CA USA, Department of Land Air and Water Resources University of California Davis CA USA] (ORCID:0000000341184590), Barnes, E. A. [Department of Atmospheric Science Colorado State University Fort Collins CO USA] (ORCID:0000000342849320), Collins, W. D. [Climate and Ecosystem Sciences Division Lawrence Berkeley National Laboratory Berkeley CA USA, Department of Earth and Planetary Science University of California Berkeley CA USA] (ORCID:0000000244639848), Dagon, K. [Climate and Global Dynamics Laboratory National Center for Atmospheric Research Boulder CO USA] (ORCID:0000000245188225), Duan, S. [Division of Physical and Life Sciences Lawrence Livermore National Laboratory Livermore CA USA] (ORCID:000000031516852X), Elms, J. [Department of Earth and Atmospheric Sciences Indiana University Bloomington Bloomington IN USA], Lee, J. [Division of Physical and Life Sciences Lawrence Livermore National Laboratory Livermore CA USA] (ORCID:0000000200167199), Leung, L. R. [Atmospheric, Climate, and Earth Sciences Pacific Northwest National Laboratory Richland WA USA] (ORCID:0000000232219467), Lu, D. [Computational Sciences and Engineering Division Oak Ridge National Laboratory Oak Ridge TN USA] (ORCID:0000000151629843), Molina, M. J. [Department of Atmospheric and Oceanic Science University of Maryland College Park MD USA] (ORCID:0000000185398916), O’Brien, T. A. [Climate and Ecosystem Sciences Division Lawrence Berkeley National Laboratory Berkeley CA USA, Climate and Global Dynamics Laboratory National Center for Atmospheric Research Boulder CO USA] (ORCID:0000000266431175), Rebassoo, F. O. [Division of Physical and Life Sciences Lawrence Livermore National Laboratory Livermore CA USA]. 2025-03-14. Recommendations for Comprehensive and Independent Evaluation of Machine Learning‐Based Earth System Models. https://doi.org/10.1029/2024jh000496
Cite the original work for its findings. Save a collection to share your selection of sources.