Emulating grid-based forest carbon dynamics using machine learning: an LPJ-GUESS v4.1.1 application

Natel, Carolina; Belda, David Martín; Anthoni, Peter; Haß, Neele; Rabin, Sam; Arneth, Almut

doi:https://doi.org/10.5194/gmd-18-4317-2025

Articles | Volume 18, issue 14

https://doi.org/10.5194/gmd-18-4317-2025

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

https://doi.org/10.5194/gmd-18-4317-2025

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

Articles | Volume 18, issue 14

Development and technical paper

|

18 Jul 2025

Development and technical paper |

| 18 Jul 2025

Emulating grid-based forest carbon dynamics using machine learning: an LPJ-GUESS v4.1.1 application

Carolina Natel, David Martín Belda, Peter Anthoni, Neele Haß, Sam Rabin, and Almut Arneth

Supplement

https://doi.org/10.5194/gmd-18-4317-2025-supplement

Data sets

ISIMIP3b bias-adjusted atmospheric climate input data Stefan Lange and Matthias Büchner https://doi.org/10.48364/ISIMIP.842396.1

LPJ-GUESS Forest Carbon Emulator (Data, Models, SHAP values) C. Natel de Moura et al. https://doi.org/10.5281/zenodo.14230951

Model code and software

LPJ-GUESS code and modifications for the emulator C. Natel de Moura https://doi.org/10.5281/zenodo.15065248

natel-c/lpjg-forestC-emulator C. Natel de Moura https://doi.org/10.5281/zenodo.14231373

TensorFlow (v2.18.0) TensorFlow Developers https://doi.org/10.5281/zenodo.13989084

A Unified Approach to Interpreting Model Predictions S. Lundberg and S.-I. Lee https://doi.org/10.48550/arXiv.1705.07874

Short summary

We developed fast machine learning models to predict forest regrowth and carbon dynamics under climate change. These models mimic the outputs of a complex vegetation model but run 95 % faster, enabling global analyses and supporting climate solutions in large modeling frameworks such as LandSyMM.