JCM v1.1: a differentiable, intermediate-complexity atmospheric model
8 affiliations
- 1Scripps Institution of Oceanography, University of California San Diego, La Jolla, USA
- 2Cecil H. and Ida M. Green Institute of Geophysics and Planetary Physics, University of California San Diego, La Jolla, USA
- 3Halıcıoğlu Data Science Institute, University of California San Diego, La Jolla, USA
- 4Department of Computer Science and Engineering, University of California San Diego, La Jolla, USA
- 5Department of Bioengineering, University of California San Diego, La Jolla, USA
- 6Google Research, Mountain View, CA, USA
- 7Scripps Institution of Oceanography (La Jolla, United States of America)
- 8University of California San Diego (San Diego, United States of America)
Abstract
In this paper we present version 1.0 of the JAX Circulation Model (JCM). JCM is an open-source, differentiable atmospheric model built in Python using the JAX numerical library. Earth system modeling is rapidly evolving, particularly through hybrid approaches that combine known dynamics with data-driven components. However, the training and validation of hybrid methods in traditional models remain difficult due to the absence of gradients and the complexity of legacy code. Differentiable models written in modern frameworks offer a path forward. JCM couples physics parameterizations to the Dinosaur dynamical core through a flexible interface that makes substitution of other schemes easy. The default parameterization scheme uses the SPEEDY (Simplified Parameterizations, primitivE-Equation DYnamics) intermediate-complexity physics scheme. This modularity supports benchmarking across physical and machine-learned schemes, with direct access to gradients for sensitivity analysis, calibration, and online learning. We show validation of JCM against the original Fortran SPEEDY code at T31 resolution. We also highlight JCM's differentiability and efficiency and outline plans for extending the framework to a differentiable Earth system model. JCM provides a lightweight yet expressive platform for accelerating research in climate modeling.
Related
- Energy Flow Shapes the Decomposition Dynamics of a Reactive Atmospheric Intermediate
- Atmospheric Semi- and Intermediate-Volatility Organic Compounds: Recent Progress and Future Perspectives.
- DiffSWE2d: a differentiable Shallow Water Equations solver for end-to-end flood and tsunami modelling
- Suêtes: An end-to-end differentiable non-hydrostatic limited-area dynamical core
- A library for differentiable signal processing and machine learning on the sphere
BibTeX
@article{davenport2026differentiable,
title = {{JCM v1.1: a differentiable, intermediate-complexity atmospheric model}},
author = {Ellen Davenport and J. Varan Madan and Rebecca Gjini and Jared Brzenski and Benjamin Crawford and Nick Ho and Tien‐Yiao Hsu and Yueshan Liang and Zhixing Liu and Veeramakali Manivannan and Eric Pham and Rohith Vutukuru and Andrew I. L. Williams and Zhiqi Yang and Rose Yu and Nicholas J. Lutsko and Stephan Hoyer and Duncan Watson‐Parris},
journal = {Geoscientific model development},
year = {2026},
month = jul,
doi = {10.5194/gmd-19-6451-2026},
url = {https://doi.org/10.5194/gmd-19-6451-2026},
note = {CC BY},
}
RIS
TY - JOUR
TI - JCM v1.1: a differentiable, intermediate-complexity atmospheric model
AU - Ellen Davenport
AU - J. Varan Madan
AU - Rebecca Gjini
AU - Jared Brzenski
AU - Benjamin Crawford
AU - Nick Ho
AU - Tien‐Yiao Hsu
AU - Yueshan Liang
AU - Zhixing Liu
AU - Veeramakali Manivannan
AU - Eric Pham
AU - Rohith Vutukuru
AU - Andrew I. L. Williams
AU - Zhiqi Yang
AU - Rose Yu
AU - Nicholas J. Lutsko
AU - Stephan Hoyer
AU - Duncan Watson‐Parris
PY - 2026
DA - 2026/07/17
JO - Geoscientific model development
DO - 10.5194/gmd-19-6451-2026
UR - https://doi.org/10.5194/gmd-19-6451-2026
AB - In this paper we present version 1.0 of the JAX Circulation Model (JCM). JCM is an open-source, differentiable atmospheric model built in Python using the JAX numerical library. Earth system modeling is rapidly evolving, particularly through hybrid approaches that combine known dynamics with data-driven components. However, the training and validation of hybrid methods in traditional models remain difficult due to the absence of gradients and the complexity of legacy code. Differentiable models written in modern frameworks offer a path forward. JCM couples physics parameterizations to the Dinosaur dynamical core through a flexible interface that makes substitution of other schemes easy. The default parameterization scheme uses the SPEEDY (Simplified Parameterizations, primitivE-Equation DYnamics) intermediate-complexity physics scheme. This modularity supports benchmarking across physical and machine-learned schemes, with direct access to gradients for sensitivity analysis, calibration, and online learning. We show validation of JCM against the original Fortran SPEEDY code at T31 resolution. We also highlight JCM's differentiability and efficiency and outline plans for extending the framework to a differentiable Earth system model. JCM provides a lightweight yet expressive platform for accelerating research in climate modeling.
SN - 1991-959X
SN - 1991-9603
ER -