NeuroDALEC: A Differentiable and Interpretable Mass-Conserving Framework for Terrestrial Ecosystem Carbon Cycle Dynamics

Wan, Meng; Liu, Tiantian; Xia, Zhixin; Nie, Ningming; Wang, Jue; Cao, Rongqiang; He, Honglin; Ren, Xiaoli; Shi, Peng; Wang, Yangang

doi:10.24963/ijcai.2026/829

NeuroDALEC: A Differentiable and Interpretable Mass-Conserving Framework for Terrestrial Ecosystem Carbon Cycle Dynamics

Meng Wan, Tiantian Liu, Zhixin Xia, Ningming Nie, Jue Wang, Rongqiang Cao, Honglin He, Xiaoli Ren, Peng Shi, Yangang Wang

Download BibTeX Cite

Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence. AI and Social Good. Pages 7455-7463.

https://doi.org/10.24963/ijcai.2026/829

Keywords

Machine Learning: Machine Learning
Multidisciplinary Topics and Applications: Multidisciplinary Topics and Applications

Abstract

Accurate simulation of terrestrial ecological carbon cycles is crucial for global climate change and ecosystem management. Process-based carbon models have high interpretability, but suffer from insufficient accuracy and slow computation due to fixed parameters. In contrast, deep-learning carbon models achieve high accuracy, but disregard physical principles, which prevents ecologists from explaining ecosystem dynamics. We propose NeuroDALEC, an interpretable framework that embeds the DALEC carbon-cycle model within a neural network, enabling differentiable computation of ecological processes. Key parameters and ensemble learning strategies are designed, and mass-conserving carbon pool state transition equations are introduced to ensure physical consistency. Experiments show NeuroDALEC outperforms existing models in both accuracy and efficiency. Moreover, it provides sufficient interpretability by predicting all components of the carbon cycle. Deployed in a real-time carbon assimilation system, NeuroDALEC supports daily carbon forecasting and decision-making. This work contributes to the United Nations' Sustainable Development Goals 13 (Climate Action) and 15 (Life on Land). The source code is available at: https://github.com/codesiena/NeuroDALEC.

Meng Wan, Tiantian Liu, Zhixin Xia, Ningming Nie, Jue Wang, Rongqiang Cao, Honglin He, Xiaoli Ren, Peng Shi, and Yangang Wang. 2026. NeuroDALEC: A Differentiable and Interpretable Mass-Conserving Framework for Terrestrial Ecosystem Carbon Cycle Dynamics. In Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence, IJCAI-26. IJCAI, 7455–7463. https://doi.org/10.24963/ijcai.2026/829

@inproceedings{ijcai2026p829,
  title     = {NeuroDALEC: A Differentiable and Interpretable Mass-Conserving Framework for Terrestrial Ecosystem Carbon Cycle Dynamics},
  author    = {Wan, Meng and Liu, Tiantian and Xia, Zhixin and Nie, Ningming and Wang, Jue and Cao, Rongqiang and He, Honglin and Ren, Xiaoli and Shi, Peng and Wang, Yangang},
  booktitle = {Proceedings of the Thirty-Fifth International Joint Conference on
               Artificial Intelligence, {IJCAI-26}},
  publisher = {International Joint Conferences on Artificial Intelligence Organization},
  editor    = {Diego Calvanese},
  pages     = {7455--7463},
  year      = {2026},
  month     = {8},
  note      = {AI and Social Good},
  doi       = {10.24963/ijcai.2026/829},
  url       = {https://doi.org/10.24963/ijcai.2026/829},
}

TY  - CPAPER
TI  - NeuroDALEC: A Differentiable and Interpretable Mass-Conserving Framework for Terrestrial Ecosystem Carbon Cycle Dynamics
AU  - Wan, Meng
AU  - Liu, Tiantian
AU  - Xia, Zhixin
AU  - Nie, Ningming
AU  - Wang, Jue
AU  - Cao, Rongqiang
AU  - He, Honglin
AU  - Ren, Xiaoli
AU  - Shi, Peng
AU  - Wang, Yangang
T2  - Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence, IJCAI-26
PB  - International Joint Conferences on Artificial Intelligence Organization
PY  - 2026
SP  - 7455
EP  - 7463
DO  - 10.24963/ijcai.2026/829
UR  - https://doi.org/10.24963/ijcai.2026/829
ER  -