Physics-Guided Machine Learning for Carbon Emission Modeling under System Disruptions: Methods and Challenges

Yihao Zhang

doi:10.54254/2753-8818/2025.DL27137

Theoretical and Natural ScienceOpen access

Physics-Guided Machine Learning for Carbon Emission Modeling under System Disruptions: Methods and Challenges

Research Article

Open Access

Physics-Guided Machine Learning for Carbon Emission Modeling under System Disruptions: Methods and Challenges

Yihao Zhang ^1*

¹ Ulster University

^*Corresponding author: Zhang-Y45@ulster.ac.uk

Published on 24 September 2025

TNS Vol.132

ISSN (Print): 2753-8826

ISSN (Online): 2753-8818

ISBN (Print): 978-1-80590-305-5

ISBN (Online): 978-1-80590-306-2

Download Cover

Abstract

Disruptions such as pandemics and energy shocks weaken the reliability of carbon-emission models. Physics-guided machine learning (PIML) offers a practical response. This review explains how PIML improves robustness when temporal continuity breaks and system structure shifts. It covers four methods: physics-informed neural networks; hybrid or embedded designs that use CTM outputs and differentiable operators; post-processing that enforces physical feasibility; and structured or graph-based models that encode conservation and transport while keeping interpretability. Evidence shows these methods are more robust with sparse data and under distribution shifts. Yet challenges remain: balancing loss terms, dealing with complex boundaries, keeping physical inputs up to date, and the high cost of multi-scale problems. On the application side, the review focuses on pandemic-driven demand contraction, energy-supply shocks with fuel switching, and policy-induced structural change. These scenarios guide evaluation and benchmarks and help improve reliability under disruptions. On the practice side, recommended tactics include reliability-weighted fusion, adaptive coupling, explicit uncertainty handling, and dynamic graphs.

Keywords:

Physics-Guided Machine Learning, Carbon Emission Modeling, Distribution Shift

View PDF

References

[1]. International Energy Agency. (2021). Net zero by 2050: A roadmap for the global energy sector. Paris, France: IEA. Retrieved from https: //www.iea.org/reports/net-zero-by-2050

[2]. Intergovernmental Panel on Climate Change. (2022). Climate change 2022: Mitigation of climate change—Contribution of Working Group III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, UK: Cambridge University Press.

[3]. International Energy Agency. (2023). CO₂ emissions in 2022 – Analysis. Paris, France: IEA. Retrieved from https: //www.iea.org/reports/co2-emissions-in-2022

[4]. Liang, X., Liu, Z., Wang, J., Jin, X., & Du, Z. (2023). Uncertainty quantification-based robust deep learning for building energy systems considering distribution shift problem. Applied Energy, 337, 120889.

[5]. Raissi, M., Perdikaris, P., & Karniadakis, G. E. (2019). Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational Physics, 378, 686–707.

[6]. Karniadakis, G. E., Kevrekidis, I. G., Lu, L., Perdikaris, P., Wang, S., & Yang, L. (2021). Physics-informed machine learning. Nature Reviews Physics, 3, 422–440.

[7]. Zhao, C., Zhang, M., Bai, J., Wu, J., & Chang, I.-S. (2025). A review of the application of machine learning in carbon emission assessment studies: Prediction optimization and driving factor selection. Science of the Total Environment, 987, 179678.

[8]. Jin, Y., Sharifi, A., Li, Z., Chen, S., Zeng, S., & Zhao, S. (2024). Carbon emission prediction models: A review. Science of the Total Environment, 927, 172319.

[9]. Liu, Z., et al. (2020). Near-real-time monitoring of global CO₂ emissions reveals the effects of the COVID-19 pandemic. Nature Communications, 11, 5172.

[10]. Wang, H., Li, D., Zhou, R., Hu, X., Wang, L., & Zhang, L. (2024). A new method for top-down inversion estimation of carbon dioxide flux based on deep learning. Remote Sensing, 16(19), 3694.

[11]. Li, L., et al. (2023). Improving air quality assessment using physics-inspired deep graph learning. npj Climate and Atmospheric Science, 6, 152.

[12]. Zhang, Y., et al. (2025). KG-FGNN: Knowledge-guided GNN foundation model for fertilisation-oriented soil GHG flux prediction. arXiv preprint arXiv: 2506.15896.

[13]. Shokouhi, P., Kumar, V., Prathipati, S., Hosseini, S. A., Giles, C. L., & Kifer, D. (2021). Physics-informed deep learning for prediction of CO₂ storage site response. Journal of Contaminant Hydrology, 241, 103835.

[14]. Chuprov, I., Derkach, D., Kychkin, A., & Efremenko, D. (2025). Application of physics-informed neural networks for solving the inverse advection–diffusion problem to localize pollution sources. arXiv preprint arXiv: 2503.18849.

[15]. Penwarden, M., Zhe, S., Narayan, A., & Kirby, R. M. (2022). Multifidelity modeling for physics-informed neural networks (PINNs). Journal of Computational Physics, 451, 110844.

[16]. Zhu, P., Liu, Z., Xu, Z., & Lv, J. (2025). An adaptive weight physics-informed neural network for vortex-induced vibration problems. Buildings, 15(9), 1533.

[17]. Beucler, T., et al. (2021). Enforcing analytic constraints in neural networks emulating physical systems. Physical Review Letters, 126, 098302.

[18]. Weir, B., Ott, L. E., Collatz, G. J., Kawa, S. R., Poulter, B., Chatterjee, A., Oda, T., & Pawson, S. (2021). Bias-correcting carbon fluxes derived from land-surface satellite data for retrospective and near-real-time assimilation systems. Atmospheric Chemistry and Physics, 21, 9609–9628.