Quantification of CO<sub>2</sub> hotspot emissions from OCO-3 SAM CO<sub>2</sub> satellite images using deep learning methods

Dumont Le Brazidec, Joffrey; Vanderbecken, Pierre; Farchi, Alban; Broquet, Grégoire; Kuhlmann, Gerrit; Bocquet, Marc

doi:https://doi.org/10.5194/gmd-18-3607-2025

Articles | Volume 18, issue 12

https://doi.org/10.5194/gmd-18-3607-2025

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

https://doi.org/10.5194/gmd-18-3607-2025

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

Articles | Volume 18, issue 12

Model description paper

|

18 Jun 2025

Model description paper |

| 18 Jun 2025

Quantification of CO₂ hotspot emissions from OCO-3 SAM CO₂ satellite images using deep learning methods

Joffrey Dumont Le Brazidec, Pierre Vanderbecken, Alban Farchi, Grégoire Broquet, Gerrit Kuhlmann, and Marc Bocquet

Data sets

Quantification of CO2 hotspot emissions from OCO-3 SAM CO2 satellite images using deep learning methods - data and weights Joffrey Dumont Le Brazidec https://doi.org/10.5281/zenodo.12788520

Synthetic XCO2, CO and NO2 observations for the CO2M and Sentinel-5 satellites G. Kuhlmann et al. https://doi.org/10.5281/zenodo.4048228

Model code and software

Quantification of CO2 hotspot emissions from OCO-3 SAM CO2 satellite images using deep learning methods - code Joffrey Dumont Le Brazidec https://doi.org/10.5281/zenodo.14013176

Short summary

We developed a deep learning method to estimate CO₂ emissions from power plants using satellite images. Trained and validated on simulated data, our model accurately predicts emissions despite challenges like cloud cover. When applied to real OCO3 satellite images, the results closely match reported emissions. This study shows that neural networks trained on simulations can effectively analyse real satellite data, offering a new way to monitor CO₂emissions from space.