Preprints
https://doi.org/10.5194/egusphere-2023-563
https://doi.org/10.5194/egusphere-2023-563
09 May 2023
 | 09 May 2023

CH4Net: a deep learning model for monitoring methane super-emitters with Sentinel-2 imagery

Anna Vaughan, Gonzalo Mateo-García, Luis Gómez-Chova, Vít Růžička, Luis Guanter, and Itziar Irakulis-Loitxate

Abstract. We present a deep learning model, CH4Net, for automated monitoring of methane super-emitters from Sentinel-2 data. When trained on images of 21 methane super-emitters from 2017–2020 and evaluated on images from 2021 this model achieves a scene-level accuracy of 0.83 and pixel-level balanced accuracy of 0.77. For individual emitters, accuracy is greater than 0.8 for 17 out of the 21 sites. We further demonstrate that CH4Net can successfully be applied to monitor two superemitter locations with similar background characteristics not included in the training set, with accuracies of 0.92 and 0.96. In addition to the CH4Net model we compile and open source a hand annotated training dataset consisting of 925 methane plume masks.

Anna Vaughan et al.

Status: final response (author comments only)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on egusphere-2023-563', Anonymous Referee #1, 22 May 2023
    • AC4: 'Reply on RC1', Anna Vaughan, 21 Aug 2023
  • RC2: 'Comment on egusphere-2023-563', Anonymous Referee #2, 01 Jun 2023
    • AC2: 'Reply on RC2', Anna Vaughan, 21 Aug 2023
  • RC3: 'Comment on egusphere-2023-563', Anonymous Referee #3, 09 Jun 2023
    • AC3: 'Reply on RC3', Anna Vaughan, 21 Aug 2023
  • AC1: 'Comment on egusphere-2023-563', Anna Vaughan, 21 Aug 2023

Anna Vaughan et al.

Anna Vaughan et al.

Viewed

Total article views: 614 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
347 243 24 614 10 11
  • HTML: 347
  • PDF: 243
  • XML: 24
  • Total: 614
  • BibTeX: 10
  • EndNote: 11
Views and downloads (calculated since 09 May 2023)
Cumulative views and downloads (calculated since 09 May 2023)

Viewed (geographical distribution)

Total article views: 632 (including HTML, PDF, and XML) Thereof 632 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 26 Sep 2023
Download
Short summary
Methane is a potent greenhouse gas responsible for around 25 % of global warming since the industrial revolution. Consequently identifying and mitigating methane emissions is an important step in combating the climate crisis. We develop a new deep learning model to automatically detect methane plumes from satellite images, and demonstrate that this can be applied to monitor large methane emissions resulting from the oil and gas industry.