Biogeosciences
Biogeosciences
BG
Articles | Volume 23, issue 9
Articles | Volume 23, issue 9
https://doi.org/10.5194/bg-23-2959-2026
© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/bg-23-2959-2026
© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.
Articles | Volume 23, issue 9
Research article
 | 
04 May 2026
Research article |  | 04 May 2026

PeatDepth-ML: a global map of peat depth predicted using machine learning

Jade Skye, Joe R. Melton, Colin Goldblatt, Angela Gallego-Sala, Michelle Garneau, and Scott Winton

Viewed

Total article views: 1,040 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
584 415 41 1,040 34 39
  • HTML: 584
  • PDF: 415
  • XML: 41
  • Total: 1,040
  • BibTeX: 34
  • EndNote: 39
Views and downloads (calculated since 18 Nov 2025)
Cumulative views and downloads (calculated since 18 Nov 2025)

Viewed (geographical distribution)

Total article views: 1,040 (including HTML, PDF, and XML) Thereof 1,026 with geography defined and 14 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 04 May 2026
Download
Short summary
We developed PeatDepth-ML, a machine learning model predicting peat depth worldwide to help estimate carbon stocks in these climate-critical ecosystems. Our model predicts median depths of 134 cm in peatlands. Using bootstrapping, we rigorously assessed how sampling bias affects predictions. This revealed predictor selection and regional accuracy can vary greatly with different data subsets, demonstrating model reliability fundamentally depends on training data quality and geographic coverage.
Read more
Share
Altmetrics
Final-revised paper
Preprint