Quantifying wildfire drivers and predictability in boreal peatlands using a two-step error-correcting machine learning framework in TeFire v1.0

Tang, Rongyun; Jin, Mingzhou; Mao, Jiafu; Ricciuto, Daniel M.; Chen, Anping; Zhang, Yulong

doi:https://doi.org/10.5194/gmd-17-1525-2024

Articles | Volume 17, issue 4

https://doi.org/10.5194/gmd-17-1525-2024

© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

Special issue:

The role of fire in the Earth system: understanding interactions...

https://doi.org/10.5194/gmd-17-1525-2024

© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 17, issue 4

Development and technical paper

|

21 Feb 2024

Development and technical paper |

| 21 Feb 2024

Quantifying wildfire drivers and predictability in boreal peatlands using a two-step error-correcting machine learning framework in TeFire v1.0

Rongyun Tang, Mingzhou Jin, Jiafu Mao, Daniel M. Ricciuto, Anping Chen, and Yulong Zhang

Supplement

https://doi.org/10.5194/gmd-17-1525-2024-supplement

Data sets

TSECfire v1.0 Rongyun Tang et al. https://github.com/tangryun/GMD

Fire_cci Burned Area dataset ESA https://geogra.uah.es/fire_cci/firecci51.php

Global Monthly GPP from an Improved Light Use Efficiency Model, 1982-2016 N. Madani and N. C. Parazoo https://doi.org/10.3334/ORNLDAAC/1789

Global GIMMS NDVI3g v1 dataset (1981-2015). National Tibetan Plateau/Third Pole Environment Data Center The National Center for Atmospheric Research https://data.tpdc.ac.cn/en/data/9775f2b4-7370-4e5e-a537-3482c9a83d88/

Anthropogenic land-use estimates for the Holocene; HYDE 3.2 C. G. M. Klein Goldewijk https://doi.org/10.17026/dans-25g-gez3

Maps of northern peatland extent, depth, carbon storage and nitrogen storage. Dataset version 1 Gustaf Hugelius et al. https://doi.org/10.17043/hugelius-2020-peatland-1

Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset (https://crudata.uea.ac.uk/cru/data/hrg/) I. Harris et al. https://doi.org/10.1038/s41597-020-0453-3

Model code and software

Tangetal2023 Rongyun Tang https://doi.org/10.5281/zenodo.10072144

Short summary

Carbon-rich boreal peatlands are at risk of burning. The reproducibility and predictability of rare peatland fire events are investigated by constructing a two-step error-correcting machine learning framework to tackle such complex systems. Fire occurrence and impacts are highly predictable with our approach. Factor-controlling simulations revealed that temperature, moisture, and freeze–thaw cycles control boreal peatland fires, indicating thermal impacts on causing peat fires.