EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Explainable machine learning highlights the role of diffuse radiation in ecosystem carbon uptake of boreal forest

Topi Markus Laanti, Aino Aarne, Maxime Durand, Steffen Manfred Noe, Victoria Miles, Dmitrii Krasnov, Markku Kulmala, Keijo Heljanko, Anna Lintunen, Ekaterina Ezhova

Environmental Data Science · 2026 · 5

DOI: 10.1017/eds.2026.10040 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Environmental Data Science, online erschienen am 06.07.2026. Verfasst von Topi Markus Laanti und 9 weiteren Autor:innen. Die Arbeit nennt 51 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Norges Forskningsråd, Atmosphere and Climate Competence Center, Eesti Teadusagentuur.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Boreal forests play a critical role in the global carbon cycle as they are one of the largest terrestrial carbon sinks globally. In this study, we employ explainable machine learning (ML) techniques to investigate the influence of environmental and vegetation variables on net ecosystem exchange (NEE), focusing specifically on the effects of diffuse radiation. We utilize a sub-hourly resolution data set including satellite and in-situ observations from three boreal or hemiboreal forest research stations across the latitudes 58–68° N to capture latitudinal variability in forest carbon uptake. Using SHAP (Shapley Additive Explanations) values, we identify key drivers influencing NEE and quantify their importance across various ML model architectures. Photosynthetically active radiation (PAR), diffuse radiation, normalized difference vegetation index, and soil temperature were identified as the variables having the largest explanatory power for NEE across the ML models. ML models using only these variables result in R 2 ≈ 0.8 $ {R}^2\approx 0.8 $ upper R 2 almost equals 0.8 and RMSE ≈ 2.3 $ \approx 2.3 $ almost equals 2.3 μ mol m − 2 s − 1 $ \mu \mathrm{mol}\;{\mathrm{m}}^{-2}\hskip0.1em {\mathrm{s}}^{-1} $ mu molm minus 2 s minus 1 . Further analysis of SHAP values indicates that higher diffuse radiation (DiffPAR) is associated with more negative NEE (stronger carbon sink). SHAP analysis highlights this effect much more clearly than raw DiffPAR measurements, because it accounts for interactions with other environmental factors. This suggests that the diffuse radiation effect emerges from interactions between DiffPAR and co-varying environmental factors (such as cloudiness and total PAR). Cases identified by SHAP ( DiffPAR SHAP < − 2 $ \mathrm{DiffPAR}\ \mathrm{SHAP}

Förderung

  • Norges Forskningsråd
  • Atmosphere and Climate Competence Center
  • Eesti Teadusagentuur
  • Research Council of Finland

Zitieren

APA

Laanti, T. M., Aarne, A., Durand, M., Noe, S. M., Miles, V., Krasnov, D., Kulmala, M., Heljanko, K., Lintunen, A., & Ezhova, E. (2026). Explainable machine learning highlights the role of diffuse radiation in ecosystem carbon uptake of boreal forest. Environmental Data Science, 5. https://doi.org/10.1017/eds.2026.10040

Vancouver

Laanti TM, Aarne A, Durand M, Noe SM, Miles V, Krasnov D, et al. Explainable machine learning highlights the role of diffuse radiation in ecosystem carbon uptake of boreal forest. Environmental Data Science. 2026;5. doi:10.1017/eds.2026.10040

BibTeX
@article{Laanti2026,
  author = {Laanti, Topi Markus and Aarne, Aino and Durand, Maxime and Noe, Steffen Manfred and Miles, Victoria and Krasnov, Dmitrii and Kulmala, Markku and Heljanko, Keijo and Lintunen, Anna and Ezhova, Ekaterina},
  title = {Explainable machine learning highlights the role of diffuse radiation in ecosystem carbon uptake of boreal forest},
  journal = {Environmental Data Science},
  year = {2026},
  volume = {5},
  doi = {10.1017/eds.2026.10040}
}
Kostenlos starten →