Articles | Volume 22, issue 1
https://doi.org/10.5194/bg-22-257-2025
https://doi.org/10.5194/bg-22-257-2025
Research article
 | 
13 Jan 2025
Research article |  | 13 Jan 2025

Explainable machine learning for modeling of net ecosystem exchange in boreal forests

Ekaterina Ezhova, Topi Laanti, Anna Lintunen, Pasi Kolari, Tuomo Nieminen, Ivan Mammarella, Keijo Heljanko, and Markku Kulmala

Related authors

Insights into new particle formation in Siberian boreal forest from nanoparticle ranking analysis
Anastasia Lampilahti, Olga Garmash, Diego Aliaga, Mikhail Arshinov, Denis Davydov, Boris Belan, Janne Lampilahti, Veli-Matti Kerminen, Tuukka Petäjä, Markku Kulmala, and Ekaterina Ezhova
Aerosol Research Discuss., https://doi.org/10.5194/ar-2025-5,https://doi.org/10.5194/ar-2025-5, 2025
Preprint under review for AR
Short summary
Comparison of shortwave radiation dynamics between boreal forest and open peatland pairs in southern and northern Finland
Otso Peräkylä, Erkka Rinne, Ekaterina Ezhova, Anna Lintunen, Annalea Lohila, Juho Aalto, Mika Aurela, Pasi Kolari, and Markku Kulmala
Biogeosciences, 22, 153–179, https://doi.org/10.5194/bg-22-153-2025,https://doi.org/10.5194/bg-22-153-2025, 2025
Short summary
Potential of carbon uptake and local aerosol production in boreal and hemi-boreal ecosystems across Finland and in Estonia
Piaopiao Ke, Anna Lintunen, Pasi Kolari, Annalea Lohila, Santeri Tuovinen, Janne Lampilahti, Roseline Thakur, Maija Peltola, Otso Peräkylä, Tuomo Nieminen, Ekaterina Ezhova, Mari Pihlatie, Asta Laasonen, Markku Koskinen, Helena Rautakoski, Laura Heimsch, Tom Kokkonen, Aki Vähä, Ivan Mammarella, Steffen Noe, Jaana Bäck, Veli-Matti Kerminen, and Markku Kulmala
EGUsphere, https://doi.org/10.5194/egusphere-2024-1967,https://doi.org/10.5194/egusphere-2024-1967, 2024
Short summary
Opinion: The strength of long-term comprehensive observations to meet multiple grand challenges in different environments and in the atmosphere
Markku Kulmala, Anna Lintunen, Hanna Lappalainen, Annele Virtanen, Chao Yan, Ekaterina Ezhova, Tuomo Nieminen, Ilona Riipinen, Risto Makkonen, Johanna Tamminen, Anu-Maija Sundström, Antti Arola, Armin Hansel, Kari Lehtinen, Timo Vesala, Tuukka Petäjä, Jaana Bäck, Tom Kokkonen, and Veli-Matti Kerminen
Atmos. Chem. Phys., 23, 14949–14971, https://doi.org/10.5194/acp-23-14949-2023,https://doi.org/10.5194/acp-23-14949-2023, 2023
Short summary
Dynamics of aerosol, humidity, and clouds in air masses travelling over Fennoscandian boreal forests
Meri Räty, Larisa Sogacheva, Helmi-Marja Keskinen, Veli-Matti Kerminen, Tuomo Nieminen, Tuukka Petäjä, Ekaterina Ezhova, and Markku Kulmala
Atmos. Chem. Phys., 23, 3779–3798, https://doi.org/10.5194/acp-23-3779-2023,https://doi.org/10.5194/acp-23-3779-2023, 2023
Short summary

Related subject area

Biogeochemistry: Greenhouse Gases
Spatiotemporal variability of CO2, N2O and CH4 fluxes from a semi-deciduous tropical forest soil in the Congo Basin
Roxanne Daelman, Marijn Bauters, Matti Barthel, Emmanuel Bulonza, Lodewijk Lefevre, José Mbifo, Johan Six, Klaus Butterbach-Bahl, Benjamin Wolf, Ralf Kiese, and Pascal Boeckx
Biogeosciences, 22, 1529–1542, https://doi.org/10.5194/bg-22-1529-2025,https://doi.org/10.5194/bg-22-1529-2025, 2025
Short summary
Eddy-covariance fluxes of CO2, CH4 and N2O in a drained peatland forest after clear-cutting
Olli-Pekka Tikkasalo, Olli Peltola, Pavel Alekseychik, Juha Heikkinen, Samuli Launiainen, Aleksi Lehtonen, Qian Li, Eduardo Martínez-García, Mikko Peltoniemi, Petri Salovaara, Ville Tuominen, and Raisa Mäkipää
Biogeosciences, 22, 1277–1300, https://doi.org/10.5194/bg-22-1277-2025,https://doi.org/10.5194/bg-22-1277-2025, 2025
Short summary
Eddy covariance evaluation of ecosystem fluxes at a temperate saltmarsh in Victoria, Australia, shows large CO2 uptake
Ruth Reef, Edoardo Daly, Tivanka Anandappa, Eboni-Jane Vienna-Hallam, Harriet Robertson, Matthew Peck, and Adrien Guyot
Biogeosciences, 22, 1149–1162, https://doi.org/10.5194/bg-22-1149-2025,https://doi.org/10.5194/bg-22-1149-2025, 2025
Short summary
Interferences caused by the biogeochemical methane cycle in peats during the assessment of abandoned oil wells
Sebastian F. A. Jordan, Stefan Schloemer, Martin Krüger, Tanja Heffner, Marcus A. Horn, and Martin Blumenberg
Biogeosciences, 22, 809–830, https://doi.org/10.5194/bg-22-809-2025,https://doi.org/10.5194/bg-22-809-2025, 2025
Short summary
Carbon sequestration in different urban vegetation types in Southern Finland
Laura Thölix, Leif Backman, Minttu Havu, Esko Karvinen, Jesse Soininen, Justine Trémeau, Olli Nevalainen, Joyson Ahongshangbam, Leena Järvi, and Liisa Kulmala
Biogeosciences, 22, 725–749, https://doi.org/10.5194/bg-22-725-2025,https://doi.org/10.5194/bg-22-725-2025, 2025
Short summary

Cited articles

Aalto, J., Anttila, V., Kolari, P., Korpela, I., Isotalo, A., Levula, J., Schiestl-Aalto, P., and Bäck, J.: Hyytiälä SMEAR II forest year 2020 thinning tree and carbon inventory data, Zenodo, https://doi.org/10.5281/zenodo.7639833, 2023. a
Abbasian, H., Solgi, E., Hosseini, S. M., and Kia, S. H.: Modeling terrestrial net ecosystem exchange using machine learning techniques based on flux tower measurements, Ecol. Model., 466, 109901, https://doi.org/10.1016/j.ecolmodel.2022.109901, 2022. a, b, c
Aubinet, M., Vesala, T., and Papale, D.: Eddy covariance: a practical guide to measurement and data analysis, Springer Science & Business Media, ISBN 978-94-007-2350-4, 2012. a
Berrar, D.: Cross-Validation, in: Encyclopedia of Bioinformatics and Computational Biology, edited by: Ranganathan, S., Gribskov, M., Nakai, K., and Schönbach, C., Academic Press, Oxford, https://doi.org/10.1016/B978-0-12-809633-8.20349-X, pp. 542–545, 2019. a
Download
Short summary
Machine learning (ML) models are gaining popularity in biogeosciences. They are applied as gap-filling methods and used to upscale carbon fluxes to larger areas. Here we use explainable artificial intelligence (XAI) methods to elucidate the performance of machine learning models for carbon dioxide fluxes in boreal forests. We show that statistically equal models treat input variables differently. XAI methods can help scientists make informed decisions when applying ML models in their research.
Share
Altmetrics
Final-revised paper
Preprint