Skip to main content
Data-driven modeling of carbon dioxide and methane fluxes across the Arctic-boreal region: recent achievements and future opportunities

Data-driven modeling of carbon dioxide and methane fluxes across the Arctic-boreal region: recent achievements and future opportunities

This is a Preprint and has not been peer reviewed. This is version 1 of this Preprint.

Add a Comment

You must log in to post a comment.


Comments

There are no comments or no comments have been made public for this article.

Downloads

Download Preprint

Authors

Anna-Maria Virkkala , Mathias Göckede, Kyle Arndt, McKenzie Kuhn, Sophia Walther, Jacob Nelson, Olli Peltola, Gerard Rocher-Ros, Annett Bartsch, David Bastviken, Grant Falvo, Alexandra Hamm, Joshua Hashemi, Sarah Ludwig, Susan Natali, David Olefeldt, Martijn Pallandt, Frans-Jan Parmentier, Brendan M. Rogers, Edward Schuur, Claire Treat, Judith Vogt, Carolina Voigt, Jennifer Watts, Isabel Wargowsky, Birgit Wild, Yili Yang, Amanda Armstrong, Valeria Briones, Elchin Jafarov, Kathleen Orndahl, Benjamin Poulter, Qing Ying, Gustaf Hugelius

Abstract

The Arctic-boreal region contains a significant portion of the global carbon reservoir, which is becoming increasingly vulnerable to atmospheric release due to climate change. Thus, it is vital to monitor and model Arctic-boreal carbon flux dynamics to understand the shifting carbon balance. Data-driven statistical and machine learning models are now commonly used to upscale carbon dioxide (CO₂) and methane (CH₄) fluxes in northern ecosystems from local to circumpolar scales. This review highlights recent progress, ongoing challenges, and new insights into data-driven Arctic-boreal carbon flux modeling. We identify five key areas for future model development: (1) developing comparable upscaling frameworks for terrestrial and freshwater ecosystems, especially for understudied systems like lakes and rivers; (2) reducing uncertainties and gaps in geospatial observations to better represent landscape heterogeneity; (3) fusing process-based with upscaling models to reduce computational demands, allowing to increase resolution and boost predictive power; (4) maximizing synergies between data-driven, process-based, and atmospheric inversion models using hybrid modeling; and (5) collaborating with field and remote sensing scientists to ingest observations in the upscaling process in near-real-time. These steps will reduce carbon budget uncertainty, advance our understanding of the carbon cycle, and support global climate policy.

DOI

https://doi.org/10.31223/X5TJ4W

Subjects

Earth Sciences

Keywords

Modeling, Arctic-boreal, Carbon flux, Machine learning, Remote sensing

Dates

Published: 2026-03-11 12:13

Last Updated: 2026-03-11 12:13

License

CC BY Attribution 4.0 International

Metrics

Views: 13

Downloads: 0