The researchers have investigated watercourses, such as rivers and streams, where carbon dioxide and methane are dissolved and can be released into the atmosphere.
The study was led by researchers at the University of Bristol, UK, in collaboration with, among others, the Swedish University of Agricultural Sciences (SLU) and LiU.
Half of the carbon is very old
Previously, it was assumed that watercourse carbon dioxide emissions mainly came from, for example, plants decomposing over the past 50鈥�100 years. Older carbon stores have previously been considered to be more stably deposited in the soil.
鈥淲hat we now know is that half of the carbon emitted is very old 鈥� in some cases it was absorbed from the atmosphere thousands of years ago, or even earlier! We had not expected it to be that old, let alone that old carbon would make up such a large proportion of what is emitted from watercourses,鈥� says Marcus Wallin, researcher at the Department of Aquatic Sciences and Assessment at SLU, in Uppsala.
Impact on future climate models
The research team has investigated over 700 watercourses in 26 different countries worldwide. They have made a detailed summary of the age of the carbon dioxide and methane emitted. By comparing measured levels of carbon-14 in watercourses with a standard reference for current atmospheric carbon dioxide, the researchers were able to date carbon dioxide and methane emissions.
鈥淭he carbon is transported from land to watercourses and then back to the atmosphere. We don鈥檛 yet know how humans affect this flow of old carbon,鈥� says lead author Josh Dean, a researcher at the University of Bristol.
Watercourses emit carbon into the atmosphere corresponding to about two gigatonnes per year This compares with emissions from human activity, which amount to between 10 and 15 gigatonnes annually. The emissions from watercourses are therefore significant at a global level. The researchers now plan to build on the results by investigating how the age of these emissions may have changed over time.
鈥淭he results can change our understanding of how future climate models should be designed,鈥� says Marcus Wallin.
Article: , Joshua F. Dean, Gemma Coxon, Yanchen Zheng, Jack Bishop, Mark H. Garnett, David Bastviken, Valier Galy, Robert G. M. Spencer, Suzanne E. Tank, Edward T. Tipper, Jorien E. Vonk, Marcus B. Wallin, Liwei Zhang, Chris D. Evans & Robert G. Hilton, Nature volume 642, pages105鈥�111 (2025), published online 4 June 2025. DOI: 10.1038/s41586-025-09023-w
The Swedish text was originally published as a press release from SLU.
