Northern wildfires may be unleashing hidden reservoirs of ancient carbon — and climate models are missing much of it.
Wildfires sweeping across the vast boreal forests of Alaska, Canada, Scandinavia, and Russia could be causing more climate damage than scientists once believed. New research led by UC Berkeley suggests these northern blazes are not only burning trees but also unlocking massive stores of carbon buried underground.
In many boreal regions, thick layers of carbon-rich soil lie beneath the forest floor. Known as peat, this material forms in cold, wet climates where plant matter does not fully decay. Over hundreds or even thousands of years, partially decomposed vegetation builds up, creating deep reservoirs of stored carbon. When fires burn into these soils, they release that long-accumulated carbon into the atmosphere.
Climate Models May Be Missing Underground Emissions
The study found that widely used wildfire emissions models are not fully capturing this belowground carbon loss. Most global models rely heavily on satellite images that detect visible flames and are largely built from data collected at lower latitudes. As a result, they often overlook fires that burn slowly and less visibly beneath the surface.
“Many of the fires that matter most for the climate don’t look dramatic from space,” said study lead author Johan Eckdahl, a postdoctoral scholar in Berkeley’s Energy and Resources Group. “Peatlands and organic soils can smolder for weeks to years, releasing enormous amounts of ancient carbon.”
The findings were published today (February 27) in Science Advances. Eckdahl and his colleagues analyzed 324 wildfires that burned across Sweden in 2018. By combining national forest records with on-site measurements, they produced a detailed map of carbon emissions. This high-resolution reconstruction showed how differences in local weather, vegetation, and soil conditions influence the amount of carbon stored and released during a fire.
When the team compared their results with six of the most commonly used global fire emissions models, they discovered major discrepancies. In some areas, emissions had been significantly overestimated. In others, especially where fires burned deep into soil layers, carbon losses were dramatically underestimated.
High Intensity vs Low Intensity Fires
In Gävleborg County, where large and intense fires burned through drier forests and were clearly visible from satellites, models tended to overstate carbon emissions. But in nearby Dalarna County, lower-intensity fires that smoldered in thick organic soils were far less noticeable from space. There, the models underestimated emissions by as much as 14 times.
“Sweden is a very large country, but it’s quite small compared to Siberia and Canada,” Eckdahl said. “We may be severely underestimating the impact of the recent extreme fire seasons in these regions.”
Measuring Carbon Loss in Burned Soils
To better understand how much carbon fires release from soil, the researchers visited 50 burned sites from 2018, including 19 high-intensity fire locations and 31 low-intensity sites. At each location, they measured the depth of the organic-rich soil layer — which can vary from a few inches to many feet — and collected soil samples. By comparing carbon levels in burned soil with those in nearby unburned forest soil, they calculated how much carbon had been emitted.
“Once you’re out there, it’s a simple task — just dig some holes — but the hard part is getting to the sites,” Eckdahl said. “Sweden has a good network of forest roads, but in Siberia, I hear it’s a real trek, which is one reason why we’re severely missing measurements from that region.”
Expanding Research to U.S. Forests
Through the Western Fire & Forest Collaborative, Eckdahl is now applying similar methods to fire-prone forests in the Western U.S. While these forests do not typically contain the same thick peat layers found in far northern boreal regions, factors such as climate conditions, vegetation types, and soil characteristics still strongly influence wildfire emissions. His ongoing work will focus on soil microbes, including bacteria and fungi, and their role in helping forests recover after fires.
“Forests in the Lower 48 and those far up north may look very different, but they share the common currency of carbon,” said Eckdahl. “By improving our understanding of how this element flows between the land and the atmosphere, we can better anticipate the impact of future fire regimes in a warming world and design smarter strategies to reduce climate risks on society.”
Reference: “Reassessing boreal wildfire drivers enables high-resolution mapping of emissions for climate adaptation” by Johan A. Eckdahl, Lars Nieradzik and Louise Rütting, 27 February 2026, Science Advances.
DOI: 10.1126/sciadv.adw5226
Co-authors of the study include Lars Nieradzik of Lund University and Louise Rütting of the Brandenburg University of Technology.
Never miss a breakthrough: Join the SciTechDaily newsletter.
Follow us on Google and Google News.
2 Comments
“In some areas, emissions had been significantly overestimated. In others, especially where fires burned deep into soil layers, carbon losses were dramatically underestimated.”
What is the difference between “significantly” and “dramatically?” Lawyer words don’t belong in scientific writing unless the author(s) explicitly define what the words are assumed to mean with regard to the magnitude or range. Do the authors have a problem with using numbers, such as percentages? Can I assume that there is negligible difference between ‘significantly’ and ‘dramatically’ and that is why the measurements from the world-wide network of CO2 monitoring sites, such as Scripp’s Mauna Loa, had not alerted modelers to the claimed discrepancy? If the net difference is negligible, then the situation becomes one of an under-appreciation of the uncertainty of the model predictions and not a net error of the models. This is one of the problems with not explicitly stating the uncertainty of estimates and measurements. One can’t expect to be able to check the mass-balance of dynamic systems if precision and variance aren’t estimated along with the arithmetic mean.
From the published article, the authors present some numbers about the FRP, or Fire Radiative Power, affecting the carbon emissions per square meter. They state, “(no FRP signal; mean: 0.191 ± 0.317 kg C m−2)” This is low-precision work where, without a statement about whether the uncertainty is 1-sigma or 2-sigma (95% probability), one can’t unequivocally state the sign (+/-) of the measurement. I doubt that three significant figures are justified. I would suggest that they should be presented as (no FRP signal; mean: 0.2 ± 0.3 kg C m−2), or a range of -0.1 to 0.5 with an unstated probability. Numbers ARE important as are definitions of what they are measuring.