Earth System Models Are Overstating Carbon Removal Because Nitrogen Fixation Is Much Lower Than We Thought
High levels of carbon dioxide in the atmosphere are a major driver of climate change. At the same time, elevated carbon dioxide can stimulate plant growth, allowing vegetation to absorb more carbon from the air. For years, this effect — often called the CO₂ fertilization effect — has been built into global climate projections. But new research suggests that this natural carbon-removal process has been significantly overestimated, and the reason comes down to one crucial element: nitrogen.
A new study published in the Proceedings of the National Academy of Sciences shows that Earth system models have been assuming far more natural nitrogen fixation than actually occurs in the real world. According to the researchers, this has led to an inflated estimate of future plant growth and carbon removal, with major implications for how we understand climate change.
Why Nitrogen Matters for Plant Growth
Plants cannot grow on carbon dioxide alone. While CO₂ provides the carbon plants need to build tissues, nitrogen is essential for proteins, enzymes, and chlorophyll. Without enough nitrogen, plants simply cannot take advantage of higher CO₂ levels, no matter how abundant carbon may be.
Most nitrogen on Earth exists in the atmosphere as nitrogen gas, a form plants cannot directly use. To become useful, nitrogen must first be converted — or fixed — into biologically available compounds. This process is carried out mainly by specialized microorganisms, often living in soil or in partnership with plant roots.
Natural nitrogen fixation happens in forests, grasslands, and other ecosystems, and it also occurs on agricultural land, especially where legumes like soybeans or clover are grown. Climate models have long included estimates of this process to calculate how much additional plant growth might occur as CO₂ levels rise.
The Core Finding: Nitrogen Fixation Is Overestimated by About 50%
The new study focused on how Earth system models represent natural biological nitrogen fixation. By comparing model assumptions with updated observational data, the researchers found a striking mismatch.
Across multiple Earth system models, nitrogen fixation on natural land was overestimated by roughly 50%. In other words, models were assuming that ecosystems were receiving about twice as much usable nitrogen from natural fixation as they actually are.
Because nitrogen availability strongly controls plant growth, this error has a cascading effect. The study found that models with higher assumed nitrogen fixation also projected stronger future plant growth. As a result, the models overestimated the CO₂ fertilization effect by about 11%.
This may sound like a modest number, but in global climate projections, an 11% difference in carbon uptake is substantial. It can affect estimates of how quickly atmospheric CO₂ will rise and how much warming the planet will experience.
Why Earth System Models Rely on Nitrogen Fixation Estimates
Earth system models are among the most important tools in climate science. They simulate interactions between the atmosphere, land, oceans, and living ecosystems. Organizations like the Intergovernmental Panel on Climate Change (IPCC) rely heavily on these models to assess future climate scenarios and inform policy decisions.
To estimate how much carbon land ecosystems can absorb, models must account for nutrient limitations. Nitrogen is one of the strongest constraints on plant growth, especially in natural ecosystems. If nitrogen supply is assumed to be high, models predict more vegetation growth and more carbon removal from the atmosphere.
The new findings suggest that many of these assumptions are simply too optimistic.
How the Researchers Reached Their Conclusions
The study was led by Sian Kou-Giesbrecht, now a professor at Simon Fraser University in Canada, and a former doctoral student at Columbia University. The research team included Columbia professor Duncan Menge and other former doctoral researchers from his lab, now working at Harvard University and Earthshot Labs.
The researchers examined how different Earth system models represent nitrogen fixation and compared those values with updated measurements from natural ecosystems. These newer estimates come from a broader and more careful reassessment of nitrogen fixation rates, correcting for biases in earlier data that tended to focus on locations with unusually high fixation.
Their analysis showed a clear pattern: models that assumed higher nitrogen fixation also predicted stronger future plant growth and carbon uptake. Once nitrogen fixation rates were adjusted downward to reflect real-world conditions, projected plant growth declined accordingly.
The Carbon Dioxide Fertilization Effect Is Not Unlimited
One of the key takeaways from this research is that the CO₂ fertilization effect has real biological limits. While rising CO₂ can stimulate photosynthesis, it does not magically remove other constraints like nitrogen availability, water stress, or soil quality.
The study reinforces a growing consensus among scientists that land ecosystems cannot indefinitely absorb increasing amounts of carbon as emissions rise. Nutrient limitations, especially nitrogen, place hard boundaries on how much additional growth is possible.
This challenges overly optimistic views that natural ecosystems will continue to offset large portions of human-caused emissions far into the future.
Natural Ecosystems Versus Agricultural Land
An interesting nuance highlighted by the research is the difference between natural land and agricultural systems. While nitrogen fixation in natural ecosystems appears to be overestimated, fixation on agricultural land is often underrepresented in models.
Crops like soybeans, peas, and other legumes host nitrogen-fixing bacteria and can contribute significantly to nitrogen inputs. However, these agricultural contributions do not compensate for the overestimation in natural ecosystems when it comes to global carbon cycling.
Since forests and grasslands cover vast areas of the planet and play a major role in carbon storage, inaccuracies in modeling these systems have outsized effects on global projections.
Why This Matters for Climate Policy
Because Earth system models inform climate targets, carbon budgets, and mitigation strategies, overestimating natural carbon removal can be risky. If policymakers believe land ecosystems will absorb more carbon than they realistically can, emission reduction targets may be set too leniently.
The authors of the study argue that climate models should be updated to reflect more accurate nitrogen fixation rates. Doing so would produce more realistic estimates of future plant growth and carbon uptake, leading to better-informed climate planning.
A Broader Lesson About Climate Modeling
This research highlights an important point: climate models are only as good as the biological processes they represent. Nitrogen cycling is complex, varies across ecosystems, and depends on microbial activity that is still not fully understood.
Improving how models simulate nitrogen fixation is not just a technical detail — it directly affects predictions of how Earth’s systems will respond to continued greenhouse gas emissions.
Research Reference
Proceedings of the National Academy of Sciences
Sian Kou-Giesbrecht et al. (2025)
Overestimated natural biological nitrogen fixation translates to an exaggerated CO₂ fertilization effect in Earth system models
https://doi.org/10.1073/pnas.2514628122
