Building trust in soil carbon as a climate solution requires stronger evidence

(Press-News.org) In a comment published in Nature Climate Change, Mark Bradford, the E.H. Harriman Professor of Soils and Ecosystem Ecology, and Yale School of the Environment research scientists Sara Kuebbing and Alexander Polussa ’25 PhD, together with colleagues Emily Oldfield ’05, ’11 MESc, ’19 PhD, of Environmental Defense Fund (EDF) and Jonathan Sanderman of the Woodwell Climate Research Center, argue that the scientific evidence supporting soil carbon’s role in mitigating climate change remains too weak to meet the standards required for policy and carbon markets.

The team notes that while models that rely on data from small-plot experiments dominate soil carbon accounting, they fail to reflect the messy, real-world conditions of working farms. Without more rigorous data collected at the scale of commercial agriculture, they caution, efforts to credit soil carbon with emissions reductions risk overstating benefits and undermining trust.

“Low confidence leads to inaction, undermining support that should be flowing to farming communities for economically and environmentally sustainable food production,” Bradford said.

The promise and the problem Soil carbon sequestration, boosting the amount of organic carbon stored in soils through practices such as cover cropping and reduced tillage, is widely viewed as a win-win strategy for climate mitigation and soil health. The public ranks soil carbon farming just behind tree planting as a preferred climate solution, the authors emphasized, citing a 2025 Communications Earth & Environment study. International frameworks, including the Intergovernmental Panel on Climate Change Greenhouse Gas Guidelines and the European Union’s Carbon Farming regulations, already incorporate soil carbon into their policies.

However, Bradford and his co-authors caution that enthusiasm has outpaced the evidence. Many current carbon accounting systems rely heavily on process-based ecosystem models that simulate how plants grow and soils cycle carbon. These models are often initialized with limited measurements, then used to predict how practices will affect carbon stocks.

Limited measurements means that validation datasets are scarce, geographically limited, and often drawn from highly controlled small-plot experiments. Such trials are excellent for understanding mechanisms but lack external validity, defined as the ability to generalize findings to diverse farms, soils, and climates.

“There is much investment happening right now in soil carbon as a natural climate solution with big expectations around its potential to mitigate climate change,” said co-author Oldfield, a senior scientist at EDF and a YSE associate research scientist. “A risk is that the results we observe among small-plot experiments that are highly controlled don’t pan out across working farms. Large-scale validation can help manage expectations and build confidence that these investments result in real progress on climate.”

Lessons from public health To underscore the challenge, the researchers draw an analogy from medicine. A vaccine may show efficacy in mouse models, but that evidence alone is insufficient to warrant large-scale human vaccination programs. Regulators demand randomized control trials and observational studies at human population scales before approving interventions.

Climate solutions, they argue, deserve the same level of scrutiny. If climate health is fundamentally important to human health, then the burden of proof for soil carbon sequestration must rest on large-scale, real-world data.

“We have high expectations for the quality of health data,” Bradford noted. “Some of the most rigorous evidence comes from trials administered in the real world, as opposed to controlled clinical conditions. Effectiveness under real-world scales builds confidence that vaccines really work. We can adapt those approaches to test the efficacy of natural climate solutions. Doing so allows us to confidently estimate the real climate benefits achieved.”

Building a stronger evidentiary foundation Collecting data at the scale of commercial agriculture is often dismissed as infeasible. Soil carbon stocks are large and variable, while changes occur slowly and unevenly across fields. The authors counter, however, that fields such as economics and epidemiology have long managed to detect meaningful signals amid noisy, real-world data.

They call for “causal inference” designs, such as randomized or controlled comparisons across many farms, that can reveal the average population-level effects of climate-smart practices. Project-scale studies involving tens to hundreds of fields could provide the gold-standard datasets needed to validate models and underpin trustworthy carbon markets and national climate goals.

Such work, they note, has historical precedent. In the early 20th century, agricultural research often combined applied statistics and farmer partnerships to test interventions across regions and seasons. Reviving that tradition, supported by public–private partnerships, could lay the foundation for more reliable soil carbon science today.

“Markets need confidence,” said co-author and vice president of science at the Woodwell Climate Research Center, Jonathan Sanderman. “Well-coordinated public-private partnerships will be critical for building the evidence base needed to continue scaling soil carbon solutions.”

Implications for policy and markets The implications of this research are wide-ranging. Many nations view soil carbon solutions as a vital part of their efforts to reduce greenhouse gas emissions, and carbon markets have already issued credits based on models validated with limited data. Climate goals and market credibility could be at risk if those credits do not reflect genuine carbon storage, the authors note.

They emphasize that stronger evidence is a scientific need and a policy imperative and encourage decision makers to take steps to ensure that soil carbon interventions deliver real, verifiable climate benefits before relying on them in accounting frameworks.

“Robust causal datasets at this scale,” they state, “would constitute a gold standard for quantifying the benefits of climate-smart interventions.”

END