In 2025, the Integrity Council for the Voluntary Carbon Market awarded its first Core Carbon Principles (CCP) labels to agricultural soil carbon methodologies. After years spent sitting between climate solution and agronomic co-benefit, soil carbon now has a recognized place in the high-integrity market.
That is a reasonable place for it to be. Soils hold nearly 3 trillion tonnes of carbon, around three times the amount held in the atmosphere, and better management of the world's cropland could remove up to 1.57 billion tonnes of carbon dioxide each year. The agronomy is well understood, the co-benefits for soil health and water retention are proven, and farmers are the largest group of land managers not yet meaningfully involved in carbon removal.
The science behind soil organic carbon (SOC) has had a long time to mature. Johan Gottschalk Wallerius laid the foundations of agricultural chemistry in 1761. Franz Karl Achard ran what is generally considered the first laboratory study of soil organic matter in 1786. Rothamsted has been measuring SOC in the same wheat field since 1865. Between 2000 and 2025 alone over 880,000 papers on the subject appear in the Google Scholar database.
The difficulty has never been science. It has been accounting. SOC is among the most cost-effective carbon removal approaches and among the hardest to certify. The shortcuts made in the past share an unfortunate characteristic: they tend to increase the number of certificates issued. Measurement errors that fall randomly on either side of the truth will largely cancel out across a portfolio. Errors that consistently favor the project do not.
Each shortcut has a well-established correction, and each correction can be applied today. Isometric’s Improved Soil Management Protocol, which was certified in July, adopts them all.
Measuring soil carbon by mass rather than by volume
Anyone who has baked from a recipe knows that a cup of flour is not a fixed quantity of flour. Press it down and the same cup holds noticeably more. Soil behaves in much the same way, and that matters a great deal for carbon accounting.
Reducing tillage is one of the most common soil management practices. It also compacts soil, raising bulk density by roughly 2% to 3%. A core sample in the third year of a project therefore lifts slightly more soil than it did at the start. More soil in the sample means more carbon in the sample, which in turn becomes more carbon per hectare, even where the amount of carbon stored in the ground has not changed at all.
Without safeguards, the project is paid for compacting soil rather than removing carbon. CarbonPlan modeled the impact of this across 142 million hectares of US cropland and found that a mid-range scenario would result in roughly 367 million certificates based on carbon removal that never happened.
The fix, known as equivalent soil mass accounting, is to compare like quantities of soil instead of like depths. Most methodologies don’t do this. Isometric requires it for every project. We calculate the soil mass from the mass of the sample and the dimensions of the corer. This removes bulk density from the calculation, along with the imprecision that measuring it independently would introduce.
Calibrating models against real measurements
Under Isometric's protocol, projects can either measure and remeasure, or, to issue more frequently, measure and model. Technologies like digital soil mapping and remote sensing make soil carbon monitoring, reporting, and verification (MRV) more affordable at scale, so it is worth being clear about the job they do. A model trained on the fields where data happened to be cheapest, validated against the same data, and then asked to identify a small increase in soil carbon against a large and noisy baseline will produce a confident-looking number. However, confidence is not the same thing as accuracy.
To fix this, Isometric requires projects using a model to physically remeasure the soil at least every five years. This also recalibrates the model, so its estimations are grounded in the actual soil it is describing. Where the model and measurement disagree, the number of certificates issued is trued up to the measured amount. If a model has overestimated, the excess is recovered from the buffer pool and must be replaced before any further certificates are issued.
This means project developers can benefit from the efficiency of modeling in the years between sampling without that model drifting away from reality.
Giving uncertainty a price
If being imprecise does not reduce the number of certificates a project receives, then precision is simply a cost with no benefit. Soil carbon is unusually exposed to this problem, because the signal a project is trying to detect is small, the natural variability it sits within is large, and the party best placed to characterize the uncertainty is also the party being paid based on the result.
The fix is to make uncertainty cost something. Isometric runs all soil organic carbon data through a Monte Carlo simulation and issues certificates at the 30th percentile of the resulting distribution, rather than at the central estimate. Where that percentile falls below zero, no certificates are issued for the period at all. A wider data distribution means fewer certificates, which puts precision in the supplier's own interest instead of something a certifier has to insist on.
Projects that invest in denser sampling, better laboratories, and tighter analytical control benefit under Isometric’s protocol.
The protocol also closes the obvious loopholes. Model validation has to draw on 25 years of published literature across four named databases, and a study "may not be excluded on the basis that its inclusion would widen the Monte Carlo input distributions or increase the uncertainty discount."
Accounting for what happens to the harvest
A soil carbon project can be measured accurately and still leave the atmosphere no better off. If a project reduces the amount of food that can be grown, that production tends to move elsewhere, along with the emissions. Yield matters in this pathway, and the market currently overlooks it.
The fix is to measure yield as carefully as carbon. Isometric requires every project to have field-level yield data. This can come from farm management records, grain elevator receipts, or crop insurance records, and must be assessed against a regionally indexed five-year pre-project baseline.
If crop yield falls by less than 3%, no action is taken, as this matches normal variability. If yield falls by 3% or more, the entire decline in yield is assessed to determine the leakage discount, not just the part beyond 3%. This removes any incentive to get close to the threshold, as crossing it means losing certificates for the entire decline, not just a slice of it.
For example, if a project results in a 5% decline in yield, Isometric will assess the full 5%, not just the 2% by which it exceeds the threshold.
Increases in yield also have a 3% threshold, but unlike the decrease threshold, only the amount above 3% is counted as a surplus. These increases can be used to offset decreases elsewhere in the project, but only for the same type of crop in the same reporting period. Finally, if the yield of a single crop in a project falls by more than 15%, the project becomes ineligible for certification for that entire reporting period.
This sets a high bar for projects, but for an important reason. It incentivizes projects that maintain or increase crop yields, an outcome that also benefits the farmers and communities working the land, and one most project developers are already aiming for.
Describing durability accurately
For most carbon removal pathways, a reversal is an accident, such as a fire, leak, or equipment failure. For soil carbon it is a management decision, and often a rational one. A farmer might return to tilling after a difficult growing season, but the stored carbon still returns to the atmosphere despite their intentions. The right way to think about the risk of reversal for soil carbon is not catastrophe, but disadoption.
The fix is to label a certificate's durability transparently. Isometric sets the durability of a soil carbon certificate at half the length of the project's commitment period. A project that commits for 40 years generates certificates that last 20. The reversal risk is priced into a buffer pool sized to each project's own circumstances, rather than a flat rate. And the commitment itself is based on outcomes rather than practices. A farmer is free to change what they do, provided the measured stock stays at or above the level already certified.
Labeling durability honestly is what makes certificates useful to buyers. If you know you are purchasing 20 years of storage, you can match it against a 20-year obligation and then price it accordingly.
What this means
Taken together, these five fixes generate soil carbon certificates that buyers can actually price: a specific quantity of carbon, stored for a stated number of years, with the cost of any remaining uncertainty incorporated into the project, rather than returned to the atmosphere.
None of these corrections require new instruments or new science. What they require is a methodology that prioritizes accuracy. Isometric’s Improved Soil Management Protocol includes all five, and we hope to see these fixes become commonplace across the market.
Like all Isometric protocols, it will continue to evolve with the science. We’re already considering how the protocol can draw distinctions between the different types of carbon stored in soils in future. There is also growing evidence that soil carbon can persist for longer at greater depths, which we’ll continue to consider as better models, quantification approaches, and measurement technologies improve.
If you’re a project developer exploring Improved Soil Management, or a buyer evaluating soil carbon certificates, get in touch to find out more.
