Soil holds more carbon than the world's vegetation and atmosphere combined, and dryland soil carbon is the single most difficult pool a carbon project can try to credit. The difficulty is not conceptual. It is that the signal is small, the variance is enormous, and the measurement is destructive.
Three Compounding Problems
Spatial heterogeneity. Soil organic carbon varies sharply over metres, driven by microtopography, texture, historical land use and vegetation patchiness. Two cores ten metres apart can differ substantially, which means the variance term in any sample-size calculation is large before you start.
Slow change. Management change typically shifts soil carbon by a small percentage per year against a large existing stock. Detecting a change of a few percent against a background that varies by tens of percent requires either many samples or many years — usually both.
Destructive sampling. You cannot re-measure the same core. Repeated sampling means new locations, which reintroduces spatial variance precisely where permanent plots would otherwise have removed it.
What the IPCC Framework Provides
The guidelines describe two accounting approaches. Stock-change compares measured carbon stocks between two points in time. Gain-loss models inputs and outputs. Stock-change is more direct and more demanding; gain-loss is more tractable and inherits model uncertainty.
Default methods reference a standard 30 cm depth. That default matters more than it looks: management effects frequently extend deeper, and sampling only to the default can miss both gains and losses in the profile below.
| Decision | Options | Consequence |
|---|---|---|
| Accounting approach | Stock-change vs gain-loss | Direct measurement vs inherited model error |
| Sampling depth | Default 30 cm vs deeper intervals | Shallow sampling can miss sub-surface change |
| Bulk density | Measured vs assumed | Assumed density is a common hidden error source |
| Stratification | By soil type and land use vs uniform | Drives required sample count directly |
The Bulk Density Trap
Soil carbon stock is carbon concentration multiplied by bulk density multiplied by depth. Concentration is usually measured carefully; bulk density often is not.
That asymmetry matters because management changes can alter bulk density itself, including compaction, tillage, and root growth. Comparing stocks across time using a fixed assumed density can produce an apparent carbon change that is really a density change. Methodologies increasingly require equivalent mass correction for exactly this reason.
What Makes a Dryland Soil Project Credible
Stratify by the variables that actually drive soil carbon, such as soil type, texture class, and land use history, rather than by field boundary. Measure bulk density alongside concentration at every depth interval. Sample deeper than the default where management could plausibly affect the profile. And set expectations on timeline honestly, recognizing that a detectable, defensible change in dryland soil carbon takes years, not one monitoring period.
Frequently Asked Questions
SustainabilityThe change is small relative to a large, highly variable stock, and sampling is destructive so permanent plots cannot be re-measured.
Sources
1. IPCC 2006 Guidelines Vol. 4 (AFOLU), soil organic carbon methods — https://www.ipcc-nggip.iges.or.jp/public/2006gl/vol4.html (accessed 16 Sep 2026)
2. UNFCCC webinar on soil organic carbon — https://unfccc.int/sites/default/files/resource/UNFCCC%20Webinar%20on%20SOC-session%203%20%28part%201%29.pdf (accessed 16 Sep 2026)
3. Verra methodologies, soil sampling requirements — https://verra.org/methodologies/ (accessed 16 Sep 2026)
Turn uncertainty into a plan
Sampling design and growth assumptions decide how much measured carbon survives conservative deduction. TREEO Carbon Simulator models sequestration scenarios and biomass growth before the field season.
