Uncertainty in carbon accounting is not a caveat at the end of a report. It is an input to the credit calculation. Standards require it to be quantified, and they apply conservative deductions scaled to it, which means the width of a project's confidence interval translates directly into the number of credits it can sell.

The Mechanism

The logic is straightforward once stated plainly. A crediting system cannot allow a project to be paid for carbon that may not exist. So where an estimate is uncertain, the system credits toward the conservative end of the plausible range.

The wider the interval, the further from the central estimate the credited figure sits. Two projects with identical true carbon and identical central estimates will issue different numbers of credits if one measured more precisely than the other.

That is the commercial case for measurement quality, and it is a much stronger argument than "rigour is good."

Where Uncertainty Comes From

Total uncertainty accumulates across every step between a field measurement and a headline tonnage.

Measurement error — diameter read slightly wrong, height estimated rather than measured, species misidentified.

Allometric model error — the equation converting dimensions to biomass carries its own regression error, often larger than measurement error.

Wood density — species-level variation, and genus-level or default substitution where identification is imprecise.

Expansion factors — below-ground biomass derived from above-ground by ratio, unmeasured pools inferred.

Sampling error — the interval arising from having measured a subset rather than the whole.

Source	Typically reduced by
Measurement error	Training, protocol, instrument choice
Allometric error	Better-matched equations, local calibration
Wood density	Species-level identification and matching
Expansion factors	Direct measurement of more pools
Sampling error	Better stratification, more plots

Propagation, Not Addition

text

Combined uncertainty (independent error terms):

    U_total = sqrt( U_measurement^2 + U_allometric^2 + U_density^2
                    + U_expansion^2 + U_sampling^2 )

Because each term is squared before summing, the LARGEST term dominates.
Halving the biggest source moves U_total far more than halving a small one.

These do not simply add. Independent errors combine in quadrature, calculated as the square root of the sum of squares, which has an important practical consequence.

Because the terms are squared before summing, the largest single source dominates. Halving a small error term barely moves the total; halving the largest one moves it substantially.

This is why "add more plots" is often the wrong optimisation. If allometric model error is the dominant term, doubling plot count improves the sampling term while leaving the dominant term untouched, effectively spending real money for a marginal gain. Identifying which term dominates is the first analytical step, and it is frequently skipped.

What This Means for Design Decisions

Three implications follow directly.

Stratify before you densify. Better strata reduce within-stratum variance, which reduces sampling error for the same plot count. It is usually the cheapest available improvement.

Match the allometry. Where allometric error dominates, which is common in mixed tropical forest using generic equations, improving equation choice or supporting local calibration yields more than any plausible increase in plot numbers.

Measure pools you are currently inferring. Expansion factors carry model error. In systems where an inferred pool is large, such as below ground biomass or soil in peat and mangrove, direct measurement can shift the total materially.

Reporting It Honestly

Standards expect uncertainty to be stated, with its basis. A project that reports a confidence interval and shows how it was derived is straightforward to assess. A project that omits it is assigned the most conservative treatment available, which is invariably worse than whatever it was avoiding disclosing.

There is also a reputational dimension. Buyers with technical reviewers read uncertainty statements as a proxy for methodological seriousness, as an honest wide interval reads better than a suspiciously narrow one with no derivation.

Frequently Asked Questions

View all FAQs

Standards credit conservatively where estimates are uncertain, so a wider interval means a larger deduction.

Sources

1. IPCC 2006 Guidelines Vol. 4 (AFOLU), uncertainty guidance — https://www.ipcc-nggip.iges.or.jp/public/2006gl/vol4.html (accessed 16 Sep 2026)

2. Verra VCS uncertainty and deduction requirements — https://verra.org/ (accessed 16 Sep 2026)

3. ICVCM Core Carbon Principles — https://icvcm.org/core-carbon-principles/ (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.

Open TREEO Carbon Simulator Contact us