A restoration project's first monitoring period is the hardest number it will ever have to defend. The carbon gain is small, the variance between plots is large, and the ratio between them is at its worst precisely when the project most needs to demonstrate progress.
The Arithmetic of a Young Stand
In year three, a planted site holds small trees of widely varying size. Some seedlings died, some are thriving, some are barely alive — and the difference between the best and worst plot can exceed the mean.
| Stage | Signal (carbon gain) | Noise (between-plot variance) |
|---|---|---|
| Years 1–3 | Very small | Very large — survival differences dominate |
| Years 4–10 | Growing | Declining as survivors converge |
| Mature | Large | Moderate, driven by site quality |
That is the reverse of what a project's finances want. Cash is needed early; credible numbers arrive later.
What Makes It Worse
Sampling too soon. A monitoring period short enough to capture almost no growth produces a change smaller than its own confidence interval — and the honest report of that is an interval spanning zero.
Ignoring mortality in stratification. If survival varied across the site and the strata do not reflect it, the within-stratum variance carries the whole difference.
Independent re-sampling. Drawing fresh plots each period reintroduces spatial variance into a comparison that permanent plots would have removed.
What Actually Helps
Permanent plots from the first measurement. Observing change in the same locations is dramatically tighter than differencing two independent samples. For a small signal, this is often the difference between measurable and not.
Longer first monitoring period. Waiting lets the signal grow relative to the noise. It costs cash flow and buys credibility.
Stratify by survival, not just by planting design. Once mortality is visible, it is frequently the strongest predictor of standing biomass — and therefore the most useful stratification variable.
Model expectations honestly. A project that forecast high early accumulation will report a shortfall. One that forecast the true shape of the curve reports on plan.
The Broader Point
This is not a flaw in restoration. It is a property of measuring small changes against variable backgrounds, and it applies equally to soil carbon and to improved forest management.
What separates projects is whether the design anticipated it. Anticipating it costs nothing at planning time and is impossible to retrofit once the first period has been measured.
Check the land before you commit
Eligibility turns on land history and forest cover at a cut-off date. TREEO Eligibility runs satellite-based forest cover detection on your area of interest — before a single seed goes into the ground.



