How does sampling uncertainty cut tree carbon credits in a CCTS afforestation project?

Question

Senseible · Accepted Answer

Tool BM-T-AR-0004 measures uncertainty as the standard error expanded to 90% confidence, divided by the mean. Up to 10% there is no discount. Above that, project tree carbon is cut by 25%, 50%, 75% or 100% of the uncertainty, in bands up to and beyond 30%. Take the standard error of the mean, multiply it by the t value for 90% confidence, and divide by the mean. The tool's own example: a mean of 45.328 t dry matter per ha from 34 plots with a standard deviation of 12.776 gives a standard error of 2.191, expanded to 3.707, an uncertainty of 8.18%. Only sampling uncertainty is counted. Up to 10%: no discount. Over 10% to 15%: 25% of the uncertainty. Over 15% to 20%: 50%. Over 20% to 30%: 75%. Over 30%: 100%. The discount lowers project estimates and raises baseline estimates, so both sides move against you. The tool's example: 60 plus or minus 9 t dry matter per ha is 15% uncertainty, a discount of 25% x 9 = 2.25, so the project mean becomes 57.75 and a baseline mean would become 62.25. Lay enough plots and stratify by species, age and site so each stratum is uniform. Baseline tree and shrub change may be taken as zero on visibly degraded land, such as eroded, saline or mine spoil land, with documents or participatory rural appraisal. Shrub strata with crown cover under 5% count as zero shrub biomass. Defaults: carbon fraction 0.47, root-shoot ratio 0.40 for shrubs.

How does sampling uncertainty cut tree carbon credits in a CCTS afforestation project?

How uncertainty is measured

Discount bands

Which way it cuts

Ways to keep it low

What Senseible does here