Carbon markets and biofinancing

An emissions allowance and an emissions credit are different objects, and conflating them is the source of most confusion about these markets. An allowance under a cap-and-trade system is a permit created by a regulator; its total quantity is a political decision and its quantity is known exactly. A credit is a claim that a tonne of carbon dioxide was emitted somewhere less, or removed, relative to what would otherwise have happened. That comparison is with a world that does not exist, so the credit’s quality is entirely a question of how well that world was modelled.

The baseline is a counterfactual, not a measurement

Additionality asks whether the activity would have happened anyway. It is unfalsifiable in the individual case: nobody can observe the forest that would have been cleared. Standards therefore approximate it with tests — regulatory surplus, financial or barrier analysis, common practice — that are procedural stand-ins for an unobservable fact. This is a real epistemic limit, not sloppiness, and it explains the recurring finding that project baselines were set higher than subsequent regional data supported. It is also why crediting has moved toward baselines set from jurisdictional or activity-wide reference data rather than project-by-project projection: a regional deforestation rate can at least be measured, even if its allocation to a specific project cannot.

Leakage is the same problem shifted in space. Protecting one parcel can move the pressure to the next parcel, so the accounting boundary has to be wide enough to catch displacement, and any boundary is a choice. Quantification frameworks such as ISO 14064-2 require baseline, boundary and leakage to be stated explicitly; the standard cannot make the estimate correct, only make it inspectable.

Permanence is a physical asymmetry

Carbon dioxide is effectively permanent in the atmosphere on management timescales, but biological carbon is a stock held in trees, soil and sediment that can burn, be harvested, be droughted or simply be managed differently by the next owner. Geological storage and mineralisation are hard to reverse; a forest is not. So a credit issued for a biological stock is a claim about a continuing condition, and the accounting has to say for how long. Hundred-year permanence periods and shorter commitments are both used, which means a nominal tonne can represent very different durations.

The instrument for this is the buffer pool: each project contributes a share of its credits, typically set in the low tens of per cent by a risk assessment covering fire, pest, political and management risk, into a shared unissued reserve. If a project reverses, credits are cancelled from the pool rather than from the buyer. It is self-insurance across a portfolio, and it holds only if pooled risks are independent. They are not fully independent — fire and drought are regionally correlated by climate — which is the structural weakness worth naming.

Where measurement is genuinely getting better

Above-ground biomass is now estimated from field plots combined with airborne and spaceborne lidar and radar, with allometric equations converting measured dimensions to carbon. Soil organic carbon is far harder, and is treated separately in this wiki. In both cases the physical measurement is improving faster than the counterfactual reasoning, and the counterfactual is what the tonne is defined against.