The Changing Face of REDD+: From Forecast to Measurement

At a Glance · How Project REDD+ Has Changed

Climate Decode considers Isometric’s protocol the most consequential change to project REDD+ since the market began, since it replaces the forecast on which every credit has rested with a measurement taken at each issuance. VM0048 corrected who sets the baseline by moving that task from the developer to Verra and its data providers, although the baseline remains a projection that is fixed until it is reassessed and is not adjusted to the deforestation observed in the meantime.

A credit issued under Isometric’s design would rest on the same comparison with control forests that researchers used to show that most legacy REDD+ credits were overstated, and the risk of over-crediting would thus sit with the developer before issuance rather than with the buyer after purchase. It is envisaged that this narrows, although it does not close, the gap in confidence between avoidance and removal credits.

The protocol text has not been published, and its value will depend on three design choices: how control forests are selected so that they are comparable yet beyond the reach of leakage, how a project demonstrates leakage below the 100% default, and how reversals are covered. Once these are settled, buyers should treat measured REDD+ credits as the reference point for avoidance and price legacy stock accordingly.

REDD+ credits compensate for deforestation that did not take place, such that every credit depends on a counterfactual that cannot be observed directly. Under the legacy methodologies, VM0007, VM0009 and VM0015, the developer selected the reference area and projected the rate of clearing for ten years, and studies that constructed matched control areas for existing projects found that most of the credited deforestation would not have occurred in any case.¹

The market has adjusted accordingly: Climate Decode’s Canopy registry data shows Verra REDD+ issuance falling from 131.4 million credits in 2021 to 14.5 million in 2025, while retirements of 35.3 million in 2025 indicate that buyers continue to draw down existing stock rather than new vintages.

Figure 1 · Verra REDD+ credits issued and retired by year. Source: Climate Decode Canopy Market Watch registry data, Verra, to May 2026.

Two responses have since been put forward at project level: Verra’s VM0048, which changes who sets the baseline, and Isometric’s proposed protocol, which changes how the baseline is established.

Table 1 · How project REDD+ crediting has changed. Sources: Verra VM0048 v1.0; ICVCM, November 2024; Isometric, September 2026; Climate Decode Canopy Market Watch registry data.

Verra replaced the legacy methodologies with VM0048 in November 2023, such that deforestation data and risk maps are now produced by Verra’s data providers and allocated to projects, and the developer no longer chooses its own baseline.² The methodology text nevertheless confirms that the baseline is estimated ex ante, applied to both ex ante and ex post calculations and reassessed only at the end of its validity period, with the sole check a one-way cap where the allocated baseline exceeds a government reference level.³ VM0048 is thus a better forecast rather than a different basis for crediting. The wider VCS Version 5 changes are covered in our nature-based solutions analysis and the CCP decision in our analysis of ICVCM’s approval of VCS v5.

Isometric announced its avoided deforestation protocol on 22 September 2026.⁴ It compares forest loss in the project with loss in matched control forests over the same period, credits only the difference and repeats the comparison at every issuance, with leakage assumed to be 100% until the developer demonstrates otherwise, against the 0 to 21% that legacy methodologies deducted.

Jurisdictional crediting under ART TREES operates at a different scale and is covered in our ART TREES analysis.

Climate Decode screens REDD+ credits by methodology, baseline approach and label, so legacy, VM0048 and Isometric credits can be compared on one basis.

The distinction between the two approaches lies in the counterfactual itself. Under VM0048 the counterfactual is a model, and any error in it carries through every issuance until the baseline is reassessed, whereas under Isometric’s design the counterfactual is observed in comparable forests, such that a fall in regional deforestation reduces credits at the next issuance rather than years later. The method is, in effect, the test that researchers applied to legacy projects, now adopted as the crediting rule.

Table 2 · How measurement changes the core REDD+ problems. Sources: Verra VM0048 v1.0; Isometric, September 2026; Climate Decode analysis.

The approach introduces a risk of its own: where clearing displaced by a project moves into the control forests, deforestation there rises and the measured benefit of the project is overstated. The selection of control forests, close enough to be comparable and far enough to lie beyond the reach of leakage, is therefore the design choice on which the credibility of the protocol will rest.

The case against avoidance credits has rested on the observation that avoided deforestation cannot be measured, whereas a tonne removed can. Crediting only the measured difference against comparable forests does not make avoidance a removal, although it moves the basis of the credit from a projection to an observed result, and Climate Decode expects this to bring back corporate buyers who withdrew from REDD+ after 2023, provided that the method holds and the ICVCM recognises it.

Finance is the principal constraint, since credits that exist only after measurement are difficult to sell forward and avoidance projects require capital at the outset. The model therefore suits well-capitalised developers first: Revalue is moving future projects to Isometric, and Imperative, with up to USD 1.25 billion in funding agreements, will certify under it.⁴ Developers should accordingly structure offtake agreements on volume ranges rather than fixed tonnes.

The move also marks a change of direction for Isometric, which built its position as a registry for durable removals, covered in our analysis of its registry, and has issued fewer than 100,000 credits across all its protocols. REDD+ offers both volume and the widest deficit in trust, and the registry’s approach is likely to command its highest premium where trust is lowest.

Four developments will determine whether measured baselines become the standard for project REDD+:

1. Isometric’s protocol text, due for consultation by November 2026, and in particular its rules on control-area matching, the path below 100% leakage and reversal cover.
2. The first VM0048 issuances, against which Climate Decode’s supply and demand model, after methodology haircuts, holds forward issuance from existing nature-avoidance projects at about 27 million credits a year.⁵
3. Any application by Isometric for CCP and CORSIA recognition of the protocol.
4. Prices for measured REDD+ credits against VM0048 and ART units, and what buyers do with legacy stock under scrutiny such as the CFTC inquiry.

Climate Decode Platform

The registry figures in this analysis come from Canopy Market Watch, which tracks issuance, retirements, ratings and labels across ten registries.

Climate Decode · Advisory

Choosing a REDD+ methodology, or a REDD+ credit

Climate Decode compares legacy, VM0048, ART and Isometric pathways on volume, risk and buyer acceptance, for developers and buyers.

Notes

Sources

Isometric’s protocol is described as announced; its text had not been published for consultation at the time of writing, and details may change.