Bringing First-Mile Intelligence to ACR Dynamic Baselining

Dynamic baselining is becoming the standard approach for Improved Forest Management projects under the American Carbon Registry. Unlike static baselines fixed at project inception, dynamic baselines are updated throughout the crediting period using observed data to reflect how forests would otherwise be managed.

Developers implementing ACR Dynamic Baselining are encountering a broader challenge facing forest carbon markets. Most of the risk, uncertainty and value creation sits at the first mile, where land is managed and harvesting decisions are made. Yet this is also where timely, granular data has historically been hardest to access.

Treefera supports ACR Dynamic Baselining by bringing first-mile visibility into a process that directly determines credit volume, credibility and capital outcomes.

What dynamic baselining changes

A baseline represents the expected harvesting and carbon outcomes in the absence of a carbon project. Under ACR’s dynamic framework, this ‘counterfactual’ is reassessed prior to each credit issuance – to reflect real changes in forest conditions and management behaviour.

Baselines are constructed by identifying legal and economic constraints on harvesting, confirming applicable management practices and developing a harvest schedule based on activity observed on comparable non-project lands. That schedule is then revisited using measured data rather than fixed assumptions.

This shift matters because baseline quality directly affects buyer confidence. When baselines are built on current, observed conditions at the first mile, credit issuance becomes more defensible and more aligned with how capital is deployed.

Why first-mile intelligence is required

Many existing workflows rely on open-source datasets such as Hansen and LCMS. While effective for identifying clear cuts, these tools struggle to detect thinning and low-intensity harvests due to coarse spatial resolution and infrequent update cycles.

These gaps create distortion at the first mile. Undetected harvesting activity inflates baseline assumptions, reduces the number of credits that can be legitimately claimed and weakens confidence in the underlying data.

Treefera’s platform is designed to address this problem. As a first-mile intelligence platform, Treefera combines satellite and sensor data with deep scientific, mathematical and environmental expertise, processed through AI-native systems built to operate at scale. This enables the detection of smaller, more frequent harvest events that conventional tools miss.

Detection outputs are weighted using regional accuracy statistics so harvested area estimates remain unbiased and scientifically defensible. The result is a baseline grounded in observed land performance rather than proxy assumptions.

Project example: Anew Climate

Anew Climate is a leading developer of large-scale forest carbon projects operating across diverse geographies and ownership structures. In implementing ACR Dynamic Baselining across its forest portfolio, Anew required timely, defensible insight into harvesting activity on comparable non-project lands.

Applying Treefera’s model increased baseline harvest intensity on comparable properties by 3.3 percent. This translated to additional credits per year, with the potential to generate significant returns over the lifespan of the project.

Using Treefera’s platform, Anew processed more than 300 properties in under 48 hours. Anew’s technical team estimated the same analysis would have taken months using prior methods. Model performance exceeded standard benchmarks, achieving an F1 score of 0.55 compared with a typical default of 0.4.

From compliance to capital confidence

Treefera’s workflow aligns directly with ACR requirements. Project developers provide project boundaries, comparable land boundaries, a defined assessment window and ground-truthed harvest data for accuracy calibration. Treefera delivers repeatable baseline updates that can be reviewed at each issuance cycle.

Recent enhancements to Treefera’s ACR Dynamic Baselining workflow have further strengthened this process. Required prerequisite analyses, including forest cover assessment, historic wind and historic fire, are now executed automatically before the final baseline is run. This reduces manual sequencing and ensures baseline outputs are generated from a consistent, fully contextualised assessment.

The workflow has also been refined to handle edge cases more robustly – such as missing forest cover years – and outputs have been improved to support clearer interpretation during validation and verification. Together, these improvements reduce friction for technical teams and increase confidence in ACR-aligned baseline results delivered through the Treefera API.

For developers like Anew, this shifts dynamic baselining from a compliance exercise to a data-driven process that supports credibility, scale and capital allocation. Accurate, trustworthy first-mile data does not just improve methodology. It determines where capital flows and which projects earn long-term confidence.

Watch the Anew case study here.