Documentary photograph illustrating drawing the Boundary
← Back to blog

Defining the Project Boundary: Three Boundaries, Not One

Methodology & dMRV•16 October 2026•TREEO Indonesia•2 min read

Ask what a project's boundary is and you get a map. The map is one of three boundaries, and the other two, which are when the project counts and which emission sources it counts, cause as many problems. The spatial one, though, is the only one that is genuinely hard to change once set.

The Three

BoundaryDefinesChanged later?
SpatialWhich land is in the projectDifficult and disruptive
TemporalCrediting period start and endFixed at validation
SystemWhich gases, pools and sources are countedOnly via deviation

The Spatial Boundary Must Be Geometrically Real

A project boundary is a set of polygons with coordinates, and it has to survive contact with a GIS.

The recurring failures are mundane and expensive: polygons that overlap, so area is double-counted; gaps between adjacent parcels that appear as unclaimed slivers; boundaries derived from a scanned map with no usable coordinate reference system; and areas that include land the project has no legal right over.

That last one is the serious one. A boundary drawn around the forest rather than around the tenure creates an unresolvable conflict at validation, and no amount of measurement precision compensates for it.

Exclusions Are Part of the Boundary

Real project areas contain things that are not the project: settlements, roads, water bodies, rock, existing cropland.

These must be excluded explicitly, with their own polygons. A project reporting 10,000 hectares that contains 800 hectares of village and river is reporting an area it will be asked about, and the correction always comes out of the project's claim.

The practical standard is that area should be derived from the GIS after exclusions, not from a title document or a concession permit. Those describe legal extent, not forested extent.

Discrete Areas and Grouped Projects

Projects spanning many non contiguous parcels, which is common in smallholder programmes, face an additional requirement: each discrete area needs its own geometry, and the methodology usually specifies how new areas may be added over time.

Grouped project structures exist precisely for this, allowing new instances to join under an approved framework. Attempting the same thing without the grouped structure means each addition is a boundary change requiring reassessment, which is slow and costly.

Developers planning to expand should choose the grouped structure at the outset. Retrofitting it is possible and unpleasant.

The Temporal Boundary Has a Trap

The crediting period start date determines which activities count. Work done before it generally does not generate credits, which catches projects that plant first and register later.

It also interacts with additionality: activities already underway before any carbon consideration are harder to argue were caused by carbon finance. The start date is a decision with both accounting and eligibility consequences, and it should be taken deliberately rather than inherited from whenever the paperwork happened to be filed.

What Good Looks Like

A shapefile with a stated coordinate system, topologically clean, with exclusion polygons, an area reconciliation against the tenure documents, and a written note explaining any difference between legal area and project area. Four things. Projects that supply them move through boundary review quickly; projects that supply a PDF map do not.

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.

Related Articles