How OSEM measures transparency in tree planting
Organizations claim millions of trees planted every year, but how much of that is actually verifiable? OSEM's scoring system quantifies how much data an organization has actually disclosed about their planting projects — down to the individual site level.
The score isn't a judgment of quality. It's a measure of transparency: how much can be independently verified from the data they've made available?
Each project is scored by examining how many data attributes are populated across its records — projects, land parcels, plantings, polygons, crops, stakeholders, and sources.
Not all attributes are equal. High-value fields like polygon boundaries, GPS coordinates, species identification, and planting dates carry significantly more weight. A polygon with adequate tree density (≥200 trees/ha) is worth 20 points; sparse polygons score less.
The project score is points scored ÷ points available, expressed as a percentage. Bonus points from stakeholder and organization data can push scores above 100%.
ProjectTable ──┐
LandTable ─────┤
PlantingTable ─┤ Score table
PolyTable ─────┤── count populated attrs ──► ┌──────────────────┐
CropTable ─────┤ × score matrix weights │ score (%) │
SourceTable ───┘ │ pointsScored │
│ pointsAvailable │
└──────────────────┘Every populated attribute scores 1 point by default. The following carry extra weight:
Polygon density matters: ≥200 trees/ha = 20 pts, ≥10 trees/ha = 2 pts, <10 = 0 pts.
Organization scores aggregate their project scores, then factor in claim verification. If an organization claims 1 million trees planted but only 50,000 are accounted for in disclosed project data, their claim percentage is ~5%.
The final organization score combines the average project quality with the proportion of claims that are actually backed by data. Organizations are also segmented by their most common stakeholder type for peer comparison.
The entire scoring methodology is open source. You can inspect the logic, challenge it, or contribute improvements.