Science
Claims travel with their evidence, or they do not travel.
The Atlas knowledge layer stores methodologies, research and case studies as citable records, and the claim-risk engine flags anything asserted beyond what the evidence supports.
Methodologies
Every methodology is versioned, peer-reviewed and linked to the projects and assets that use it.
| METHODOLOGY | VERSION | STATUS | SECTORS |
|---|---|---|---|
| Soil Organic Carbon | v2.1 | ACTIVE | AgricultureClimate |
| Blue Carbon MRV | v1.0 | ACTIVE | OceanClimate |
| Biodiversity Uplift Index | v1.2 | ACTIVE | AgricultureOceanClimate |
| Watershed Hydrology | v1.0 | ACTIVE | ClimateAgriculture |
| Circular Material Recovery | v1.0 | ACTIVE | Circular |
| Population Wellness Index | v0.9 | DRAFT | Health |
Knowledge library
- METHODOLOGY
Soil organic carbon sampling and uncertainty bounds
Atlas MRV Library · v2.1 · demo record
- CASE STUDY
Cooperative-led restoration at 10,000 ha scale
Regenerative Valley · demo record
- POLICY
Anti-greenwashing claim review procedure
Values & Policy Engine · demo record
- RESEARCH
Mangrove blue carbon permanence under storm regimes
Atlas Science Index · demo record
Evidence levels
Every data point displayed on the platform carries one of these labels. Aggregates never hide the weakest link in their evidence chain.
- VERIFIED
Independent third-party review completed. Evidence package meets methodology standard.
- UNVERIFIED
Data submitted and validated by platform. Awaiting independent review.
- ESTIMATED
Modelled or interpolated from adjacent data. Uncertainty range disclosed.
- PREDICTED
Forward projection from model. Clearly labelled as forecast, not outcome.
- USER-PROVIDED
Self-reported by project operator. Not yet validated by platform or verifier.
Anti-greenwashing engine
Automated claim-risk detection runs on every published statement, with evidence displayed beside the claim.
- Unsupported environmental claims
- Missing baselines
- Weak evidence
- Double counting
- Inconsistent methodologies
- Suspicious data
- Conflicting measurements
- Unsupported AI-generated claims
Flagged claims are held from publication until evidence is provided or the claim is revised. The engine never suppresses evidence — it requires it.
Atlas knowledge graph
Every entity in the platform is connected in a queryable knowledge graph. Claims can be traced from assertion back to raw data through every transformation.
- Projects → Ecosystems
- Ecosystems → Interventions
- Interventions → Metrics
- Metrics → Scientific evidence
- Evidence → Organizations
- Organizations → People
- People → Locations
- Locations → Assets
- Assets → Funding
- Funding → Outcomes