Standardising understanding and collecting evidence of earth observations.
Project Facts
GreenSight is a satellite-enabled urban heat monitoring and intervention planning platform built on the Vectorio Stream Processing Framework. It transforms free Copernicus Sentinel-2 (10m multispectral) and Sentinel-3 (thermal) satellite data into decision-ready heat stress maps, intervention impact estimates, and forensic evidence packs for European municipalities.
| Proposal Acronym | VectorSight |
| Full title | VectorSight: Satellite-Enabled Urban Heat Monitoring for Municipal Climate Action |
| Programm | FIERCE GreenSight: Fostering the Innovative potential of Earth observation for Responsible Climate action in Europe |
| Call | FIERCE Open Call for Scale-Ups |
| Funding Authority | European Union Copernicus Programme (downstream application) |
| Duration | 6 Months |
| Date | March 2026 |
The reusable platform enables municipal climate officers to:
- Monitor urban heat islands at 10-metre resolution across entire city areas
- Compare heat stress before and after interventions (tree planting, cool roofs, water features)
- Model intervention scenarios before implementation
- Produce cryptographically verifiable evidence packs suitable for regulatory reporting (EU Taxonomy, CSRD)
GreenSight is configuration-driven: the same pipeline architecture serves additional environmental domains (drought monitoring, fire/burn risk, flood mapping, crop health) through profile changes with no code modifications required.
Technical Approach
| Component | Description |
|---|---|
| Data Sourced | Copernicus Sentinel-2 (10m, 13 bands, 5-day revisit), Sentinel-3 SLSTR (1km thermal) |
| Processing | Vectorio pipeline: ingest, quality masking, spectral indices (NDVI, Albedo, LST), temporal compositing, spatial fusion, scenario modelling, map tile generation |
| Output | Heat stress maps at 10m resolution, intervention impact estimates, forensic evidence packs with cryptographic integrity proofs |
| Platform | Cross-platform (macOS, Windows, iPad), offline-first, .NET MAUI + Blazor |
| Performance | One Sentinel-2 scene (800 MB, 110×110 km) processed in ~2 minutes on Apple Silicon |