⚙️ How It Works

Our system transforms complex satellite datasets into actionable, location-specific insights using geospatial analytics and AI-powered natural language generation.

🟢 Green Access Analysis

Objective:
Detect vegetation-deficient neighborhoods and recommend new micro-parks.
Pipeline:
  • 1
    Retrieve Sentinel-2 surface reflectance (Bands 8 & 4).
  • 2
    Compute
    NDVI = (NIR – Red) / (NIR + Red)
  • 3
    Classify NDVI into low (< 0.2), moderate (0.2–0.5), and high (> 0.5) greenness zones.
  • 4
    Overlay with OpenStreetMap (OSM) data for roads, parks, and accessibility.
  • 5
    Identify underserved zones beyond a 10-minute walking radius from existing parks using OSMnx isochrones.
  • 6
    Feed contextual data (population, soil, terrain) into the AI engine for micro-park recommendations.
Output:
Interactive NDVI maps, green gap overlays, and AI-generated "Suitability & Recommendation" briefs.

🔴 Urban Heat Island Analysis

Objective:
Detect heat stress zones for cooling interventions.
Pipeline:
  • 1
    Acquire MODIS day/night and ECOSTRESS Land Surface Temperature (LST) data.
  • 2
    Compute z-score anomalies relative to city means.
  • 3
    Apply DBSCAN clustering to identify contiguous heat clusters (> 2σ).
  • 4
    Integrate land cover, NDVI, and population exposure data.
Output:
Severity-coded UHI maps (Elevated / High / Severe) with tables showing population exposure and recommended cooling measures (e.g., green roofs, reflective surfaces, tree corridors).

🟣 Air Quality Analysis

Objective:
Locate air pollution hotspots and vulnerable communities.
Pipeline:
  • 1
    Aggregate Sentinel-5P NO₂ and GEOS-CF PM₂.₅ reanalysis data.
  • 2
    Standardize via z-scores and apply DBSCAN clustering to delineate hotspots.
  • 3
    Enrich with contextual layers (population, schools, hospitals, industrial zones).
Output:
Pollution cluster maps, exposure summaries, and AI briefs linking pollution with land use and policy interventions.

NASA & ESA Data Sources

NASA Data

  • Landsat Program
    Historical reference and fallback for NDVI composites
  • SMAP (Soil Moisture Active Passive)
    Soil moisture context for vegetation and UHI modeling
  • MODIS (Moderate Resolution Imaging Spectroradiometer)
    Land surface temperature and aerosol data
  • ECOSTRESS
    High-resolution thermal data for intra-urban heat mapping
  • SEDAC (Socioeconomic Data and Applications Center)
    Population and exposure context

ESA / Partner Data

  • Copernicus Sentinel-2
    NDVI-based vegetation monitoring
  • Copernicus Sentinel-5P
    Air quality (NO₂, O₃, SO₂) monitoring
  • JRC Global Surface Water
    Water body reference and blue-green corridor detection
  • ISRIC SoilGrids
    Soil data for suitability assessment
  • OpenStreetMap
    Roads, green polygons, land use
  • WorldPop / GHSL / Copernicus DEM
    Population and elevation context