ABOUT ASTRASENSE

How Satellite AI Detects
Environmental Hazards

ASTRASENSE is an advanced environmental intelligence platform that fuses multi-spectral satellite data with machine learning models to provide early warning of natural hazards including floods, droughts, vegetation loss, and land degradation — anywhere on Earth, in near real-time.

SYSTEM PIPELINE
01

Satellite Data Ingestion

  • ESA Sentinel-2 MSI (10-60m resolution, 5-day revisit)
  • NASA MODIS Terra/Aqua (250-500m, daily coverage)
  • Landsat-8/9 OLI-TIRS (30m, 16-day revisit)
  • VIIRS NPP thermal band data
  • ERA5 atmospheric reanalysis integration
02

Environmental Index Computation

  • NDVI = (NIR - Red) / (NIR + Red) — vegetation greenness
  • NDWI = (Green - NIR) / (Green + NIR) — surface water
  • LST derived from TIR band + emissivity correction
  • EVI = 2.5 × (NIR - Red) / (NIR + 6×Red - 7.5×Blue + 1)
  • SAVI, MNDWI, BSI supplementary indices
03

AI Risk Prediction

  • Multi-index fusion via gradient boosted ensemble models
  • LSTM time-series models for temporal anomaly detection
  • Threshold-based flood inundation probability engine
  • Drought index composite: SPI + PDSI + VCI correlation
  • Geospatial clustering for hotspot delineation
INDEX REFERENCE GUIDE

Spectral indices derived from satellite multispectral bands

NDVI
Normalized Difference Vegetation Index
(NIR − Red) / (NIR + Red)
Range: −1 to +1Healthy: 0.4 – 0.8

Measures live green vegetation. High values indicate dense healthy plants; low or negative values indicate bare soil, water, or unhealthy vegetation.

NDWI
Normalized Difference Water Index
(Green − NIR) / (Green + NIR)
Range: −1 to +1Healthy: −0.3 – 0.0 (land)

Highlights water bodies and moisture. Positive values indicate open water; negative values indicate vegetation or dry land.

LST
Land Surface Temperature
Derived from TIR band + ε correction
Range: −50°C to +70°CHealthy: 15°C – 30°C

Satellite-derived thermal emission from Earth's surface. Extreme values indicate heat stress, drought, or unusual thermal anomalies.

EVI
Enhanced Vegetation Index
2.5 × (NIR − Red) / (NIR + 6·Red − 7.5·Blue + 1)
Range: −0.2 to +1.0Healthy: 0.3 – 0.7

Corrects for atmospheric and canopy background distortions. More sensitive than NDVI for dense tropical forests and high biomass areas.

TECHNOLOGY STACK
Next.js 16React Framework
React LeafletInteractive Maps
OpenCage APIGeocoding
RechartsData Visualization
Framer MotionAnimations
TypeScriptType Safety

Start Analyzing Now

Search any location to receive instant AI-powered environmental risk analysis

Search a LocationView Global Alerts