πŸ›°οΈ Sentinel-2 Β· Remote Sensing Β· AI

Monitor bird habitats
from space.

A geospatial intelligence platform that turns Sentinel-2 multispectral imagery into actionable ecology: quantify vegetation health, detect water and urbanization, and estimate bird-population-decline risk for any location on Earth.

πŸš€ Launch Dashboard 🌍 Explore examples
0
Spectral bands (B02–B11)
0
Indices Β· NDVI Β· NDWI Β· NDBI
0
Habitat score range 0–100
🌍
Global coverage

Everything you need to assess a habitat

Click any point on the map and get a full ecological readout in seconds.

πŸ—ΊοΈ

Interactive Map Analysis

Click anywhere on a high-resolution satellite basemap to fetch and analyze live Sentinel-2 imagery for that location.

🌿

NDVI Β· NDWI Β· NDBI

Compute vegetation, water and built-up indices from multispectral bands, rendered as clear color heatmaps.

πŸ¦…

Habitat & Risk Scoring

Convert spectral signals into a Habitat Health Score (0–100) and a Bird Population Decline Risk level.

πŸ“ˆ

NDVI Time-Series

Track vegetation health across months to spot seasonal cycles and long-term habitat degradation.

🌦️

Live Weather

Pull real-time temperature, humidity, wind and cloud cover to contextualize habitat conditions.

πŸ“„

PDF Reports

Generate a documented habitat analysis report for any location, ready to share or archive.

How it works

A fully serverless pipeline β€” no persistent servers, deploys anywhere.

1

Pick a location

Click a point on the interactive Mapbox satellite map.

2

Fetch Sentinel-2

A serverless function pulls a low-cloud multispectral tile (B02–B11).

3

Compute indices

NDVI, NDWI and NDBI are derived and rendered as color composites.

4

Score & classify

Habitat health, decline risk and land-cover are derived from the indices.

5

Report

View metrics, weather and heatmaps β€” then export a PDF.

Read the full methodology β†’

Technology

Built to run entirely on serverless infrastructure.

Sentinel-2 MSI Mapbox GL JS Python Serverless NumPy Matplotlib tifffile Chart.js OpenWeather Vercel
Land-cover is derived deterministically from spectral indices β€” grounded in remote-sensing conventions rather than an unverified model.

Ready to explore a habitat?

Open the dashboard and analyze any location on Earth in seconds.