Lapwing

Sentinel-2 satellite imagery downloader and analyst for agricultural field boundaries. Give it a shapefile, get back clipped GeoTIFFs — RGB true-colour, NDVI, NDWI, False Color, and elevation (DEM).

A companion to Perch — Perch lands your drone imagery, Lapwing watches from above.

---

Features

• Downloads all Sentinel-2 L2A scenes for a field boundary (all time, cloud-filtered)
• Clips each image exactly to your field boundary
• Produces GeoTIFFs ready to open in QGIS
• Output types:
  • RGB — true-colour (TCI)
  • NDVI — vegetation health index (colourized + raw float32)
  • NDWI — water index
  • False Color — NIR composite (healthy vegetation = bright red)
  • DEM — elevation + slope maps (cascades: NRCan 1m → OpenTopography LiDAR → Copernicus GLO-30)
• Historical NDVI average — composites all years within a seasonal window
• GUI (gui.py) and CLI (finder.py) interfaces
• Parallel downloads; optional S3 direct access for speed

Supported input formats

• Zipped shapefile (.zip)
• Loose shapefile (.shp + siblings)
• GeoPackage (.gpkg)

Requirements

• Python 3.10+
• Free Copernicus Data Space account
• (Optional) OpenTopography API key for higher-res DEMs

Installation

pip install -r requirements.txt

Configuration

Copy .env.example to .env and fill in your credentials:

CDSE_USERNAME=your@email.com
CDSE_PASSWORD=yourpassword

# Optional — faster S3 downloads
CDSE_S3_KEY=
CDSE_S3_SECRET=

# Optional — higher resolution DEMs
OPENTOPO_KEY=

S3 credentials are generated at dataspace.copernicus.eu → Account → S3 credentials.

Usage

GUI

python gui.py

CLI

python finder.py field.zip
python finder.py field.zip --cloud-cover 5 --output my_images --workers 4

All outputs are GeoTIFF (.tif) — open directly in QGIS.