Observing global economic activity with free satellite data.
An open-source pipeline for monitoring economic activity from space using freely available satellite data. Ship detection, nighttime light analysis, port activity tracking, and more — no machine learning, no paid imagery, no GPU required.
Currently focused on Tokyo Bay as the first target, with plans to expand to major ports and economic zones worldwide.
| Source | Resolution | Cost | Use Case |
|---|---|---|---|
| Sentinel-2 (ESA) | 10m | Free | Ship detection, land use change, port monitoring |
| AIS (AISStream.io) | Real-time | Free | Ship tracking, vessel identification, trade flow |
| VIIRS (NASA/NOAA) | 500m | Free | Nighttime light, economic activity proxy |
src/
├── collectors/ # Data acquisition
│ ├── sentinel2.py # Sentinel-2 image download (global)
│ ├── ais.py # AIS ship tracking via WebSocket
│ └── viirs.py # VIIRS nighttime light data
├── analyzers/ # Analysis pipelines
│ ├── detect_ships_sentinel2.py # Ship detection (NDWI + anomaly)
│ ├── analyze_viirs_nightlight.py # Nighttime light analysis
│ └── visualize_sentinel2.py # Satellite image visualization
└── utils/
└── config.py # Configuration & API settings
Detects vessels in Sentinel-2 imagery using statistical anomaly detection. No training data needed — works on any coastal area worldwide.
- Water mask: NDWI + NIR absorption + visible light thresholding
- Coastal filtering: 150m erosion to remove port structures
- Anomaly detection: Multi-band brightness (mean + 3σ), ≥2 of 4 bands
- Shape filtering: Length 40-500m, aspect ratio ≥ 1.3
Results on Tokyo Bay (first validation):
- 2026-03-07: 358 vessels detected (Large: 54, Medium: 109, Small: 195)
- 2026-03-12: 320 vessels detected (Large: 63, Medium: 103, Small: 154)
- Validated against MLIT statistics (estimated 270-560 vessels instantaneous)
git clone https://github.com/satlens/satlens.git
cd satlens
pip install -r requirements.txt
cp .env.example .env
# Edit .env with your API keys| Service | Key | Sign up |
|---|---|---|
| Copernicus Data Space | COPERNICUS_CLIENT_ID / SECRET |
https://dataspace.copernicus.eu/ |
| AISStream.io | AISSTREAM_API_KEY |
https://aisstream.io/ |
| NASA Earthdata | EARTHDATA_USERNAME / PASSWORD |
https://urs.earthdata.nasa.gov/ |
All services offer free API keys.
# Ship detection on Sentinel-2 imagery
python -m src.analyzers.detect_ships_sentinel2
# AIS data collection (5 minutes)
python -m src.collectors.ais --duration 300
# VIIRS nighttime light analysis
python -m src.analyzers.analyze_viirs_nightlight- 海外ファンドは衛星データで何を見ているのか
- Sentinel-2で東京湾の船を数えてみた (coming soon)
MIT
無料の衛星データを使って、世界の経済活動を宇宙から観測するオープンソースパイプラインです。
Sentinel-2（光学衛星画像）、AIS（船舶位置データ）、VIIRS（夜間光データ）を組み合わせ、船舶検知・夜間光解析・港湾活動のモニタリングを行います。機械学習不要、有料画像不要、GPU不要。
現在は東京湾を最初の検証対象として分析を進めており、今後は世界の主要港湾・経済圏に展開予定です。
詳しくは note で実験記録を公開しています。