ClimateVision

---

What is ClimateVision?

ClimateVision is an open-source machine learning platform that detects environmental change from satellite imagery. It uses deep learning (U-Net, Siamese networks) to monitor deforestation, arctic ice melting, and flooding — giving conservation NGOs and researchers automated alerts without manual analysis. Built on Sentinel-2 and Landsat data via Google Earth Engine, it runs as a REST API with a React dashboard for real-time monitoring.

---

Installation

git clone https://github.com/Climate-Vision/ClimateVision.git
cd ClimateVision
pip install -r requirements.txt

---

Quickstart

Start the API server:

uvicorn climatevision.api.main:app --reload --host 0.0.0.0 --port 8000

Run a deforestation analysis:

curl -X POST http://localhost:8000/api/predict/json \
  -H "Content-Type: application/json" \
  -d '{
    "bbox": [-60.0, -15.0, -45.0, -5.0],
    "start_date": "2023-01-01",
    "end_date": "2023-12-31",
    "analysis_type": "deforestation"
  }'

Launch the dashboard:

cd frontend && npm install && npm run dev
# Visit http://localhost:5173

Explore the API docs: http://localhost:8000/docs

---

Features

• Multi-type climate analysis — deforestation, arctic ice melting, flood detection (drought and wildfire detection planned)
• Deep learning inference — U-Net semantic segmentation and Siamese network change detection on Sentinel-2 imagery
• Automated data pipeline — Google Earth Engine integration with cloud masking, normalization, and 256×256 tiling
• NGO management — register organisations, subscribe to regions, receive threshold-based alerts via email or webhook
• REST API — FastAPI backend with paginated run history, stats endpoint, and full OpenAPI docs
• React dashboard — interactive map with bbox region selector, Recharts analytics, confidence gauges, and run history
• MLflow experiment tracking — log training runs, hyperparameters, and model checkpoints
• ONNX export — optimised model export for fast production inference

---

Documentation

Full documentation: github.com/Climate-Vision/ClimateVision/wiki

• Getting Started
• API Reference — http://localhost:8000/docs when running locally
• Project Structure
• Training Guide
• Colab Notebook

---

Models & Analysis Types

Analysis Type	Status	Classes	Satellite Bands
Deforestation	Active	forest, non-forest	B02, B03, B04, B08
Arctic Ice Melting	Active	sea-ice, open-water, land	B02, B03, B04, B11
Flood Detection	Active	water, flooded, dry-land	B03, B08, B11
Drought Monitoring	Planned	normal, stressed, severe	B04, B08, B11, B12
Wildfire Detection	Planned	unburned, burned, active-fire	B04, B08, B11, B12

Performance benchmarks (baseline U-Net on held-out test sets):

Metric	Value
Forest segmentation IoU	in progress
Change detection F1	in progress
Inference time (256×256 tile)	~45ms on CPU
API response time	<100ms

Benchmarks will be updated as the team completes training runs. See MLflow tracking for experiment history.

---

Contributing

We welcome contributions — bug reports, new analysis types, model improvements, documentation, and translations.

# Fork the repo, then:
git checkout -b feature/your-feature-name
pip install -r requirements.txt
pytest tests/
# Submit your PR against the develop branch

Read the Contributing Guide and Code of Conduct before getting started.

Good first issues are labelled good first issue on GitHub.

---

License & Citation

This project is licensed under the MIT License — see LICENSE for details.

If you use ClimateVision in your research, please cite:

@software{climatevision2026,
  author  = {ClimateVision Contributors},
  title   = {ClimateVision: Open-Source AI Platform for Environmental Monitoring},
  year    = {2026},
  url     = {https://github.com/Climate-Vision/ClimateVision},
  version = {0.2.0}
}

---

⭐ Star us on GitHub  ·  Contribute  ·  Report a Bug