# Satlas Marine Infrastructure: Monthly Offshore Wind Turbine and Oil and Gas Platform Detections from Sentinel-2 (Ai2)

## Facts

- Provider: Allen Institute for AI (Ai2)
- License: ODC-By-1.0 (https://github.com/allenai/satlas/blob/main/GeospatialDataProducts.md)
- Stored at: Google Cloud Storage, gs://satlas-explorer-public/outputs/marine (us)
- Access: public, readable without an account
- Category: Earth Observation & Geospatial
- Source: https://github.com/allenai/satlas
- Page: https://scigantic.com/archive/3a5961e0-07fe-5077-a207-ce1a63ea9deb
- Keywords: Earth Observation & Geospatial, satellite imagery, Google Cloud Storage

## About

Satlas Marine Infrastructure is a global, monthly record of offshore wind turbines and offshore platforms (oil and gas rigs and similar fixed structures at sea) detected by Ai2's deep learning models in Sentinel-2 satellite imagery at 10 m per pixel. Each detection is a point with a category attribute, offshore_wind_turbine or offshore_platform. The outputs/marine/ folder holds one snapshot per month starting in January 2016, each in three formats (YYYY-MM.geojson, YYYY-MM.kml and YYYY-MM.shp), plus latest.* files for the most recent month, an index.txt listing every file with its hash and modification time, and marine.geojson, a single file in which each structure carries start and end attributes giving when it was built or removed. Snapshots are refreshed monthly, and earlier months can be revised when temporal smoothing or improved models change the estimates. Use it to track offshore wind farm build-out, count oil and gas platforms by region, or as labels for marine object detection. Read with GeoPandas, GDAL/OGR or any GIS. License: ODC-By (Open Data Commons Attribution). Cite Satlas, Allen Institute for AI (satlas.allen.ai).

## Load the data

```python
import gcsfs
fs = gcsfs.GCSFileSystem(token='anon')
fs.ls('satlas-explorer-public/outputs/marine/', detail=False)[:10]
```

Lists the files with anonymous access. Runs on any machine, with no cloud account.

## Data structure

Example file: `outputs/marine/2016-01.geojson`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/3a5961e0-07fe-5077-a207-ce1a63ea9deb
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/3a5961e0-07fe-5077-a207-ce1a63ea9deb
- Over MCP: `get_schema_card` with id `3a5961e0-07fe-5077-a207-ce1a63ea9deb` at https://api.scigantic.com/mcp

## Related topics

- [Earth Observation & Geospatial datasets](https://scigantic.com/topics/earth-observation-and-geospatial)
- [Satellite imagery datasets](https://scigantic.com/topics/satellite-imagery)
