GRAIN is an OpenStreetMap (OSM) based dataset of the world's irrigation canals. OSM country-scale data is retrieved from Geofrabrik.de, processed using OSMium CLI Toolkit, and QGIS to filter for waterways and to convert the OSM PBF file to the lightweight cloud optimised GeoParquet format.
A Random Forest classifier, trained on 20,000 samples of rivers and canal geometry, is used to differentiate between man-made canals and natural rivers to correct tagging issues prevalent in OSM data. Canal use type is determined using ESA CCI Land Cover dataset. Validation is done on in-stu canal network data obtained from national and regional scale government datasets, and manually delineated canal vector data.
Dataset is available for download at Zenodo - GRAIN v.1.0
GRAIN is made available as country level files in GeoParquet and ESRI shapefile formats. All GRAIN files are projected to EPSG:4326, based on the WGS-84 datum. Users are recommended to download the GeoParquet files due to the significant smaller file size.
The methodology and validation is described in : Suresh, S., Hossain, F., Mishra, V., Hossain, N., 2025. GRAIN - Global Registry of Agricultural Canal Networks, Earth System Science Data *Submitted for review
Users are encouraged to refer to the GRAIN Documentation website for information on reading and visualizing GRAIN, general methodology, recreating GRAIN data, GRAIN attribute schema etc.
We welcome feedback, bug reports, and suggestions for improvements.
Issue Tracker: Please use the GitHub Issues page to report:
- Bugs or data errors
- Missing canal networks in specific regions with prevalent OSM data
- Feature requests (e.g., new attributes, processing steps)
- Questions about data usage
When reporting a bug or data issue, please include:
- The country/region affected
- The GRAIN version you are using
- A clear description of the issue
- Screenshots, shapefiles, or links to reference data (if available)
GRAIN is designed as a community-driven dataset.
We encourage contributions from researchers, GIS specialists, and citizen scientists.
- Fork the repository and create a new branch for your changes.
- Make your changes:
- Improve code in
/src - Add new validation data by providing a link to the data in a text file in
/assets/validation
- Improve code in
- Test your changes locally to ensure nothing is broken.
- Submit a pull request with:
- A short description of the changes
- Links to any supporting data or references
- Clear notes if the changes affect dataset structure or methodology
- Adding new regional validation datasets.
- Improving the classification pipeline or adding new ML features.
- Creating notebooks that showcase GRAIN applications.
For large contributions (e.g., major processing changes, new metadata fields), please open an issue first so we can discuss and coordinate.
For general inquiries or collaboration requests, please contact:
Sarath Suresh
Graduate Research Assistant, SASWE Research Group
University of Washington
✉️ Email: [email protected]
🔗 LinkedIn