Skip to content
← Directory

software · Global

Prithvi-EO-2.0-300M-TL-Sen1Floods11

Hugging Face model ibm-nasa-geospatial/Prithvi-EO-2.0-300M-TL-Sen1Floods11

Visit source ↗

Claim this listing

Represent Prithvi-EO-2.0-300M-TL-Sen1Floods11? Send a claim for review so we can discuss listing updates. Claims do not automatically grant ownership or verification.

Explore paid sponsorship →

What's in it

Data offered docs-verified
flood extent segmentation (water/flood vs no water) from Sentinel-2 imagery · six-band input: Blue, Green, Red, Narrow NIR, SWIR, SWIR 2
Coverage docs-verified
Global (training data spans 6 continents, 14 biomes, 357 ecoregions)
Formats llm-inferred
GeoTIFF (Sentinel-2 L1C .tif input) · PyTorch model weights
Access docs-verified
Free and open source under Apache-2.0; run locally via the repo inference script or TerraTorch fine-tuning config

Useful for

Use cases llm-inferred
Post-event flood extent mapping from Sentinel-2 imagery to validate modelled flood outlines · Historic flood extent baseline from Sentinel-2 archive for FRA calibration or evidence · Satellite-derived water mask input for hydraulic model calibration in flood studies
Audience llm-inferred
flood consultants · hydrologists · GIS analysts · remote sensing specialists

Limitations

Constraints llm-inferred
Fine-tuned for single-timestamp segmentation; multi-temporal use is not demonstrated in the docs · Training set is 446 labelled 512x512 chips across 11 flood events, so performance on UK conditions is not documented · Input is Sentinel-2 L1C; Sentinel-1 SAR is not supported by this checkpoint · Class -1 (no data/clouds) must be handled in outputs · Not deployed by any Hugging Face Inference Provider; requires a local Python/TerraTorch environment · Pixel resolution of training chips is not stated in the docs
Licence
apache-2.0
Authentication
No API key required (repository README, auto-extracted; check upstream requirements)
GitHub stars
Unknown
Last commit
Unknown
Last successful page check
14/10/2026
Page health
healthy
Package
Not recorded
Verification
Not independently verified

Provenance & evidence

Verified via HF API: exists; 2766 downloads/30d, 12 likes.

Metadata observed 14/10/2026 · research sweep: manual-2026-10-15 (Hugging Face models)