Derived products and code for compact Prithvi floodwater segmentation on Sen1Floods11

Derived numerical tables, split manifests, a snapshot of the public Python helpers, and a sample of 11 chips (one flood event each) for Accuracy and efficiency of compact foundation models for surface-water mapping during floods (Farajpoor and Narimani, 2026). The sample includes true-colour Sentinel-2 excerpts, visualisations of the Sen1Floods11 hand labels, and study-generated GeoTIFFs (predicted water, probability, error, entropy) for fine-tuned Prithvi-Tiny and U-Net. This is not a re-host of the original Sen1Floods11 GeoTIFF archive; obtain that dataset from gs://sen1floods11 and cite Bonafilia et al. (2020). True-colour panels contain modified Copernicus Sentinel-2 data. Code: github.com/MohammadrezaNarimaniUCDavis/Prithvi_Tiny_Floodwater_GeoAI.