Dataset Viewer
Auto-converted to Parquet Duplicate
text
stringlengths
9
13
sample_0001
sample_0002
sample_0003
sample_0004
sample_0005
sample_0006
sample_0007
sample_0009
sample_0010
sample_0011
sample_0012
sample_0013
sample_0016
sample_0018
sample_0020
sample_0021
sample_0022
sample_0023
sample_0024
sample_0025
sample_0026
sample_0027
sample_0028
sample_0029
sample_0030
sample_0032
sample_0034
sample_0037
sample_0038
sample_0040
sample_0042
sample_0044
sample_0045
sample_0046
sample_0048
sample_0049
sample_0051
sample_0052
sample_0053
sample_0054
sample_0055
sample_0056
sample_0057
sample_0058
sample_0060
sample_0062
sample_0064
sample_0065
sample_0066
sample_0067
sample_0069
sample_0071
sample_0072
sample_0074
sample_0075
sample_0076
sample_0077
sample_0078
sample_0079
sample_0080
sample_0081
sample_0082
sample_0084
sample_0086
sample_0087
sample_0088
sample_0089
sample_0090
sample_0091
sample_0093
sample_0094
sample_0096
sample_0097
sample_0098
sample_0100
sample_0104
sample_0105
sample_0106
sample_0108
sample_0109
sample_0110
sample_0111
sample_0112
sample_0115
sample_0116
sample_0117
sample_0118
sample_0120
sample_0121
sample_0122
sample_0123
sample_0124
sample_0125
sample_0126
sample_0128
sample_0129
sample_0130
sample_0131
sample_0132
sample_0133
End of preview. Expand in Data Studio

WaterBench: Evaluating Geospatial Foundation Models for Coastal and Marine Tasks

Overview

WaterBench is a benchmark and evaluation protocol for assessing GFMs on high-resolution (10m) radar and optical satellite imagery (Sentinel-1, Sentinel-2) across two downstream task families: image-level regression and classification (e.g., water quality, bathymetry, oil-slick detection) and pixel-level segmentation (e.g., mangroves, seagrass). We also include 300m Sentinel-3 ocean-color imagery as contextual information. WaterBench spans multiple coastal regions and seasons, defining fixed splits for in-distribution (ID) evaluation and held-out tests that are spatial (new geographies) and, where applicable, temporal (new years).

Total size: ~175 GB (44 tar archives across 6 tasks)

Tasks

Task Resolution Modalities Target / Classes Metrics Train Val Test Spatial Temporal
OceanState 10 km S1, S2, S3 Continuous (SST, SSS, SSH, CHL, O2) RMSE, RΒ² 1,400 200 400 200 200
GEBCO 500 m S1, S2 Continuous (depth) RMSE, RΒ² 5,000 800 1,600 800 800
Oil Slick 10 m S1, S2, S3 Binary (oil/non-oil) Acc, F1 900 150 300 150 ---
ACA 10 m S1, S2 Multi-class (benthic: 7, geomorphic: 12) Acc, mIoU 8,211 2,719 2,658 879 ---
Seagrass 10 m S1, S2 Binary (presence/absence) Acc, mIoU 832 114 222 124 ---
GMW 10 m S1, S2 Binary (mangrove extent) Acc, mIoU 11,012 1,835 3,670 1,820 ---

Modalities

Modality examples

Task Descriptions

Image-Level Tasks

  • OceanState (global, 2017-2023): Multi-target regression predicting five ocean variables (SST, SSS, SSH, CHL, O2) from Sentinel-1/2/3 imagery. Labels from Copernicus Marine Service (CMEMS). Spatial OOD: Polar biome. Temporal OOD: 2022-2023.

  • GEBCO (global, 2025 grid): Depth regression from satellite imagery in shallow coastal waters. Labels from GEBCO 2025 global bathymetry grid. Spatial OOD: Mediterranean Sea. Temporal OOD: 2022-2023.

  • Oil Slick (global, 2017-2023): Binary classification of oil slicks vs. lookalikes. Labels from GlobalOSD-SAR dataset. Spatial OOD: Mediterranean Sea.

Pixel-Level Tasks

  • ACA (global, 2018-2020): Coral reef segmentation into geomorphic zones and benthic classes following Allen Coral Atlas taxonomy. Spatial OOD: Great Barrier Reef + Bermuda.

ACA targets

  • Seagrass (Maldives, 2021): Binary segmentation of seagrass meadows in shallow lagoons. Spatial OOD: Northern/Southern atolls.

Seagrass targets

  • GMW (global, 1996-2020): Binary segmentation of mangrove extent along tropical/subtropical coasts from Global Mangrove Watch. Spatial OOD: Sundarbans + Everglades + Niger Delta.

GMW targets

Data Format

All images are stored as GeoTIFF files with LZW compression, packaged in uncompressed .tar archives.

Sentinel-2 (images_s2)

  • 12 bands: B02, B03, B04, B05, B06, B07, B08, B8A, B11, B12, B01, B09
  • 224x224 pixels at 10m resolution (resampled from native resolutions)
  • Surface reflectance values scaled to [0, 1]

Sentinel-1 (images_s1)

  • 2 bands: VV, VH (backscatter)
  • 224x224 pixels at 10m resolution

Sentinel-3 (images_s3) β€” OceanState, Oil Slick only

  • 16 OLCI bands (Oa01-Oa12, Oa16-Oa18, Oa21)
  • 120x120 pixels at 300m resolution

Labels

  • OceanState/GEBCO: Continuous values in metadata.csv columns
  • Oil Slick: Binary label in metadata.csv
  • ACA: Per-pixel GeoTIFF masks (benthic: 7 classes, geomorphic: 12 classes)
  • Seagrass/GMW: Per-pixel binary GeoTIFF masks

Repository Structure

data/
β”œβ”€β”€ OceanState/
β”‚   β”œβ”€β”€ metadata.csv
β”‚   β”œβ”€β”€ metadata.json
β”‚   β”œβ”€β”€ splits/{random,geographic,temporal}/{train,val,test}.txt
β”‚   β”œβ”€β”€ OceanState-images_s2.tar
β”‚   β”œβ”€β”€ OceanState-images_s1.tar
β”‚   └── OceanState-images_s3.tar
β”œβ”€β”€ GEBCO/
β”‚   β”œβ”€β”€ metadata.csv, metadata.json, splits/
β”‚   β”œβ”€β”€ GEBCO-images_s2-{00,01}.tar
β”‚   └── GEBCO-images_s1.tar
β”œβ”€β”€ OilSlick/
β”‚   β”œβ”€β”€ metadata.csv, metadata.json, splits/
β”‚   β”œβ”€β”€ OilSlick-images_s2-{00,...}.tar
β”‚   β”œβ”€β”€ OilSlick-images_s1-{00,01}.tar
β”‚   └── OilSlick-images_s3.tar
β”œβ”€β”€ ACA/
β”‚   β”œβ”€β”€ metadata.csv, metadata.json, splits/
β”‚   β”œβ”€β”€ ACA-images_s2-{00,...}.tar
β”‚   β”œβ”€β”€ ACA-images_s1.tar
β”‚   β”œβ”€β”€ ACA-labels_benthic.tar
β”‚   └── ACA-labels_geomorphic.tar
β”œβ”€β”€ Seagrass/
β”‚   β”œβ”€β”€ metadata.csv, metadata.json, splits/
β”‚   β”œβ”€β”€ Seagrass-images_s2.tar
β”‚   β”œβ”€β”€ Seagrass-images_s1.tar
β”‚   └── Seagrass-labels.tar
└── GMW/
    β”œβ”€β”€ metadata.csv, metadata.json, splits/
    β”œβ”€β”€ GMW-images_s2-{00,...}.tar
    β”œβ”€β”€ GMW-images_s1-{00,01}.tar
    └── GMW-labels.tar

Download

Full dataset

huggingface-cli download ayushprd/WaterBench --repo-type dataset --local-dir WaterBench

Single task (e.g., ACA)

huggingface-cli download ayushprd/WaterBench --repo-type dataset --include "data/ACA/*" --local-dir WaterBench

Extract archives

cd WaterBench/data/ACA
for f in *.tar; do tar xf "$f"; done

Splits

Each task provides pre-defined train/val/test splits as text files listing sample IDs (one per line).

  • splits/random/ β€” Standard i.i.d. split
  • splits/geographic/ β€” Spatial OOD: held-out geographic regions in test set
  • splits/temporal/ β€” Temporal OOD: held-out years in test set (OceanState, GEBCO only)

Loading splits in Python

from pathlib import Path

task_dir = Path("WaterBench/data/OceanState")
train_ids = task_dir.joinpath("splits/random/train.txt").read_text().strip().split("\n")
val_ids = task_dir.joinpath("splits/random/val.txt").read_text().strip().split("\n")
test_ids = task_dir.joinpath("splits/random/test.txt").read_text().strip().split("\n")

License

This dataset is released under the CC-BY-4.0 license.

Downloads last month
47