17 datasets
Earth science data spans satellites, oceans, the solid earth and water. NASA's EOSDIS is the main gateway for Earth observation products, and 38-Cloud provides Landsat 8 scenes with pixel-level cloud masks. Ocean data comes from IMOS (roughly 30TB of measurements), BODC, the Alabama Real-Time Coastal Observing System and NOAA's estuarine monitoring program. Geohazard records include the USGS earthquake archives and the Smithsonian Global Volcano and Eruption Database.
Water and resources are covered by FAO's AQUASTAT, the Global Wind Atlas, the UK Oil and Gas Authority's well, seismic survey and pipeline data, and a US water hardness table. NASA Earthdata downloads need a free Earthdata login. Satellite and ocean products are often HDF5, NetCDF or GeoTIFF with their own processing levels, so read the product guide to see whether values are raw, calibrated or gridded before you combine them.
38-Cloud (Cloud Detection)github.com
170Contains 38 Landsat 8 scene images and their manually extracted pixel-level ground truths for cloud detection. They are cropped into multiple 384*384 patches. It has 8400 patches for training and 9201 patches for testing.
Global water resources and uses
EOSDIS - NASA's Earth Observation System Data and Information Systemearthdata.nasa.gov
NASA's Earth Observation System Data and Information System
Earth Modelsearthmodels.org
Global Wind Atlasglobalwindatlas.info
The Global Wind Atlas is a free, web-based application developed to help policymakers, planners, and investors identify high-wind areas for wind power generation virtually anywhere in the world, and then perform preliminary calculations. The Global Wind Atlas facilitates online queries and provides…
Maintained public index of ~400 researched US hot springs (names, GPS coordinates, reported temperatures, access type). Picks up where NOAA's national thermal springs dataset left off after its 2025 decommissioning, with per-record status curation. JSON, free to use.
roughly 30TB of ocean measurements
Alabama Real-Time Coastal Observing SystemDauphin Island Sea Lab
National Estuarine Research Reserves System-Wide Monitoring ProgramNOAA National Estuarine Research Reserve System
long-term estuarine monitoring data
Oil and Gas Authority Open DataOil and Gas Authority, UK
The dataset covers 12,500 offshore wellbores, 5,000 seismic surveys, and 3,000 pipelines over more than five decades.
OpenWaterAtlas Spot-Climate-Species-Routes DatasetOpenWaterAtlas
2,928 named dive, kite, surf and freedive sites joined to 112,548 OBIS + GBIF marine species occurrences within 5 km of each site (with a family-level terrestrial-noise filter), per-spot 5-year daily climate aggregates (Open-Meteo + ERA5), and 34,876 OpenFlights direct routes. One citable bundle…
Radiance GeoJSON — Global Light PollutionRadiance GeoJSON (project repository)
Global nighttime light pollution dataset derived from VIIRS Nighttime Lights imagery, offering 0.5 km resolution radiance values, MPSAS sky brightness, and Bortle scale classifications for 173 countries. Available as compressed GeoJSON files (.json.gz) and a PostgreSQL/PostGIS database dump (3.2GB…
US Water Hardness DatasetTapWaterData
Measured drinking-water hardness for 16,561 US cities (mg/L as calcium carbonate and grains per gallon), each value carrying its source, tier, and sample date. Built from public primary sources - utility Consumer Confidence Reports, the EPA/USGS Water Quality Portal, and USGS classification bands…
USGS Earthquake Archivesearthquake.usgs.gov
This dataset contains 4148 simulation results, i.e., 4148 pairs of predictor/target. bkt.npy contains the breakthrough curves from all 6 injection wells recorded at the pumping well. pz.npy contains the 2D coordinates of the backtracked particles' end points, used to delineate the WHPA.