Climate & weather

AI Weather Prediction (AIWP) Model Reforecasts

This is an archive of pure AI-based weather prediction reforecasts produced collaboratively between the Cooperative Institute for Research in the Atmosphere (CIRA) and the NOAA Global Systems Laboratory (NOAA-GSL). Currently, FourCastNetv2-small, Pangu-Weather, and GraphCast are included, with more models to come. Each of these models has been initialized with both NOAA GFS (directories with no extension) and ECMWF IFS initial conditions (directories ending in "IFS"). The datasets are updated with near-real-time data twice per day (00Z and 12Z initializations). FourCastNetv2-small and Pangu-Weather are available from 10/2020 to...

Listed for discovery. Access is provided by the original publisher under its terms.

Data coverage

This is an archive of pure AI-based weather prediction reforecasts produced collaboratively between the Cooperative Institute for Research in the Atmosphere (CIRA) and the NOAA Global Systems Laboratory (NOAA-GSL). Currently, FourCastNetv2-small, Pangu-Weather, and GraphCast are included, with more models to come. Each of these models has been initialized with both NOAA GFS (directories with no extension) and ECMWF IFS initial conditions (directories ending in "IFS"). The datasets are updated with near-real-time data twice per day (00Z and 12Z initializations). FourCastNetv2-small and Pangu-Weather are available from 10/2020 to present, while GraphCast is available from 01/2022 to present. These periods are selected to avoid overlap with training and fine-tuning periods for each model. Data may be missing and is not guaranteed to be available at any given time. If there is data missing that you require, please email [email protected]. For real-time visualizations, visit our visualizations page. The directory structure is as follows\: s3://noaa-oar-mlwp-data/MMMMvNNNIII/YYYY/mmdd/MMMMvNNNIIIYYYYmmddhhfXXXfYYYZZ.nc; where MMMM = Four character model code (FOUR, PANG, or GRAP) NNN = Three digit version number (currently 100 or 200) YYYY = Four digit year mm = Two digit month dd = Two digit day III = Three character initial condition system (GFS) hh = Two digit initialization time (00, 06, 12, or 18) XXX = Three digit first forecast hour of file (currently 000 for all files) YYY = Three digit last forecast hour of file (currently 240 for all files) ZZ = Two digit forecast step (currently 06 for all files) Variables available by model\: - FourCastNet v2-small: - 10-m u wind component - 10-m v wind component - 100-m u wind component - 100-m v wind component - 2-m temperature - Surface pressure - Mean sea level pressure - Total column integrated water vapor - Geopotential at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - Temperature at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - U wind component at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - V wind component at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - Relative humidity at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - Pangu-Weather: - 10-m u wind component - 10-m v wind component - 2-m temperature - Mean sea level pressure - Geopotential at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - Temperature at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - U wind component at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - V wind component at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - Specific humidity at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - GraphCast Operational: - 10-m u wind component - 10-m v wind component - 2-m temperature - Mean sea level pressure - Geopotential at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - Temperature at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - U wind component at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - V wind component at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - Specific humidity at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - Vertical velocity at: [1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 50] hPa - 6-hr accumulated precipitation

Access & permitted use

Review the publisher’s documentation for availability, eligibility and permitted uses, including commercial analysis, model training and redistribution.

Open Data. There are no restrictions on the use of this data.

Need help sourcing data?

Tell us your required fields, population, geography and time period. We can review suitable sources and licensing options.

Request sourcing help

Related data sources

Climate and weatherPublic download

NOAA Global Daily Climate Observations

Daily precipitation, temperature, snowfall and other surface observations from the Global Historical Climatology Network across international stations.

NOAA National Centers for Environmental Information
Climate & weatherExternal source; review access terms

(EXPERIMENTAL) NOAA FourCastNet Global Forecast System (FourCastNetGFS) (EXPERIMENTAL)

The FourCastNet Global Forecast System (FourCastNetGFS) is an experimental system set up by the National Centers for Environmental Prediction (NCEP) to produce medium range global forecasts. The model runs on a 0.25 degree latitude-longitude grid (about 28 km) and 13 pressure levels. The model produces forecasts 4 times a day at 00Z, 06Z, 12Z and 18Z cycles. Major atmospheric and surface fields including temperature, wind components, geopotential height, relative humidity and 2 meter temperature and 10 meter winds are available. The products are 6 hourly forecasts up to 10 days. The data format is GRIB2. The FourCastNetGFS...

NOAA
Geospatial dataExternal source; review access terms

10m Annual Land Use Land Cover (9-class)

This dataset, produced by Impact Observatory, Microsoft, and Esri, displays a global map of land use and land cover (LULC) derived from ESA Sentinel-2 imagery at 10 meter resolution for the years 2017 - 2024. Each map is a composite of LULC predictions for 9 classes throughout the year in order to generate a representative snapshot of each year. This dataset was generated by Impact Observatory, which used billions of human-labeled pixels (curated by the National Geographic Society) to train a deep learning model for land classification. Each global map was produced by applying this model to the Sentinel-2 annual scene...

Impact Observatory