UT-GraphCast Hindcast Dataset: A Global AI Forecast Archive from UT Austin for Weather and Climate Applications
This dataset addresses the problem of accessible, long-term weather forecasting data for researchers and practitioners in meteorology and climate science, though it is incremental as it applies an existing model to new data.
The researchers tackled the need for a comprehensive global weather forecast archive by generating the UT GraphCast Hindcast Dataset from 1979 to 2024 using the Google DeepMind GraphCast Operational model, providing daily 15-day deterministic forecasts on a 25 km grid for 45 years with forecasts delivered in under one minute.
The UT GraphCast Hindcast Dataset from 1979 to 2024 is a comprehensive global weather forecast archive generated using the Google DeepMind GraphCast Operational model. Developed by researchers at The University of Texas at Austin under the WCRP umbrella, this dataset provides daily 15 day deterministic forecasts at 00UTC on an approximately 25 km global grid for a 45 year period. GraphCast is a physics informed graph neural network that was trained on ECMWF ERA5 reanalysis. It predicts more than a dozen key atmospheric and surface variables on 37 vertical levels, delivering a full medium range forecast in under one minute on modern hardware.