WeatherNext Graph (GraphCast) is a deterministic weather forecasting model developed by Google DeepMind and Google Research. Built on a Graph Neural Network (GNN) architecture implemented as an encoder-processor-decoder system, it represents the Earth's atmosphere as a graph and learns to predict the next state at each node. GraphCast outperformed ECMWF's HRES (the leading operational deterministic forecast) in over 90% of tested variables, showing particular skill in predicting tropical cyclones, atmospheric rivers, and extreme temperatures. It was named a runner-up to Science's Breakthrough of the Year in 2023 and won the MacRobert Award in 2024. For more, see the DeepMind blog post.
At a glance
| Attribute | Details |
|---|---|
| Coverage | Global |
| Spatial Resolution | 0.25° (~30 km at the equator) |
| Temporal Resolution | 6 hours |
| Forecast Horizon | 10 days |
| Initialization Frequency | Every 6 hours (00, 06, 12, 18 UTC) |
| Forecast Type | Deterministic (single forecast) |
| Architecture | Graph Neural Network — encoder-processor-decoder (paper) |
| Training Data | ERA5 / HRES-fc0 |
| Initialization Inputs | HRES-fc0 |
| Historical Forecasts | 2020 – present |
Supported variables
Surface variables
| Variable Name | Description | Units |
|---|---|---|
2m_temperature |
2 metre temperature | K |
10m_u_component_of_wind |
10 metre U (eastward) wind component | m/s |
10m_v_component_of_wind |
10 metre V (northward) wind component | m/s |
mean_sea_level_pressure |
Mean sea level pressure | Pa |
total_precipitation |
Total precipitation (6-hour accumulated) | m |
Atmospheric variables (pressure levels)
These variables are available at 13 pressure levels: 50, 100, 150, 200, 250, 300, 400, 500, 600, 700, 850, 925, 1000 hPa.
| Variable Name | Description | Units |
|---|---|---|
{level}_geopotential |
Geopotential height | m2/s2 |
{level}_specific_humidity |
Specific humidity | kg/kg |
{level}_temperature |
Temperature | K |
{level}_u_component_of_wind |
U (eastward) wind component | m/s |
{level}_v_component_of_wind |
V (northward) wind component | m/s |
{level}_vertical_velocity |
Vertical velocity | Pa/s |
Replace {level} with the pressure level in hPa. For example,
500_geopotential for geopotential at 500 hPa.
See the Glossary for definitions of meteorological terms.
How to access
WeatherNext Graph datasets on BigQuery and Earth Engine were deprecated on July 29, 2026. Users previously on the allowlist for Graph automatically have access to WeatherNext 3 datasets. For deterministic or summary forecasts, use the WeatherNext 3 mean and statistics datasets. See the deprecation page for migration details and WeatherNext 3 for current operational forecasts.