# Google opens two datasets used in WeatherNext 3

> SatWeave provides hourly satellite mosaics, while PARDIG maps precipitation every 15 minutes. Google says both datasets are openly available for research and development.

Oossa · 2026-11-13 · https://oossa.com/en/google-opens-two-datasets-used-in-weathernext-3

Event date: 2026-11-13

Google DeepMind says it is releasing SatWeave and PARDIG, two datasets used in its WeatherNext 3 forecasting model, for researchers and developers. SatWeave combines satellite imagery into a global mosaic across 17 visible, water-vapor and infrared channels, updated hourly. PARDIG estimates global precipitation at roughly 10-kilometer resolution every 15 minutes.

Google says PARDIG improved hourly rain-detection skill by more than 30% over NASA’s IMERG-Final for light-to-moderate precipitation in its evaluations. The datasets’ live feeds and archives are available on Google Cloud Storage; PARDIG’s historical data, covering 1999 to 2026, is also on Google Earth Engine.

## The facts

- PARDIG estimates precipitation at roughly 10-kilometer resolution every 15 minutes, with average latency of about five hours.
- Google says PARDIG’s hourly precipitation detection skill was over 30% higher than IMERG-Final for light-to-moderate rain.

## Why it matters

Researchers studying floods or droughts in places without dense radar coverage can access PARDIG’s global estimates, though its usefulness for particular local applications remains to be assessed.

## Sources & references

1. [Opening the datasets behind our AI weather models](https://deepmind.google/blog/opening-the-datasets-behind-our-ai-weather-models/) – Google DeepMind, 2026-11-10

Last updated: 2026-11-13