# A global forecast that can refresh with the hour.

Google DeepMind announced WeatherNext 3 with hourly refreshes, higher-resolution precipitation and clean-energy variables, and integration into Google products.

Canonical: https://brightaifuture.com/discoveries/weathernext-3
Format: discovery
Source publication: 2026-09-03
Bright publication: 2026-09-19
Substantive update: None recorded
Evidence and review: Deployed; confidence: medium; approved; ai-assisted. AI-assisted editorial comparison with the cited primary sources, explicit evidence limits, and held alternatives. Publication authorized by the site owner on 2026-09-19; no human source review or independent replication is claimed.

## The human problem

Weather-dependent decisions often happen between traditional forecast cycles and at scales global forecasts do not resolve cleanly.

## The prior constraint

Generating frequent global probabilistic forecasts with useful local variables is computationally demanding.

## AI’s actual role

The model incorporates real-time satellite data to generate frequently refreshed forecasts and derived variables.

## The documented result

Google reports integration into Search, Gemini, Maps, Maps Platform, and Cloud. The announcement does not provide independent operational-skill evidence for every claimed use.

## Why it may matter

More frequent forecasts could help planners revisit a decision as new observations arrive, provided local authorities and uncertainty remain visible.

## Limitations

Performance and deployment claims are vendor-reported.

Product availability is not evidence of equal local accuracy.

It is not a public-warning system and does not replace meteorological agencies.

## Unresolved questions

How does skill vary by region, hazard, and lead time?

What independent benchmark results will be published?

How are forecast changes and uncertainty communicated to users?

## Provenance and history

{
  "dates": {
    "eventDate": "2026-09-03",
    "publicationDate": "2026-09-03",
    "captureDate": "2026-09-19",
    "lastReviewedDate": "2026-09-19"
  },
  "provenance": {
    "origin": "editorial",
    "externalId": "https://blog.google/innovation-and-ai/models-and-research/google-deepmind/introducing-weathernext-3/"
  },
  "revisions": [
    {
      "id": "revision:sept26-wn3-01",
      "recordedAt": "2026-09-19",
      "summary": "Initial vendor-bounded WeatherNext 3 draft.",
      "sourceIds": [
        "source-wn3-google"
      ]
    }
  ],
  "corrections": []
}

## Original sources

- [Introducing WeatherNext 3, our most advanced and accurate global weather AI model](https://blog.google/innovation-and-ai/models-and-research/google-deepmind/introducing-weathernext-3/)

## Continue exploring

- [Planet](https://brightaifuture.com/worlds/planet)
- [How much earlier could we understand what is coming?](https://brightaifuture.com/threads/weather)