Tomorrow.io says its Sky 1 model gives Microsoft and Google weather AI a better start
Tomorrow.io says its Sky 1 model, fed with observations from its own constellation of 15 microwave sounder satellites, gives better starting conditions to two leading AI weather models: Microsoft's Aurora 1.5 and Google DeepMind's WeatherNext 2.
AI weather models produce global forecasts in minutes but still need an estimate of the current atmosphere to start from, and many rely on systems that update that picture every six hours. Sky 1 uses machine learning to update it hourly, the company says.
In 28 test cases, adding the satellite observations cut initial water-vapor error by about 5 to 8% compared with Sky 1 without them. Across 112 Aurora 1.5 forecasts, the observations lowered average near-surface temperature error at weather stations at every lead time out to five days, with modest gains that were strongest in the first two days. WeatherNext 2 showed a similar pattern.
Started from Sky 1 instead of raw GFS data, Aurora 1.5 gained about a day of lead time on near-surface temperature. Tomorrow.io notes that this result comes from Sky 1 itself, not from its satellites. The study is the company's own and was published on the Tomorrow.io blog.
We hope you enjoyed this article
Consider subscribing to one of our newsletters like Daily AI Brief.
Also, consider following us on social media:
Daily AI Brief
Daily report covering major AI developments and industry news, with both top stories and complete market updates
Market report
2025 Generative AI in Professional Services Report
Thomson Reuters
This report by Thomson Reuters explores the integration and impact of generative AI technologies, such as ChatGPT and Microsoft Copilot, within the professional services sector. It highlights the growing adoption of GenAI tools across industries like legal, tax, accounting, and government, and discusses the challenges and opportunities these technologies present. The report also examines professionals' perceptions of GenAI and the need for strategic integration to maximize its value.
Read moreYou may also like
Microsoft and Terradot adapt Aurora weather model to speed up rock-based carbon removal
Colorado State Develops AI System for Earlier Hail Forecasts
Envision Tianji Tops ECMWF AI Weather Quest
Microsoft releases seven MAI models at Ignite, led by MAI Code 2, and opens licensed weights
Ai2 releases OlmoEarth v1.3 for satellite mapping
Daily AI Brief: the AI news that matters, in your inbox.