Weather intelligence company Tomorrow.io has announced DeepSky, which it describes as the world’s first AI-native, space-based atmospheric and oceanic sensing constellation, aimed at significantly increasing global observation density to support next-generation weather forecasting.
The company said DeepSky is designed to make the Earth’s atmosphere and oceans continuously observable in near real time, enabling faster and more accurate decision-making. The announcement comes one week after Tomorrow.io completed full deployment of its first satellite constellation, which is equipped with Ka-band radar and microwave sounders.
As artificial intelligence plays an expanding role in weather forecasting, Tomorrow.io said forecast performance is increasingly limited by the availability and frequency of observations rather than computing power or algorithms. According to the company, modern AI-based models require dense, high-frequency and diverse data inputs that existing satellite infrastructure cannot consistently provide.
DeepSky is intended to address this gap. The system is a proliferated low Earth orbit (pLEO) constellation made up of satellites carrying multiple proprietary sensing instruments. Operating at scale, the constellation is designed to deliver higher revisit rates than traditional satellite systems, supporting faster refresh cycles for global and regional forecast models and improved prediction of rapidly evolving and extreme weather events.
Tomorrow.io said DeepSky is designed to complement existing government geostationary and low Earth orbit assets, extending their value through higher observation cadence and additional sensing modalities rather than replacing them.
Each satellite in the constellation will feature multimodal sensing across a wide portion of the electromagnetic spectrum for atmospheric and ocean observation. This approach will significantly increase temporal density while enabling new classes of sensors that have previously been limited to bespoke scientific missions due to cost or revisit constraints.
“Operational resilience now depends on treating atmospheric data with the same rigor as any other mission-critical infrastructure,” said Nikhil Ahuja, senior director, planning and supply chain at Amazon. “The advancement in sensing and rapid refresh frequency DeepSky enables creates a new class of AI-driven decision systems that are more adaptive and localized.”
DeepSky is intended to support a range of operational users, including civilian meteorological agencies, severe weather and hurricane forecasting centers, defense and national security organizations and international partners.
Matt Garland, CTO at BNSF, said, “Modern supply chains can no longer rely on static planning or historical averages. True resilience comes from continuously sensing operating conditions and translating that intelligence into network-wide decisions.”
Development is underway, with deployment planned in phases and further details to be shared as the program progresses.
In related news, Google launches its most advanced AI forecasting model – WeatherNext 2