Nvidia (NASDAQ:NVDA) on Monday introduced a new suite of open-source models and tools for AI-driven weather prediction at the American Meteorological Society’s Annual Meeting.
Branded as Earth-2, the platform is described as the world’s first fully open, accelerated software stack for weather and climate forecasting, designed to broaden access for researchers, startups, developers, enterprises and public-sector agencies worldwide.
Earth-2 is built to speed up the entire forecasting workflow — from ingesting raw observational data to producing 15-day global outlooks or short-term local storm forecasts. By relying on AI rather than traditional physics-based simulations run on supercomputers, the system can significantly reduce both computing time and costs.
The newly launched models include Earth-2 Medium Range, powered by the Atlas architecture; Earth-2 Nowcasting, built on StormScope; and Earth-2 Global Data Assimilation, which uses HealDA. These additions expand an existing Earth-2 lineup that already features models such as Earth-2 CorrDiff and Earth-2 FourCastNet3.
A growing number of organizations are already deploying the technology. AI weather firm Brightband is using Earth-2 Medium Range to generate daily global forecasts, while the Israel Meteorological Service has adopted Earth-2 CorrDiff in operational settings, reporting a roughly 90% cut in computing requirements versus conventional approaches.
Energy groups including TotalEnergies, Eni, GCL and Southwest Power Pool are assessing or implementing Earth-2 models to enhance forecasting for operational planning. Financial institutions such as S&P Global Energy and AXA are applying the technology to risk analysis and scenario modeling.
Earth-2 Medium Range and Earth-2 Nowcasting are currently available through NVIDIA Earth2Studio, Hugging Face and GitHub, while Earth-2 Global Data Assimilation is expected to be released later this year.