NOAA Launches AI-Driven Global Weather Models to Cut Costs and Boost Accuracy
NOAA officially deploys a new suite of artificial intelligence weather prediction models that use drastically less computing power while delivering faster and more accurate forecasts. The innovative systems include a hybrid approach that consistently outperforms both traditional and AI-only models.
NOAA officially launches a groundbreaking suite of artificial intelligence-driven global weather prediction models, marking a major leap forward in forecast speed and efficiency. These new systems provide meteorologists with faster delivery of highly accurate weather guidance while using up to 99.7% less computing power than traditional models. NOAA administrators highlight this strategic use of AI as a new paradigm that drastically reduces computational expenses while improving large-scale weather and tropical track accuracy.
The operational suite consists of three distinct applications designed to work together and enhance overall prediction capabilities. The Artificial Intelligence Global Forecast System (AIGFS) serves as the primary AI model that matches traditional systems in output but generates forecasts much faster. Meanwhile, the Artificial Intelligence Global Ensemble Forecast System (AIGEFS) provides a range of probable outcomes and extends forecast skill by an additional 18 to 24 hours compared to older methods.
The most innovative addition is the Hybrid-GEFS (HGEFS), which is a first-of-its-kind operational system that blends the AI-based AIGEFS with NOAA's traditional physics-based ensemble model. Initial testing reveals that this hybrid "grand ensemble" consistently outperforms both the AI-only and physics-only systems. By combining the strengths of both methodologies, NOAA achieves unprecedented levels of forecast reliability for meteorologists and the public.