NOAA Launches AI-Powered Global Weather Models

NoAA has introduced an innovative range of artificial intelligence (AI)-powered global weather models, significantly enhancing the speed, accuracy, and efficiency of weather forecasts. This advancement marks a transformative step in how meteorologists and the public receive critical weather information.

AI-Powered Global Weather Models by NOAA

Neil Jacobs, Ph.D., the NOAA administrator, stated that this strategic incorporation of AI signifies a paradigm shift in American weather modeling. These new AI models enhance forecast accuracy for both large-scale weather patterns and tropical weather tracks, delivering results more rapidly while minimizing computational costs.

Key Features of the New Models

The latest suite includes three specific applications:

• AIGFS (Artificial Intelligence Global Forecast System): Utilizes AI to produce faster weather forecasts, requiring up to 99.7% less computing power compared to traditional models.
• AIGEFS (Artificial Intelligence Global Ensemble Forecast System): An ensemble model offering a variety of potential forecast outcomes, showing improved performance that extends forecast skill by an additional 18 to 24 hours over the existing GEFS.
• HGEFS (Hybrid-GEFS): Combines AI-based forecasts with NOAA’s classic ensemble model to consistently outperform both standalone systems, a groundbreaking achievement for operational weather centers.

AIGFS Details

The AIGFS system stands out for its efficiency. For example:

• A single 16-day forecast requires only 0.3% of the computing resources utilized by its predecessor, the GFS.
• This translates to faster data availability for forecasters, taking roughly 40 minutes to complete.

However, the initial version of AIGFS demonstrates some limitations in tropical cyclone intensity forecasts, which NOAA aims to enhance in future updates.

AIGEFS Insights

The AIGEFS operates as a 31-member forecast ensemble, offering multiple weather scenarios rather than a single prediction. Key points include:

• Forecast skill comparable to the current operational GEFS.
• Efficiency using only 9% of the resources needed by the GEFS.

Improvements are ongoing to further refine the modeling capabilities.

HGEFS Overview

The HGEFS, a unique innovation, merges the traditional GEFS and the new AIGEFS, creating a robust 62-member ensemble.

• This hybrid system significantly enhances performance by better representing forecast uncertainties.
• Continuous efforts are underway to enhance HGEFS’s hurricane intensity forecasts.

Collaborative Efforts for Forecasting Innovation

This suite of AI-driven models is part of Project EAGLE, a partnership involving NOAA’s National Weather Service and various research labs. Jacobs highlighted the ongoing collaboration with academia and private sectors to advance forecasting technologies.

NOAA also utilized Google DeepMind’s GraphCast model as a foundation, refining it with its own Global Data Assimilation System. This method effectively improved performance, especially with GFS-based initial conditions.

NOAA’s initiative in launching these AI-powered global weather models is poised to revolutionize meteorological practices, promising better protection for lives and properties worldwide through enhanced forecasting capabilities.