NOAA will shift critical weather forecasting services to public cloud in modernization push

The National Oceanic and Atmospheric Administration (NOAA) has announced plans to modernize critical weather prediction capabilities as part of a deal with Google Cloud.

Under the terms of the deal, Google Cloud will serve as the primary high-performance computing (HPC) infrastructure provider for NOAA’s Weather and Climate Operational Supercomputing System (WCOSS).

This will see WCOSS move from traditional, on-premises hosting to a fully public cloud approach. The scientific agency said this will deliver a weather modelling system that is “more flexible and easy to update”.

NOAA Administrator Neil Jacobs said the shift to cloud-based HPC will “eliminate traditional bottlenecks” associated with on-prem systems and accelerate future research capabilities.

“Not only will this new approach provide access to the latest chips, it will advance weather forecasting by aligning operational and research compute and provide capabilities for elastic on-demand model development,” he said.

NOAA will begin migrating workloads from services such as the Global Forecast System and Global Ensemble Forecast System to the new cloud environment in early 2027.

The organization said it expects this transition to be completed by December next year.

Powering NOAA activities

Google Cloud H4D VMs will form the “primary computing backbone” for NOAA in its collaboration modernization push, the duo noted.

Powered by fifth-generation AMD Epyc processors, these are designed specifically for compute-intensive workloads such as numerical weather prediction (NWP).

Google Cloud said these will provide low-latency networking and the significant compute power required to handle “massive, tightly-coupled simulations” used in weather prediction.

NOAA touted three key advantages of switching to a cloud-based approach, including:

• Enabling “rapid updates and adaptations” in weather modelling
• Improving resilience and reducing downtime
• “Instantly increasing compute power” when required, such as during tropical storm seasons

“We are charting a course to global leadership in cloud-based numerical weather prediction,” said Ken Graham, director of NOAA’s National Weather Service.

“This is a significant achievement for NOAA; one that will ensure our critical modeling operations can seamlessly adapt to emerging science, and strengthen our mission to protect life and property.”

Long-term goals

The collaboration with Google Cloud builds on a long-running digital transformation strategy at the agency.

In December last year, NOAA announced plans to ramp up AI-powered weather modelling capabilities, which included the launch of the Artificial Intelligence Global Forecast System (AIGFS).

This, the organization said at the time, is a system that uses AI to speed up weather forecasting capabilities. NOAA revealed the cloud-hosted system uses up to 99.7% less computing resources compared to traditional counterparts.

Elsewhere, NOAA announced a series of contracts earlier this year to develop cloud-based technology for NWS forecasters at more than 120 local offices and centers across the United States.

NOAA and Google Cloud have long-standing ties dating back to 2011. The agency used Google DeepMind and Google Research’s WeatherNext model to predict landfall for Hurricane Melissa five days in advance, according to Google Cloud.

FOLLOW US ON SOCIAL MEDIA

Follow ITPro on Google News and add us as a preferred source to keep tabs on all our latest news, analysis, views, and reviews.

You can also follow ITPro on LinkedIn, X, Facebook, and BlueSky.