The National Oceanic and Atmospheric Administration (NOAA) is turning to artificial intelligence (AI), cloud, and high-performance computing technologies to better forecast and disseminate weather and climate information, NOAA Administrator Rick Spinrad said on June 4.
At a hearing before the House Science, Space, and Technology Committee’s environment subcommittee, Spinrad said growing concerns about severe weather are requiring the agency to “optimize some additional capabilities.”
NOAA recently predicted an above-normal 2024 Atlantic hurricane season after recording an unprecedented number – 28 – of billion-dollar weather and climate disasters last year.
In response, the agency is transforming a key National Weather Service computer system to a cloud framework, Spinrad said. NOAA is also rolling out AI applications – on a test basis – that translate weather watches and warnings into multiple languages.
Asked by Rep. Maxwell Frost, D-Fla., if technology allows NOAA to do more without necessarily needing increased funding, Spinrad responded: “Certainly with things like AI. For example, right now if we do a translation of a forecast it can typically take a half hour or 45 minutes for human intervention. But AI allows that to be done in just 3-5 minutes.”
“So at no real additional cost,” he said, “we can improve the service.”
The hearing focused on NOAA’s fiscal year 2025 budget request of $6.6 billion, a $224 million increase over the fiscal year 2024 enacted funding level. Spinrad said the proposed budget, which would increase funding for the agency’s satellite systems, would “help NOAA better forecast weather events, issue accurate warnings, and observe environmental phenomena connected to climate change-related impacts and patterns.”
The use of AI and cloud continues a focus on technology that has been building since NOAA in 2020 released a strategy to coordinate efforts to use AI across the agency.
AI use cases at NOAA, officials have said, include using the technology to identify fish species from satellite imagery and predicting precipitation sizes, such as hail size.
At a House hearing in March, former acting NOAA administrator Neil Evans said the agency is looking to AI and cloud to improve weather forecasting. “Cloud is the answer,” he said, to address the growing volume of new weather data.
At the June 4 budget hearing, Spinrad said the weather service – which is part of NOAA – is planning to transform the Advanced Weather Interactive Processing System (AWIPS) to a cloud framework. AWIPS is an interactive computer system that integrates meteorological and hydrological data with satellite and radar imagery, helping forecasters prepare and issue more accurate and timely forecasts and warnings.
Spinrad said the move will give forecasters “secure, remote access to provide in-person, impact-based decision support services to decision makers any time, anywhere.”
He called the change among a series of investments that are “absolutely necessary,” given the increase in severe weather and prediction for “another very busy hurricane season” this year.
NOAA is also investing in high-performance computing, along with AI applications, to improve weather modeling, Spinrad said. The subcommittee chairman who led the hearing, Rep. Max Miller, R-Ohio, sponsored a bill to require NOAA and the Department of Energy to use high performance computing and other technologies to conduct research focused on advancing weather and climate predictions.
The legislation passed the House in May 2023 and is awaiting action in the Senate.