Over the course of a calendar year ending in May 2025,
the United States absorbed nearly $1 trillion in damages due to extreme weather. This amount, representing 3% of U.S. gross domestic product, was driven by rising insurance costs and a series of disasters primarily concentrated in the Ten Across geography, such as Hurricanes Helene and Milton and the fires in Los Angeles.
More than ever before, timely and detailed forecasts are needed to properly prepare—and in some cases to evacuate—communities ahead of such extreme events. Leaders across sectors are further in need of advanced weather modeling to support larger-scale mitigation and adaptation efforts.
The data that influence such public and private decision-making mainly stem from the National Weather Service’s
six billion daily weather observations. The NWS recently
shed 600 of its 4,000 positions, prompting
a public warning from five former agency directors that understaffing could undermine the quality and delivery of forecasts, potentially putting many Americans at greater risk.
At the same time, advanced artificial intelligence capabilities are contributing to a trend toward increased commercial ownership of U.S. weather forecasting. However, today's guest, Dr. Amy McGovern, points out that while today's AI can create and curate efficient weather models better than a conventional supercomputer, its monitoring capabilities are not comparable to the collective experience and proficiency of NWS scientists.
Listen in as Ten Across founder Duke Reiter and Dr. McGovern, an expert in the integration of AI in meteorological science, explore the current forecasting landscape and how the emergence of private sector AI-powered modeling is influencing its evolution.
Related articles and resources: Read about
Brightband’s Extreme Weather Bench, led by Amy McGovern
NOAA stops tracking cost of extreme weather and climate disasters (
UtilityDive, May 2025)
Most Americans use federal science information on a weekly basis, a new poll finds (
NPR, May 2025)
Former Weath