AI boosts hurricane forecasting accuracy, yet human expertise remains crucial for weather prediction

Artificial intelligence is significantly enhancing the precision of weather forecasting, particularly for hurricanes, despite ongoing criticism regarding the massive energy consumption of data centers, which contributes to climate change. The U.S. National Hurricane Center (NHC) began utilizing Google DeepMind's AI models last year and is currently collaborating on next-generation versions. DeepMind's research scientist, Ferran Alet, noted that the AI leverages existing weather models to their limits, asserting that tackling extreme weather is a primary scientific goal. A study published in Nature in August of this year indicated the model could improve forecast accuracy by a day compared to previous models. In predicting Hurricane Isaias, the AI identified its potential as a tropical cyclone a full week before its official formation. Dan Brown, head of the NHC business division, confirmed the AI models entered operational use in late June 2025, with DeepMind's models integrated into the NHC's path and intensity forecasting systems in 2026. While the AI excels at predicting complex steering currents, Brown emphasized that human experts are indispensable. He stated that experts must synthesize predictions from various models, as no single model provides consistent judgments across different forecast periods, and only human judgment can incorporate real-time aircraft and satellite data to assess a storm's actual structure.