AI Tools Revolutionize Typhoon Predictions in Taiwan

Taiwan's weather forecasters are now using AI to predict typhoons, outperforming traditional methods. With AI-powered tools like Nvidia's FourCastNet and Huawei's Pangu-Weather, forecasters successfully predicted the path of Typhoon Gaemi. AI's accuracy in weather prediction has led to better preparedness, although experts believe further testing is needed.

As tropical storm Bebinca barrels towards waters off northern Taiwan gathering strength into a possible typhoon, weather forecasters in Taipei are using a new and so far successful method to help track its path - artificial intelligence (AI).

AI-generated forecasts, powered by software from tech giants like Nvidia, whose chips are made by Taiwan's semiconductor champion TSMC, have outperformed traditional methods in predicting typhoon tracks. In July, AI-based weather models helped Taiwan better predict the path and impact of Typhoon Gaemi, which was the strongest to strike the island in eight years.

The new technology impressed Taiwan forecasters by predicting a direct hit eight days before Gaemi made landfall, outperforming conventional methods. "People are starting to realize AI indeed delivered some stunning performances compared to conventional models," said Chia Hsin-sing, director at the Taiwan Integrated Disaster Prevention of Technology Engineering Consulting Company Ltd. Forecasters, including Lin Ping-yu from Taiwan's Central Weather Administration (CWA), now track Bebinca using the same AI tools, which have given them greater confidence in their predictions.

(With inputs from agencies.)