Science & Tech · Science & Environment · 31 mins ago

AI models forecast Typhoon Champi's path better than conventional systems

AI models forecast Typhoon Champi's path better than conventional systems

AI models forecast Typhoon Champi's track more accurately than conventional physics-based systems, which calculate weather using data and physical laws.

Champi formed over the Pacific on Oct.

15 and initially moved west; by Oct.

19, conventional forecasts mostly expected it to approach Taiwan and then China.

AI models instead predicted a northward turn near Okinawa, and the typhoon later curved sharply after approaching Japan's Nansei Islands.

Japan's Meteorological Agency mainly uses conventional models but also refers to external AI forecasts, while private service Weathernews provides AI track forecasts to paying subscribers.

Weathernews says its forecasts for last year's first 26 typhoons had about 30 percent smaller track errors than physics-based models.

AI can help people prepare earlier: an agricultural business in Ibaraki used Weathernews forecasts to install flood barriers and arrange staff shifts about a week before Typhoon Dujuan approached the Kanto region in September.

Predicting storm strength remains a challenge, making it harder to forecast rainfall and wind speeds, and experts say the agency must consider how AI can support disaster-prevention information.

1 source

Related

What people are saying

Munehiko YamaguchiDepartment of Typhoon and Severe Weather Research researcher
AI excels at forecasting large-scale atmospheric flows, including the Pacific high-pressure system and westerlies that affect typhoon tracks, and accurately predicted the timing of the turn,
via Mainichi EN
cannot be ignored. We need to consider how to use them to help provide disaster prevention information,
via Mainichi EN
Kohei SakamotoWeathernews forecasting center official
We can improve accuracy by making the most effective use of AI,
via Mainichi EN