Thu, Aug 21, 2025 page2

• Rogue wave system to roll out next year

  UNTAMED SEA: Development of the detection system began in 2017, using cameras to monitor the coast, combined with CWA data on waves and an AI prediction model

  • By Tsai Yun-jung and Hollie Younger / Staff reporter, with staff writer
  • • 
    • 
    • 

  The Central Weather Administration (CWA) and National Cheng Kung University have developed an artificial intelligence (AI)-powered system to predict rogue waves, which is expected to be rolled out nationwide by the end of the year, the agency said yesterday.

  Rogue waves, also known as “freak waves,” are large, unpredictable surface waves that can pose a serious safety risk.

  Many factors can cause rogue waves, although the precise mechanisms of their development, timing and locations are under-researched, university Department of Hydraulic and Ocean Engineering professor Doong Dong-jiing (董東璟) said.

  

  A rogue wave crashes against the shore in New Taipei City’s Longdong Bay in an undated photograph taken with a Central Weather Administration camera.

  Photo courtesy of the Central Weather Administration

  The system, the first of its kind in Taiwan, is a pioneering technology, he added.

  The system is being used in New Taipei City and national scenic areas along the north and northeast under the Tourism Administration’s jurisdiction, CWA Deputy Administrator Fong Chin-tzu (馮欽賜) said.

  Over the past 25 years, Taiwan has experienced an average of 17 rogue waves annually, sweeping an average of 30 people per year out to sea, CWA data showed.

  The agency began the project by monitoring the occurrence of rogue waves and building a database, which was used to train an AI model to give more accurate predictions of their formation, Fong said.

  The system, which began development in 2017, uses cameras to monitor the coast and identify abnormal waves, combining information with data provided by the CWA’s wave models and an AI prediction model to assess the likelihood of a rogue wave occurring within 24 hours.

  As one of the main causes of rogue waves are swells, swell data is also incorporated into the machine learning, CWA Marine Meteorology and Climate Division head Pan Chi (潘琦) said.

  The agency has established at least one abnormal wave monitoring station in 15 coastal cities and counties across Taiwan, and expects to add another by the end of the year, he said.

  If an area needs additional monitoring systems, it can request assistance from the CWA, he said.

  The New Taipei City Fire Department and the Rueifang District (瑞芳) Office have already budgeted for an additional warning system at Bianfudong (蝙蝠洞), or the “bat cave,” in Rueifang, an area prone to rogue waves, Pan said.

  If a dangerous wave is detected, the system would immediately notify disaster prevention personnel to safely evacuate visitors from the area, he said.

  The agency has not yet decided on whether to send cellphone alerts of rogue waves, which would require following the Disaster Prevention and Protection Act (災害防救法), he said.

  The CWA currently provides forecasting information, while on-site responses are handled by local authorities and relevant agencies, he added.

  Early warning information is to be made available on the CWA Web site and weather app, Pan said.
• Most Popular

  • 1

    Can Taiwan keep the talent it invites?

  • 2

    China’s AI models pose risks, NSB says

  • 3

    Taiwan and Japan have ties that run deep, premier says

  • 4

    Taiwan expresses regret after CPTPP does not address its accession

  • 5

    PLA could attack with little warning: US