The Ministry of Oceans and Fisheries National Institute of Fisheries Science announced on the 23rd that it has developed a 'temperature prediction algorithm' by combining satellite-observed surface temperature data and artificial intelligence (AI) technology to improve the accuracy of sea temperature prediction technology and plans to expand its services in earnest starting in May.
Generally, sea temperatures in major ocean areas have been predicted qualitatively using statistical indicators. In contrast, the National Institute of Fisheries Science has developed a high-resolution algorithm predicting temperatures in wide areas at 1 km intervals and is continuously improving prediction accuracy by incorporating the characteristics of the East Sea, West Sea, South Sea, and East China Sea into the AI. In fact, prediction accuracy has improved from 90% last year to 94% currently.
Since November of last year, the National Institute of Fisheries Science has been providing a pilot prediction service for domestic waters and plans to expand the service to include the northwestern Pacific Ocean starting in May.
Additionally, by utilizing this temperature prediction algorithm, the institute plans to proactively respond to fishery disasters caused by abnormal phenomena such as high and low temperatures and cold water masses, and to establish a scientific basis to reduce the damage.
Choi Yong-seok, president of the National Institute of Fisheries Science, noted, 'The National Institute of Fisheries Science continues to strive to produce marine and fishery information that reflects the valuable opinions of citizens' and added, 'We will strengthen research to produce advanced information in the marine and fishery sector by actively introducing new technologies such as AI.'