Artificial intelligence is deeply reshaping human capabilities for forecasting and understanding complex marine phenomena. On June 6, the Fourth China Digital Earth Conference officially opened in Qingdao, Shandong. At this conference, the "Langya" 2.0 version (short for "Langya") of the global oceanic phenomenon intelligent forecasting large model independently developed by the Institute of Oceanology, Chinese Academy of Sciences was officially released. This achievement marks that global ocean forecasting technology is rapidly moving towards a new intelligent stage that is faster, more precise, and interactive.

As a landmark product of the Institute of Oceanology, Chinese Academy of Sciences, focusing on the intersection of artificial intelligence and ocean science, the iterative upgrade of the "Langya" large model is clear. In the "Langya" 1.0 version released in 2024, the model had already successfully achieved high-precision forecasting of basic state variables such as global ocean temperature, salinity, and currents, successfully verifying the feasibility of AI technology in the field of ocean forecasting at the fundamental level.

The newly launched "Langya" 2.0 has made a leap from "forecasting state variables" to "intelligent forecasting of complex marine phenomena." The research team focused on the practical challenges of marine disaster prevention and shipping safety, developing six specialized vertical sub-models for six core marine phenomena: typhoons, precipitation, storm surges, internal solitary waves, mesoscale eddies, and sea ice.

In specific application scenarios, these vertical models have shown significant practical value. For example, the typhoon forecasting model significantly improves the accuracy of 24-hour path and intensity forecasts for typhoons with extreme uncertainties such as "rapid intensification" and "sudden turns"; while the accompanying precipitation forecasting model provides a stronger computing power support for early warnings of typhoon rains and other extreme precipitation events.

The release of "Langya" 2.0 indicates that the forecasting of complex marine phenomena is moving beyond traditional and obscure professional variables, truly entering industrial-level application scenarios that are perceptible, applicable, and decision-making oriented. This new intelligent support system will play an irreplaceable role as an intellectual advisor in key areas such as global marine disaster prevention and mitigation, and the security of golden shipping routes in the future.