From Basic Forecasting to Predicting Typhoons and Storm Surges
The intersection of artificial intelligence and earth sciences has yielded a powerful new tool for understanding and predicting the complex dynamics of the world’s oceans. Xinhua reports that researchers unveiled LangYa 2.0, a highly advanced AI marine forecasting model, at the Fourth China Digital Earth Conference in Qingdao on June 6, 2026. Developed by the Institute of Oceanology under the Chinese Academy of Sciences (IOCAS), this new iteration represents a significant leap forward from its predecessor, expanding its capabilities from basic ocean variables to the prediction of complex, high-impact marine phenomena.
The original LangYa 1.0, launched in late 2024, demonstrated the potential of AI in oceanography by forecasting basic variables, such as temperature, salinity, and currents, for 1 to 7 days with remarkable efficiency, reportedly 10,000 times faster than conventional methods. However, LangYa 2.0 moves beyond basic metrics to tackle the most challenging and destructive events in the marine environment. The new model is designed to predict six key marine phenomena: typhoons, extreme rainfall, storm surges, internal solitary waves, mesoscale eddies, and sea ice.
Wang Fan, president of IOCAS and the project’s spokesperson, highlighted the profound differences between the two versions. “Version 1.0 tells you what the sea temperature is,” Wang explained. “Version 2.0 tells you where a vortex might form and when a storm surge will arrive.” He further illustrated the advancement with a medical analogy: “It’s the difference between a tool that helps a doctor read a scan and an AI that reads the scan and gives you the diagnosis directly.” This shift from descriptive data to actionable, predictive intelligence is a game-changer for disaster preparedness, maritime operations, and climate research.
Six Specialized Sub Models, Each Trained on Distinct Marine Phenomena
The architecture of LangYa 2.0 is sophisticated, comprising six specialized vertical sub-models. Each of these sub-models has been trained on unique datasets and physical mechanisms specific to the phenomenon it predicts. This specialized approach allows the model to capture the nuanced dynamics of different marine events. For instance, the internal solitary waves module can forecast the evolution of these massive, underwater waves over a 30-day period, allowing users to query their speed and amplitude for the coming week. This capability is crucial for the safety of underwater operations, such as submarine navigation and deep-sea drilling.
Top-Ranked Arctic Sea Ice Forecasting at 3km Resolution
One of the most impressive achievements of LangYa 2.0 is its ability to predict Arctic sea ice. The model ranked first among multiple global models in the 2025 seasonal forecast for the September Arctic sea ice extent, a validation provided by the Sea Ice Prediction Network. As the Arctic region becomes increasingly accessible and strategically important due to climate change, accurate sea ice forecasting is vital for shipping routes, resource exploration, and environmental monitoring. LangYa 2.0 provides monthly-scale predictions at a high resolution of 3 kilometers, offering unprecedented detail for this critical region.
Implications for China’s Maritime Ambitions
The development of LangYa 2.0 underscores China’s commitment to leveraging AI for strategic scientific and economic advantages. The ability to accurately predict typhoons and storm surges has direct implications for coastal infrastructure, shipping, and disaster mitigation, potentially saving lives and billions of dollars in economic losses. Furthermore, the model’s capabilities align with China’s broader ambitions in marine research and resource utilization, supporting the country’s growing maritime economy.
The name “LangYa” itself reflects a blend of cultural heritage and scientific ambition. Drawn from the ancient Cihai encyclopedia, the term means “precious, jade-like treasure.” It also references the ancient LangYa terrace, located 20 kilometers from Qingdao, a site of historical significance. This naming convention highlights the project’s roots in Chinese scientific tradition while aiming for cutting-edge technological achievement.
Looking ahead, the IOCAS team has ambitious plans for the LangYa project. The roadmap includes developing interdisciplinary models that integrate climate and ecological data to provide a more holistic understanding of the Earth system. Additionally, the researchers aim to explore pathways for marine intelligent agents, potentially creating autonomous systems capable of conducting complex oceanographic research and monitoring tasks.
The launch of LangYa 2.0 is a testament to the rapid progress in AI models and research in China. By combining vast amounts of oceanographic data with advanced machine learning techniques, Chinese researchers are creating tools that rival or surpass traditional forecasting methods. As these models continue to evolve and integrate with other scientific disciplines, they will play an increasingly central role in our ability to understand, predict, and adapt to the dynamic forces of the natural world. The success of LangYa 2.0 not only enhances China’s scientific prestige but also provides a valuable resource for the global community in addressing the shared challenges of climate change and marine conservation.