Invited Talk
Data-driven Global Atmosphere-ocean-land Coupled Model
Yoo-Geun HAM, Seoul National University

Biography

Dr. Yoo-Geun Ham is an associate professor of Department of Environmental Managements, Seoul National University, Republic of Korea. He received his B.S. from Seoul National University in 2003 and Ph.D. in Atmospheric Sciences from Seoul National University in 2009. His research field is data assimilation, climate modeling, and tropical and mid-latitude climate dynamics. He pioneered the use of deep learning algorithms to global oceanic/atmospheric/sea-ice data assimilation, global Earth system modeling, predicting tropical and mid-latitude climate variabilities, and analyzing the impact of the global warming on extreme weather events.

Professor Ham is a member of the Young Korean Academy of Science and Technology (YKAST), and WWRP/WCRP S2S Machine Learning working group. He is a recipient of 2024 Ministerial Commendation for contributions to the promotion of basic research from Ministry of Science and ICT in South Korea, 2024 Frontiers in Science Award (FSA) from International Congress of Basic Science (ICBS) in China, and 2020 Young Scientist Award, and 100 Best Performances in National R&D in 2020 best achievement in pure basic from Ministry of Science and ICT in South Korea.

Deep learning-driven data models have recently gained significant attention in global weather forecasting within 2 weeks. However, the current approaches introduced so far have limitations to extend its forecasts beyond two weeks due to their inability to accurately simulate atmosphere-ocean-land interactions. This study develops a deep learning-based atmosphere-ocean-land coupled model to improve long-range climate simulation/forecast accuracy. To properly simulate coupling strength between the atmosphere, ocean, and land modules, the one module’s encoders to predict its own components are separately trained from the encoders to be coupled to other modules (so called ‘coupled-feature generation (CG) module’). After the pre-training of each sub-coupled modules (i.e., atmosphere + ocean CG modules, ocean + atmosphere CG modules, etc.), those are fine-tuned within a fully coupled model framework. In addition, the coupled model is designed to predict tendency (i.e., difference between the future and current values) to less rely on the autoregressive feature of predictands and replay buffer mechanism is applied to increase model’s ability to predict long-term climate. The proposed model is compared to a data-driven coupled model in which all atmosphere-ocean-land variables are trained and predicted as a single model. Inference for a period of century exhibited stable climate states without any significant drift, confirming its ability to simulate long-term climate. The simulation quality of El Nino-Southern Oscillation, which is one of most prominent atmosphere-ocean coupled process, is compared with the observations. The proposed deep learning coupled model exhibited higher performance for predicting the heatwaves in regions with strong atmosphere-land coupling.