Japan Geoscience Union Meeting 2024

convener:Hisashi Yashiro(National Institute for Environmental Studies), Masuo Nakano(Japan Agency for Marine-Earth Science and Technology), Takuya Kawabata(Meteorological Research Institute), Miyakawa Tomoki(Atmosphere and Ocean Research Institute, The University of Tokyo)

High-performance computing (HPC) is one of the important research infrastructures supporting Today's weather, climate, and environmental science studies. The computational performance of supercomputers such as the Earth Simulator, the K computer, and Fugaku makes it possible to achieve higher resolution, a wider computational domain, more ensemble calculations, and the use of more sophisticated physical processes. On the other hand, due to changes in computer trends, large-scale computations of weather and climate require closer collaboration with the computational science field. From FY2022, feasibility studies on Japan's next flagship machine to follow Fugaku have begun, and the specific requests for next-generation computing infrastructure are reported from the weather/climate research fields. HPC is also showing its power in data science, and research on data assimilation methods using high-frequency/high-density observational big data and the combined use of AI technology has made remarkable progress in recent years. Furthermore, the 'digital twin' concept supported by these computational results is attracting attention as a large-scale system that includes data infrastructure development and social implementation. Co-hosted with the Meteorological Society of Japan, this session calls for research topics in weather, climate, and environmental science that focus on "computation," including numerical modeling, big data analysis, data assimilation, and AI technology. The participants can share future perspectives on atmospheric science research using HPC and the challenges to be addressed.