| Session Details | |
| Section | IG - Interdisciplinary Geosciences |
| Session Title | Data-driven Modeling in Geoscience |
| Main Convener | Dr. Hiromichi Nagao (The University of Tokyo, Japan) |
| Co-convener(s) | Dr. Shinya Nakano (The Institute of Statistical Mathematics, Japan) Dr. Tatsu Kuwatani (Japan Agency for Marine-Earth Science and Technology, Japan) Prof. Katsuaki Koike (Kyoto University, Japan) Dr. Jun'ichi Fukuda (University of Tokyo, Japan) |
| Session Description | Numerous observational/experimental “big data” obtained by recent extensive observation systems and/or high-sensitive sensors have substantially contributed to spatiotemporal modeling in various fields of geoscience. However, even such big data are still insufficient to comprehensively understand the whole static and dynamic systems of the Earth, so that modeling techniques, such as sparse modeling, data assimilation, emulation and other mathematical methods, play important roles to extract essential information from high-dimensional data. This session especially focuses on methodologies beneficial to data-driven modeling, and highly welcomes papers related to such as (1) information-science methods that extract essential information from high-dimensional data and/or uncovering unobservable structures, (2) actual modeling from real observational/experimental data based on statistical methods, (3) fundamental mathematics that generates new algorithms or innovates previous algorithms useful in data analyses, and (4) miscellaneous methodologies for data-driven modeling. |