[M-GI29] Data-driven geosciences

[MGI29-05] Exploration in applications of "Data Descriptive Science" to fields of energy resources and earth sciences

★Invited Papers

*Anna Suzuki¹, Keiichiro Goto¹, Kazuya Ishitsuka², Takeshi Tsuji³, Shotaro Akaho⁴, Yasuaki Hiraoka² (1.Tohoku University, 2.Kyoto University, 3.The University of Tokyo, 4.National Institute of Advanced Industrial Science and Technology)

Keywords:Mathematics, Data science, Energy resources engineering, Topological data analysis, Flow, Fracture structures

Data-driven science has made remarkable progress in recent years, while the true value encompassed by big data cannot be fully realized if AI technology is black boxed. To effectively use such big data, it is essential to describe the essential structure of data in an appropriate mathematical language, to give meaning to phenomena using that descriptive language, and to understand their mechanisms. Under the leadership of Hiroaki Hiraoka, we have started a Grant-in-Aid for Scientific Research on Innovative Areas (A) "Data Description Science" in 2022. In this area, we aim to solve problems by developing descriptors (mathematical description language that expresses the essential structure of data) that focus on the "shape" and "movement" of data, using state-of-the-art mathematical and data science methods. In particular, this field consists of the trinity of mathematics, data science, and applications, creating a new fusion field "data description science" that transcends the boundaries of existing academic disciplines. In this presentation, I would like to introduce our attempt of data description science in the field of energy resources and earth sciences.