Japan Geoscience Union Meeting 2023

Presentation information

[J] Online Poster

S (Solid Earth Sciences ) » S-TT Technology & Techniques

[S-TT44] Seismic Big Data Analysis Based on the State-of-the-Art of Bayesian Statistics

Mon. May 22, 2023 10:45 AM - 12:15 PM Online Poster Zoom Room (6) (Online Poster)

convener:Hiromichi Nagao(Earthquake Research Institute, The University of Tokyo), Aitaro Kato(Earthquake Research Institute, the University of Tokyo), Keisuke Yano(The Institute of Statistical Mathematics), Takahiro Shiina(National Institute of Advanced Industrial Science and Technology)

On-site poster schedule(2023/5/21 17:15-18:45)

10:45 AM - 12:15 PM

[STT44-P02] Minimum information dependence modeling for analyzing the dependence in mixed-domain data: Application to the analysis of the dependence between source depths and focal mechanisms

*Keisuke Yano1, Tomonari Sei2 (1.The Institute of Statistical Mathematics, 2.The University of Tokyo)

Keywords:mechanism solutions, dependence, statistical model

In this presentation, we will introduce a statistical model and an estimation method for the dependence between variables with different formats, proposed by Sei and Yano, arXiv. As a demonstration of the proposed method, we present an example of analyzing the dependence between source depths and focal mechanisms using the JMA catalog.
The proposed statistical model for mixed-domain data specifies the joint distribution of d variables with unknown parameters that define the dependence between variables and the marginal structures. The proposed model is called the minimum information dependence model. By estimating the dependence parameter from data, we can quantitatively evaluate what dependencies exist between each variable and how much dependencies exist.
As a demonstration of the minimum information dependence model, we present an analysis of the dependence between the mechanism solution and the source depth. In this application, 158 earthquakes from the JMA catalog for 2021 were used. We attempt to quantitatively evaluate the dependence between these two using the minimum information dependence model, where the P- and T-axes are considered as an orthogonal frame in three-dimensional space, and the depth of the epicenter is considered as a real value.