Spatial Distribution of Line Faults and Its Applications

• Activity Number: 39
• Type: Contributed
• Date/Time: Sunday, August 9, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistical Computing
• Author(s): Toshinari Kamakura*
• Companies: Chuo University
• Keywords: spatial point process ; directional data ; missing data ; line length

Abstract

Active faults play important roles for predicting earthquakes and evaluating their risks in seismology. However, it is very difficult to get joint distribution of line faults with three dimensions of locations, lengths and directions. Another difficulty comes from that the observations are sometimes incomplete. Firstly, we investigate the estimators of line length parameter based on the simulation study, generating the random segments on the field (0,1)*(0,1) with a midpoint of the line with 2-dimensional Poisson process. We can observe the heavy bias if we ignore the edge information; most of data set are observed in the bounded windows and this gives rise to incomplete observations. In the article we propose a simple and approximating method for the line lengths and also insists that three dimensional distributions are needed for the real data of active faults in Japan.