2005 Volume 5 Issue 1 Pages 1-16
We present an effective method to model shallow-subsurface S-wave structures and apply it to Osaka Bay areas. In our approach, the geological structure is represented as a multi-layered model by referring dense geological logging data, and it is considered to be a structure model of physical properties. An S-wave velocity in each layer is modeled so that the resulting structure model can well reproduce the theoretical amplifications calculated by one-dimensional velocity structures of the velocity logging data. In this study, the geological structure of Osaka Bay areas is represented as a model having four layers:(1) filling, (2) alluvial layers including Ma13, (3) a diluvial gravel layer (Dg1), and (4) a diluvial clay layer (Ma12). The theoretical amplifications are calculated using the four-layer model. S-wave velocity in each layer is modeled through an inversion process to fit the model amplifications with the theoretical ones calculated using the detailed velocity logging data. We assume that the S-wave velocity assigned to the top layer varies locally, while those assigned to the other layers are common in the target area. As the results, S-wave velocities in (2)-(4) layers are modeled as 143, 343, and 195 m/s, respectively. The velocity structure consisting of (1)-(4) layers with the above S-wave velocities reflects general characteristics of the detailed velocity structures from velocity logging data. The theoretical amplifications calculated with resulting model structures correspond well with those calculated from the velocity structures of velocity logging data.
