Figure 7. Slip rate and the corresponding frequency content for Scenario 3 (a and b) and Scenario 4 (c and d). Frequency contents at 8.3 km (e) and at 22.0 km (f) are compared among these two scenarios and Scenario 1.
3.3 Roles of depth-varying upper-plate rigidity
We examine the role of depth-varying upper-plate rigidity in Scenario 5 (heterogeneous velocity structure and homogeneous friction with Lof 0 and d of 0). Similar to Scenario 1, Scenario 5 exhibits a large stress drop up to the trench (Figure 8a) and the large slip is concentrated near the trench (Figure 8b). In particular, this scenario produces a maximum slip > 50 m at the trench (Figure 8b), which is much larger than all other scenarios, including Scenario 1. This result is intuitive because as the wall rock becomes less rigid, the trenchward portion becomes more compliant. Thus, more slip is generated under the same amount of stress drop. Comparing with the other scenarios, this scenario suggests that low-velocity rock layers in the upper plate dominates total amount of shallow slip, if the shallow portion of a subduction plane is velocity-weakening.
The rupture propagation features show that, except for the initial increase in rupture velocity at the deep part of the subduction plane and the trench portion, rupture velocity generally ranges from 2-3 km/s (Figure 8d and 8e). The near-trench rupture velocity significantly exceeds Vs in both hanging wall and footwall (i.e., supershear rupture), indicating that an employment of low- velocity upper plate layers do not fully cap the rupture velocity, due to the effects of free surface and shallow-dipping fault geometry. Comparing with other scenarios (e.g., rupture velocity along the central profile in all scenarios), we can find that the upper-plate low-velocity layers contribute significantly to slow rupture at a narrow range of shallow depth (e.g., 1-3 km depth).