Figure 4. Numerical results of Scenarios 1 (left panels) and 2 (right panels). (a) and (f) show the shear stress change distribution (negative as stress drop); (b) and (g) show the total slip distribution. (c) and (h) show the rupture time contour with an interval of 5 seconds; the red curves indicate that the rupture front reaches the trench at 53 s and 54 s, respectively. (d) and (i) show the rupture velocity distribution. (e) and (j) show the along-dip profiles of rupture velocity along the central line (red line in (d) and (i)) and along 50 km away along-strike from the central line (black line in (d) and (i)). Blue curves in (e) and (j) are shear wave velocity depth profiles in the hanging wall (solid) and footwall (dashed). The crosses in (d) and (i) indicate the locations of the two on-fault stations.
We select two on-fault stations at 8.3 km and 22 km depth (downdip distance of 32 km and 85 km, respectively) along a depth profile at 50 km along-strike distance (Figure 4d or 4i) to examine slip rates and their frequency contents at different depths. Peak slip rate is comparable between the two stations in Scenario 1 (Figure 5a), while it is significant smaller at the shallow station than at the deep station in Scenario 2 (Figure 5c). This contrast is consistent with rupture propagation and slip distribution in the two scenarios analyzed above. Both scenarios show depletion in high frequency content at the shallow station compared with that at the deep station, with Scenario 2 exhibiting a larger amount of depletion (Figure 5b and 5d). Scenario 1 shows high-frequency depletion at the shallow station above ~1 Hz, while that occurs above ~0.2 Hz in Scenario 2, suggesting strong effects in high-frequency depletion at shallow depth form either heterogeneous fault friction, or heterogeneity velocity structure, or both. To direct compare high-frequency depletion at each station from the two scenarios, we plot the amplitude spectra for the shallow station in Figure 5e and the deep station in Figure 5f. By comparing the slopes of the spectra, we can see that Scenario 2 has a larger amount of high-frequency depletion than Scenario 1 at the shallow station for most frequencies above ~0.2 Hz, though there is some complexity at ~2Hz. At the deep station, it appears Scenario 2 radiates more high frequency signals. We will further unravel the roles of heterogeneous friction and rigidity individually on high-frequency depletion in the next two subsections.