Numerical simulation of ambient noise tomography at 3D structures using 3D finite-difference method
Abstract
Demand on the development of non-invasive measurement methods for shallow S-wave velocity (Vs) structure is increasing. Active and passive surface wave method will play important role in such measurements. Passive surface wave method, microtremor array measurements, or ambient noise tomography particularly receives large attention since the method can penetrate several hundreds to several kilometers easily. Applicability of the ambient noise tomography to complex velocity structures with horizontal velocity change is the one of the issues to be figured out to apply the method to site investigations. We performed numerical simulation of the ambient noise tomography at 3D structures using 3D finite-difference method to evaluate the effect of complex structures on the analysis of the ambient noise tomography. Figure 1 shows an example of the simulation. The model consists of three layers with Vs of 400, 600 and 800 m/s respectively. There are two low velocity zones with Vs of 200 m/s at the depth of 3 m and 6 m. Receivers are deployed in a 100 x 100 m square array with 4 m spacing. The number of receivers is 676. Figure 1a shows the true Vs structure beneath the receiver array. The dimensions of the model are 550 x 550 x 150 m. Sources were randomly distributed outside of the receiver array to simulate ambient noise field. A 3D viscoelastic finite-difference method with 4th order velocity-stress staggered grid scheme was used to calculate seismic wave field. Cell size is 1 m, time step is 0.25 ms and data length is 65 s. Ten 65 s records were calculated with different source distribution. Ambient noise data were processed by CMP-SPAC method. CMP interval (bin size) was 10 m and the number of bins was 100. A dispersion curve was calculated for each bin. An 1D inversion was applied to each dispersion curve with horizontal constraint. The 1D velocity profiles were interpolated to a 3D velocity model. The 3D velocity model obtained from the ambient noise tomography (Figure 1b) is generally consistent with the true model (Figure 1a) and the simulation shows the applicability of the ambient noise tomography to complex velocity structures.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFM.S31B..08H
- Keywords:
-
- 7203 Body waves;
- SEISMOLOGYDE: 7255 Surface waves and free oscillations;
- SEISMOLOGYDE: 7260 Theory;
- SEISMOLOGY