*4.4. Shear Wave Velocity Profile*

According to the pre- and post-processing steps mentioned above, we obtained virtual shot gathers with each receiver as the virtual shot point. We used phase-shift measurement for dispersion analysis and then used the genetic algorithm [35] to invert the underground shear wave velocity structure.

The genetic algorithm is an efficient, parallel, random global optimization search algorithm based on biological genetic mechanisms and natural selection. It searches for the optimal solution by giving a hypothetical solution and simulating the process of natural evolution. This method has been widely promoted and successfully applied in many fields, and it is also a commonly used Rayleigh wave dispersion curve inversion method. The initial model parameters (i.e., hypothetical solution) and the inversion parameters are shown in Table 1.


**Table 1.** Initial model parameters and the parameters used for inversion by genetic algorithm.

In Figure 9, we show the inversion result of the dispersion curve in Figure 8. Figure 9a shows the initial picked dispersion curve (red dotted line) and the inverted dispersion curve (blue line), and Figure 9b shows the given upper and lower limits of the initial shear wave velocity model (red dotted line) and the inverted underground shear wave velocity structure (blue line). The high-frequency component of the surface wave mainly reflects the shallow underground information, while the low-frequency component mainly reflects the deep underground information. As there were few or no dispersion data recorded by the passive seismic data in the low-frequency range (1–2.5 Hz), the inversion results of the shallow part (0–200 m in Figure 9b) of the shear wave velocity structure are more accurate, while the inversion results of the deep part (200–300 m in Figure 9b) are not necessarily the same. Therefore, the depth range of the finally inverted shear wave velocity profile is 0–200 m underground.

**Figure 9.** The inversion result of the dispersion curve in Figure 8. (**a**) The initial picked dispersion curve (red dotted line) and the inverted dispersion curve (blue line). (**b**) The given upper and lower limits of the initial shear wave velocity model (red dotted line) and inversed shear wave velocity structure (blue line).

We then selected the virtual shot gathers at 30 receivers, used phase-shift measurement for dispersion analysis, and obtained the inverse shear wave velocity structures of those 30 receivers. Finally, the shear wave velocity structure profile in this area was obtained by Kriging interpolation [36], as shown in Figure 10; the red inverted triangle indicates the position of the inverted shear wave velocity structure, and the section length of the final inversion is 1300 m. As the data quality of the following (1300–1500 m) is poor, it cannot avoid bringing errors to the whole inversion, so it was abandoned. Its underground structure is clear, and there is a fault, F, at approximately 900 m.

**Figure 10.** Shallow shear wave velocity profile from passive surface wave inversion.

### **5. Discussion**

Although the noise or signal removal algorithm based on SS-CWT proposed by Mousavi et al. [25] has achieved outstanding results in this case study, there are still some problems that need further discussion. In this section, we highlight two issues based on our experience of processing the Sichuan passive seismic dataset: (1) the applicability of this method in other passive seismic datasets, and (2) the efficiency of this method in processing passive seismic datasets.
