**5. Field Example**

Per the interpreter's request, the primary focus is on suppressing internal multiples below 6 s to observe the original seismic profile more clearly. The seismic profile below 4 s is shown in Figure 9. From Figure 9a, it can be seen that the events below 6 s are similar in wave shape to the strong energy primaries, and the energy and frequency of these events are significantly different in the horizontal direction. Based on the relevant characteristics of multiples and multiples identification methods, these events are determined as internal multiples generated by the three sets of strong reflection interfaces of Middle Cambrian. Proof by facts, these internal multiples are challenging to suppress in pre-stack data using the filtering methods, such as the Radon transform method. Therefore, on top of the pre-stack suppression of multiples, we suppressed those internal multiples that are difficult to suppress in post-stack data using the iterative virtual events internal multiples suppression method.

**Figure 9.** Profile comparison before and after the internal multiple suppression. (**a**) Original stacked profile; (**b**) stacked section after the internal multiple suppression.

Figure 9 shows the profile comparison before and after the internal multiple suppression. Figure 9a displays the partial stack profile before internal multiple suppression, while Figure 9b shows the corresponding stack profile after the internal multiple suppression. As seen in Figure 9b, the internal multiples below 6 s are mostly suppressed, and the horizontal difference in energy and frequency of the events below 6 s is smaller after the internal multiples are suppressed, particularly at the position indicated by the red arrow. To more clearly observe the effect of internal multiple suppression before and after, the part in the yellow box in Figure 9 is enlarged, as shown in Figure 10. In Figure 10, the event, which is similar to the upper strong reflection waveform with the opposite phase, periodic appearance, and significant lateral variation, is essentially suppressed. This is consistent with the analysis and identification results of multiples discussed earlier (Figures 2 and 3). The results demonstrate that the iterative virtual event multiple suppression method can effectively suppress the internal multiples in field data.

**Figure 10.** Part of profile comparison before and after the internal multiple suppression. (**a**) Original stacked profile; (**b**) stacked section after the internal multiple suppression.

To further analyze the suppression effect and accuracy of internal multiples, some sections before and after internal multiple suppression are selected for spectrum analysis, as shown in Figure 11. Figure 11a is the partial superimposed section before internal multiple suppression, Figure 11b is the partial superimposed section after internal multiple suppression, and Figure 11c shows the frequency distribution curve before and after internal multiple suppression. It can be seen from the figure that after suppressing internal multiples, the frequency spectrum is significantly broadened, and the spectral energy of the seismic profile is increased. This indicates that while using the iterative virtual event technology to suppress the internal multiples, the effective wave is protected, and the resolution of seismic data is improved. The method can provide reliable seismic data for subsequent studies of hydrocarbon source rocks.

**Figure 11.** Spectrum analysis before and after the internal multiple suppression. (**a**) Partial stacked sections before the internal multiple suppression; (**b**) partial stack section after the internal multiple suppression; (**c**) frequency distribution curve before and after the internal multiple suppression.
