Viscoacoustic generalized screen propagator in constant-Q model

Abstract

When seismic waves propagate through the geological formation, there is a significant loss of energy and a decrease in imaging resolution, because of the viscoacoustic properties of subsurface medium. This profoundly impacts seismic wavefield propagation, imaging and interpretation. To accurately image the true structure of subsurface medium, the consensus among geophysicists is to no longer treat subsurface medium as ideal homogeneous medium, but rather to incorporate the viscoacoustic properties of subsurface medium. Based on the generalized screen propagator using conventional acoustic wave equation (acoustic GSP), our developed method introduces viscoacoustic compensation strategy, and derives a one-way wave generalized screen propagator based on time-fractional viscoacoustic wave equation (viscoacoustic GSP). In numerical experiments, we conducted tests on two-dimensional multi-layer model and the Marmousi model. When comparing with the acoustic GSP using the acoustic data, we found that the imaging results of the viscoacoustic GSP using the viscoacoustic data showed a significant attenuation compensation effect, and achieved imaging results for both algorithms were essentially consistent. However, the imaging results of acoustic GSP using viscoacoustic data showed significant attenuation effects, especially for deep subsurface imaging. This indicates that we have proposed an effective method to compensate the attenuated seismic wavefield. Our application on a set of real seismic data demonstrated that the imaging performance of our proposed method in local areas surpassed that of the conventional acoustic GSP. This suggests that our proposed method holds practical value and can more accurately image real subsurface structures while enhancing imaging resolution compared with the conventional acoustic GSP. Finally, with respect to computational efficiency, we gathered statistics on running time to compare our proposed method with conventional Q-RTM, and it is evident that our method exhibits higher computational efficiency. In summary, our proposed viscoacoustic GSP method takes into account the true properties of the medium, still achieves migration results comparable to conventional acoustic GSP.