Inversion compensation of attenuation and transmission loss based on wave field extrapolation for seismic resolution enhancement

Abstract The low-resolution characteristics of deep seismic signals fail to meet the requirements of high-precision seismic exploration. Conventional inverse Q-filtering does not fully compensate for the viscoelastic attenuation of amplitudes, and this incompleteness increases with depth. In addition, deep seismic reflection waves suffer obvious transmission losses due to the numerous impedance interfaces they traverse, further reducing the energy of deep seismic signals. In order to achieve optimal compensation for both attenuation and transmission loss under a common theoretical framework, we propose an inversion compensation scheme based on frequency-domain wave field extrapolation. Seismic signals that have not experienced energy losses are extrapolated from the surface to the underground. During this extrapolation process, seismic signals are modified according to the physical mechanisms of seismic attenuation and transmission loss. Based on the post-stack migration imaging conditions, we obtain the mathematical relationship between seismic signals before and after experiencing energy losses (attenuation or transmission). Then, the inversion schemes for compensating the energy losses are proposed by combining the total variation regularization. Finally, the compensation steps for enhancing seismic resolution using our proposed method are provided. The superior performances of our method are illustrated by both synthetic and field data examples. We find that our inversion compensation method can achieve maximum compensation for effective signals, meanwhile avoiding the amplification of random noise. In this case, our method can provide high-resolution and high-fidelity seismic data for subsequent reservoir prediction works.