The difficulty of prestack three parameter inversion is the stability of inversion and the uniqueness of the results. In order to solve this problem, there are many ways have been tried, in addition to discard density terms. According to Bayesian probability theory, It is an ideal method to add a priori constraints to reduce the uncertainty of inversion. So that we need to add more prior information such as geological priori information and well log information to reduce the uncertainty. From the seismic exploration theory and practical experience, It is confirmed that the seismic reflection (diffraction) wave field from the subsurface sedimentary strata is characteristic or compressible. Therefore, the parameters of the layered strata model are also compressible or sparse. Utilizing the sparse prior of seismic data, this paper introduces the compressed sensing theory to construct a sensing matrix to reduce the dimensionality of inversion matrix. Then we carry out the reflection coefficient in compressed sensing domain by sparse reconstruction method which named POCS algorithm to ensure the stability of the three parameter inversion. Finally, Synthetic examples and the field seismic data are tested. The result demonstrate that proposed compressed sensing three parameters AVO inversion strategy can improve the stability of elastic parameters estimation and be capable of provide more credible density information.
Presentation Date: Wednesday, September 27, 2017
Start Time: 9:45 AM
Location: 370D
Presentation Type: ORAL
