Separating Highly-Blended 3D Land Data by Sparse Inversion
J. Song, P. Li, W. Wang, Z. Qian, P. Sun and H. Xue
Event name: 81st EAGE Conference and Exhibition 2019
Session: Poster: Multi-Component Seismic A / Simultaneous Sources A
Publication date: 03 June 2019
Info: Extended abstract, PDF ( 4.1Mb )
Price: € 20
Simultaneous shooting greatly improves seismic acquisition productivity, but poses great challenge to the separation of interfering sources in the highly-blended land data. We proposed an inversion-based deblending method applied on 3D common receiver gathers. First, seismic data is transformed into frequency-wavenumber-wavenumber(FKK) domain. Then, coherent signals are estimated by imposing sparse constraint on the transformed data. After that, interference noises are predicted using the estimated signals and then subtracted from the original input. Driven by data residuals, signals are iteratively updated with shrinking thresholds until fully separation of the signal and noise. Test on 3D real land data shows that the proposed method can separate highly-blended seismic data in an accurate, stable and fast way.