A joint matrix minimization approach for seismic wavefield recovery

Area of Science:

• Geophysics
• Seismic Imaging
• Signal Processing

Background:

• Seismic wavefield reconstruction is crucial for seismic image processing due to incomplete observational data.
• Existing methods often process seismic traces individually, potentially missing interdependencies.

Purpose of the Study:

• To propose a novel joint matrix minimization model for seismic wavefield recovery.
• To address the challenge of reconstructing wavefields from sub-sampled seismic data.

Main Methods:

• A joint matrix minimization model using collective representation of sub-sampled traces.
• Formulation of an l2,p-regularized joint matrix minimization (0 < p ≤ 1).
• Development of a unified algorithm to solve the matrix optimization problem with convergence analysis.

Main Results:

• The proposed method effectively reconstructs seismic wavefields by considering interrelations between multiple observations.
• Numerical experiments on synthetic and field data demonstrate efficient performance.
• The technique achieves good reconstruction accuracy and computational cost.

Conclusions:

• The joint matrix minimization approach offers a powerful strategy for seismic wavefield recovery.
• The developed algorithm efficiently handles computational challenges associated with the l2,p regularization.
• The method shows significant potential for practical applications in seismic data processing.