Least squares reverse time migration imaging with illumination preconditioned based on improved PRP conjugate...

Area of Science:

• Geophysics
• Seismic Imaging
• Computational Seismology

Background:

• Least Squares Reverse Time Migration (LSRTM) is a leading seismic imaging technique.
• The Polak-Ribiere-Polyak Conjugate Gradient (PRPCG) method offers good numerical performance but suffers from weak convergence in LSRTM.
• Insufficient iterative gradient information can lead to low-resolution seismic images.

Purpose of the Study:

• To enhance the convergence speed and accuracy of LSRTM.
• To address the limitations of traditional PRPCG in LSRTM.
• To improve the resolution and detail in seismic imaging results.

Main Methods:

• Constructed an optimization factor to improve the gradient iteration direction, creating an Improved PRPCG (IPRPCG) method.
• Integrated an illumination preconditioned (IP) operator with IPRPCG for LSRTM (IPRPCG-IP-LSRTM).
• Evaluated the method's performance through experimental analysis.

Main Results:

• The IPRPCG-IP-LSRTM method significantly improved detail characterization and event continuity in imaging results.
• The iterative curve converged substantially faster compared to standard methods.
• Normalized data residual was reduced by an average of 6.55%.

Conclusions:

• The proposed IPRPCG-IP-LSRTM method effectively enhances seismic migration imaging accuracy.
• The optimization factor and preconditioning operator successfully address convergence and resolution issues.
• This approach offers a more detailed and accurate representation of subsurface structures.