Adaptive finite difference for seismic wavefield modelling in acoustic media

Area of Science:

• Geophysics
• Computational Seismology
• Numerical Methods

Background:

• Efficient numerical seismic wavefield modeling is crucial for advanced seismic imaging techniques like reverse-time migration and full-waveform inversion.
• Finite difference methods are a widely adopted numerical approach for forward modeling in seismology.

Purpose of the Study:

• To introduce a novel finite difference scheme for seismic wavefield modeling.
• To enhance the accuracy of finite difference operators in seismic imaging applications.

Main Methods:

• A time-to-space wavelet mapping technique is introduced to implement finite difference.
• Finite difference coefficients are computed by minimizing the discrepancy between mapped wavelet spatial derivatives and the finite difference operator.
• Coefficients are adapted based on velocity and source wavelet bandwidth to maximize operator accuracy.

Main Results:

• The novel finite difference method demonstrates superior accuracy compared to standard finite difference approaches.
• The method's performance is comparable to existing optimized finite difference schemes, such as Zhang's scheme.
• Numerical examples validate the effectiveness of the proposed adaptive coefficient computation.

Conclusions:

• The developed time-to-space wavelet mapping finite difference scheme offers improved accuracy for seismic wavefield modeling.
• This adaptive approach provides a more precise finite difference operator, beneficial for seismic imaging.
• The method presents a valuable alternative for accurate and efficient seismic data processing.