2D SEISMIC MODELING SIMULATION USING FINITE DIFFERENCE METHOD IN CASE OF ACOUSTIC OVERTHRUST AND VERTICAL TRANSVERSELY ISOTROPIC(VTI) SYNTHETIC MODELS

• Main Author: Zakia Permata Aji, Dinda
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/86350
• Institution: Institut Teknologi Bandung

One of the applications of the seismic method is to detect the characteristics of subsurface layers based on their seismic velocity distribution. This is done by triggering seismic waves from a shot point source, which then propagate through the medium. The wavefield generated from a shot at the source will be received by receivers in such a way that it depends on the velocity distribution within the medium during wave propagation. The seismic wave propagation pattern can be simulated by solving the wave equation numerically, one method of which is using the finite difference (FD) scheme in both the spatial and temporal domains. Due to the simplicity of FD application for wave propagation, forward modeling (FD) simulation is performed to evaluate the resolution accuracy of the wave propagation results. In this study, the FD method is implemented using a second-order scheme in the temporal domain, while for the spatial domain, a fourth-order scheme is used, as there is the potential for numerical instability when using higher orders in the temporal domain. Seismic wave propagation simulation is applied to a synthetic acoustic Overthrust (OT) model and a Vertically Transversely Isotropic (VTI) model with a grid size of 25m × 25m. To evaluate the characteristics of the seismic wave propagation, a Ricker wavelet source with three frequencies—4 Hz, 8 Hz, and 12 Hz—is used. The simulation results show that higher frequencies provide better resolution while still meeting the CFL condition. The simulation for the OT and VTI models is quite stable, fulfilling the CFL conditions with a calculated grid size of 21.91857m for the OT model and 12.4212m for the VTI model, meaning that the chosen grid size of 25m satisfies the CFL condition. The simulations for both models show no aliasing because the ratio between the spatial increment and the wavelength is less than 0.5. The simulation results demonstrate good vertical resolution to distinguish between two layers. The thinnest resolvable layer in the OT model is 91.043m, while in the VTI model it is 237.963m.