Acoustic finite difference parameter analysis and modelling in MATLAB

David Cho, Chad M. Hogan, Gary F. Margrave

The modelling of seismic energy is a valuable tool in seismology. It is useful in the understanding of seismic wave propagation as well as seismic imaging and inversion problems. Modelling will generally involve a tradeoff between cost and accuracy due to certain aspects such as grid dispersion and compute time. Thus a balance must be found between the two and parameters chosen that are best suited for a specific situation. In addition, since most industry applications use SEG-Y for writing seismic data, it is useful to output any modelled results into SEG-Y for a more convenient transition between software applications.

By using MATLAB and the CREWES MATLAB software package, an analysis was performed on various modelled data sets to determine an appropriate sampling for the spatial grid. Different bandwidth filters were also applied to the seismograms to determine their effect on the data. Increasing the sampling of the spatial grid enhances the results but requires a more intensive computation process. In addition, by applying an appropriate filter, erroneous data can be rejected by removing parts of the high frequency numerically dispersive signal.

After parameter analysis, a data set was modelled and loaded into VISTA for processing. The processing results did suggest that the modelled data was successfully written to SEG-Y.