Low-frequency, high-amplitude surface wave (ground roll) is an old problem in land-based seismic field records.
Conventional processing modules are to suppress ground- rolls, such as frequency and velocity filters, which use the
Fourier transform. All of these techniques assume that the frequency contents of seismic trace are stationary, but in
reality seismic traces have time varying frequency contents.
In this study an alternative method to the Fourier transform known as the wavelet transform was used to suppress ground rolls. The key advantage of the wavelet transform over the Fourier transform is that it does not assume that the trace is stationary, and it can also localize the target information in the time-scale domain. The wavelet transform decomposes the seismic trace using basis functions that have finite extent in both frequency and time.
In this research, an algorithm was prepared to take the wavelet transform of seismic traces of a shot record. Then a filter for ground roll suppression was designed based on the results of wavelet transform. Finally, the efficiency of this filter was compared to that of the band-pass and F-K filters.