Some Application of Wavelet Transform In Seismic Data Processing

Ground-roll, swell noise, guided waves and random noise are just some of most persistent types of noise in land, marine and OBC data and can be hard to remove with traditional methods. Many different techniques based on Fourier transforms are used to suppress noise in exploration seismology. Wavelet Transform presents a new tool to aid seismic processing and they have been successfully applied for compressing and de-noising purposes. For pre-stack filtering, 1D and 2D Wavelet Transform-based filters have been proposed by Deigahn and Watts (1998), Yu et al. (2002), Yu and Garossino (2005) and Abdul-Jauward and Khene (2000). In this paper we reproduce some of the algorithms used in the research community, and further test and develop them. We use both 1D and 2D Stationary Wavelet Transform-based filters to suppress ground-roll, acquisition footprints and random noise. We also consider resolution enhancement in wavelet transform domain, with some observations and preliminary results. Because most of these techniques are applied on post-migration data they present very inexpensive and computationally efficient solutions. Any new information about structures and subtle geological information that can be derived after application of this type of technique can lead to new discoveries (Countiss, 2002).