Ground roll noise attenuation by local time-frequency transform

In recent years, one of the techniques which has been studied in geophysics and has had many usages in seismic data processing and interpretation is timefrequency analysis. The most common applications of time-frequency methods have been in seismic interpretation. However, in recent years, geophysicists have used this analytical method for processing purposes (Elboth et al., 2010; Askari and Siahkoohi, 2008). In this paper, we use a new time-frequency method for ground roll noise attenuation. This method, provided by Liu and Fomel (2010) and known as local time-frequency transform, aims at depicting the nonstationary character of seismic data. The proposed transform uses a Fourier basis to match the target signal using regularized least-squares inversion. It provides an invertible time-frequency representation, which makes the method suitable for analyzing nonstationary data. The obtained results by applying the method on both synthetic and real data show that local time-frequency transform is efficient in depicting nonstationarity character of seismic data and also attenuating ground-roll noise.