Extracting and suppressing surface waves based on high resolution linear Radon transform

Abstract:

Surface waves are strong noise in petroleum seismic exploration. It is necessary to suppress this kind of noise to improve the signal-to-noise ratio of seismic data. In the time-offset (t-x) domain, surface waves and body waves interfere with each other, which make it difficult to separate them. Due to the characteristics of low velocity, low frequency and strong energy, surface-wave suppression is possible in the frequency-velocity (f-v) domain. We propose a novel strategy to separate surface waves from body waves using high-resolution linear Radon transform (LRT). The proposed high-resolution LRT shows the advantage that the shot gather can be transformed between the t-x domain and the f-v domain with amplitude of seismic data preserved. Because of that, we can obtain the energy distribution relation of surface and body waves in the f-v domain by high-resolution LRT. After extracting the energy of surface waves in the f-v domain, we transform the selected energy back to the t-x domain to retrieve surface waves. Finally, we subtract them from the raw shot gather arithmetically to suppress surface waves. Compared with the conventional the f-k method, results of synthetic and real-world examples demonstrate that the proposed method produces more accurate results in surface-wave suppression. 

Keywords:surface waves and body waves; Radon transform; time-offset domain; frequency-velocity domain