Velocity images from stacking depth-slowness seismic wavefields

Abstract

We present synthetic and real data examples processed using a depth imaging method for long-offset multichannel seismic data via the intercept-time-slowness (τ-p) domain. The refracted and reflected information contained in the wavefield is processed simultaneously. Our scheme uses common midpoint data that is transformed to the τ-p domain and mapped to the depth-slowness (z-p) domain using a downward continuation algorithm. The velocity function for downward continuation may be obtained iteratively from the diving ray trajectory within the slowness-depth wavefield or estimated using an independent method. Horizontal events within the z-p wavefield are isolated by applying a mute function corresponding to the downward continuation velocity function in the z-p domain. Stacking the resulting wavefield gives an image trace that may be superimposed on the velocity function to form a velocity image. Following testing on synthetic data the method is applied to two long-offset marine seismic streamer data sets with offsets up to 18 km, from the northeast Atlantic margin. These examples demonstrate the potential of the method to obtain velocity images in difficult geological locations, where conventional processing has been less successful. © 2006 The Authors Journal compilation © 2006 RAS.