Suppression of seismic surface waves based on adaptive weighted super-virtual interferometry

For land seismic surveys, the surface waves are the dominant noises that mask the effective signals on seismograms. The conventional methods isolate surface waves from the effective signals by the differences in frequencies or apparent velocities, but may not perform well when these differences are not obvious. Since the original seismic interferometry can only predict inter-receiver surface waves, we propose the use of super-virtual interferometry (SVI), which is a totally data-driven method, to predict shot-to-receiver surface waves, since this method relieves the limitation that a real shot should collocate with one of the receivers for adaptive subtraction. We further develop the adaptive weighted SVI (AWSVI) to improve the prediction of dispersive surface waves, which may be generated from heterogeneous media at the near surface. Numerical examples demonstrate the effectiveness of AWSVI to predict dispersive surface waves and its applicability to the complex near surface. The application of AWSVI on the field data from a land survey in the east of China improves the suppression of the residual surface waves compared to the conventional methods.