Vector-dip filtering of seismic data in the time-frequency domain

摘要

Methods and systems for separating surface wave velocity information from surface wave noise in seismic data are described. The seismic data can be comprised of irregularities in spatial sampling, non-stationarity in time and non stationarity in frequency. The methods and systems can then be adapted to create a multi-component dip filter that removes high amplitude dispersive noise from the seismic data based on the use of slant stacking. Transforming the squired seismic data associated with a plurality of traces into time-variant spectral estimates T__ r704 Extracting a local velocity associated with the plurality of traces from the time-variant spectral estimates T (706 Calculating an amplitude and a phase associated with one of the plurality of traces using the local velocity T__ _ _ _ _ _ 7 0 8 Determining the noise based on the amplitude and phase S tet710 Subtracting the determined noise from the acquired seismic data