Gabor deconvolution applied to a Blackfoot dataset

Victor Iliescu, Gary F. Margrave

A new deconvolution technique was applied to the data gathered over the Blackfoot field owned by PanCanadian Petroleum Ltd. In this paper, the Gabor method for deconvolution, which is a nonstationary, transform-based technique is compared with the familiar spiking deconvolution, which is operator based. The Gabor method compensates for the time-variant frequency attenuation, which is the main cause of the nonstationarity of seismic signals. The nonstationary Gabor deconvolution is a robust and flexible method that successfully replaces the stationary, classical Wiener deconvolution. This paper shows, using real data examples, that the Gabor method can be applied with confidence as an alternative to the Wiener deconvolution and the time-variant spectral whitening (TVSW) techniques. Within the design gate of Wiener deconvolution, Gabor deconvolution produces similar results. Outside this gate, Gabor deconvolution is superior.