A robust approach to estimate Q from surface seismic data and inverse Q filtering for resolution enhancement

Abstract

Accurate estimation of seismic quality factor (Q) is important in seismic data processing to correct for the velocity dispersion effects and to compensate for absorption losses through inverse Q filtering. To estimate it, often logarithmic spectral ratios of the non-stationary seismic signal between two depth levels are linearly inverted. As these ratios are usually derived from the standard Fourier transform (FT) which has a poor time-frequency resolution, this can lead to biased Q estimation. We have calculated spectral ratios from the high-resolution time-frequency spectrum using the Stockwell transform (ST). We then non-linearly inverted these improved spectral ratios to estimate Q using the Levenberg-Marquardt (LM) method. For a synthetic wedge model, results demonstrate a reasonable improvement in the accuracy of Q estimation with the combined ST-LM approach. Application on a real P-wave reflection seismic dataset revealed an anomalous attenuation zone of very low Q (29–75) values. Inverse Q filtering using the estimated Q profile has enhanced the seismic resolution below it and minimized the differential viscous losses. Frequency slice filtering on Q compensated data has improved the S/N ratio and hence the amplitude fidelity of reflectors.