A comparison of methods for estimating Q

Quality factor (Q) as a quantitative measurement of seismic absorption can be used as a hydrocarbon indicator, to implement deconvolution, or in forward modeling. However, quality factor remains difficult to reliably measure from seismic data. A lot of approaches in either the time or frequency domain have been proposed in the past. A good performance was validated in either VSP, crosswell, or sonic waveform data. However, accurate Q extraction needs amplitude or spectrum preservation processing. So, it is, especially for pre-stack data, still challenging because of much more noise interference and more difficult interface tracking. Detailed analysis and comparison of those methods, especially for commonly used centroid frequency shift method, spectral ratio method and peak frequency method, are necessary when estimating Q from reflection data. Beside, in this paper some data preprocessing techniques which aim to improve the accuracy of Q estimation were discussed.