Attenuation and multicomponent data CREWES Research Report  Volume 14 (2002) 1 The impact of attenuation on the resolution of multicomponent seismic data Richard A. Bale and Robert R. Stewart ABSTRACT In this paper, we undertake a comparative analysis of the expected effect of constant Q absorption on different modes, illustrating these effects by modelling absorption for homogeneous and layered models. We find that when S- and P-wave attenuation filters are compared in depth, they are exactly equal for the same Q value, in the homogeneous case. Higher wavenumbers for given frequencies in the source wavelet give an initial advantage to S-wave resolution in depth, which may be lost to attenuation if S-wave Q is less than P-wave Q, and/or if there are very low shear velocities in the near surface. Finally, dispersion, which inevitably accompanies attenuation, will differ for P and S modes with different Q values, resulting in event correlation errors. One, perhaps surprising, implication of this work is the need for better low-frequency recording to enhance shear-wave resolution. Additionally we provide relationships between interval and effective parameters including a Dix type inversion formula which could be used to derive shear-wave Q values from converted wave data. INTRODUCTION An important practical question for multicomponent seismic surveys is how absorption impacts shear or converted-wave resolution compared with that of P-waves. Converted-waves (specifically, those converting from P to S upon reflection) have the potential for providing higher resolution than P-waves, due to the shorter wavelengths associated with the same temporal frequencies. In practice, it is often observed that converted-wave resolution does not reach this ideal, particularly at depth. One possible reason suggested for this is the stronger effect of Q attenuation upon converted-waves. When considering the effect of absorption on converted-waves, we must consider two different values Q P and Q S , in much the same way as a medium has two different velocities V P and V S . Rock physics provides a theoretical relationship, derived from complex elastic moduli, connecting the P-wave, S-wave, and bulk compressional Q values, Q P , Q S , and Q K , as follows (e.g. Winkler and Nur, 1979): K P S S P S P Q V V Q V V Q 1 3 4 1 1 3 4 1 2 2                 - +         = From this equation, and assuming an infinite value for Q K , Udias (1999) argues that Q S can be expected to be 4/9 of Q P (if S P V V 3 = ). Whilst this is theoretically true for a dry medium, Winkler and Nur (1979) showed that for saturated or partially saturated media, one could equally find P K Q Q ≤ , in which case we would have P S Q Q ≥ .