Anisotropic velocity analysis for lithology discrimination

Abstract

A new velocity analysis technique is presented for analyzing moveout of signals on multichannel surface seismic or VSP data. An approximate, skewed hyperbolic moveout formula is derived for horizontally layered, transversely isotropic media. This formula involves three measurement parameters: the average vertical velocity and horizontal and skew moveout velocities. By extending Dix‐type hyperbolic moveout analysis, we obtain improved coherence over large source‐geophone offsets for more accurate moveout correction. Compared with the stacking velocity obtained by simple hyperbolic analysis methods, the three velocity parameters estimated by this technique contain more physically meaningful geologic information regarding the anisotropy and/or velocity heterogeneity of the subsurface. Synthetic P‐wave model experiments demonstrate that the skewed hyperbolic moveout formula yields an excellent fit to time‐distance curves over a wide range of ray angles. Consequently, the measurement parameters are shown to reflect adequately the characteristics of velocity dependence on ray angle, i.e., velocity anisotropy. The technique is then applied to two field offset VSP data sets to measure and analyze the velocity parameters. The results show that the apparent anisotropy, defined as the ratio between the horizontal moveout velocity and average vertical velocity, correlates reasonably well with lithology. Highly anisotropic shale and chalk exhibit higher horizontal‐to‐vertical velocity ratios and sandstones show lower ratios.