Velocity-independent time-domain seismic imaging using local event slopes

Introduction

The conventional approach to seismic data analysis (Yilmaz, 2000) consists of two main steps: estimating seismic velocities (the subsurface macromodel) and seismic imaging (mapping of the reflected seismic energy to the reflector positions). The two steps can be repeated when the velocity model gets refined by imaging. Estimating velocities remains one of the most labor-intensive and time-consuming tasks in seismic data processing although several approaches have been developed to partially automate and simplify it (Lambaré et al., 2004b; Adler and Brandwood, 1999; Siliqi et al., 2003). In time-domain imaging, effective seismic velocities are picked from semblance scans. More than just a single parameter (such as seismic velocity) needs to be picked to use higher-order traveltime approximations such as those suggested by the multifocusing approach (Hertweck et al., 2004; Landa et al., 1999).

The idea of velocity-independent time-domain seismic imaging belongs to Ottolini (1983), who considered decomposing seismic data into a range of local slopes. Wolf et al. (2004) observed that it is possible to perform moveout analysis by estimating local data slopes in the prestack data domain using an automatic method such as plane-wave destruction (Fomel, 2002). In this paper, I extend Ottolini's idea of velocity-independent imaging and show that extracting local event slopes in prestack data is sufficient for accomplishing all common time-domain imaging tasks, from hyperbolic and non-hyperbolic normal moveout corrections to dip moveout and prestack time migration. Rather than being a prerequisite for seismic imaging, seismic velocities turn into data attributes that can be extracted from the input data simultaneously with imaging.

The idea of using local event slopes estimated from prestack seismic data goes back to the work of Rieber (1936) and Riabinkin (1957). It was used later by Sword (1987b) and extended in the method of stereotomography (Billette et al., 2003; Lambaré et al., 2004a; Lambaré, 2004; Billette and Lambaré, 1998). In the depth imaging context, local data slopes were also utilized by Baina et al. (2003) for anti-aliased Kirchhoff migration and by Hua and McMechan (2003) in the method of parsimonious depth migration. In this paper, I extend and formalize the application of these ideas to time-domain imaging. By analogy with the oriented wave equation, which describes wave propagation in the space of local orientations (Fomel, 2003a), I use the term oriented when referring to local slopes.

The analysis proceeds from normal moveout (NMO) to dip moveout (DMO) and prestack time migration. I derive analytical expressions for mapping the data attributes to the imaging domain and demonstrate their use with synthetic and field data tests.

Velocity-independent time-domain seismic imaging using local event slopes