next up previous [pdf]

Next: ANGLE-DOMAIN COMMON-IMAGE GATHERS Up: Liu et al.: Stacking Previous: Liu et al.: Stacking

Introduction

The image of geology produced by Kirchhoff migration or wave-equation migration often suffers from artifacts, especially in the case of irregular acquisition or complex wave propagation (Qin et al., 2005; Tang, 2007). Artifacts sometimes bury seismic reflectors, especially in subsalt areas, and make interpretation and model building difficult. Developments in acquisition technology (Howard and Moldoveanu, 2006; Michell et al., 2006) have provided richer coverage and better illumination of the subsurface; however, areas with poor illumination still exist in complex geological environments, such as subsalt zones.

Prucha et al. (2000); Kuhl and Sacchi (2003) used preconditioned inversion in the reflection angle domain to improve imaging in complex media. Hoop and Ursin (2003); Xu et al. (2001) proposed focusing on dip and AVA compensation in common-image gathers to suppress artifacts and improve signal-noise ratio in Kirchhoff migration. To compensate irregular illumination at image points, Kessinger (2004) presented illumination-angle compensation in Kirchhoff migration, in which inclusion of an amplitude-weighting term can dramatically improve the migrated image. Audebert et al. (2005); Bloor et al. (1999) proposed a method of regularization of illumination in the multiangle domain for Kirchhoff migration, in which a hit-count technique is applied to suppress migration artifacts and balance migration amplitudes. Qin et al. (2005) presented interactive dip-gather stacking for attenuating artifacts of Kirchhoff migration. Tang (2007) analytically demonstrated artifacts in angle-domain common-image gathers (ADCIGs) by sparsely-sampled wavefields and proposed selective stacking on the basis of local smoothing of the envelope function in shot-profile wave-equation migration. Manning et al. (2008) proposed the MAZ-stack method for weighting signal in areas of poor illumination for multiazimuth seismic data, in which weights of stacking were chosen as binary (zero or one).

Local similarity is a local attribute measured between two signals (Fomel, 2007a) that has been applied to multicomponent seismic image registration (Fomel et al., 2005; Fomel, 2007a) and time-lapse image registration (Fomel and Long, 2009). In an earlier work (Liu et al., 2009), we also applied local similarity as a weight for stacking common-midpoint gathers in order to improve signal-noise ratio of seismic data.

In this paper, we present another application of local similarity in stacking ADCIGs in order to normalize illumination. Our method applies local similarity between initial image and ADCIGs using a soft threshold to normalize stacking. It can attenuate migration artifacts and restore migration amplitudes. This method can be regarded as true-amplitude illumination compensation arising from the so-called Beylkin determinant (Audebert et al., 2005; Albertin et al., 1999; Beylkin, 1985).


next up previous [pdf]

Next: ANGLE-DOMAIN COMMON-IMAGE GATHERS Up: Liu et al.: Stacking Previous: Liu et al.: Stacking

2013-07-26