Introduction

Focusing-effect AVO or AVA is the phenomenon of velocity and/or absorption lenses creating substantial amplitude variations, but only small traveltime anomalies Kjartansson (1979). The patterns thus created can interfere significantly with AVO/AVA caused by lithological contrasts at the reflector. To render amplitude analysis feasible, these patterns need to be removed from the image. I will use the acronym ``FEAVO'' to refer to focusing-effect AVO or AVA in general, reserving ``FEAVA'' only for specific references to the angle domain. These terms refers only to amplitudes focusing through heterogeneities smaller than the Fresnel zone, as formalized by Spetzler et al. (2004), and which do not cause energy to be lost by sending it outside the finite spatial extent of the seismic survey (i.e., ``illumination problems''). Focusing can be positive (usual meaning of term) or negative (i.e., in the case of absorption).

Vlad and Biondi (2002), Vlad (2002), Vlad et al. (2003) and Vlad (2004) have conjectured that the key to removing FEAVO is creating an accurate velocity model that contains the lenses which cause the focusing, then performing one-way wavefield extrapolation migration with this velocity model. I will present a qualitative proof of this statement. I will also analyze the peculiarities of modeling FEAVO effects with one-way or two-way wavefield extrapolation algorithms.