The effects of three-dimensional structure on two-dimensional images of crustal seismic sections and on the interpretation of shear zone morphology

Crustal scale seismic images provide information on the geometry of subsurface structure. In this paper we examine shear zones as they provide geometrical constraints on the evolution of the crust and as they provide pathways for the migration of mineral-rich fluids from the lower crust. However, they typically appear in seismic images of the deep crust as laterally continuous bands of discontinuous reflections with individual reflections often having high amplitudes. Geological mapping of exposed shear zones show them to have a complex 3-D structure yet crustal-scale seismic reflection surveys use single or at the most only a few profiles, and therefore only create 2-D images of these structures. The processing and imaging of the multifold common midpoint (CMP) data assumes that the seismic energy comes entirely from within the plane of the section. In this paper, we use full-waveform 3-D synthetic data to consider the effects that 3-D topography on a reflector has on reflection character on a 2-D profile. We base our synthetics on an observed shear zone and test models with both a single layer and anastomizing layers. We show that topography on the reflector out of the plane of the section may cause spurious events both above and below the expected target depth. We derive the basic understanding using a simple isotropic homogeneous model, however, we then demonstrate that this is a robust phenomenon and is endemic on all 2-D sections even if the overburden is not homogeneous. We demonstrate that we obtain similar results with a velocity gradient or, for a more extreme case, with a 2-km-deep basin filled with low-velocity sediment. For crustal scale seismic profiles, in particular, the effect is pervasive as neither stacking nor migration can discriminate against out-of-plane energy and the 2-D stack represents the 3-D seismic response of a broad swath centred on the profile. However, we conclude that using the modelled data it is possible to identify qualitatively where there is significant contamination from out-of-plane topography and show examples from a shear zone in the Archaean Yilgarn Block in Western Australia.