Source-receiver, reverse-time imaging of dual-source, vector-acoustic seismic data

Abstract

Novel technologies in seismic data acquisition allow for recording full vector-acoustic (VA) data: pointwise recordings of pressure and its multicomponent gradient, excited by pressure only as well as dipole/gradient sources. Building on recent connections between imaging and seismic interferometry, we present a wave-equation-based, nonlinear, reverse-time imaging approach that takes full advantage of dual-source multicomponent data. The method's formulation relies on source-receiver scattering reciprocity, thus making proper use of VA fields in the wavefield extrapolation and imaging condition steps in a self-consistent manner. The VA imaging method is capable of simultaneously focusing energy from all in-and outgoing waves: The receiver-side up-and downgoing (receiver ghosts) fields are handled by the VA receiver extrapolation, whereas source-side in-and outgoing (source ghosts) arrivals are accounted for when combining dual-source data at the imaging condition. Additionally, VA imaging handles image amplitudes better than conventional reverse-time migration because it properly handles finite-aperture directivity directly from dual-source, 4C data. For nonlinear imaging, we provide a complete source-receiver framework that relies only on surface integrals, thus being computationally applicable to practical problems. The nonlinear image can be implicitly interpreted as a superposition of several nonlinear interactions between scattering components of data with those corresponding to the extrapolators (i.e., to the model). We demonstrate various features of the method using synthetic examples with complex subsurface features. The numerical results show, e.g., that the dual-source, VA image retrieves subsurface features with "super-resolution", i.e., with resolution higher than the limits of Born imaging, but at the cost of introducing image artifacts not present in the linear image. Although the method does not require any deghosting as a preprocessing step, it can use separated up-and downgoing fields to generate independent subsurface images.

Vector-acoustic reverse time migration of Volve ocean-bottom cable data set without up/down decomposed wavefields

“…Note that here we use a monopole-type source, but with the recent introduction of dual-source technology that combines pressure with gradient/dipole marine sources , handling of ingoing and outgoing waves at the source array will also be possible in the near future by combining the extrapolated receiver fields together with monopole and dipole source side propagators as explained by Vasconcelos (2013). An imaging condition is finally used to map the interaction of source and receiver wavefields in the subsurface and create a representation of the earth's physical property contrasts:…”

“…Shot-profile VARTM (see Appendix A for a vector-acoustic derivation of receiver-profile migration) solves the injection limitations of standard RTM without requiring any preprocessing of the data (Vasconcelos, 2013). Simultaneous injection (indicated by the & symbol in the following equation) of the full pressure and (negative) normal particle velocity with two different types of injection sources, dipole and monopole respectively, enables "on-the-fly" wavefield separation into the up-and downgoing components of the recorded field, with upgoing waves that are injected only downward and downgoing waves that back-propagate only upward:…”

“…To perform mirror imaging, the free surface is removed and the receiver extrapolation domain is extended to a depth of −4.5 km with a wavespeed model in the negative depths that is a mirrored version of the migration velocity model with respect to the free surface (Vasconcelos, 2013). This allows the arrivals related to the ghost wavefield in Figure 3b to continue propagating upward toward negative depths; the final ghost receiver wavefield at negative depths is reflected to positive depths prior to imaging using the same source wavefield as that used for the upgoing images.…”

“…In the context of source-receiver interferometric imaging (Oristaglio, 1989;Halliday and Curtis, 2010), Vasconcelos (2013) formulates migration in a VA fashion and shows that VA reverse time migration (VARTM) can produce images deprived of ghost reflectors, even when taking as input non-decomposed data. The The University of Edinburgh, School of GeoSciences, Edinburgh, UK.…”

Marchenko imaging

“…However, for nonlinear applications, it can be useful to use the redatumed data R ∪ ðx i ; x 0 i ; tÞ and R ∩ ðx j ; x 0 j ; tÞ at two different depth levels ∂D i and ∂D j (with x 3;j > x 3;i ). For example, these responses can be used as input for a new nonlinear imaging method, as introduced by Fleury and Vasconcelos (2012), Ravasi and Curtis (2013), and Vasconcelos (2013) or for full-waveform inversion (Virieux and Operto, 2009), to resolve the parameters of the target zone between these depth levels. Van der Neut et al (2013) discuss an alternative method to combine imaging from above with imaging from below.…”